Section 7:



The section opens with Gareth Olds’ paper, “What Should be Done, and How?”.” Olds suggests that experience with different policy approaches to support entrepreneurs and programs to train entrepreneurs raise questions about their effectiveness and ways to increase firm formation. While liquidity and credit constraints are the most frequent target for policy, these may not be the most productive foci. Rather, insurance mechanisms may play a greater role in entrepreneurial propensity. Olds distinguishes between the policy objectives of creating more entrepreneurs or more entrepreneurial ventures and suggests that perhaps a broader goal—such as overall job growth—without a bias toward one path or another, would allow for greater public and private experimentation.

In “New Economy, Old Challenges Facing Entrepreneurs,” Aparna Mathur discusses the obstacles to entrepreneurship that have existed in the past and are even more difficult in our current economic environment. Labor market challenges, she explains, persist even six years into the recovery from the Great Recession, particularly unemployment and underemployment and with young workers. Mathur explains that the new gig economy offers us new opportunities, but a weaker safety net than we had in the past. Her policy prescriptions to help guide the transition to an economy that offers more stability and opportunity for entrepreneurs include reforming unemployment insurance; expanding training programs and apprenticeships for the young; reforming—and simplifying—the tax system; and rolling back occupational licensing requirements to assist more low-income entrepreneurs.

Usman Ahmed’s essay, “Technology-Enabled Entrepreneurship and an Enabling Environment,” discusses the way the Internet has transformed how new businesses get funding, attract customers, and manage finances. The Internet allows businesses to access markets beyond the local or traditional. For example, 95 percent of businesses on eBay export to other countries. Ahmed recommends that business training programs, both public and private, need to focus on educating entrepreneurs to better leverage the Internet. Furthermore, he suggests that regulation-making processes be made “entrepreneur first,” and that commercial programs run by governments should be tailored and directed to entrepreneurs more effectively.

Brink Lindsey outlines a series of policy recommendations to rekindle economic dynamism in his paper, “Accelerating Growth by Curbing Regressive Regulation.” First, he explains, we need to roll back “regressive regulation”—regulatory barriers that keep out new entrants and result in upward wealth redistribution. In addition, we need to lift restrictions on the exchange of ideas posed by overly protective copyright and patent law, particularly in terms of copyright criminal liability and patent protection on software and business methods. Occupational licensing also needs to be overhauled to make it more justifiable and more aligned with the Supreme Court ruling in North Carolina State Board of Dental Examiners v. Federal Trade Commission. And, permanent resident visas should be granted to foreign students who graduate from American universities with STEM degrees and all graduate degrees. Dependents should not be counted against the cap. Finally, Lindsey argues that land use regulation should be reoriented toward economic dynamism in a variety of ways, including tax rebates to neighbors of new developments and a shift from property taxes to land value taxation.

Focusing on tax policy, Leonard Burman and John Iselin discuss their recommendations for enhancing entrepreneurial activity in their paper, “Tax Compliance Costs and Business Formation.” Parts of the federal tax code can help encourage and support business creation, they explain, but the fixed costs of compliance also present a barrier to entrepreneurial entry. At the point of entry, the tax code can shape decisions made by business owners in terms of entity choice and structure. Therefore, they recommend reducing fixed-tax compliance costs to promote more entry and simplify the choices entrepreneurs face. The authors explain that there are several options for lowering these compliance costs for young and small businesses: simplify parts of the tax code; create simpler accounting methods for small companies; and reduce the complexity created by overlapping state and local tax burdens.

Scott Winship’s paper, “Modernizing Federal Fiscal Policy for Economic Growth,” looks at federal fiscal policy, arguing that changes in this area that reduce the deficit can help increase productivity and renew entrepreneurial growth. He recommends eliminating the corporate income tax and replacing it with accrual taxation of capital gains. The revenue losses could be offset with various tax increases that would fall mostly on upper-income filers. He also suggests eliminating most tax expenditures and reducing individual tax rates. And he suggests shrinking the federal deficit through entitlement reform: reducing Social Security benefit growth among retirees, raising the retirement age, and incentivizing saving through private accounts. Finally, to bring the spending pace down, he recommends that Medicare be changed to a system based on “premium support” and competitive bidding by insurers.

In “How Law Schools Can and Should be Involved in Building Ecosystems that Foster Innovation, Entrepreneurship, and Growth,” Tony Luppino discusses the ways the law shapes the entrepreneurial process, including entry requirements, labor law, and land use regulation, among others. Law schools, he suggests, can help foster entrepreneurial ecosystems by offering direct service to entrepreneurs, providing education on law and entrepreneurship, conducting policy research, collaborating with different government organizations, and devising technologies that reduce legal barriers. Transactional law clinics at law schools across the country, he explains, should continue to find ways to close gaps in legal services between urban and rural entrepreneurs and should recruit established attorneys to help entrepreneurs on a pro bono basis. Law schools also should work with other schools on university campuses—including computer science, business, and public administration—to identify barriers to entrepreneurship and work with government to lower them.

Amar Bhide takes the very specific case of the response to the HIV-AIDS outbreak to draw larger lessons about innovation policy in “How Can We Do Better? Lessons from HIV-AIDS.” This response, he explains, serves as a lesson in how to spur a higher level of multi-player innovation across the economy, as well as a stunning exception to the slow pace of innovation in medicine more generally. Multi-player innovation, he explains, is a key component of entrepreneurial growth; adoption, use, testing, and experimentation by entrepreneurs (and consumers) are just as important as technological breakthroughs. By contrast, a more closed and “artisanal” approach—as in medicine—does not lead to high rates of innovation. Despite many improvements in medical care, the character of innovation in most of medicine is more homogenous and uniform than in other fields, yielding a potentially lower likelihood of breakthroughs. The approach to innovation in medicine and other closed fields should be multi-player, pluralistic, decentralized, and continuously accretive. In medicine, this means having the FDA focus on safety rather than efficacy, and scaling back intellectual property protections. Other fields that are similarly characterized by closed innovation can draw comparable lessons.

What Should Be Done, and How?

By Gareth Olds Assistant Professor in the Entrepreneurial Management Unit Harvard Business School
Gareth Olds
Gareth Olds


Recent efforts to spur entrepreneurship reflect a shift in national priorities over the past few decades. Mounting evidence that entrepreneurship is a key driver of both job creation1 and economic mobility2 have led policymakers to extend capital and training resources in an attempt to increase the total number and overall success of entrepreneurs. These efforts are undertaken largely in the hopes that an untapped wave of entrepreneurial talent lies just below the surface, locked in wage jobs for lack of credit or expertise. And while there is some evidence that self-employment can have positive effects on job satisfaction and wellbeing, these programs generally are couched as part of a larger economic agenda focused on job creation, economic mobility, and growth, rather than social welfare.

What has been done?

Policy interventions around increasing the number of small businesses have fallen into three main groups: improving capital availability, reducing bankruptcy risk, and providing training. Most notable among the first is the Small Business Administration’s program to guarantee up to 85 percent of the value of some small business loans (the UK has a similar program). This incentivizes banks to extend capital to more would-be entrepreneurs, which theoretically should increase the number of new business starts if households face credit rationing in the market for loans.

Regarding bankruptcy, many states have expanded so-called “homestead exemptions” for Chapter 7 filings, shielding a larger set of assets from creditors and making it less risky to take on personal debt to start a small business. These exemptions increase the demand for small business loans, but have a non-monotonic effect.3 Further, personal bankruptcy protection also affects the supply of loanable funds by increasing bank risk, which can push up interest rates and tighten lending.4 Further, because these laws apply only to debt accrued by owners of unincorporated firms, it remains a difficult policy lever for encouraging entrepreneurship.

More recently, public and private organizations have turned their attention to human capital development and training. Starting a small business requires a diverse and particular set of skills that might be difficult for aspiring entrepreneurs to acquire. With this in mind, many attempts have been made to communicate best practices on a low- or no-cost basis. The Kauffman Foundation’s FastTrac program, for example, provides training and networking for entrepreneurs in a number of countries, and the Goldman Sachs 10,000 Small Businesses Initiative aims to accelerate small firms by providing mentorship and training. The Small Business Administration’s e200/Emerging Leaders initiative focuses on underserved communities, and the administration claims the training increases hiring and revenues for enrolled firms.

Small Business Development Centers (SBDCs), a loose network of more than 1,000 nonprofit organizations affiliated with the Small Business Administration, offer classes in starting and growing a business in both urban and rural settings across all fifty states. In a landmark study called Project GATE (“Growing America Through Entrepreneurship”), the Small Business Administration and the Department of Labor randomly assigned individuals who were interested in entrepreneurship to take classes at SBDCs after a thorough intake interview to determine their needs. In total, more than 4,000 people participated in the experiment, half of whom received training in fourteen SBDCs and nonprofit community organizations across seven cities. To date, Project GATE remains the only large-scale randomized control trial to evaluate the effectiveness of entrepreneurship training.

The results were largely negative. Recipients of training were more likely to run a business in the short-term, but no more likely than the control group to be entrepreneurs in the long-term.5 Effects on job satisfaction and income generally were insignificant, though unemployed participants seemed to show relatively larger increases in self-employment, consistent with more recent work on the entrepreneurial potential of unemployed households.6

Less clear is why the GATE experiment failed to increase the rate of entrepreneurship. One possibility is that the skills required to be a successful entrepreneur simply cannot be taught. Less pessimistically, it could simply be that the “best practices” applicable to a venture’s success are so contextual—varying so greatly by industry, geography, or personality—that a universal set of “entrepreneurship skills” simply does not exist. This heterogeneity in the needs of small business owners would demand a different pedagogical approach than class-based learning.

Open questions

There are a number of crucial ingredients to building a policy agenda around spurring entrepreneurship, but three key questions stand out as being unresolved.

  1. What is the biggest barrier faced by potential entrepreneurs?

The resources available to entrepreneurs have expanded enormously in recent decades. Besides stronger bankruptcy protection and a wide array of training programs and centers, the total amount of capital available to new firms has grown markedly, to the point where credit constraints are often non-binding for entrepreneurs in the vast majority of the income distribution.7 The relaxation of this constraint is, in small part, driven by a more aggressive and competitive market for equity financing and the encroachment of angel networks into traditional venture capital territory. However, since the vast majority of new firms are debt- rather than equity-financed, this shift must be largely the result of more sophisticated screening mechanism in assessing creditworthiness (and the accompanying ability to price loans for a larger set of households), as well as the growth of non-loan forms of debt such as credit cards.8

So what is the central barrier to starting a business? Recent work in development economics suggests that low-income households lack consumption-smoothing mechanisms, and that providing insurance facilitates entry into self-employment.9 Uninsured risk also restrains existing small firms from taking advantage of potentially profitable expansion opportunities, whereas liquidity constraints are less binding even in a developing world context.10

Neoclassical views of the world focus on the importance of risk aversion, which is assumed to be negatively related to entrepreneurial ability and potential enterprise value. This assumes people sort based on taste for risk, with more risk-loving individuals more inclined to take on debts and enter self-employment. In reality, the ability to self-insure against income shocks is heterogeneous in the population, and when we observe the binary decision about whether to start a firm, we are collapsing risk tolerance with insurance availability. Doing so tends to underestimate the risk-aversion of entrepreneurial households and overestimate it for those who stay in wage employment: richer households or those with stronger family networks are more likely to become entrepreneurial conditional on the same level of (relative) risk aversion as a poorer one, for the simple reason that they have insurance mechanisms in place to smooth consumption in case the firm fails. This operates independently of a traditional wealth effect, but has similar effects in that it further compounds the asymmetry in entrepreneurship across the wealth distribution.11

However, the fact that uninsured risk from consumption shocks is the most immediate concern in contexts where credit pricing is difficult and capital often scarce does not necessarily mean it is the most important consideration for would-be entrepreneurs in the industrialized world. While capital constraints may be even less binding than in a developing context, social insurance mechanisms are also more sophisticated in richer countries. More work needs to be done to understand what types of risk potential concern entrepreneurs most, and whether these are public goods that can be provided only by governments or whether some policy-driven market-making can coordinate the efforts of private actors.

Demand composition and international trade provide another potential source of asymmetry between entrepreneurs in developed and developing economies, and may explain some of the decline in self-employment in the past several decades. Whereas small manufacturers in low-labor-cost countries often have a choice between selling to the domestic market or exporting abroad, entrepreneurial individuals in richer countries rarely have export opportunities, and, though they have a larger domestic market, demand is increasingly for service-sector goods. While service-oriented entrepreneurs have more insulation from outsourcing than manufacturers do, they are necessarily restricted to the industries where service goods make up the bulk of production.

Further, polarization in the types of service-industry jobs—driven by increased professionalization and stronger barriers to entry in higher-end services and a larger supply of lower-end providers—can drive down the average revenue of a small service-oriented firm, making entrepreneurship less attractive in industrialized countries. Intervention on this front would be much more difficult: in many cases, subsidizing local export-oriented manufacturers amounts to violating free trade agreements, and preferential tax treatment for firms who contract locally for services may also invite regulatory implications.

Finally, human capital acquisition may still play an important role, but in order to accurately place it on a list of policy priorities, we need to understand the nuances of entrepreneurial training and the nature of the unmet demand for skills.

  1. How important is learning by doing in entrepreneurial success?

Assessing the role for entrepreneurial training in a policy agenda requires a deeper understanding of the mechanisms through which individuals acquire the skills necessary to start a venture. One way to analyze this skill formation is through learning-by-doing models, in which steep upfront costs of acquiring human capital rapidly decline with time, but experiential learning is a necessary component of skill development. In other words, if skills cannot be easily transferred between individuals—perhaps because each venture is too idiosyncratic and resists a general treatment or because the general skills needed are simply not relatable between people—then supporting entrepreneurial development means helping people move down the learning curve by building experience rather than providing classroom training.

This framework fits nicely with a large and growing body of work focused on firm age and its effect (if any) on productivity and survival. Experiments that introduce randomness into firm age are few and far between, but extending work on firm age dynamics to cover small, independent businesses in addition to larger firms may lend insight into how important learning by doing is for entrepreneurs.

There is another group whose experience might be valuable to understanding this channel: serial entrepreneurs. Do we need more entrepreneurs or more entrepreneurial ventures? If entrepreneurial skill development isn’t transferrable between individuals but does carry between ventures, serial entrepreneurs can lend insight into how this type of human capital accumulates. The types of skills that serial entrepreneurs develop in the course of their careers can be used as a focal point for encouraging young entrepreneurs to improve their likelihood of survival.

This has important policy implications, since a policy agenda focused on supporting serial entrepreneurs will look very different from one whose goal is to increase the total number of self-employed individuals. If we wish to encourage serial entrepreneurship, then graduated personal bankruptcy protection with stronger protection for firm owners that declines over time might encourage would-be entrepreneurs to “fail early and often.” If we feel there simply need to be more entrepreneurs regardless of experience, we could strengthen the unemployment insurance system and provide benefits to unemployed workers as they get their businesses up and running (seven states already have these programs in place). Understanding how experience affects survival will help hone policies to be more targeted and effective.

  1. How important and heterogeneous is entrepreneurial quality?

Another question whose answer strongly informs the structure of policy interventions deals with how closely entrepreneurial “quality” is linked to success. In one extreme case, where individual ability is orthogonal to firm outcomes, policies that increase the total number of entrepreneurs—but that do not attempt to sort them by quality—will have the greatest success in boosting revenues and job creation. This could be the case if venture quality or geographic location were the most important ingredients to firm survival, with entrepreneurs acting more as stewards of the business opportunity than as captains of their own vessels. In another extreme, managerial excellence and capacity for risk-taking are the key determinants of survival, in which case policymakers should try to design mechanisms for sorting out and supporting high-quality entrepreneurs at the expense of the others.

Some recent trends might shed some light on this quantity-quality problem. While it is widely remarked that entrepreneurship—as measured by the self-employment rate—has been declining for years, a more complicated picture emerges when “entrepreneurship” is defined as ownership of an incorporated firm. Figure 1 plots the proportion of the working-age population who report being self-employed at least part-time (blue line) and who report being self-employed in an incorporated firm (red line) over the last few decades. While the self-employment rate has dropped by 25 percent in this time period, the fraction of people running incorporated firms has risen by nearly 20 percent, with an even larger increase when considering the run-up to the Great Recession. Not only that, but the proportion of businesses that are incorporated (the ratio of the blue line to the red one) has risen even more dramatically, from fewer than one in six to nearly one in four.12 Incorporation often is used as a proxy for high-quality firms, since businesses that incorporate are more likely to be growth-oriented and hire employees.

The secular decline in self-employment may in fact be a good thing, if two conditions are met: (1) a decline in the quantity of entrepreneurs is being met with an increase in average quality, and (2) there exists appreciable heterogeneity in entrepreneurial quality, and it is positively predictive of firm survival, revenues, and job creation. To the extent that predictors of quality can be identified,13 policies devoted to increasing the total number of firms may be misallocating resources away from higher-quality firms.

Moving the goalposts.

Even after resolving these questions and designing a policy agenda around them, there remains the problem of developing metrics to evaluate success. What constitutes a “win” when it comes to promoting entrepreneurship? Do we want more entrepreneurs, more small business loans, or higher quality firms? Should we be happy if we give up growth in the base of small businesses for a higher fraction of serial entrepreneurs?

Since many of the questions that would inform these metrics—such as the nature of entrepreneurial skill development and the role of quality versus quantity—remain unresolved, one possibility is to take an adaptive approach to success and “move the goalposts” as we go. For example, policymakers could choose an ultimate outcome measure of interest, such as employment growth, and then experiment with several policies to find out which interventions are cost effective at boosting it. Then the intermediate outcome measures that the most successful policies had an effect on could be used as evaluation metrics for future policies. It could be that the best way to boost employment is to reduce the number of entrepreneurs and focus attention on high-quality ones, in which case future interventions can be judged on how well they encourage survival of high-potential firms. Alternatively, serial entrepreneurs might be the key to job growth, in which case the average number of firms that entrepreneurs found becomes the focal point.

This adaptive approach also has the advantage of being agnostic as to policy structure and scale. Even relatively cheap and non-invasive interventions such as informing people about the returns to entrepreneurship and making them aware of available resources could have large effects, particularly if there are strong information asymmetries. By focusing on more general equilibrium-level priorities such as job growth, instead of more intermediate outcomes such as the self-employment rate or average revenues, policymakers can encourage a wider array of attempts to promote entrepreneurship that are unbound by narrower views on what constitutes progress.

About the Author
Gareth Olds is an Assistant Professor in the Entrepreneurial Management Unit at HBS, teaching the first-year Entrepreneurial Manager MBA course. His dissertation work examined the effect of public healthcare and income support programs. Professor Olds earned his Ph.D. in Economics at Brown University in 2014, where he received the Merit Dissertation Fellowship, the National Science Foundation IGERT Fellowship, and the Graduate Teaching Award. He also holds an A.M. from Brown University and a B.A. from Whitman College in Walla Walla, Washington.


  1. Kane, Tim J. 2010. “The Importance of Startups in Job Creation and Job Destruction.” Available at:
  2. Fairlie, Robert W. 2005. “Entrepreneurship and earnings among young adults from disadvantaged families.” Small Business Economics 25.3: pp. 223–236.
  3. Georgellis, Yannis, and Howard J. Wall. 2006. “Entrepreneurship and the Policy Environment.” Federal Reserve Bank of St Louis Review, Vol. 88, No. 2.
  4. Berkowitz, Jeremy, and Michelle J. White. 2004. “Bankruptcy and Small Firms’ Access to Credit.” RAND Journal of Economics, 35 (1), pp. 69–84.
  5. Fairlie, Robert W., Dean Karlan, and Jonathan Zinman. 2012. “Behind the gate experiment: Evidence on effects of and rationales for subsidized entrepreneurship training.” Working Paper No. 17804. National Bureau of Economic Research.
  6. Hombert, Johan, Antoinette Schoar, David Sraer, and David Thesmar. 2014. “Can Unemployment Insurance Spur Entrepreneurial Activity?” Working Paper No. 20717. National Bureau of Economic Research.
  7. Hurst, Erik, and Annamaria Lusardi. 2004. “Liquidity Constraints, Household Wealth, and Entrepreneurship.” Journal of Political Economy, Vol. 112, No. 2, pp. 319–347.
  8. Chatterji, A., and R. Seamans. 2012. “Entrepreneurial Finance, Credit Cards and Race.” Journal of Financial Economics, 106(1): pp. 182–195.
  9. Bianchi, Milo, and Matteo Bobba. 2010. “Liquidity, Risk, and Occupational Choices.” Available at:
  10. Karlan, Dean, Chris Udry, Robert Osei, and Isaac Osei-Akoto. 2014. “Agricultural Decisions after Relaxing Credit and Risk Constraints.” Quarterly Journal of Economics, 129(2).
  11. Of course, if relative risk aversion is constant, absolute risk aversion declines with wealth.
  12. This is a naive calculation that ignores the possibility that individuals are involved with multiple businesses.
  13. Levine, Ross, and Yona Rubinstein. “Smart and Illicit: Who Becomes an Entrepreneur and Does it Pay?” NBER Working Paper No. 19276.

New Economy, Old Challenges Facing Entrepreneurs

By Aparna Mathur Resident Scholar In Economic Policy Studies American Enterprise Institute
Aparna Mathur
Aparna Mathur



The U.S. economy is facing many challenges today. Economic growth and dynamism are down. This shows up in the lower rates of entrepreneurship that have been observed over the last decade, as well as in the extended periods of slack in the labor market, with high levels of long-term unemployment and youth unemployment. At the same time, many of the traditional factors that affect entrepreneurial dynamism, such as policy uncertainty, taxes, sales, and financing, are even more challenging in this economic environment. The new economy, or the gig economy, presents many opportunities. On the one hand, it offers individuals who otherwise might be left out of the entrepreneurial process an opportunity to behave in an entrepreneurial manner by sharing their skills, assets, and time. However, this new economy lacks much of the safety net that perhaps comes with traditional jobs. In this paper, I discuss these new and old challenges and offer some policy ideas that might help the transition to an economy that offers more stability, security, and opportunity for American entrepreneurs and their families.

I. Introduction

Entrepreneurship is always challenging. The only difference is that sometimes the economy presents new challenges that may change the nature of entrepreneurship itself, and sometimes the economy presents age-old challenges that change the outcomes for entrepreneurial ventures, but not the nature of entrepreneurship itself. With regard to the former, I discuss the new sharing economy, which has allowed individuals across all demographics to be more entrepreneurial by “sharing” their skills, assets, and time. At the same time, older challenges, such as taxes, regulations, and uncertainty persist, which ultimately will affect the entrepreneurial success of these ventures. In the future, as the new economy takes over and spreads widely, it is likely that these two challenges will become one as the new entrepreneurs are faced with the same old set of regulations and tax policies that the older entrepreneurs have faced. But, for now, it seems like the two entrepreneurial economies are moving in parallel: one, the economy that shows up in official statistics, and the second that is visible but largely hidden in official statistics. This paper attempts to shed light on both.

The Great Recession took a big toll on entrepreneurship and overall business creation. The Kauffman Index of Entrepreneurial Activity captures the rate of new business creation at the individual owner level.1 The businesses captured include all incorporated or unincorporated businesses, as well as employer and non-employer businesses. Data from this index show that entrepreneurial activity increased over the course of the recession, but has recently started to decline. The increase in entrepreneurial activity at the start of the recession most likely was a consequence of the weak labor market, because a significant share of individuals starting businesses were unemployed. The latest data from 2013 suggest that entrepreneurial activity has declined between 2012 and 2013, but a smaller share of owners is coming out of the pool of unemployed. In other words, “necessity” entrepreneurship may have declined marginally. The share of “opportunity” entrepreneurship, which captures individuals coming out of other labor market states, declined significantly over the recession and only recently has shown a slight uptick.

Data from other sources suggest similar trends. Establishment births declined over the recession, as shown by data from the Business Employment Dynamics database, but recently have started returning to pre-recession levels.2 However, job creation by these businesses is still fairly low. New establishments that are coming up are smaller on average and are staying small, which is affecting overall employment levels. The average employment size of a five-year-old establishment born in 2007 was almost one-third lower than one born ten years earlier. As a consequence, the five-year employment level from establishments born in 2007 is 1.2 million less than the five-year mark for establishments born in 1997.3

These data suggest that there is a clear correlation between economic conditions and entrepreneurship. While economic growth may, on the one hand, lead to fewer business startups because unemployed workers are not forced to start businesses out of necessity, economic growth also may lead to higher rates of entrepreneurship as the improvement in economic conditions may offer better prospects for businesses’ survival, as well as ease of access to finance and other factors that may facilitate startups. Further, entrepreneurship itself may lead to economic growth if startup businesses are successful, grow quickly, and employ workers.

Therefore, we need to better understand the factors that affect entrepreneurial activity, small business formation and survival, and large employer businesses. In this paper, I will use that information to suggest ways in which we can help promote “opportunity” entrepreneurship by addressing the new and old challenges faced by these businesses.

II. New Economic Challenges Facing Entrepreneurs

Economic growth and labor market dynamism

As mentioned in the introduction, entrepreneurship and economic dynamism may be highly correlated. Per a recent paper by Roger Gordon, the growth rate of real per capita GDP has been slowing down and will slow down further in the next few decades.4 This is a consequence of many factors, but primarily the decline in the labor force participation rates for prime age adults and youth, the decline in educational attainment as captured by the high dropout rates in high school and college, and the lack of real wage growth at the bottom of the income distribution. Nor will this slowdown be offset by rapid changes in the rate of technological innovation.

The slow recovery in the labor market shows up in more than the decline in labor force participation. Today, nearly 6.5 million persons are employed part-time involuntarily because they could not find full-time jobs. Meanwhile, another 1.9 million want jobs, but have not searched for work in the previous few months because they are too discouraged to look. Another 2.1 million workers have been jobless for twenty-seven weeks or more.5 This labor market slack has resulted in weak wage growth. A recent study from the Federal Reserve Bank of Atlanta suggests that the large supply of unemployed and underemployed workers, which includes part-time workers who want full-time jobs, could be restraining wage growth.6 Further, while full-time workers experienced year-over-year median wage gains of 3 percent in 2014, part-time workers experienced wage growth of about half as much as that of full-time workers from 2011 to 2013. Hence, an elevated share of part-time workers may have depressed overall hourly wage growth.

The labor market also is seeing an increasing polarization in work. Polarization can be a consequence of automation, which happens when machines or computers replace workers in routine tasks. It can also occur due to international trade or the offshoring of jobs to countries where wages are lower. Jobs in the middle part of the skill distribution are thought to be particularly vulnerable to these two factors, because these more generally involve routinized tasks rather than high-skill or low-skill jobs.7, 8 Data show that, over the Recession, middle-skill jobs experienced a sharper and more long-lasting employment decline than high- or low-skill jobs did. The paper finds that middle-skill workers (primarily in manufacturing, construction, or clerical work) with no college degrees typically leave unemployment and exit the labor market, rather than finding low-skill or high-skill jobs. This can be explained by the fact that, when transitioning out of a job, a middle-skill worker needs to invest in more education to get a high-skill job. About 85 percent of high-skill workers have taken at least some college courses, as opposed to half of middle-skill workers. On the other hand, middle-skill workers earn considerably more than low-skill workers, which affects their willingness to take up low-skill jobs. In general, participation appears to be a function of education. Workers with the lowest education levels have the lowest participation rates. Results from this study suggest that falling labor force participation among prime-age males can be primarily explained by a lack of demand for middle-skill workers, providing a strong link between polarization and labor force participation.

Another challenge facing the economy is the high levels of youth unemployment. Per research compiled by the U.S. Chamber of Commerce, one-half to two-thirds of millennials are interested in entrepreneurship, and more than 25 percent are already self-employed. In 2011, millennials launched almost 160,000 startups each month, and 29 percent of all entrepreneurs were twenty to thirty-four years old.9 It is important to encourage this trend, because startups are essential to job creation in the United States, and there has been a large dip in new business creation since the recession.

How do we address these challenges? While there are many policies that potentially could improve economic dynamism and labor market slack, I will focus on a couple that relate directly to entrepreneurship and the new economic challenges.

First, how do we encourage entrepreneurship for a generation that is burdened by student loans, has low levels of work experience, and is frequently turned down by banks for startup loans?

From my perspective, the real issue is that millennials very often lack the necessary expertise and experience to start and run a business. Work skills can be earned while still in school through vocational training and apprenticeship programs, or can be obtained after school through work and on-the-job training. For millennials, the recession and subsequent unemployment resulted in fewer opportunities to gain these skills after school. At the same time, they regret that they did not get enough work experience while still in school. This is evident from another Pew survey that asked college graduates if, in retrospect, they could have better prepared for the types of jobs they wanted. About half of all college graduates said getting more work experience while still in school would have put them in a better position to get the kinds of jobs they wanted. Another three in ten said they should have started looking for jobs sooner or picked different majors.10

Confirming this view, a Kauffman Foundation poll of 872 millennials found that 92 percent support increased access to education and training needed to run a small business as a way to encourage people to become entrepreneurs. A smaller number, 81 percent, supported student loan relief and 83 percent believe that Congress should increase the availability of startup loans.11

To help prepare this younger generation for the job market, we need better training programs at high schools and colleges and more apprenticeships in partnership with businesses and industries around the country. These programs could be run as tax credit programs for employers that offer these kinds of training programs. This would incentivize businesses to participate by offsetting some of the program costs. Such a program is currently being tried in South Carolina through the Apprenticeship Carolina program, which has doubled the number of apprenticeship offerings in the state.

Second, to help the unemployed and, especially, the long-term unemployed, we need to reform the unemployment insurance system. One idea is to use unemployment insurance (UI) funds to help startups by the unemployed. Research suggests that programs such as these in France and Germany have been successful at helping the unemployed become entrepreneurs. For instance, in France, the reform allowed unemployed individuals who started their own businesses to keep their access to unemployment insurance for three years in case their business ventures failed.12 Before the new rule, an entrepreneur would lose his unemployment insurance once he declared self-employment. The main objective of the program was to provide insurance against failure and shortfalls in cash flow during the first three years. The study suggests that the program was successful. The monthly entry of new businesses increased by 12 percent post-reform. More importantly, the quality of new firms did not deteriorate; there were no significant differences in the failure rate, hiring rate, or growth rate of young firms in the industries where the reform had the most impact. In addition, the unemployed entrepreneurs who started these businesses were ambitious about the growth prospects and were interested in hiring new workers. The overall positive benefits included shorter unemployment spells and the reallocation of labor to more productive and higher-paying jobs. This would involve reforming our current Self-Employment Assistance programs that are available only in a few states and to a limited pool of UI recipients. One reason for the low participation is that the program only offers support to workers who are permanently separated and likely will exhaust their unemployment benefits. This likely affects the quality of workers who are eligible for these programs. A redesigned program could make support available to all unemployment insurance recipients across the country and could provide some level of downside insurance for a limited period of time for a new business.

Another idea is to use UI funds to match workers directly to jobs through a wage subsidy. UI funds could supplement incomes for workers as they undergo training at an employer firm. This could help these workers attain skills that would improve their productivity on the job, and also benefit employers. The more skilled the workers become, the more likely they would be to contribute to productivity growth and respond to technological changes.

The new sharing economy

The new sharing economy, or gig economy, is posing its own sets of challenges and opportunities and creating a new group of “micropreneurs.” On the one hand, it provides flexibility and offers opportunities to individuals across all demographics. In these times of uncertainty, it provides an additional avenue for earning an income through assets such as cars, houses, time, and abilities that people already possess.13 People with an extra room can offer a room for rent. People with a car can offer rides to other passengers. Sixty-two percent of New York Airbnb hosts say that being able to provide this service helped them stay in their homes.14 Anecdotal evidence suggests that people doing two to three tasks a day through TaskRabbit can earn almost $3,500 a month. Full-time workers can earn as much as $6,000–$7,000 a month.15

On the downside, these are not traditional jobs that provide benefits and security of tenure, but, in many ways, they allow people to supplement incomes, earn enough for their families, or boost retirement savings. But then, it’s not clear that we need benefits to be tied to jobs, either. With the Affordable Care Act, workers can buy health insurance individually on the exchanges instead of getting it through their jobs, which improves their ability to be mobile across jobs. There are new technologies that allow these 1099 employees to keep track of their taxes and do withholdings automatically. TaskRabbit has introduced a wage floor, making it impossible for workers to earn less than $12.80 an hour.16 That’s higher than any state minimum wage in the United States. Guilds like and Freelancers Union are creating ways for independent contractors to pool bargaining power to access discounted health insurance and telecom plans.

In fact, a recent study by Alan Krueger of Uber shows that independent contractors represented 7.4 percent of the workforce in 2005, up slightly from 6.7 percent in 1995. Eighty-two percent of these workers in 2005 reported that they preferred their work arrangements to traditional jobs. This was despite the fact that they knew that workers in those traditional relationships were more likely to have health insurance coverage.

At Uber, 19 percent of the drivers are under the age of thirty, and 25 percent are aged fifty or older. They are highly educated, with 48 percent having college degrees. Sixty-six percent still have full-time jobs, and 8 percent had been unemployed prior to becoming Uber drivers. About 25 percent are actively looking for full-time jobs, and another 25 percent are looking for part-time jobs. About 38 percent of people do this as their main job, working more than thirty-five hours per week.

About 91 percent said that this job helps them earn more income to better support their families, 87 percent value the flexibility, and 74 percent said that the job helps them maintain a steady income because other sources of income are unstable or unpredictable. Women drivers were more likely to state flexibility as an important benefit. It is also important for the drivers to earn an income while looking for another job. Sixty-one percent said that it strengthened their sense of financial security.

Uber provides driver-partners with access to a service called Stride Health to help them select health insurance coverage that is appropriate for their situations. So, it does seem that, going forward, there will be fewer formal jobs, but more sharing in work, a more entrepreneurial approach to work, and cooperation in the production of goods and services.

III. Older, Persistent Problems Facing Entrepreneurs

Moving beyond the new economic challenges typified by the loss in business dynamism and labor market slack, there are older, more long-term challenges that have persisted for entrepreneurs. These are the subject of much research. For instance, we know that the highest rates of job creation come from young and small businesses.17 However, per the most recent (April 2015) report from the National Federation of Independent Businesses, small business owners are, on average, less optimistic than they were before the recession.18 They still face tremendous uncertainty about the future, coming from poor expectations of sales, hiring, and government regulations. Nearly two-thirds of owners in a Bank of America survey said that their businesses still have not recovered completely from the Great Recession. In another survey by Pepperdine University’s Graziado School of Business and Management and Dun & Bradstreet Credibility Corp, 63 percent of small businesses said it was difficult to get loans, while only 36 percent of larger businesses said it was difficult.19 These results are also reflected in the National Federation of Independent Business data. Per the latest NFIB survey, 11 percent of small business owners cited weak sales as their top business problem. A large majority of business owners cited government regulations and red tape as the biggest hurdle (23 percent), followed by taxes (22 percent) and quality of labor (12 percent). A much smaller fraction cited cost and availability of insurance, competition from other businesses, and cost of labor as factors.20 I will address each of these in turn and recommend some policy ideas to overcome these challenges.

Policy uncertainty

There is a rapidly growing body of literature that focuses on the effect of uncertainty on business or economic activity. A new paper by Baker et al. investigates whether uncertainty about taxes, government spending, and other policy matters deepened the 2007–2009 recession and slowed the recovery. The authors develop a new index of policy-related economic uncertainty and estimate its dynamic relationship to output, investment, and employment. The index averages several components that reflect the frequency of news media references to economic policy uncertainty, the number of federal tax code provisions set to expire in future years, and the extent of forecaster disagreement over future inflation and federal government purchases. Vector autoregressive (VAR) model estimates show that an increase in policy uncertainty equal to the actual change between 2006 and 2011 foreshadows large and persistent declines in aggregate outcomes, with peak declines of 2.2 percent in real GDP, 13 percent in private investment, and 2.5 million in aggregate employment.

Another paper by Bachmann et al. (2010) uses microdata from the Federal Reserve Bank of Philadelphia’s Business Outlook Survey and Germany’s IFO Business Climate Index to investigate how measures of business uncertainty, which are derived from managers’ business expectations, are related to economic activity. They find that increases in business uncertainty are associated with prolonged declines in economic activity. Rodrik (1991) shows how policy uncertainty can act as a tax on investment and cause firms to forego investments until its resolution.

A 2011 paper by economists at the Federal Reserve Bank of Cleveland also finds that policy uncertainty specifically affected small firm hiring and capital investments over the period 1986 to 2011.21

While it is impossible to argue that there should be no uncertainty about the policy climate facing small businesses, it is important not to ignore the idea that uncertainty matters for small business hiring and expansion. It also is important to recognize that temporary policies can create uncertainty, because business owners need to be able to forecast costs and revenues. Finally, legislation that is complex and hard for businesses to unravel creates its own issues. An example of this is the Affordable Care Act, which I will discuss in the next section. However, an issue with that legislation is that not only was there uncertainty about when and whether it would be passed and would come into full effect, and about what costs would be associated with it, but the legislation itself is complex and hard for businesses to understand and incorporate in their decision making. This is likely to have affected business owners’ decisions relating to investment and hiring, which are particularly important in today’s economy.

Health care costs

New government regulations, such as the Affordable Care Act, likely are holding back entrepreneurial growth. Many business owners are confused about what their health care liabilities are likely to be when they are forced to provide health insurance to employees or pay a penalty. The employer mandate, which took effect in 2014, applies to all firms with fifty or more employees. If an employer fails to provide “qualifying health insurance,” the employer has to pay a per-employee excise tax fine. The tax is $2,000 per employee ($3,000 if an employee receives coverage through an exchange). If the employer has a waiting period to get into the plan, there is an additional tax of $400 to $600. A small employer with 100 employees could easily find himself paying a tax of $300,000 per year. Even if an employer provides health insurance, it can be deemed “unqualified” by the U.S. Department of Health and Human Services.22

Further, many self-employed people choose not to have health insurance—because either they would rather self-insure, they have a mini-med plan, or they want to retain capital in their small businesses. The Affordable Care Act strips this choice. Starting in 2015, everyone was required to obtain “qualifying” health insurance or face an excise tax of at least 2.5 percent of adjusted gross income. For those who are employees of their own firms of at least fifty employees, the firm also would be subject to the employer mandate excise tax described above.

 A number of papers have focused on the effect of health care costs on entrepreneurship. Gruber (1992) finds that health insurance mandates reduce employee coverage in small firms by as little as 1 percent. This is similar to the finding by Gabel and Jensen (1989), though in a 1992 survey, they showed that 19 percent of sampled small firms did not offer coverage due to state-mandated benefits. Of the papers linking health insurance and entrepreneurship, Gruber and Poterba (1994) analyze the impact of the Tax Reform Act of 1986, which allowed self-employed individuals to deduct a certain percentage of their health insurance costs from their taxable income, thus bringing them closer to the tax treatment afforded to employer-provided health insurance. Their paper suggests that a 1 percent increase in the cost of health insurance coverage would reduce the probability for coverage for self-employed households by 1.8 percent. Perry and Rosen (2001) find a statistically negative effect of self-employment on the probability of being insured.

A paper that I published (Mathur 2009) focused specifically on state health insurance mandates and their impact on job creation by small firms. Health insurance is regulated at the state level by state-mandated health benefits.23 These are regulations issued by the state that mandate minimum levels of certain benefits as part of policies offered (e.g., chiropractic services, mental illnesses, etc.).24 The cost effect of mandates varies due to differences in state laws. For example, mandated benefits accounted for about 12 percent of claims costs in Virginia in 1993, 22 percent of claims in Maryland in 1988, and 5 percent of claims in Iowa in 1987.25 The studies that reported the highest costs were those for Maryland and Massachusetts, which have more mandated benefits than most other states do. Unlike a lot of earlier studies that focused on the 1980s, the study focuses on the 1990s, when there was a tremendous increase in the number of mandates passed by states. The number of states with six or more mandated benefits increased dramatically between 1988 and 1997. The evidence strongly suggests that, while some mandates matter more than others in the job creation decision of small firms, the most significant impact on small firms is simply in terms of the total number of mandates in a state: the larger the number of mandates, the lower the probability of employment generation. Studying the predicted probabilities for different levels of mandated benefits, the data show a clear negative relationship between the size of the firm and the total mandated benefits. The predicted probability of owning a business with more than one employee goes down from 0.45 to 0.34 (i.e., nearly ten percentage points) as the number of mandates goes up from zero to sixteen. The probability of owning a firm with more than two employees goes down by nearly 50 percent for the same change in mandated benefits, and by about 35 percent for firms with six or more employees.

The fact that the ACA employer penalty is based on the number of employees could create a disincentive for firms to add workers. This is aggravated by the fact that a recent study of small employers with fewer than ten employees found that 56 percent of the employers misunderstood portions of the ACA’s employer penalty, 32 percent believed they would be required to provide health insurance, and 24 percent believed they would have to pay a penalty.26 A related concern is that firms may try to cut worker hours below the threshold for full-time, which is thirty hours a week under ACA. Perhaps one way to avoid these losses is to change the definition of full-time to forty hours per week.

Some ACA provisions aim to reduce the burden on small firms.27 For example, small business health care tax credits offset the costs of providing insurance to employees. According to the GAO, many business owners have complained that the size of the credit is too small to create the incentive to begin offering insurance, and the rules are too complex. One way to fix these issues is to expand the credit and simplify the rules, as suggested in a recent Congressional Research Service (CRS) report.28 Other ideas include a deduction for health insurance costs for the self-employed for themselves and their families.


The recent NFIB survey rated taxes, aside from regulations, as the most important problem facing small businesses.29 The problem is not only high taxes, but also the complexity of the tax code. Many small employers pay taxes using the individual tax brackets, and a large share of small employer profits are taxed in the top bracket. According to the Tax Foundation, many small employer businesses face top marginal rates between 44 percent and 48 percent.30 These businesses account for 67 percent of business income. These high tax rates discourage investment and employment creation. Therefore, reforming the tax code to lower the marginal rates on these employer businesses would help the economy.

The most recent tax hike comes from the Affordable Care Act, which imposes new taxes on businesses. Under the ACA, the first tax increase on small employers is a Medicare payroll tax hike. The Medicare payroll tax for wages and self-employment earnings in excess of $250,000 for couples ($200,000 for singles) will rise from 2.9 percent to 3.8 percent. This is a direct tax hike in the marginal income tax rate paid by the self-employed and general partners.

Starting in 2013, the Affordable Care Act also imposes a 3.8 percentage point hospital insurance tax on investment income of more than $250,000. Some of this so-called “investment income” is actually small business profits. Notably, investors in small businesses (limited partners and passive shareholders in Subchapter-S corporations) will face this tax. Active trade or business income is excluded, but, of course, most of that will face the higher Medicare tax described above. This provision will make it harder for employers to raise capital to create jobs and expand business operations.

What are the effects of high taxes on entrepreneurship? There is vast academic literature studying this topic. Many papers have studied transitions from wage and salary employment to entrepreneurship as a function of the tax rates faced by individuals and firms. This option is valuable to the extent that personal income is taxed at a higher rate than corporate income is. In recent years in the United States, the corporate tax rate for a small firm could be as low as 15 percent, which is below the marginal personal (plus payroll) tax rate faced by, effectively, all individuals. As a result, a firm generating tax losses will prefer to be non-corporate so that the entrepreneur can deduct these losses against other personal income, saving on personal income taxes. In contrast, when and if the firm generates profits, for tax purposes, the entrepreneur will prefer to incorporate so that these profits are taxed at the lower corporate tax rate. The paper by Cullen and Gordon (2002) shows that reducing the minimum corporate income tax by five percentage points leads to a doubling of entrepreneurial activity in different quintiles and in the aggregate. If personal income tax rates were cut by five percentage points, this would lead to a nearly 30 percent drop in entrepreneurial activity, with larger percentage drops in the highest-earning quantiles. Finally, a flat tax of 20 percent would increase self-employment activity by 15 percent. Such a tax cut reduces the taxes saved from deducting business losses, while profits remain largely taxed at the corporate tax rate. As a result, risk-taking is discouraged. In addition, as emphasized by Domar and Musgrave (1944), a lower personal tax rate implies less risk-sharing with the government, in itself making self-employment less attractive to risk-averse individuals. The potential tax savings from going into business simply to reclassify earnings as corporate rather than personal income for tax purposes also falls when personal tax rates fall.

In another paper, Gentry and Hubbard (2000) show that the less progressive the income tax schedule is, the greater the incentive to entrepreneurial entry. Gentry and Hubbard (2000) emphasize a different effect of the tax system on risk-taking that arises even if investors are risk-neutral. If the marginal tax rate under the personal income tax is an increasing function of taxable income, then entrepreneurs are able to save little in taxes on any losses they incur but can owe substantial taxes on any profits. The more progressive the tax schedule, therefore, the more risk-taking lowers the expected after-tax return from the project. As a result, a progressive rate schedule discourages risk-taking.

Bruce (1998) similarly finds that taxes have significant effects on the probability that an individual will leave a wage and salary job to become self-employed. Estimates indicate that a five percentage point increase in the difference between an individual’s expected marginal tax rates in wage and salary employment and self-employment reduces his transition probability by about 2.4 percentage points.

A different strand of literature focuses on the effect of the entrepreneur’s own taxes on his or her ability to hire workers and expand investment. Carroll et al. (1998) analyze the income tax returns of a large number of sole proprietors before and after the Tax Reform Act of 1986 and determine how the substantial reductions in marginal tax rates associated with that law affected the decision to hire labor and the size of those individuals’ wage bills. The authors find that raising the entrepreneur’s “tax price” (one minus the marginal tax rate) by 10 percent raises the probability of hiring workers by about 12 percent. Further, conditional on hiring employees, taxes also influence total wage payments to workers. A 10 percent increase in the tax price would increase the median wage bills of entrepreneurs by 3 percent to 4 percent. These effects are more pronounced for high-income sole proprietors. Therefore, raising tax rates on high-income entrepreneurs could result in lower wages for workers employed at these firms.

Using a similar dataset, Carroll et al. (1998) also study capital investment decisions by entrepreneurs. Taxes affect the demand for investment through their impact on the user cost of capital. An increase in the personal tax rate raises the user cost and negatively affects investment. Another channel through which taxes affect investment is liquidity constraints. An increase in taxes reduces the entrepreneur’s cash flow. To the extent that liquidity constraints are present, this leads to a reduction in the demand for capital. The authors investigate both channels and find that the substantial reductions in marginal tax rates for the relatively affluent had quantitatively significant influences on their investment decisions. A five percentage point increase in marginal tax rates reduced the proportion of entrepreneurs who made new capital investments by 10.4 percent and decreased mean expenditures by 9.9 percent.

In another closely related paper, the authors find that income taxes exert a statistically and quantitatively significant influence on firm growth rates. Raising the proprietor’s tax price by 10 percent increases gross receipts by about 8.4 percent. This finding is consistent with the view that raising income tax rates discourages the growth of small businesses.

Another challenge for large corporations is corporate tax reform. Today, the United States has the highest corporate tax rate in the Organisation for Economic Co-operation and Development. This discourages investment in the United States, as capital flows to low-tax jurisdictions. In addition, the U.S. rules for international taxation are based on the worldwide system of taxation, which place firms at a competitive disadvantage, encouraging them to engage in corporate tax inversions and creating incentives to retain foreign earnings overseas.

We need to reform corporate taxes, personal taxes, and the system of international taxation. Perhaps lowering the headline rate, eliminating some business tax deductions, and moving toward a system of territorial taxation would be a step in the right direction.

Other ideas include increasing the small business expensing limit so businesses are able to immediately recover costs instead of claiming depreciation deductions over specified periods, repealing the AMT, and making healthcare costs deductible, same as for large employers.

Financing issues

In the NFIB survey, 53 percent of owners said that they did not want to borrow.31 This weak loan demand is indicative of the pessimism of small business owners who are reluctant to take loans that they may not be able to repay. So the fact that only 2 percent of owners reported that financing was their top business problem does not provide the complete picture. Over the course of the recession, between 2008 and 2011, the number of owners with a business loan declined from 44 percent to 29 percent.32 As per a more recent study, small business loans (less than $1 million) dropped significantly between 2008 and 2012 to a level 17 percent below that prior to the recession.33 In contrast, lending to larger businesses has rebounded faster.

Small business loans are down because of weak sales. Despite increases in recent and expected future revenues, revenues still have not reached levels seen prior to the recession.

To the extent that small businesses are facing a credit crunch, encouraging the growth of alternative lending programs such as online lending may improve access. Some studies suggest that loan approval rates for small businesses are much higher from alternative lending sources than from traditional, commercial banks.34 Online lending platforms offer applications that can be completed faster than all the paperwork required by large banks, and they can be approved and the loans obtained faster, as well.

Occupational licensing

Differences in state licensing requirements can make it difficult for entrepreneurs and workers to find opportunities and jobs, creating stale labor markets and under-employment. Mutual recognition of other state licenses would improve worker mobility and, thereby, boost economic dynamism. Stephen Slivinski, a senior economist at the Goldwater Institute, finds that states with strict occupational licensing standards have lower levels of low-income entrepreneurship.35 The average low-income entrepreneurship rate is higher than the national average, at 380 entrepreneurs per 100,000 low-income residents. Yet, some states have higher rates than average and some have lower rates. To explain that difference, the study matches data from the Kauffman Foundation and the Institute for Justice for the first time ever and discovers that the higher the rate of licensure of low-income occupations, the lower the rate of low-income entrepreneurship. The states that license more than 50 percent of the low-income occupations had an average entrepreneurship rate 11 percent lower than the average for all states. The states that licensed less than a third had an average entrepreneurship rate that was about 11 percent higher. Colorado, Vermont, and New Mexico have the highest rates of low-income entrepreneurship in the country, while Kentucky, Wisconsin, and Mississippi have the lowest.36

A separate Aspen Institute study found that half of all low-income entrepreneurs leave poverty within five years.37 To encourage entrepreneurship among low-income workers, the report recommends reforming state occupational licensing laws so they aren’t barriers to entry. Those reforms should include eliminating some existing licensing requirements and creating a private-sector-directed certification system that would replicate what exists in many industries already.

IV. Conclusion

The U.S. economy is facing many challenges. Economic growth and dynamism are down. This shows up in the lower rates of entrepreneurship that have been observed over the last decade, as well as in the extended periods of slack in the labor market, with high levels of long-term unemployment and youth unemployment. At the same time, many of the traditional factors that affect entrepreneurial dynamism, such as policy uncertainty, taxes, sales, and financing, are even more challenging in this economic environment. The new economy, or the gig economy, presents many opportunities. On the one hand, it offers individuals who otherwise might be left out of the entrepreneurial process an opportunity to behave in an entrepreneurial manner by sharing their skills, assets, and time. However, this new economy lacks much of the safety net that perhaps comes with traditional jobs. In this paper, I discuss these new and old challenges and offer some policy ideas that might help the transition to an economy that offers more stability, security, and opportunity for American entrepreneurs and their families.

About the Author
Aparna Mathur is a resident scholar in economic policy studies at the American Enterprise Institute. She received her Ph.D. in economics from the University of Maryland, College Park in 2005. She has testified before Congress numerous times and published articles in top scholarly journals and the popular press on issues of policy relevance. Her work has been cited in academic journals and leading news magazines such as the Economist, the Wall Street Journal and Businessweek. She has been an adjunct professor at Georgetown University’s School Of Public Policy and has taught economics at the University of Maryland.


Bachmann, R., S. Elstner, and E. Sims. 2011. Uncertainty and Economic Activity: Evidence from Business Survey Data. American Economic Journal: Macroeconomics, 5(2):217–249. 

Baker, S., N. Bloom, and S. Davis. 2013. Measuring Economic Policy Uncertainty. Chicago Booth Research Paper No. 12-02.

Bruce, D. 1998. Effects of the United States Tax System on Transitions into Self-Employment. Syracuse University.

Carroll, R., D. Holtz-Eakin, M. Rider, and H. Rosen. 1998. Entrepreneurs, Income Taxes, and Investment. NBER Working Paper No. 6347.

Cullen, J., and R. Gordon. 2002. Taxes and Entrepreneurial Activity: Theory and Evidence for the U.S. NBER Working Paper No. 9015.

Domar, E.D., and R.A. Musgrave. 1944. Proportional Income Taxation and Risk-Taking, Quarterly Journal Economics 58: 388–422.

Gentry, W., and R. G. Hubbard. 2000. Tax Policy and Entrepreneurial Entry.

Gruber, J. 1994. The Efficiency of a Group-Specific Mandated Benefit: Evidence from Health Insurance Benefits for Maternity. NBER Working Paper No. 4157.

Gruber, J., and J. Poterba. 1994. Tax Incentives and the Decision to Purchase Health Insurance: Evidence from the Self-Employed. The Quarterly Journal of Economics: 109(3): 701–733.

Jensen, G., and J. Gabel. 1989. State Mandated Benefits and the Small Firm’s Decision to Offer Insurance. Journal of Regulatory Economics: 4(4):379–404.

Krueger, A., and J. Hall. 2015. An Analysis of the Labor Market for Uber’s Driver-Partners in the United States. Princeton University. Industrial Relations Section; 587. 

Mathur, A. 2009. Health Insurance and Job Creation by the Self-Employed. AEI. 

Perry, C., and H. Rosen. 2001. The Self-Employed Are Less Likely To Have Health Insurance Than Wage Earners. So What? NBER Working Paper No. 8316.

Rodrick, D. 1991. Policy Uncertainty and Private Investment in Developing Countries. NBER Working Paper No. 2999. 


  5. The most recent data from the Bureau of Labor Statistics is available at
  6. Federal Reserve Bank of Atlanta. “Wage Growth a Missing Piece of the Full Employment Puzzle.” 2014 Annual Report. 
  7. Low-skill workers typically include cooks, waiters, security guards, etc.
  8. Foote, Christopher L., and Richard W. Ryan. 2015. “Labor market polarization over the business cycle.” National Bureau of Economic Research Working Paper No. 21030.
  12. Hombert, Johan, Antoinette Schoar, David Sraer, and David Thesmar. “Can Unemployment Insurance Spur Entrepreneurial Activity? Evidence From France.” 
  23. There are mandated providers, as well, but we have only included mandated benefits in our study.
  24. For example, the mental health illness mandate in Montana specifies that firms must offer minimum thirty days of inpatient services.
  25. General Accounting Office. 1996.

Technology-Enabled Entrepreneurship and an Enabling Environment

By Usman Ahmed Head of Global Public Policy at PayPal and Adjunct Professor of Law at Georgetown University Law School
Usman Ahmed
Usman Ahmed


The Internet has changed nearly aspect of our lives: how we communicate, work, travel, and engage with the world. So it is no surprise that the Internet also has fundamentally changed the way a new business secures funding, finds customers, and manages finances. It is essential to reflect on how the process of entrepreneurship has been profoundly shifted by technology and on the tremendous socio-economic benefits that have resulted from this trend. And, it is also worth noting that the policy world has struggled to keep up with these developments. Fundamental policy changes need to be made in order to create a further enabling environment for today’s technology-enabled entrepreneurs.

Traditionally, when an entrepreneur wanted to start a business, he or she would solicit friends and acquaintances, and perhaps a bank, in order to secure funds; would find initial customers through personal contacts and/or a local storefront; and complete transactions in cash or by check with transaction histories and accounting done on paper ledgers. Each of those processes has been fundamentally shifted through digital technology. Fledgling businesses now secure funding through crowdfunding platforms and online working capital lenders. They create web storefronts from day one and use online marketplaces and marketing platforms to secure customers from around the world. They manage transactions using online and electronic payments mechanisms, as well as digital accounting services that simplify processes.

The sweet spot of both eBay and PayPal since their inception has been entrepreneurs and young micro firms that are seeking customers beyond their local markets. These are manufacturers, retailers, and service providers of nearly every type. These are often traditional businesses that are using the Internet to enhance their offerings. The Internet enables these businesses to enjoy the benefits of more efficient data processing, enhanced customer service, and the ability to instantly connect with customers around the world.

Data from both eBay and PayPal help to demonstrate how the Internet has changed every aspect of entrepreneurship. This paper will proceed in two parts: Part I will describe, through data, how technology has fundamentally altered several aspects of the entrepreneurship process. Part II will suggest policy changes that need to be made to create an enabling environment for technology-enabled entrepreneurs.

I. Technology Has Enhanced the Process of Entrepreneurship

Entrepreneurs are often early technology adopters, as they are excited by novel products and willing to take risks on innovation.1 Entrepreneurs have seen tremendous benefits from using these new technologies. This section will focus on three traditional problem areas for entrepreneurs and how technology is opening up new pathways to success.
  1. Access to Finance
    According to World Bank Group Enterprise Surveys, access to finance is the number one barrier faced by small and medium-sized enterprises (SMEs) around the world.2 Yet, data from the Federal Financial Institutions Examination Council demonstrates that loans to firms with revenues of $1 million or less have gone down, and the value of those loans also has decreased.3 Moreover, data from the Federal Deposit Insurance Commission demonstrates that the number of bank branches in the United States has been declining for the past three years.4 The Federal Reserve Banks of New York, Atlanta, Cleveland, and Philadelphia find in their 2014 Joint Small Business Credit Survey Report that more than 50 percent of businesses with less than $1 million in revenues are unable to secure financing.5

    New online mechanisms for securing funds are stepping in where existing financial institutions have fallen short. Crowdfunding platforms like Kickstarter, Indiegogo, and Kiva provide entrepreneurs of all types with instant access to financing from people all around the world.6 Moreover, dedicated online working capital loan offerings from companies like OnDeck, Kabbage, and PayPal offer entrepreneurs and young firms the opportunity to quickly scale up through smaller loans that have a much more favorable payback process than traditional financing mechanisms do.7

    The example of PayPal Working Capital is quite illustrative. An application for PayPal Working Capital can be completed in five minutes, as compared with traditional financial institution loan applications, which can take several hours. The PayPal Working Capital loan product is a single fixed fee, with no periodic interest, requirement for a credit check, or pre-payment penalties. Traditional financial institution loans often include periodic interest, require an onerous credit check, and utilize fees and penalties.8 This might explain why 90 percent of PayPal Working Capital users gave the product a nine or ten (on a scale of zero to ten) in a recent survey.9 An entrepreneur now can secure funding in a matter of minutes and focus on building a great product instead of focusing on onerous loan applications.

  2. Finding Customers
    Traditionally, entrepreneurs have struggled to find new customers because of lack of brand recognition, industry contacts, and scale. To acquire customers, a traditional entrepreneur would need to spend large sums of money on marketing or open a physical store in a high-traffic area. Yet, even these mechanisms were quite limited in the scale effects they could achieve. The Internet changes this calculus.

    In 2005, Hal Varian, the noted Professor of Economics at University of California, Berkeley, coined the term “Micro-Multinational” to describe how the Internet enables small firms to reach well beyond their traditional customer bases.10 Professor Varian proposed that information technology and the Internet would dramatically lower the costs for entrepreneurs and small businesses to find customers around the world. He projected that small, technology-enabled enterprises could do business across borders because services built on top of the Internet would provide communication and trust mechanisms, enabling cross-border transactions to take place. This revolutionary phenomenon, the Micro-Multinational, the globally connected entrepreneur, is no longer a vision for the future.

    There are more than three billion people connected to the Internet; these are three billion potential customers for the modern entrepreneur.11 Simply putting up a website can bring in customers from down the street or across the world. This revolution in scale, trust, and brand promotion have resulted in significant economic gains for entrepreneurs around the world.

    Businesses that once were limited to their local towns are now using the Internet to sell their products and services around the world. Data from the U.S. Commerce Department demonstrates that less than 1 percent of U.S. small businesses export their products because of the difficulties associated in finding customers in other markets.12 Data from eBay marketplaces shows that more than 95 percent of the U.S. entrepreneurs and SMEs on the platform export.13 For those traditional businesses that could find customers in another country (perhaps because the business was near the border or there was a family or historical connection to a particular country), the business was usually limited to one export market.14 The findings for technology-enabled entrepreneurs are quite different. Research from the eBay marketplace demonstrates that more than 279,000 small firms in the United States reached consumers on at least two continents last year, and a remarkable 190,000 exported to consumers on four or more continents in 2014 alone.15 Moreover, these trends are not merely limited to entrepreneurs in the United States. Research on entrepreneurs in India, Chile, Italy, Australia, and South Africa found similar trends of high export rates and markets.16 The global startup is now a reality.

  3. Managing Financial Services
    Getting paid and managing finances are often major issues for entrepreneurs. Entrepreneurs need to get paid quickly because they often need to turn around and reinvest revenues in the business to achieve greater growth. The traditional mechanisms for business-to-business transactions—the check—sometimes can take days to clear, which greatly limits the ability of a lean entrepreneur to reinvest and innovate. Moreover, the traditional mechanism for business-to-consumer transactions—cash —has its own set of concerns with theft, logistics of money transfers, and tender time delays.17 A Tufts University study estimates that the cost of cash to U.S. business is $40 billion annually.18 Moreover, managing back-end finances is also a disproportionately difficult task for small entrepreneurs, as they cannot afford accounting or legal departments.

    Innovations from a wide variety of Internet-based service providers are helping entrepreneurs and small firms deal with the complexities of managing the finances of their fledgling businesses without creating unreasonable overhead costs. Payment service providers like Stripe, Dwolla, and Braintree enable a startup to accept online payments by inserting just a few lines of code.19 Intuit’s QuickBooks cloud service enables a startup to have a one-stop solution for managing back-end finances. Moreover, these solutions are often interoperable so that entrepreneurs’ payments can easily be linked in and tracked using an accounting service.20 These straightforward technology tools open up possibilities for startups by removing the burden of having to hire internal or external manpower to deal with finances or accounting.

    Technology cannot resolve every problem that today’s entrepreneur faces. Public policy needs to be shifted to meet the needs of entrepreneurs and to solve problems that technology alone cannot.

II. Policy Changes Need to be Made to Create an Enabling Environment for Technology-Enabled Entrepreneurship

The policy world struggles to keep up with the needs of entrepreneurs and innovators, and is often stuck fighting yesterday’s battles. Partially as a result of this, the ratio of new firms to all firms, or the startup rate, in the United States has steadily declined since 1980.21 Forward-looking policy changes—like those discussed below—could help to revitalize the enabling environment for startups.
  1. Educate Entrepreneurs on how to Leverage the Internet and Innovative Financial Services Products
    While entrepreneurs may be early adopters of technology, they also may not know the best ways to use that technology for business purposes. Data from the Canadian Chamber of Commerce, which surveyed small businesses using technology to engage in commerce, demonstrates that, while 96 percent of Canadian companies had websites they used for business purposes, only 27 percent were able to accept online payments.22 This demonstrates that, while some businesses may understand the basics of Internet technology, they are not yet leveraging the technology to its fullest extent. A Boston Consulting Group study on small businesses revealed that businesses in the developing world that are heavy technology users saw more revenue and job growth than businesses that lagged behind in technology usage.23 Technology solutions provide significant advantages for entrepreneurs, but most entrepreneurs remain unaware of these products.

    Commerce departments around the world need to foster the use of business technology by hosting webinars and live seminars, as well as providing networking opportunities and advice for entrepreneurs and small firms on how they can leverage technology to improve their businesses. Not every entrepreneur is a tech entrepreneur, but every entrepreneur can benefit from technology services.

  2. Think “Entrepreneur First” When Reviewing Regulatory Policies and Commercial Programs
    The Organization for Economic Cooperation and Development reports that proportionate compliance costs can be ten to thirty times greater for small firms than for larger firms.24 Governments often solicit comments from large enterprises when they are considering new or reviewing existing legislation, but the concerns of small businesses are rarely represented. Those that are going to be the most effected by the regulations are often overlooked. Labor, financial services, trade, tax, and nearly any other type of regulatory policy can have tremendous effects on entrepreneurs. Also, governments around the world offer domestic businesses loan guarantees, export promotion programs, and business consulting services. These initiatives are often tied to regulations that make it very difficult for entrepreneurs to take advantage of them.25 Moreover, entrepreneurs are rarely aware of these programs because of legacy government marketing systems.

    Policymakers need to carefully consider how any regulatory action might stem the ability of an entrepreneur to start and grow a business. Moreover, policymakers need to review existing regulations to ensure that they do not disproportionately harm entrepreneurs. Finally, governments must tailor their commercial programs to the needs of entrepreneurs and advertise these programs through the digital channels that are viewed by the modern entrepreneur.


Digital technology has fundamentally altered the landscape for today’s entrepreneur. But, the outlook for entrepreneurs remains uncertain. Policy can play an important role in improving the environment for entrepreneurs. These policies must think entrepreneur first and leverage technology to reach and benefit entrepreneurs. Technology is merely a tool, but it is a powerful one, and one that policymakers ought to uuse to boost entrepreneurs’ chances of success.

About the Author
Usman Ahmed is Head of Global Public Policy at PayPal, where he researches the policy implications of the transformational impact that the Internet is having on business and society. Ahmed is also an Adjunct Professor of Law at Georgetown University Law School, where he teaches courses on international law and policy issues related to the Internet. Ahmed earned his JD from the University of Michigan, his MA from Georgetown University, and his BA from the University of Maryland.


  1. “Early Adopters Are Great, But They Aren't Most Customers.”, Nov. 14, 2014.
  2. IFC: Assessing Private Sector Contributions to Job Creation and Poverty Reduction, IFC Jobs Study (2013).
  3. FFIEC. Small loans to businesses and farms 2005–2014.
  4. Federal Deposit Insurance Corporation. Number of Institutions, Branches, and Total Offices FDIC-Insured Commercial Banks Balances at Year End, 1934–2014.
  5. Federal Reserve Banks of New York, Atlanta, Cleveland, and Philadelphia. 2014. Joint Small Business Credit Survey Report.
  6. Zhou, Jennifer. 2012. Crowdsourcing: A New Revolution for Startups honors thesis. New York University Leonard N. Stern School of Business, pp. 7–19.
  7. Weddell, Greg. “What Banks Can Learn About Lending from PayPal and Square.American Banker, June 3, 2014.
  8. PayPal Working Capital Website.
  9. Esch, Darrell. Thousands of Merchants Benefit from PayPal Working Capital Program, Now Extended (March 20, 2014).
  10. Varian, Hal. “Technology Levels the Business Playing Field.” The New York Times, Aug. 25, 2005.
  11. Internet used by 3.2 billion people in 2015. BBC News, May 26, 2015.
  12. International Trade Administration. Exporting is good for your bottom line.
  13. eBay Public Policy Lab. 2015 U.S. Small Business Global Growth Report
  14. International Trade Administration. Exporting is good for your bottom line.
  15. eBay Public Policy Lab. 2015 U.S. Small Business Global Growth Report.
  16. eBay Public Policy Lab. Commerce 3.0–A Global Phenomenon Blog Post
  17. Tufts University Institute for Business in the Global Context. 2013. The Cost of Cash in the United States, pp. 25–29.
  18. Ibid.
  19. Wohlsen, Marcus. “Stripe’s Next Big Step in Moving Money Everywhere.” Wired, March 23, 2015.
  20. Reader, Ruth. “Intuit overhauls QuickBooks Online and inks new deals with Box and PayPal.” Venture Beat, Oct. 22, 2014.
  21. Litan, Robert E., and Ian Hathaway. 2014. “Declining Business Dynamism in the United States: A Look at States and Metros.” Brookings Institution.
  22. The Canadian Chamber of Commerce. 2010. Power Up the Network: A Report on Small Business Use of E-business Solution in Canada.
  23. BCG. 2014. Ahead of the Curve: Lessons on Technology and Growth from Small-Business Leaders.
  24. Arzeni, Dr. Sergio, Director of the OECD Centre for Entrepreneurship, SMEs and Local Development, in a keynote address to The G20 Agenda for Growth: Opportunities for SMEs Conference, Melbourne, June 20, 2014; as quoted in the Conference Report of Proceedings.
  25. Owen, Judy. “Why Government Rarely Helps Entrepreneurs.” Forbes, August, 22 2012.

Accelerating Growth by Curbing Regressive Regulation

By Brink Lindsey Vice President for research Cato Institute
Brink Lindsey
Brink Lindsey


As the twenty-first century unfolds, it has become clear that the U.S. economy is facing formidable challenges. The sluggish recovery from the recession of 2008-09 now appears to be the “new normal” for the foreseeable future. Demographic factors are primarily to blame, but there also are alarming signs that U.S. economic dynamism is in long-term decline.1

At the same time, the longstanding connection between overall growth and rising real incomes has broken down for large numbers of Americans. Even as the economy grew robustly during the past few decades, wages for the less educated failed to keep pace. The old adage, “a rising tide lifts all boats,” is thus under double assault—the waters are rising more slowly, and yachts are levitating while dinghies are sinking.

In this paper, I will focus on the first problem—the apparent decline in the U.S. economy’s long-term growth rate.2 In particular, I want to make the case for an agenda of pro-growth policy reforms that can command assent across the ideological spectrum. In a happy coincidence, these policy changes also would chip away at the second problem—namely, the growing economic divide between the highly educated and everybody else. The reform agenda I propose here would, therefore, serve not only to accelerate growth, but also to make growth more inclusive.  

To pursue these dual objectives, I propose a wide-ranging effort to roll back policies whose primary effect is to inflate the incomes and wealth of the rich, powerful, and well-established by shielding them from market competition. Such policies can be grouped under the heading of “regressive regulation”—regulatory barriers to entry and competition that redistribute income and wealth up the socioeconomic scale.3

My proposals target four major examples of regressive regulation, two at the federal level, and two at the state and local level:

  • Excessive monopoly privileges granted under copyright and patent law,
  • Protection of incumbent service providers through occupational licensing.
  • Restrictions on high-skilled immigration.
  • Artificial scarcity created by land-use regulation.

In all four examples, government policy works to erect barriers to entry.

In the first two cases, the restriction is on entry into a product market: Businesses are not allowed to sell products deemed to infringe on a copyright or patent, and individuals are not allowed to sell services without a license. In the other two cases, actual physical entry into a geographic area is limited—on the one hand, immigration into the country and, on the other, the development and the purchase or rental of real estate.

All of these policy barriers interfere with the “creative destruction” that is at the heart of economic growth—namely, the introduction of new wealth-creating products and production methods, and the reallocation of labor and capital to bring those innovations to scale.

Copyright and patent laws are supposed to promote innovation and growth by creating temporary monopolies that raise the return on producing new ideas. But copyrights and patents impose costs as well as confer benefits. First, they raise the price of protected goods and, thereby, burden consumers with a deadweight loss. Second, they help some innovators, but end up hurting others. Innovation frequently occurs through the borrowing and adaptation of others’ ideas. By restricting access to those ideas, copyrights and patents actually raise the cost of innovation and, thus, push in the direction of slower growth.

In recent years, the costs imposed by copyright and patent law have escalated dramatically. In the case of copyright, this has occurred through the wildly excessive extension of copyright terms, criminalization of copyright violations, and ongoing hostility to new copying technologies that may be used to reproduce copyrighted material. In the case of patents, costs have escalated because patent protection has been extended to vaguely defined innovations in software and business methods, and also because of the rise of “patent trolls” that buy up patent portfolios and monetize them through litigation. Thanks to such ill-advised developments, these laws now cause a significant drag on innovation and growth.

Occupational licensing also has expanded dramatically in recent decades. In 1970, about 10 percent of all U.S. jobs were subject to licensing requirements; today, the figure stands at around 30 percent. These laws are justified in the name of consumer protection, but research shows that the laws’ restrictions do little or nothing to benefit consumers. They do, however, slow new business formation and job creation: Comparing occupations licensed in some states but not others, employment growth is 20 percent lower in the restrictive states. These laws amount to a frontal assault on entrepreneurship, reducing entry by new businesses that frequently are the vessels of new ideas. Furthermore, by conditioning entry on passing exams that test knowledge of current ways of doing things, they help cement the status quo in place, and discourage development of new business models.

Immigrants are a major catalyst of U.S. entrepreneurship and innovation. According to one study of a large sample of engineering and technology companies founded between 1995 and 2005, 25 percent of those companies had at least one foreign-born founder. Yet current immigration laws make it very difficult for such highly talented individuals to live and work in our country. Out of roughly one million permanent resident visas awarded each year, only about 70,000 go to individuals based on their work skills or economic value. Temporary visas allow about 650,000 high-skilled workers to reside in the United States at any given time—a mere 0.4 percent of the workforce.

Zoning laws have been widespread in the United States for nearly a century, but their increasing restrictiveness during the past few decades has caused major distortions in where people decide to live. The distortions are especially pronounced on both coasts. The “regulatory tax” of land-use restrictions now constitutes about 20 percent of home values in Baltimore, Boston, and Washington, D.C.; more than 30 percent in Los Angeles and Oakland; and roughly 50 percent in Manhattan, San Francisco, and San Jose. Because of inflated housing costs, far fewer people live in America’s large coastal cities—i.e., the most productive and highest-income areas of the country—than would otherwise be the case. The economic impact may well be enormous. According to a recent study, reducing land-use controls in the most restrictive U.S. cities to the level of the median city would boost overall U.S. output by a whopping 9.5 percent.

While stifling economic growth that can lift living standards for everyone, regressive regulation does a very good job of boosting the living standards of its privileged beneficiaries. Copyright and patent laws support great corporate empires in media, entertainment, software, and pharmaceuticals, and contribute greatly to the vast fortunes amassed by the leaders and stars of those industries. Occupational licensing inflates earnings of licensed incumbents by an estimated 15 percent. Highly skilled U.S. professionals, such as doctors and engineers, benefit from tight restrictions on immigration, while low-skilled, native-born workers are exposed to a huge influx of low-skilled workers from abroad. Zoning laws, meanwhile, effect a huge wealth transfer to existing homeowners, who generally have both higher incomes and net worths than renters.

Because regressive regulation both inhibits entrepreneurship and growth (very unpopular with conservatives), and exacerbates economic inequality (very unpopular with progressives), it occupies an ideological no-man’s land in the contemporary national political debate. Conservatives tend to cheerlead for deregulation, but are less likely to focus on cutting regulations that help big business and the well-off. Progressives, meanwhile, tend to defend regulation as necessary and beneficial, but that impulse weakens considerably when regulation’s obvious effect is to entrench privilege and deepen disadvantage. As a result, although the policy areas discussed above are frequently the subjects of vigorous debate, the contending sides are not divided along left/right or Republican/Democratic lines.

Meanwhile, it’s hard to find disinterested policy experts anywhere on the ideological spectrum who defend the status quo in any of these policy areas. There is a strong expert consensus at present that copyright and patent protections have become excessive, that the costs of occupational licensing outweigh its benefits, that more openness to high-skilled immigration would be salutary, and that land-use regulation is too restrictive.

When you reflect on the unusual politics of regressive regulation—a lack of ideological or partisan conflict combined with a strong expert consensus in favor of reform—it becomes apparent that a campaign to spur growth by curbing such regulation has distinctive political appeal and a rock-solid economic rationale. Here is an opportunity to break out of the highly polarized red-versus-blue stalemate in which so many important policy debates now are trapped. Instead of the opposing forces being arrayed along the left/right axis, the contest would pit experts across the political spectrum against interest groups that benefit from existing policy. Rather than yet another left/right showdown, the debate this time could be framed as a choice between the public interest and vested interests.

Without a doubt, lobbies that profit from regressive regulation pose a formidable obstacle to reform. Any political effort to overhaul the policy areas discussed here can count on well-organized, well-financed, tenacious opposition. From an intellectual perspective, reforming regressive regulation represents the low-hanging fruit of pro-growth policy changes: There is little debate among the informed and disinterested that reform could make a significant, positive difference. Yet from a political perspective, the prospect of taking on determined and politically muscular interest groups is daunting. In this case, the low-hanging fruit is guarded by dragons.

Although progress will be difficult, incremental gains surely are possible. To that end, let me offer here a sampling of concrete reform ideas. These reform proposals vary in ambition and, consequently, in the likelihood of their enactment. Note that some of the more incremental proposals are premised upon failure to enact more ambitious reforms.

For copyright and patent law:
  • End criminal liability for copyright violations.
  • End the use of civil-asset forfeiture in copyright criminal prosecutions.
  • End any copyright liability for noncommercial copying.
  • Reduce copyright terms prospectively.
  • End patent protection for software and business methods.

For occupational licensing:

  • Pass state-level legislation establishing a right to engage in any lawful occupation unless:
    • The government can demonstrate a compelling interest in protecting public health and safety.
    • Occupational licensing is the least-restrictive means of advancing that compelling interest.
  • Establish state-level commissions to recommend specific occupations for which licensing requirements should be eliminated, followed by “fast-track” legislative votes on the commissions’ recommendations.
  • Initiate Federal Trade Commission actions against state licensing boards whenever their composition and control fail to meet the standards of “state-action” antitrust immunity set recently by the Supreme Court in North Carolina State Board of Dental Examiners v. Federal Trade Commission.
  • Create a federal program, along the lines of “Race to the Top,” in which grants are awarded to states that streamline occupational-licensing policies.

For high-skilled immigration:

  • Grant permanent resident visas to all college graduates with science, technology, engineering, and mathematics (STEM) degrees, and all holders of graduate degrees.
  • Expand the number of permanent resident visas that are awarded based on skills and economic value. Do not count dependents that come with the visa applicant against the relevant cap.
  • Remove the requirement that no country account for more than 7 percent of visas awarded based on skills and economic value.
  • Expand the number of H-1B temporary work visas.
  • Allow states to award a certain number of permanent and temporary visas on their own.

For land-use regulation:

  • Establish metropolitan-area zoning budgets that target aggregate growth in housing units and prohibit downzonings until the target is reached.
  • Establish a series of temporary tax rebates paid to neighbors of new developments out of increased tax revenue yielded by developments, thereby giving neighbors incentive not to oppose development.
  • Shift away from traditional property taxes (which primarily are taxes on structures, not improvements) to land-value taxation, thus creating incentives for higher-value development of land.

The idea of a left/right coalition to push deregulation may sound farfetched, but it is not without precedent. Consider the country’s last major episode of pro-market regulatory reform during the late 1970s and early 1980s. During this brief period, price-and-entry regulation of airlines, trucking, and railroads was dismantled. Price controls on oil and gas were lifted, interest-rate caps for savings and checking accounts were removed, and the AT&T monopoly was ended, paving the way for competition in long-distance telephony.

Those too young to remember can be forgiven for associating all of this with Ronald Reagan but, in fact, Jimmy Carter signed most of the relevant legislation. Edward Kennedy spearheaded airline deregulation on Capitol Hill, assisted by his aide, Stephen Breyer. And while the rise of Chicago School economics—especially the law-and-economics movement—supplied momentum for these policy changes, so did the activism of Ralph Nader.

History never repeats itself, but sometimes it rhymes. With luck, contemporary reformers can follow their predecessors’ good examples.

About the Author
Brink Lindsey is vice president for research at the Cato Institute. This paper is adapted from his recent Cato Institute white paper, “Low-Hanging Fruit Guarded by Dragons: Reforming Regressive Regulation to Boost U.S. Economic Growth.”


  1. For my own analysis of the growth slowdown, see Brink Lindsey, “Why Growth Is Getting Harder,” Cato Institute Policy Analysis no. 737, October 8, 2013
  2. I have addressed the second problem – rising inequality along educational lines – at length elsewhere. See Brink Lindsey, Human Capitalism: How Economic Growth Has Made Us Smarter – and More Unequal (Princeton, NJ: Princeton University Press, 2013)
  3. For a full exposition of this reform agenda, see Brink Lindsey, “Low-Hanging Fruit Guarded by Dragons: Reforming Regressive Regulation to Boost U.S. Economic Growth,” Cato Institute white paper, June 22, 2015

Tax Compliance Costs and Business Formation

By Len Burman Director of the Urban-Brookings Tax Policy Center Urban Institute and Tax Policy Center
John Iselin Research Assistant Urban Institute and Tax Policy Center
Len Burman
Len Burman
John Iselin
John Iselin


The papers in this volume explore how entrepreneurship might play a role in mitigating growing economic inequality. Most experts believe that technology is a key factor in driving inequality. (Council of Economic Advisers 1997) For example, more and more jobs are being replaced by technology: automated cashiers at grocery stores, robots on assembly lines, and, possibly not too far in the future, driverless Uber cars (Kozner 2015) The secret to success in this new world is to be able to do things that computers can’t. Starting a business is likely to be in that category for the foreseeable future, as entrepreneurship is a quintessential example of an activity that relies heavily on human creativity and manual dexterity. While successful startups ultimately may reach a scale where automation is profitable, that’s unlikely to be true in the development phase. And Holtz-Eakin, Rosen, and Weathers (2000) find evidence that entrepreneurship is a key channel for economic mobility.

This paper focuses on an oft-neglected channel through which the tax system can affect entrepreneurship and inequality. A large body of research has focused on the direct effects of taxation on business formation, risk-taking, and employment.1 While there’s no doubt that our complex, loophole-ridden tax code could be vastly improved, its current effects on business formation and entrepreneurship are ambiguous. By insuring against down-side risk and allowing losses to be offset, it can help encourage entrepreneurship, but it also can limit the return. However, tax compliance costs create a clear and unambiguous barrier to entry for new businesses. That is because a significant portion of compliance costs is essentially fixed: businesses of any size must perform certain basic functions to meet their tax obligations. While compliance costs rise with business size—Walmart spends much more on tax compliance than the independent corner grocery does—taxes are a smaller share of costs for large businesses than for small ones. This gives large businesses an unnatural advantage over small businesses and could deter business formation.

This brief paper discusses ways in which the tax system in general and tax complexity in particular affects small businesses, and considers some options to simplify tax compliance for small enterprises. We end with an agenda for further research.

How do taxes affect small businesses?

The tax system complicates decisions facing potential new business owners, as they must consider a menu of different tax rates, deductions, and credits when deciding on the size and legal structure of the firm. Potential entrepreneurs must decide not only whether to head out on their own, but also must navigate a host of decisions regarding whether to incorporate and how to finance investment.

For example, the tax system may discourage risk-taking and raise the cost of access to equity markets. Gentry and Hubbard (2000) argued that the progressive income tax structure, which taxes successful business at relatively high rates, but limits deductibility of losses, discourages risk-taking (because risky investments are more likely to produce losses). Another factor is that the corporate income tax raises the cost of access to capital markets (Gale and Brown 2013; Rosenberg and Marron 2015). Companies must incorporate to sell shares on public financial markets. However, once incorporated, equity capital invested in these firms face two taxes: the corporate income tax on profits and taxes on dividends and capital gains that is levied on shareholders. Austin, Burman, and Rosenthal (2015) find that the vast majority2 of corporate equity issued in the United States is not subject to shareholder-level taxation, but companies clearly are sensitive to the relative taxation of corporations and pass-through entities. S-corporations, partnerships, limited liability companies, and sole proprietorships have grown dramatically over the past three decades. While C-corporations (subject to corporate income tax) accounted for almost 80 percent of business income in 1980, their share fell to 46 percent by 2011 (Cooper et al. 2015).

There is rich literature on the subject of incorporation decisions at both the federal and state levels. Some studies have used the Tax Reform Act of 1986 as a natural experiment to study the effects of changing incentives to structure as a pass-through or a C-corporation. Gordon and MacKie-Mason (1990) concluded that businesses with losses tended to move to the corporate sector, while those with positive profits became pass-through entities. Carroll and Joulfaian (1997) examined a panel of corporate income tax returns spanning 1985 to 1990 and found that the 1986 tax reform increased the probability that C-corporations switched to the S-corporation form. At the sub-national level, both Goolsbee (2002) and Luna, Murray et al. (2010) found evidence that state corporate taxes explained a significant share of the variation in probability of incorporation among states. By discouraging companies from incorporating, the tax system makes it more difficult for many firms to access capital markets.

Not all tax provisions deter business formation. Gordon and Cullen (2002), Gordon (1998), and Bruce and Schuetze (2004) argue that the strategic use of the corporate and individual income tax rate structures and business owners’ ability to shelter more income from taxes than wage earners can actually make starting a business more attractive than it would be in the absence of taxes. In addition, it is much more difficult for the IRS to verify income for small businesses than for wage earners who receive W-2s; as a result, some small businesses also benefit from opportunities to successfully evade tax obligations.

A number of targeted tax provisions aid small firms, which generally face lower effective tax rates than larger firms do (Gale and Brown 2013; Rosenberg and Marron 2015; Marron 2014). For example, section 179 expensing of investments allows smaller businesses to immediately deduct some or all investment costs, but the maximum deduction phases out based on the amount of qualifying investment.3 Incorporated small businesses also may use cash accounting, as opposed to the less-generous accrual accounting method required of larger businesses; face lower average tax rates under the corporate income tax; are more easily organized in pass-through form; and are exempt from the corporate alternative minimum tax and certain mandates under the Affordable Care Act.4

Compliance cost and firm size

Survey evidence finds that the costs of compliance are proportionately larger for small firms than large ones, which is consistent with the existence of nontrivial fixed costs. This indicates that small firms face a large tax compliance burden, both in the aggregate and relative to large businesses (Marron 2014).

While small businesses are not necessarily entrepreneurial,5 new innovative firms almost all start small, and large fixed costs can slow their growth and raise the odds of failure. Also, if the costs are known in advance, they could discourage some would-be entrepreneurs from trying to start new businesses.

The research into compliance costs mainly utilizes surveys of businesses regarding their size, composition, and time and money spent filing their taxes. The studies then estimate the relationship between firm-specific factors, including size, on compliance costs. The research has faced significant challenges, including sometimes-low response rates, issues in defining firm size, and assumptions necessary to estimate the cost of time spent on tax compliance. This final issue is critical, as most of the cost of tax compliance—at least 60 percent of total compliance costs—arises from the monetized value of the time spent meeting legal tax obligations (Contos et al. 2012).

Evidence on compliance costs is based on surveys, most of which have yielded relatively low response rates. As Slemrod and Venkatesh (2002) point out, the direction of the non-response bias is uncertain. Companies that find compliance especially burdensome might choose to vent their frustration on a compliance survey, or their aversion to tax preparation might translate into a disinclination to complete and return a complex survey instrument.

In 1992, Slemrod and Blumenthal (1996) surveyed 1,329 large corporations from the IRS Coordinated Examination Program, receiving 365 replies, a response rate of 27.5 percent. The authors find that larger firms have large compliance costs, but these costs rise more slowly than firm size, indicating higher costs per unit at smaller sizes. This result is robust with respect to different measures of firm size (Slemrod and Blumenthal 1996).

International evidence from the 1990s is similar. Studies from Southeast Asia, New Zealand, Australia, the United Kingdom, and the Netherlands conducted in the late 1990s all found compliance costs to be proportionately larger for small firm sizes than for large ones. Again, the findings are not sensitive to the measure of firm size (Slemrod and Venkatesh 2002).

Slemrod and Venkatesh (2002) surveyed medium-sized firms, defined as firms with at least $5 million in assets, in 2001. They sent out 4,323 surveys to both firms and tax professionals, and received 443 completed surveys, a 10.25 percent response rate. The low response rate raised questions about validity, which the authors recognized. Weighting the survey responses to reflect the population, the survey suggested compliance costs, once again, are proportionately lower for larger firms.

DeLuca et al. (2007) surveyed firms that had less than $10 million in assets, meeting the IRS definition of a small business. They contacted 7,000 small businesses by both mail and telephone, in an effort to increase the response rate. This yielded a 38 percent response rate. Across three measures of business size—assets, employees, and receipts—the authors find evidence of high fixed costs of compliance.

Contos et al. (2012) contacted more than 22,000 corporations and partnerships that filed a federal income tax return in 2009. The response rates were relatively high—31.1 percent for small firms, 35.6 percent for medium firms, and 34.1 percent for large firms. This paper showed that the estimates of compliance cost were very sensitive to assumptions about how labor costs varied with firm size. Previous research had assumed a fixed wage rate across firm size. Contos et al. (2012) argued that a lower hourly wage is appropriate for small firms because they typically pay lower wages than large firms do. Using these new data and methodology, the authors confirmed the general conclusions of the earlier studies. Compliance costs create economies of scale; however, the barrier to entry for small firms is much lower, assuming lower costs for time spent on tax compliance.

The average cost of tax compliance, assuming constant labor costs for very small firms (under $100,000 of assets), was estimated at $7,100, assuming time is valued at a constant rate. (See Table 1.) The estimate falls by one-quarter when assuming that labor in small firms is only worth eight dollars per hour. By contrast, the average compliance cost for large firms (more than $500 million in assets) increases by more than 500 percent under the assumption that they face very high labor costs.

Table 1. Estimated Tax Compliance Cost by Firm Size, Type of Firm, and Labor Cost per Hour, Tax Year 2009

Source: Contos et al. (2012)

When considering entrepreneurial new firms, however, it might be more appropriate to apply a significantly higher labor cost than the $8-per-hour average assumed for small firms in Panel B of Table 1. No studies, to our knowledge, have examined this question.
State and local taxes create additional complexities. Most states levy income and sales taxes on businesses, and localities levy property taxes. Because of the different levels of government, tax rates may vary substantially within small geographic areas. An urban taxpayer may face a state and local option sales tax, as well as property taxes from county, city, school district, and a special taxing district (e.g., water utility). Firms with employees in multiple states with income taxes must remit withheld income taxes in each state. Some cities also levy commuter taxes, requiring additional reporting and accounting costs.

Unfortunately, unlike federal tax compliance, there is relatively little evidence on state and local tax compliance burdens. A notable exception is Slemrod and Venkatesh (2002), which surveyed medium and large firms in 2001 to estimate the share of compliance costs they devoted to federal, state, and local taxes, as well as rules related to foreign-source income. Overall, 26 percent of internal compliance costs and about 23 percent of the external costs (paid to tax professionals) were attributable to state and local taxes. For mid-size firms, with less than $5 million of assets, about 16 percent of internal costs and 24 percent of external costs were attributable to state and local taxes. Thus, state and local taxes are a significant component of compliance costs.

While these costs may create additional barriers to entry for new firms, some state and local programs are intended to offset part or all of the compliance burdens. Many firms are attracted to a location by tax incentives. So-called incubators set up by states and cities are intended to encourage business formation within their boundaries. One example is START-UP NY, which exempts new businesses (or expansions of current businesses) that set up within specific space affiliated with a university. Firms and their employees within these programs are exempt from most taxes, as well as provided with assistance setting up their businesses.6

Sales taxes also create compliance burdens. States, recognizing that businesses act as collection agents, have tried to address the costs associated with collecting the tax. Most states allow a vendor discount or pay a commission of a small percentage of the sales tax collections to compensate the business for compliance and remittance.

A related issue has to do with the taxation of mail order and Internet commerce. The Supreme Court ruled in the Quill decision that states may not require out-of-state retailers to collect and remit use taxes unless the retailer has nexus in that state (i.e., a physical presence), because the requirement to remit taxes to hundreds or even thousands of taxing jurisdictions that have different rules regarding tax bases and rates would be an undue burden and, hence, a restraint of interstate commerce that violates the Commerce Clause of the Constitution.7 The Court noted that Congress has the ability to pass legislation to allow states the ability to require remote vendors to collect use taxes. To address Quill and Congress’ lack of action, several states have entered into a compact to create simplified uniform tax rules. Collection by remote vendors, however, is still voluntary until Congress acts with legislation like the Marketplace Fairness Act.8

Of course, all federal, state, and local taxes and compliance costs for new business should be considered compared to what the person would face if only earning wage income. In other words, there are compliance costs for all taxpayers, and the relevant research has not explicitly identified the incremental costs associated with operating a business. Plus, there are benefits of different types of business organization—sole proprietorship, partnership, corporation—that may outweigh the costs of compliance.

In addition, some of the recordkeeping businesses do for tax purposes would be necessary for internal management, even absent any taxation. That is why it is difficult to measure accurately the marginal costs of tax compliance, and surveys may confound tax compliance with otherwise necessary accounting costs.

Taxes also are not the only—or necessarily the most important—source of fixed costs for businesses. Both government regulation and federal contracting represent areas where small businesses might face disadvantages because of their size. Complying with regulations poses a burden on businesses, although the boundary between the fixed and variable component of regulatory compliance may be less clear than it is for taxes. Still, there is evidence that regulation reduces the number of smaller firms, which suggests that regulations have a significant fixed cost component (Calcagno and Sobel 2014).

Reform Options

Reducing or eliminating fixed tax compliance costs would benefit new and small businesses and likely promote entry into entrepreneurial activity. A straightforward way to do this would be to allow a fixed dollar tax credit for all businesses. This could reduce or eliminate the fixed cost component of tax compliance, but it would be very difficult to design such a credit in a way that would be both effective and not prone to abuse. A non-refundable tax credit would do little to ease the immediate cash flow challenges of startups that might not see taxable profits for several years.9 A refundable credit, however, could encourage the establishment of fraudulent entities created solely for the purpose of claiming the tax credit. Firms could be required to prove that they actually had incurred labor and other costs related to tax compliance, but this would, ironically, significantly complicate tax compliance for small businesses.

A more appealing option is to simplify the tax law. However, there’s a trade-off between stemming evasion and minimizing compliance costs. That is, some portion of tax complexity arises from trying to get businesses to pay the tax that they owe (and, nonetheless, more than half of small business income is not properly reported to the IRS10). Removing tax expenditures and simplifying choices of accounting regimes would reduce compliance costs, however, without undermining enforcement.

Beyond that, the ideal approach would be to design a tax system where taxpayers can signal willingness to comply in exchange for a simplified compliance regime. President George W. Bush’s tax reform commission (The President’s Advisory Panel on Federal Tax Reform, 2005) proposed one reform in this category. It would have created new accounts that small cash-basis businesses could use to fully account for income and expenses. Annual cash flow would be as simple as comparing end-of-year and beginning-of-year balance in the special account. The Obama administration has proposed something similar (U.S. Treasury 2012).

In addition, simplifying and coordinating state and local tax regimes could significantly reduce compliance costs. One possibility would be to encourage states to conform their income taxes to more closely match federal definitions of taxable income. In addition, more states could be encouraged to simplify their sales taxes along the lines agreed to by twenty-four states in the Streamlined Sales and Use Tax Agreement.

Future research

There are a several avenues for future research in this area. In particular, do the costs of compliance rise or fall as firms age? In other words, does learning about the tax system make compliance less costly, or do firms naturally become more complex, with concomitant new compliance burdens as they age? The evidence of how costs vary by size provides an indirect answer to this question, but if rising compliance costs deter growth, then we might be missing an important link in the connection between tax complexity and entrepreneurial activity. 

It would be very useful to have an all-in measure of the costs of tax compliance, including federal, state, and local. In addition to creating a barrier to entry, costs of coordinating among multiple levels of government and multiple jurisdictions could affect firms’ decisions to expand in size or change location.

About the Authors
Len Burman is director of the Urban-Brookings Tax Policy Center at the Urban Institute, the Paul Volcker Professor of Public Administration and International Affairs at the Maxwell School of Syracuse University, a research associate at the National Bureau of Economic Research, and a senior research associate at Syracuse University’s Center for Policy Research. He cofounded the Tax Policy Center, a joint project of Urban and the Brookings Institution, in 2002. Burman has held senior-level positions in both the executive and legislative branches and is an author of Taxes in America: What Everyone Needs to Know, with Joel Slemrod, The Labyrinth of Capital Gains Tax Policy: A Guide for the Perplexed, and numerous papers. Burman holds a BA from Wesleyan University and a PhD from the University of Minnesota.

John Iselin, is a Research Assistant at The Urban / Brookings Tax Policy Center, his topics include: federal, state, and local public finance. Iselin graduated from Reed College in 2013.

Author Acknowledgements

We thank Norton Francis, Kim Rueben, and Eric Toder for helpful comments on an earlier draft and Bill Gale, Donald Marron, and participants in the Ewing Marion Kauffman Foundation New Entrepreneurial Growth Conference for helpful discussions. Views are those of the authors and should not be attributed to any of the organizations with which we are affiliated or their funders.


Austin, Lydia, Leonard Burman, and Steve Rosenthal. 2015. “The Dwindling Share of U.S. Corporate Stock Held in Individual Taxable Accounts.”

Bruce, Donald, and Herbert J. Schuetze. 2004. “Tax Policy and Entrepreneurship.” Swedish Economic Policy Review 2 (11): 233–265.

Calcagno, Peter T., and Russell S. Sobel. 2014. “Regulatory Costs on Entrepreneurship and Establishment Employment Size.” Small Business Economics 42 (3): 541–559. doi:10.1007/s11187-013-9493-9.

Carroll, Robert, and David Joulfaian. 1997. “Taxes and Corporate Choice of Organizational Form.” US Department of the Treasury, OTA Paper 73.

Contos, George, John Guyton, Patrick Langetieg, Allen H. Lerman, and Susan Nelson. 2012. “Taxpayer Compliance Costs for Corporations and Partnerships: A New Look.” In . Washington, DC: Internal Revenue Service and the Urban-Brookings Tax Policy Center.

Cooper, Michael, John McClelland, James Pearce, Richard Prisinzano, Joseph Sullivan, Danny Yagan, Owen M. Zidar, and Eric Zwick. 2015. “Business in the United States: Who Owns It and How Much Tax They Pay.” NBER Chapters. National Bureau of Economic Research, Inc. 

Council of Economic Advisers. 1997. Economic Report of the President. Washington, DC: The White House: Council of Economic Advisers.

DeLuca, Donald, John Guyton, Wu-Lang Lee, John O’Hare, and Scott Stilmar. 2007. “Estimates of U.S. Federal Income Tax Compliance Burden for Small Businesses.” In National Tax Association Proceedings. Columbus, OH: National Tax Assocation.

Fox, William F., and Matthew N. Murray. 1997. “The Sales Tax and Electronic Commerce: So What’s New?” National Tax Journal 50 (3): 573–592.

Gale, William G., and Samuel Brown. 2013. “Small Business, Innovation, and Tax Policy: A Review.” Innovation, and Tax Policy: A Review (April 8, 2013).

Gentry, William M., and R. Glenn Hubbard. 2000. “Tax Policy and Entrepreneurial Entry.” The American Economic Review 90 (2): 283–287.

Goolsbee, Austan. 2002. “The Impact and Inefficiency of the Corporate Income Tax: Evidence from State Organizational Form Data.” National Bureau of Economic Research.

Gordon, Roger. 1998. “Can High Personal Tax Rates Encourage Entrepreneurial Activity?” IMF Staff Papers 45 (1): 49–80.

Gordon, Roger H., and Julie Berry Cullen. 2002. “Taxes and Entrepreneurial Activity: Theory and Evidence for the US.” National Bureau of Economic Research.

Gordon, Roger H., and Jeffrey K. MacKie-Mason. 1990. “Effects of the Tax Reform Act of 1986 on Corporate Financial Policy and Organizational Form.” National Bureau of Economic Research.

Holtz-Eakin, Douglas, Harvey S. Rosen, and Robert Weathers. 2000. “Horatio Alger Meets the Mobility Tables.” Small Business Economics 14 (4): 243–274.

Hurst, Erik, and Benjamin Wild Pugsley. 2011. “What Do Small Businesses Do?” National Bureau of Economic Research. 

Kozner, Anthony Wing. 2015. “Google Cabs And Uber Bots Will Challenge Jobs ‘Below The API’—Forbes.” Accessed June 11.

Luna, LeAnn, Matthew N. Murray, and others. 2010. “The Effects of State Tax Structure on Business Organizational Form.” National Tax Journal 63 (2): 995–1022.

Marron, Donald B. 2014. Tax Issues Facing Small Business. Washington, DC.

Rosenberg, Joseph W., and Donald B. Marron. 2015. “Tax Policy and Investment by Startups and Innovative Firms.” Available at SSRN 2573259

Slemrod, Joel B., and Marsha Blumenthal. 1996. “The Income Tax Compliance Cost of Big Business.” Public Finance Review 24 (4): 411–438.

Slemrod, Joel B., and Varsha Venkatesh. 2002. “The Income Tax Compliance Cost of Large and Mid-Size Businesses.” Ross School of Business Paper, no. 914.

The President’s Advisory Panel on Federal Tax Reform. 2005. “Simple, Fair, and Pro-Growth: Proposals to Fix America’s Tax System.”

U.S. Treasury. 2012. “The President’s Framework for Business Tax Reform.” Joint Report. Washington, DC: The Department of the Treasury and The White House.


  1. Firms that set up entirely in tax-free areas are exempt from income, sales, and select other taxes for ten years. Employees are exempt from commuter taxes for five years. See
  2. em>See Fox and Murray (1997) for a discussion of the issues raised by the Quill decision.
  3. See
  4. Unused credits may be carried over to later years and might ultimately be valuable when the company becomes profitable.
  5. See
  6. Marron (2014), Rosenberg and Marron (2015), and Gale and Brown (2013) provide nice surveys of the evidence.
  7. Austin, Burman, and Rosenthal (2015) estimate that only 16 percent of corporate equity issued in the United States is subject to the U.S. individual income tax. Most of it is held by foreigners, nonprofits, life insurers, or in tax-free retirement accounts.
  8. In 2015, the maximum deduction is $25,000 and the amount that may be expensed phases out between $200,000 and $225,000 of qualifying investment. Between 2010 and 2014, the maximum deduction was $0.5 million, phasing out between $2 million and $2.5 million of investment. (The higher deduction limit and threshold for phase-out have been extended several times since 2012 and might be extended retroactively again to apply to 2015.)
  9. Gale and Brown (2013) for a full list of provisions that benefit small businesses.
  10. Hurst and Pugsley (2011) concluded that “[m]ost firms start small and stay small throughout their entire lifecycle … [and] do not innovate along any observable margin.”

Modernizing Federal Fiscal Policy for Economic Growth

By Scott Winship Walter B. Wriston Fellow Manhattan Institute for Policy Research
Scott Winship
Scott Winship


Increasing economic growth rates has proved more difficult than in the past. In an essay he wrote at the Kauffman Foundation, Brink Lindsey noted that there are diminishing returns to copying others’ innovations and to doing more of what has always been done. In his view, we have entered the era of “frontier economics,” where “older, easier sources of growth are drying up and, as a result, the prospects for continued dynamism and prosperity hinge more than ever before on the pioneering entrepreneurial upstarts that explore and extend the technological frontier.”1

How can we promote entrepreneurial growth in today’s mature American economy? A number of private and public policies would potentially increase growth rates, but federal fiscal policy offers an important opportunity to inject new life into the economy. The policies offered below would make our tax policy more efficient, increasing investment in American industries, leveling the playing field for startups, and minimizing the distortions introduced by tax expenditures. The policies also would put spending on a trajectory that will, in time, reduce deficits. In the shorter run, addressing entitlement spending will reassure private investors that the federal government can act responsibly and that government borrowing will not crowd out private investment and lending.2

Prospects for Growth

Elsewhere, Lindsey has pointed out that we are experiencing diminishing growth in all the determinants of gross domestic product (GDP)—in hours worked per person, education levels, capital investment per worker, and the output produced holding these inputs constant.3 Some analysts, such as Northwestern University economist Robert Gordon, go so far as to lament the “demise of U.S. economic growth.”4

It is easy, however, to overstate the extent to which the tides of growth have receded. Between the business peaks of 2000 and 2007, GDP per person rose by 1.5 percent annually. This was far below the 3.1 percent growth rate between 1959 and 1969, but also behind the 2.0–2.1 percent rates of the 1948–59, 1969–79, 1979–89, and 1989–2000 cycles. Nevertheless, even by this measure of growth, what we see is not an extended period of slow economic growth; rather, it is the past fifteen years that look disappointing, with little variation between 1948 and 2000, excepting the outlier decade of the 1960s.5

Growth in GDP per capita indicates the extent to which there is more income available for each person over time. That can change for reasons good and bad. If more working-age men drop out of the labor force because they cannot find jobs, that could lower GDP per capita, but it also could decline if more older workers are affluent enough to retire earlier than in the past. The increase in GDP per capita can be expressed in terms of three components: the rise in labor productivity (output per hour of work), the change in hours per worker, and the change in workers per capita. When broken out this way, three conclusions emerge. First, productivity growth declined in the 1960s, 1970s, and 1980s, but it accelerated thereafter, through the 2000–2007 cycle. Productivity growth was strong in the pre-recession aughts; in the non-farm business sector, it grew by 2.6 percent annually, compared with 2.8 percent in the 1950s and 1960s. Productivity growth has slowed again since 2007, but that hardly suggests a “new normal” of “secular stagnation.”

Second, growth in the share of the population that works has been declining since the 1970s business cycle, to the point where there was no growth in workers per capita during the aughts. We can expect that the share of Americans who work actually will decline moving forward; this is primarily a story about the life cycle of the baby boomers. That puts a drag on measured economic growth, but it doesn’t reflect diminishing improvement in living standards. The fact that fewer people soon will be working is (mostly) a positive or benign development driven by the retirement of an unusually large group of older workers. On the other hand, labor-force participation among working-age men has fallen steadily for at least sixty-five years.

Third, annual hours per worker have fallen since 1948, except for a pause during the 1980s and 1990s. This long-term trend, too, is mostly positive or benign. As more and more women have entered the labor force over this period—plateauing in the mid-1990s—the number of part-time and part-year workers has risen, and affluence has allowed many older workers to transition into retirement by working less than full-time year-round.

These trends suggest two key areas of concern in thinking about economic growth: labor productivity growth and reduced work among non-elderly men. I have written elsewhere on the decline in employment among working-age men.6 It is a long-term trend driven by an increase in the number of men who are out of the labor force and tell government surveyors that they do not want a job. A major enabler of this growth in voluntary joblessness is the Social Security Disability Insurance (SSDI) program, which has ballooned since the 1980s. Part of that increase is due to the aging of the baby boomers and the increase in work—and therefore eligibility—among women, but perhaps half the rise is due to policy changes that have made it easier to qualify for SSDI benefits.

The rest of this essay is addressed to describing policies that might increase growth by increasing productivity. I offer five fiscal policy reforms to increase productivity growth, eliminate distortions that hurt growth, unleash private investment, and promote economic efficiency generally.

Proposal 1: Swap Corporate Taxes for Direct Investor Taxation

Tax debates between liberals and conservatives most often revolve around whether to raise or lower top individual income tax rates. The truth of the matter is that we don’t really know much about the effects on growth of different tax rates.7 In part, that is because, like many empirical questions, this one is difficult to answer convincingly. But the answer also surely depends on the details: will a tax cut be paid for elsewhere, or will it increase the deficit? Will tax-rate cuts be accompanied by measures that broaden the base of taxable income? How high are tax rates to begin with? Which tax rates will be targeted?

What we do know is that taxes on saving and investment tend to reduce saving and investment, which is harmful to growth. We also know that the top U.S. corporate income tax rate—39 percent when state taxes are included—is the highest in the world and that that discourages domestic and foreign investment. It also encourages tax avoidance through mergers with foreign companies to relocate headquarters abroad and by declaring that profits from intangible assets derive from foreign branches.

We know that shareholder earnings are effectively taxed twice—once when corporate profits are taxed and again when shareholders realize capital gains or receive dividends. This double taxation incentivizes businesses to retain earnings rather than distribute profits to shareholders, which locks up capital inefficiently.

The deductibility of corporate interest payments leads businesses to prefer financing investment through borrowing rather than raising capital from the sale of equities. That leads to overleveraging and also benefits incumbent firms over startups, because the former have had time to demonstrate their financial worthiness.8 And, both the corporate and individual income tax codes are riven with tax expenditures and tax breaks benefiting specific companies and industries, which distorts economic activity in any number of ways. The complexity of the tax codes leads to wasteful economic activity around tax avoidance and compliance, including inefficiently large tax accounting and lobbying industries.9

In one fell swoop, tax reform could eliminate the complexity of the corporate tax code, scrap the cronyist provisions in it, liberate capital from being locked up in retained earnings, remove the current code’s bias in favor of debt financing and the problems that flow from it, and resolve the international taxation issues that put American companies at a competitive disadvantage. It also could rid shareholders of inefficient incentives to hold equities until they die—which resets the baseline from which their heirs’ capital gains are calculated—instead of reallocating their capital elsewhere.

The solution involves a grand swap based on a recent proposal by Eric Toder of the Urban Institute and Alan Viard of the American Enterprise Institute.10 The corporate income tax would be eliminated, along with the inefficiencies it creates. In its place, capital gains on stock in publicly traded corporations would be taxed on an accrual basis, and they and dividends would be taxed at ordinary income rates. Accrual taxation is an idea that goes back at least to 1937. Rather than gains being taxed when they are realized—when assets are sold—they would be taxed annually as they accrue, whether realized or not. Not only would this swap address the myriad inefficiencies of the corporate tax code; it would increase foreign investment in U.S. companies because only U.S. citizens and residents would pay U.S. corporate income taxes.11

This swap would cost roughly $2.2 trillion over ten years and would lower taxes primarily on upper-income filers.12 While this reduction could be justified as a partial mitigation of the high tax rates currently imposed by double taxation, the proposal here opts for deficit neutrality instead of fully eliminating the higher rates caused by that double taxation. (But see Proposal 2, below, for a deficit-neutral option to reinstate preferred tax rates on these investments.) Following Toder and Viard, my proposal would pay for this revenue shortfall through offsetting tax increases that primarily would fall on upper-income filers.

Specifically, the deductibility of state and local income and property taxes on individual returns would be eliminated, as would the exclusion of investment income from life insurance and annuities (another move in the direction of accrual taxation).13 In addition, the maximum earnings subject to the Social Security payroll tax would be doubled, putting the share of earnings in the economy subject to the tax roughly where it was in 1983 (and where, absent rising income concentration, it would have stayed).14 As an additional benefit, by increasing revenues to the Social Security Trust Fund, its solvency would be extended, probably by a decade.15

Proposal 2: Scrap Individual Income Tax Expenditures and Lower Rates

Tax expenditures are provisions in the tax code that result in revenue losses in order to benefit individuals and firms behaving in specific ways. They make the tax code unnecessarily complex, distort economic activity, reward politically connected companies and industries, and subsidize actions that would have been undertaken even in the absence of the tax break. All these consequences reduce economic growth.

Proposal 1, above, would eliminate corporate tax expenditures by getting rid of the corporate tax code. It would drop several individual income tax expenditures as well, including the preferential treatment of capital gains, qualified dividends, and built-up gains in life insurance policies and annuities, and the deductibility of state and local taxes. By moving to taxing gains on an accrual basis, the so-called step-up basis for capital gains at death is rendered irrelevant for gains from corporate stock.

But most tax expenditures in the individual income tax code would remain. The current proposal would eliminate nearly all of them and lower individual tax rates. The lower rates might raise economic growth above and beyond the benefits of tax simplification. The only tax expenditures that would be retained in this proposal would be the Child Tax Credit, the Earned Income Tax Credit, and certain tax expenditures related to closely held businesses that report income on individual tax returns rather than on corporate returns.16 Because the corporate tax swap would not affect these businesses—except that realized gains would be taxed at ordinary income rates—tax provisions related to, for example, depreciation, might be relatively economically efficient.

The rate reductions should result in roughly the same distribution of taxation across income levels that would exist if the tax expenditures were retained, and they should not increase future deficits. Notwithstanding the corporate tax swap proposed above, it would be desirable if the rate reduction here produced preferential tax rates on accrued capital gains and on qualified dividends in order to incentivize investment. It also would be desirable to permanently “fix” or eliminate the Alternative Minimum Tax, which was intended to ensure that upper-income tax filers were not able to unduly lower their taxes by using tax expenditures, but which now affects more and more upper-middle-class filers.

The result of implementing these two tax-related proposals would be a greatly simplified income tax system, and most Americans would be able to file a one-page tax return. The reduction in tax rates would be significant—it is estimated that eliminating all individual tax expenditures would allow rates to be cut in half.17 The reduction would be smaller here, as a small number of tax expenditures would be retained, and the cost of eliminating others already is accounted for in the corporate tax reform proposal above. There would be immediate benefits to growth, and the long-term effect likely would involve faster growth rates rather than a one-time boost in the level of growth. Ultimately, a progressive consumption tax system would be even better for growth than the individual income tax system that the proposals here would create.18 By simplifying the individual income tax code and eliminating the corporate tax code, these proposals would lay the groundwork for future adoption of a progressive consumption tax.

Proposal 3: Social Security Pension Reform for Deficit Reduction, Increased Investment, and Economic Security

Federal debt held by the public was 39 percent of GDP in fiscal year 2008, before rising to 74 percent by 2015.19 The only other time that debt levels were so high in American history was during and immediately after World War II. Smaller deficits in the next few years are projected to temporarily lower debt-to-GDP, but the Congressional Budget Office (CBO) expects it to rise again by 2019. In 2039, federal debt held by the public is expected to be more than 100 percent of GDP—nearly an all-time high (and rising). If we started this year, it would take non-interest spending cuts equivalent to 13 percent of projected outlays or tax increases totaling 14 percent of projected revenue over twenty-five years to get the debt level below its 2008 level by 2040.20

Alternatively, imagine we were to reduce the deficit by $480 billion in fiscal year 2016 (2.6 percent of GDP) and then continued to reduce deficits by 2.6 percent of GDP each year until 2040. In that case, the public-debt-to-GDP ratio would still be higher than its average from 1957 to 2007 and higher than its level in 2007. Reducing the deficit by 2.6 percent of GDP over even a decade would be equivalent to a ten-year budget proposal saving $6 trillion; for comparison, the Bush tax cuts collectively are projected to cost less than that over the twenty year-period from 2001 to 2020.21 The necessary changes only get larger, the longer we delay them.

In the long run, high debt levels crowd out investment in capital goods, which lowers productivity growth and, therefore, wage growth. They would raise interest rates, hurting borrowers. They would result in large debt payments, which would require spending cuts or tax increases just to avoid further exacerbating debt growth. They would leave us unable to cope with severe economic downturns or to pay for war. They could produce a sovereign debt crisis or a broader financial crisis.

The long-term debt problem is a function of high and rising expenditures rather than low and falling revenues. From 1975 to 2014, federal revenues amounted to 17.3 percent of GDP, and they were 17.7 percent of GDP in 2015. By 2040, they will rise to 19.4 percent of GDP, exceeded only in the war year of 1944 and in 2000 at the height of the tech-stock bubble.22 Federal spending was 20.5 percent of GDP from 1975 to 2014 and 20.9 percent in 2015, but it is projected to rise to 25.3 percent by 2040, higher than any year except 1943–1945, during World War II.

Rising federal expenditures, in turn, will be driven by rising health care costs and by entitlement spending from the aging of the baby boomers. Social Security spending on retired workers and survivors will account for more than one-fifth of additional spending between 2015 and 2040, as will Medicare expenditures, and other health care programs (Medicaid, Children’s Health Insurance Program [CHIP]), and subsidies on the new Obamacare exchanges) will account for more than 10 percent. Interest on the debt will make up about one-fifth of spending over these years, so Social Security for non-disabled Americans and health care programs will amount to over 70 percent of non-interest spending between 2015 and 2040.23

It therefore makes sense to focus deficit-reduction efforts on these programs. In fact, without too much exaggeration, it may be said that what happens to spending levels in other parts of the federal budget is minimally relevant to deficit reduction. The problem with tackling these entitlements—particularly those for retirees—is that political considerations necessitate a relatively slow phase-in of reforms. That means that comprehensive reforms to slow spending on these programs will not have immediate effects on deficits. However, by putting the federal budget on a clear path to sustainability and reassuring investors, these reforms might be expected to increase growth even in the short run.

The first entitlement reform proposed here would slow the growth in Social Security spending for retired workers. The model for Social Security reform would be House Speaker Paul Ryan’s proposal from his “Roadmap for America’s Future Act of 2010.” The 2010 version of the Roadmap would have reduced to the rate of inflation the benefit growth among retirees who have the maximum lifetime earnings subject to the Social Security payroll tax. Retirees with lower lifetime earnings would see less of a cut in benefit growth, with roughly the bottom 30 percent enjoying the same benefit growth as under current law. Moreover, retirees with the lowest lifetime earnings—initially, up to twice the 2009 minimum wage for a full-time worker—who worked twenty to thirty years would be guaranteed monthly benefits between 100 and 120 percent of the federal poverty line for an aged individual, depending on the lifetime earnings and the number of years worked.24

The Roadmap would have accelerated the increase in the Social Security “normal” retirement age (when full benefits may be received) and then indexed its growth. If life expectancy at the retirement age outpaced the duration of working lives, the retirement age would automatically increase. This provision would promote economic growth by encouraging the healthy elderly to work longer. A safety net for those who cannot work or cannot find work would remain in the Social Security Disability Insurance (SSDI) and Supplemental Security Income programs.

Ryan also proposed to subject costs for employer-sponsored health insurance to the Social Security payroll tax. And, the Roadmap proposed a system of voluntary personal accounts to increase the retirement security of future retirees. Workers younger than age twenty-five could choose to contribute a share of their payroll taxes to one of a number of private individual accounts offered through a new Personal Social Security Savings Board. Workers could choose to enroll anytime through age twenty-five and could switch back and forth upon experiencing a qualifying life event. Contributions would not be taxed. Once fully phased in over twenty years, the average worker could direct 5.1 percentage points of the 12.4 percent payroll tax to an individual account. Choosing this option would mean a reduction in retirement benefits from the traditional Social Security program—to zero for workers fully participating in individual accounts over their career. Nevertheless, the proposal would have guaranteed that owners’ account balances would have been at least as large at retirement as their contributions, adjusted for inflation, though the Social Security Administration (SSA) concluded that very few workers would require this guarantee.

All told, the chief actuary of the SSA estimated that, in its first ten years, the Ryan proposal would have reduced the federal debt held by the public by $540 billion (in 2015 dollars). After twenty years, the debt would be lower by $1.2 trillion, and after thirty years by $2.3 trillion.25 According to the SSA chief actuary, it is at this point that the real debt reduction is likely to kick in; forty years out, the debt would be lower by roughly $4 trillion, and then by more than $9 trillion ten years later.

SSA estimates do not take into account the effects of policy changes on economic growth, so these deficit-reduction estimates are likely understated. Another key feature of the proposal is the way it would boost growth, apart from any effect on the national debt. By incentivizing saving through private individual accounts, the proposal would increase investment in the private economy.

Proposal 4: Compassionate Health Care Markets for Deficit Reduction and Efficiency

The Social Security retirement program will bust deficits primarily because it is already so large today. Federal health care spending actually will grow dramatically over time. Spending on Social Security pensions will rise from about 4.1 percent of GDP in 2015 to 5.4 percent in 2040.26 Federal spending on Medicare, Medicaid, CHIP, and the new exchanges created by the recently enacted Affordable Care Act (ACA) will grow from 5.2 to 8.0 percent.

The ACA, while officially projected to reduce budget deficits, will almost surely have the opposite effect and, thereby, hurt growth. When the legislation passed in 2010, CBO estimated that it would save $118 billion over its first ten years and lower spending by one-quarter to one-half of GDP in the subsequent decade.27 However, those estimates depended heavily on the assumption that provider cuts specified in the bill and in current law at the time would proceed without modification. In conveying the estimates, CBO director Douglas Elmendorf tactfully noted:

The reconciliation proposal and H.R. 3590 would maintain and put into effect a number of policies that might be difficult to sustain over a long period of time. Under current law, payment rates for physicians’ services in Medicare would be reduced by about 21 percent in 2010 and then decline further in subsequent years; the proposal makes no changes to those provisions. At the same time, the legislation includes a number of provisions that would constrain payment rates for other providers of Medicare services. In particular, increases in payment rates for many providers would be held below the rate of inflation.28

Here is what Medicare’s chief actuary, Paul Spitalnic, had to say about the provider cuts in the July 2014 annual report of the Medicare trustees:

By the end of the long-range projection period, Medicare prices for many services would be less than half of their level without consideration of the productivity price reductions. Before such an outcome would occur, lawmakers would likely intervene to prevent the withdrawal of providers from the Medicare market and the severe problems with beneficiary access to care that would result. Overriding the productivity adjustments, as lawmakers have done repeatedly in the case of physician payment rates, would lead to substantially higher costs for Medicare in the long range than those projected in this report.29

In April 2015, Congress passed legislation—subsequently signed by President Obama—that reduced the cuts over the next twenty years to provider reimbursement that the ACA depended on for deficit reduction. However, the cuts that are now assumed are still unrealistic. Starting in 2048, the provider cuts established in the new law are projected to be bigger than those assumed by the ACA.30

In December 2015, Congress passed—and the President signed—legislation delaying by two years the so-called “Cadillac Tax” on benefits from the most generous employer health plans. This tax will not be any more popular in 2020, when it is now scheduled to take effect. Nothing has changed, then, since the ACA passed: health care reform likely will increase deficits because the provider reimbursement cuts and benefit taxes on which it depends are unlikely to be implemented.

Not only is the ACA likely to increase deficits; it has made our already-inefficient health care system even more so. The ACA exchanges regulate the number and types of health plans available to consumers, what benefits health plans must offer, who they must serve, how much more they can charge high-risk enrollees versus low-risk ones, how much cost sharing they can impose, and how much profit they can make relative to their revenues. It (further) restricts how much health care providers can charge Medicare patients. It orders Americans to purchase health insurance coverage or pay a fine, and it levies a penalty on employers that do not offer coverage to their workers. Even where many Americans are willing to make one or more of these trade-offs in the name of compassion or justice, there are more efficient ways to realize those values in our health care system.

In any kind of health care reform, of primary concern should be how two groups fare: the poor and the sick. Obamacare’s approach is to create risk-pooling exchanges to protect the sick, to mandate comprehensive benefit packages, to subsidize the insurance of the poor and much of the middle class, to set the prices that health care providers can charge via Medicare, and to mandate that everyone on some level be “in the system” to prevent the whole contraption from unraveling. This is an expensive and needlessly complex way to protect the poor and the sick.

A primary source of inefficiency in the American health care system is the convention of misapplying principles of insurance to pay for routine and predictable health care services. We don’t believe that homeowner insurance should cover the cost of painting the house. We don’t rely on car insurance to pay for gas; we use it in the event that we are in an accident and incur large expenses. Imagine that we could purchase insurance policies that would pay for home and car maintenance. Further imagine that we purchased it in groups—at work or in a regulated exchange—where everyone was charged the same amount for the insurance, and imagine that someone else—an employer, or a taxpayer, or a Chinese bank—paid for much of the coverage.

Several consequences would follow. First, because everyone is paying the same amount for the coverage, the tendency would be to increase the amount of home and car maintenance of which we partake. After all, if my coworker is painting his house and we both pay the same amount for our comprehensive homeowner insurance, aren’t I paying for his home improvement unless I paint my house, too? Second, since any amount I spend would only partially come out of my pocket, I would be tempted to go ahead and get those rims that I’ve been eyeing for the wheels of my car. I wouldn’t have splurged if I had no insurance, but it definitely would be cheap enough to buy with my employer footing a chunk of the bill. Third, the local hardware store would start charging me more for house paint, since I would care less whether he charges a dollar more here or there than he would have if I weren’t being subsidized. Fourth, more and more jobs would go toward serving customers seeking maintenance-related purchases. That would leave fewer people to sell other products and services that would be in greater demand absent comprehensive insurance. Finally, people would start to believe, over time, that it’s not enough to help poor homeowners or nearsighted drivers obtain coverage for trees falling on their house or for three-car pileups. Rather, more and more people would come to believe that everyone should be covered for re-carpeting expenses and oil changes.

For all these reasons, purchases of goods and services related to home maintenance would increase, and the cost of home and car maintenance would rise above the prices that would have been charged absent insurance coverage. Obamacare is estimated to have raised 2014 premiums in the individual market by 49 percent in the average county.31 And there is no such thing as a free lunch, in the end—I may think that my employer is footing the bill, but he’s really just taking it out of my pay before he cuts the check. It’s nice to have other taxpayers and Chinese banks helping me overconsume, but that eventually creates growth-sapping deficits or necessitates that my taxes go up. And, as home and car maintenance gets more and more expensive, I’ll get more frustrated with the cost of living, and it will be harder and harder to keep the poor from being priced out of these purchases.

Alternatively, if homeowner and car insurance just covers catastrophic expenses, I’ll avoid all this overconsumption, inflation, and misallocation of resources. I’ll be protected from incurring costs that I can’t afford and that I can’t predict, and I’ll be able to pay the costs of home and car maintenance out of pocket.

But what about the poor? How will they pay for home and car maintenance? We should take seriously the limited purchasing power of low-income people and families, but does it really make sense to rely on comprehensive insurance and incur all the distortions it creates? Aren’t there simpler and more efficient ways to help the poor? There are; and, in the context of health insurance, the same is true of the sick.

The proposal offered here is based on a plan created by a team of physicians and health policy experts and sponsored by the National Research Initiative at the American Enterprise Institute.32 It would create a single federal exchange that would offer a wide array of plans from participating insurers, subject to fewer regulations than even the pre-Obamacare status quo. It would thus circumvent inefficient state regulations around, for instance, health benefit requirements. There would be no community rating, meaning that insurers could charge enrollees rates based on their specific health status, age, and other characteristics. There would be no tax preferences favoring employer-sponsored health insurance, though employers would be free to provide their workers with group coverage. Subsidies would be provided to individuals on the basis of income and health status, as estimated by expected medical expenditures, to provide help with premiums and cost sharing. The acute-care portion of Medicaid would be folded into the new system (and states would be required to contribute revenues equivalent to what they would have contributed), but Medicare would not.

Every insurer participating in the exchange would have to offer a quote for a standardized “basic plan” to every person seeking coverage. The basic plan, featuring a high deductible, would serve to establish the subsidy given to consumers, who would receive premium support equal to some fraction of the second-lowest quote, which would be different for each person or family, with the fraction varying by income. Insurers also could provide quotes for other, more comprehensive, plans, including ones involving long-term contracts.

Consumers could choose any plan they wanted, but the subsidy they would receive would be tied to the second-lowest-cost plan quoted to them, so they would bear the additional expense of choosing more comprehensive plans. Individuals and families on the exchange would receive contributions to health savings accounts that would vary with income and health status. The health savings accounts could be used toward the cost of more generous health plans or for expenses before deductibles have been reached.

While catastrophic coverage would be less generous than the policies that many insured Americans currently enjoy, it should not be assumed that most people will be unable to afford the protection provided by more generous policies. The premium support and health savings account subsidies would shore up purchasing power. Notably, accumulations in health savings accounts roll over from year to year, so even many low-income consumers would be able to afford plans more generous than the basic plan. Many people would see the cost of coverage that they formerly had on their jobs decline because of individual ratings, and workers who used to have employer coverage but no longer do would see their take-home pay rise. They could use that discretionary income to purchase coverage along the lines of what they used to have. Furthermore, because of the elimination of most of the economic distortions of the current system, health care inflation ought to slow, making the new, market-based system more attractive relative to today’s with each passing year.

There would be a “safety-net tax” that those without coverage would have to pay to make up the cost of uncompensated care. Providers of such care would be reimbursed with the revenue from the tax. In this system, the cost of the sick would be absorbed by healthier taxpayers through direct federal subsidy of high premiums. But, because catastrophic coverage would be affordable to everyone (healthy young people, for instance, would enjoy low premiums even before any premium support), because those not covered would pay the safety-net tax, and because the cross-subsidization of the sick is not confined to exchange participants, there would be no free-rider incentives for people to go without insurance until they are sick.

While eliminating tax breaks for employer coverage might sound radical, there is a strong case to be made for doing so. The United States is uniquely dependent on employers to provide health insurance and other benefits. The system made sense in earlier decades, when lifetime employment was the expectation by employer and employee alike and when few married women worked.

Today, however, employer benefits have a number of disadvantages that hurt economic growth. Employer-provided benefits burden employers with risks that other rich countries have spread more broadly, potentially hurting America’s international competitiveness. They impede job mobility and leave workers vulnerable when they are out of work. They also discourage cost containment by obscuring trade-offs faced by employees and insulating consumers from their choices. In addition, information about the receipt of employer-provided benefits is typically absent from major social surveys. That inspires economic anxiety by obscuring the extent to which rising non-wage compensation contributes to improved living standards and instead showing misleadingly small gains in wages and salaries. Finally, employer-provided benefits limit choice on the part of workers (who can choose only among the options provided to them by their employers) and limit competition among insurers (which would have to compete for individuals if they did not have to compete simply for firms).

Employer-sponsored health benefits still would be available to workers, but they no longer would be tax-favored. The proposals outlined in this growth agenda would eliminate the health care-related breaks in the individual income tax code, and doing away with corporate income taxes would eliminate health care subsidies to employers. Because these proposals are roughly revenue-neutral—replacing the individual tax breaks with lower marginal rates, for instance—eliminating these tax incentives would not raise taxes in the aggregate among rich or poor filers. The Social Security reform plan outlined above would remove the exclusion of employer-provided health insurance from payroll taxes and thereby would fund the transition to individual accounts.

While some have expressed concern that removing employment-based subsidies for group health insurance might cause employers not to offer such coverage, it is unclear why that should be the case. If employees want group health coverage, there is no reason that a firm should not be able to offer it, at the expense of higher wages and salaries. These sorts of trade-offs between pay and benefits currently exist, but employees are shielded from the full cost of the trade-off by the cross-subsidies embedded in the existing tax breaks.

This proposal would be designed to be deficit neutral from a current law baseline. Premium support would be offered primarily to lower-income individuals and families and would phase out as income increases, with no subsidy for those with incomes above a certain threshold. It would be funded primarily through the savings from replacing Medicaid acute-care coverage with the new system that integrates the poor into the broader health care system enjoyed by everyone else.33 Because the ACA is likely to cost significantly more than CBO projects it will—as evidenced by CBO’s skepticism about the viability of the legislation’s provider cuts—this proposal, in combination with the Medicare reform discussed next, should reduce actual real-world deficits. Even if it did not, it would promote economic growth and the welfare of Americans by making our health care system much more efficient.

Proposal 5: Improve Medicare and Reduce Deficits through Competition

Medicare simultaneously is much larger than the other health care programs and will grow at a faster rate than Social Security, rising from 3.0 percent of GDP to 5.1 percent between 2015 and 2040.34 The trust fund on which the program relies for hospital-based spending is projected to become exhausted in less than fifteen years, at which point taxes will need to be raised or spending will need to be reduced.

These estimates, it should be emphasized, take into account the features of the ACA intended to control costs. These include the creation of the Independent Payment Advisory Board (IPAB), the Center for Medicare and Medicaid Innovation, value-based purchasing for hospital services, the Patient-Centered Outcomes Research Institute, and other initiatives. A much bigger ACA cost-saving measure was the inclusion of the hundreds of billions of dollars in provider cuts already discussed. Because those cuts are unlikely to be enacted, Medicare’s effect on longer-run deficits—and thereby economic growth—almost surely are understated by official projections. The shaky fate of the Cadillac Tax also points to this conclusion.

In theory, Medicare could be incorporated into the health care reform model just outlined, where participating private insurers would be required to offer seniors quotes on a standard benefit package but could charge whatever premiums they want, with federal subsidies for premiums and cost sharing. The approach proposed here would depart from this model in key ways, largely to retain the entitlement to existing Medicare benefits and in recognition of the potential cost to the federal government of subsidizing individually rated premiums in an older, poorer, and sicker population.

Medicare should be transformed into a system based on “premium support” for beneficiaries and competitive bidding among insurers. House Speaker Paul Ryan has explored Medicare reforms along these lines for several years, and the proposal here essentially endorses the reforms embodied in recent House budget resolutions.35

Specifically, those becoming eligible for Medicare in 2025 or later would see wholesale changes in the Medicare program. Rather than most of them receiving benefits under traditional, “fee-for-service” Medicare, new beneficiaries would choose from offerings on a “Medicare Exchange.” Private plans on the exchange would bid on a standard benefit package that would be actuarially equivalent to today’s Medicare package and would compete with one another—and traditional Medicare—for beneficiary dollars. Plans also could offer packages more generous than the standard one, though none less generous.

Those beneficiaries would see their premiums subsidized by an amount equal to the second-lowest bid for the standard package in their region.36 Unlike the proposal offered above for the non-elderly population, this bid would be the same for all individuals in a region. People could choose more expensive plans, but they would have to pay the difference out of their own pockets. Choosing the least costly plan would allow them to keep the difference between its cost and that of the second-cheapest plan.

Payments to private plans on behalf of Medicare beneficiaries would be adjusted based on age and health status. Upper-income beneficiaries would face higher premiums, as in today’s Medicare programs, for physician services and prescription drugs. Otherwise, premiums would not vary with individual characteristics, as they do in the proposal above for the non-elderly. Furthermore, a risk-adjusted mechanism would be instituted so that plans with more or fewer costly enrollees would see transfers among one another on a budget-neutral basis. Plans would have to provide coverage to anyone who could afford it, and low-income beneficiaries would receive subsidies to partly defray out-of-pocket costs.

CBO estimates that in 2020, a premium support system beginning in 2018 and including all Medicare beneficiaries (not just new ones) would reduce Medicare spending by 6 percent, or $45 billion, if premium support were pegged to the cost of the second-lowest bid.37 It would reduce spending on hospital and physician benefits by 11 percent for those not also eligible for Medicaid. These effects, CBO surmises, would not grow for a decade, but probably then would increase over time. CBO suspects that, if the provider cuts scheduled under current law are not implemented, the effect of the premium support system on Medicare spending would be greater than implied by these figures. However, the proposal here would phase in and exempt older seniors, so it would take longer to realize the full savings from switching to premium support. CBO estimates that, even after ten years, only 45 percent of the Medicare population would be participating in premium support if it were phased in to include only new beneficiaries.

Premium support would result in lower premiums because competitive pressures would lead plans to reduce the amount of excessive health care services delivered to enrollees and to reduce their profits and administrative costs. CBO also assumes that, even after risk adjustment, private plans would have healthier enrollees than traditional Medicare does. In addition, cost shifting—where providers charge higher rates to private insurers to make up for the cost controls they face in traditional Medicare—would diminish as traditional Medicare’s share of beneficiaries declines. These effects would be greater than the effect of upward overall pressure on provider rates that would result from the shrinking of traditional Medicare, with its single-payer design, and the effect of expanded, less-efficient provider networks in private plans to handle the increasing number of enrollees.

CBO estimates that the annual premium charged by traditional Medicare in 2020, averaged across regions, would be $1,600 higher than the average of the plans with the second-lowest bid and $1,500 higher than what it would have been in Medicare, absent premium support (nearly twice as high). In most regions, the premium bid from traditional Medicare would be significantly higher than the second-lowest bid; but in some regions (including those in which there were no private plans participating), Medicare would be the second-lowest bid or would be close to or lower than that bid. In some regions, then, many more people would opt for traditional Medicare than in other regions.

In this system, half of Medicare beneficiaries would opt for traditional Medicare, according to CBO. They would pay annual premiums $900 higher than under current law, while those opting for private plans (not necessarily the second-lowest-cost one) would pay $200 more than under current law—$700 less than those choosing to remain in traditional Medicare. Overall, beneficiaries would pay premium costs $500 more than under current law. Their out-of-pocket expenses would decline, and total spending by beneficiaries would be 11 percent higher in 2020 than under current law.

Because of the political costs of making dramatic changes to Social Security and Medicare, it is difficult to implement deficit-reduction policies that significantly alter the benefits of today’s seniors. That makes it challenging to reduce federal deficits through entitlement reform in the next twenty years. Increases in revenues or spending cuts on the order required to make a bigger, quicker dent in our fiscal problem would be large enough to potentially have short- or long-term costs to economic growth.


Any comprehensive strategy to increase entrepreneurial growth should include reforms of our inefficient tax code and of the entitlement programs that otherwise will create growth-sapping deficits. That is, it should seek to better use federal fiscal policy to promote growth. The tax and spending policies proposed here constitute five major reforms along these lines. Taxes would not increase much—employer-provided health insurance would be subject to payroll taxes, but other revenue changes would net out and have no effect on the distribution of taxes. Spending reduction would be focused on the major entitlements for retirees and related to health care, where cuts are possible while promoting growth and improving the well-being of beneficiaries.

If this agenda were enacted, the federal budget would look very different, and it would be much friendlier to growth. “Frontier economics” requires modern fiscal policies that are more conducive to growth than the inefficient ones we adopted, in many cases, decades ago. We no longer have the luxury of living with bad policy if we want to escape the disappointingly low growth rates we have seen in recent years.

About the Author
Scott Winship is the Walter B. Wriston Fellow at the Manhattan Institute. Previously, he was a fellow at the Brookings Institution. Winship’s research interests include living standards and economic mobility, inequality, and insecurity. Winship writes a column for; his research has been published in City Journal, National Affairs, National Review, The Wilson Quarterly, and Breakthrough Journal. He holds a B.A. in sociology and urban studies from Northwestern University and a Ph.D. in social policy from Harvard University.


  1. Lindsey, Brink. 2011. “Frontier Economics: Why Entrepreneurial Capitalism Is Needed Now More than Ever.” Ewing Marion Kauffman Foundation.
  2. In a companion paper, I offer an upward mobility agenda to complement this growth agenda. See Scott Winship. 2015. “Up: Expanding Opportunity in America,” in Policy Options for Improving Economic Opportunity and Mobility (Washington, DC, and New York: Manhattan Institute for Policy Research, Center on Budget and Policy Priorities, and Peter G. Peterson Foundation).
  3. Lindsey, Brink. 2013. “Why Growth Is Getting Harder.” Cato Institute, Policy Analysis No. 737. 
  4. Gordon, Robert J. 2014. “The Demise of U.S. Economic Growth: Restatement, Rebuttal, and Reflections.” NBER Working Paper No. 19895 (Cambridge, MA.: National Bureau of Economic.
  5. All GDP and productivity figures are the author’s computations from Bureau of Economic Analysis data.
  6. Winship, Scott. 2015. “Nice Non-Work If You Can Get It.” National Review 67(1). Scott Winship. 2015. “How to Fix Disability Insurance.” National Affairs 23. 
  7. Gale, William G., and Andrew A. Samwick. 2014. “Effects of Income Tax Changes on Economic Growth.” Working Paper (Washington, DC: The Brookings Institution). 
  8. Salam, Reihan. 2013. “A Plan for Tackling Corporate Debt Bias.” The Agenda Blog, February 22, 2013. 
  9. Toder, Eric, and Alan D. Viard. 2014. “Major Surgery Needed: A Call for Structural Reform of the U.S. Corporate Income Tax.” Working paper. 
  10. Ibid.
  11. See ibid. for discussions of taxation of non-publicly traded businesses and other technical details of the proposal.
  12. Brookings-Urban Tax Policy Center estimates provided to Toder and Viard indicated that this swap would cost $168 billion in 2015, or 47 percent of what corporate tax revenues would have been. Using the CBO estimates for corporate tax revenue from 2016 to 2025 and multiplying by 0.47, the ten-year cost of this proposal would amount to $2.16 trillion. See, Table 4-1. I thank Eric Toder for suggesting this approach to estimating ten-year costs.
  13. Eliminating the deductibility of state and local taxes would raise roughly $1.2 trillion over ten years, based on CBO figures at I begin with the first year after the phase-in and adjust the CBO’s ten-year estimate accordingly. Including investment income from life insurance and annuities would raise roughly $220 billion over ten years, based on CBO figures at, which would raise $24 billion in the first year after the phase-in. I adjust the ten-year revenue estimate accordingly.
  14. See for ten-year estimates. I begin with the first year after the phase-in and adjust the CBO’s ten-year estimate accordingly. See for the rationale for linking the maximum taxable level to economy-wide earnings.
  15. Congressional Budget Office. 2010a. “Social Security Policy Options.” To be sure, this extension is a fairly meaningless convention of trust fund accounting. Raising payroll taxes will push out the solvency of the trust fund by adding revenues from which Social Security benefits may be paid, allowing the rest of the federal government to borrow more from the trust fund than would have otherwise been possible. But the federal government’s greater debt to the trust fund eventually must be paid back. In the long run, the debt owed the public will be unaffected because borrowing will be necessary either way—to pay Social Security benefits or to pay back the Social Security Trust Fund.
  16. Elsewhere, I have proposed expanding the CTC and the EITC to promote work and delayed, planned, and marital childbearing. See Scott Winship. 2015. “Up: Expanding Opportunity in America,” in Policy Options for Improving Economic Opportunity and Mobility (Washington, DC, and New York: Manhattan Institute for Policy Research, Center on Budget and Policy Priorities, and Peter G. Peterson Foundation). 
  17. Berman, Leonard E., and Marvin Phaup. 2012. “Tax Expenditures, the Size and Efficiency of Government, and Implications for Budget Reform,” in Jeffrey Brown, ed., Tax Policy and the Economy (Chicago: University of Chicago Press), NBER series vol. 26: 93–124.
  18. Viard, Alan D., and Robert Carroll. 2012. Progressive Consumption Taxation: The X Tax Revisited (Washington, DC: American Enterprise Institute).
  19. Congressional Budget Office. 2015a. “Long-Term Budget Projections,” from “The 2015 Long-Term Budget Outlook.”
  20. Congressional Budget Office. 2015b. “The Long-Term Budget Outlook.”
  21. Congressional Budget Office. 2015a. Regarding the Bush tax cut computation, the total cost from 2001 to 2010 was no more than $2.5 trillion (see and For 2011 and 2012, the cost was roughly $350 billion ( The 2013 to 2022 cost is projected to be $2.8 trillion (see
  22. Office of Management and Budget. 2015. “Fiscal Year 2016 Budget of the U.S. Government.” Historical Table 1.2.  and Congressional Budget Office. 2015b.
  23. Congressional Budget Office (2015b).
  24. Social Security Administration Office of the Chief Actuary. 2010. Letter to the Honorable Paul D. Ryan, April 27, 2010. 
  25. Ibid. I put these into 2015 dollars using the CPI series implied in Table 1c.
  26. Congressional Budget Office (2015b).
  27. Congressional Budget Office. 2010b. Letter to the Honorable Nancy Pelosi, March 20, 2010.
  28. Ibid.
  29. Boards of Trustees, Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds. 2014. “2014 Annual Report.” 
  30. See
  31. Roy, Avik. 2014. “3,137-County Analysis: Obamacare Increased 2014 Individual-Market Premiums by Average of 49%.” The Apothecary Blog, June 18, 2014.
  32. Bhattacharya, Jay, et al. 2013. “Best of Both Worlds: Uniting Universal Coverage and Personal Choice in Health Care” (Washington, DC: American Enterprise Institute).
  33. Medicaid (excepting long-term care expenses) cost roughly $330 billion in 2011 (see Assuming a rough ten-year cost of $3.5 trillion, that would provide the deficit-neutral budget for the subsidies in the new system. In the more generous specification proposed in the AEI report, the ten-year cost would have been closer to $7.5 trillion, but it would have fully covered the second-lowest-cost premiums of every participant on the exchange regardless of income and would have subsidized deductibles and cost sharing much higher up the income scale than proposed here. The tax-credit-based proposals offer alternative cost estimates for comparison. The Burr-Coburn-Hatch proposal would have spent roughly $2.4 trillion over ten years (, and a proposal offered by James Capretta would have cost $240 billion in 2020 alone. The cost of repealing the ACA (roughly $100 billion over ten years—see would be offset by revenues from the interaction of two policies proposed above: increasing the cap on earnings subject to Social Security payroll taxes; and eliminating the exclusion of employer-sponsored health insurance from payroll taxation. The Social Security reform proposal offered above assumed that the revenues from the latter change would be fully spent to pay for reform, but there would be about $100 billion more over ten years available for other purposes because the Social Security reform did not rely on the additional revenue from lifting the earnings cap. According to CBO, eliminating the exclusion of employer-sponsored health insurance from payroll taxes would cost roughly $1.2 trillion over ten years (, p. 23). The earnings-cap proposal here would raise the amount of earnings taxed by 8 percent (; 8 percent of $1.2 trillion is roughly $100 billion.
  34. Congressional Budget Office (2015b).
  35. See; and
  36. The last Ryan budget did not specify how the premium support level would be set, but it cited a CBO report that would set it at the average bid rather than the second-lowest.
  37. Congressional Budget Office. 2013. “A Premium Support System for Medicare: Analysis of Illustrative Options.”

How Law Schools Can and Should be Involved in Building Ecosystems that Foster Innovation, Entrepreneurship and Growth

By Anthony J. Luppino Professor of Law and Director of Entrepreneurship Programs, Senior Fellow with the UMKC/Regnier Institute for Entrepreneurship & Innovation University of Missouri-Kansas City School of Law
Anthony J. Luppino
Anthony J. Luppino



Law, lawyers, and legal processes have pervasive effects on innovation and entrepreneurship, and corresponding potential to either inhibit or facilitate economic growth.1 Statutes, regulations, and associated judicial and administrative processes impact myriad aspects of the planning, launch, and operations of startup ventures, including: identification; protection and use of intellectual property; business entity formation and taxation; capital raising; establishing rights and obligations among the enterprise, its employees, and other goods, services, or credit providers; defining the contours of fair business practices in relation to the physical environment and competitive markets; and many other areas of general or industry-specific regulation.

Lawyers often are portrayed as creators of complexity, and as either obstacles to be overcome or necessary evils to be tolerated in the pursuit of an entrepreneur’s aspirations. In reality, lawyers are more appropriately treated as messengers and navigators of complexity. Moreover, good lawyers take on additional roles and have been described as enterprise architects, transaction costs engineers, deal quarterbacks, and problem-solvers who offer solutions, options, and suggestions. Their ability to serve those functions is born of training as both advocate and devil’s advocate, and is informed by experiences with many businesses and negotiations among contracting parties and with regulators.2

Recognizing that lawyers are among the professionals who have potential to advance entrepreneurship, the Ewing Marion Kauffman Foundation asked me to undertake a 2007 study of what law schools in the United States were doing to help produce more lawyers capable of realizing that potential. The resulting report, entitled Can Do: Training Lawyers to be Effective Counselors to Entrepreneurs, revealed that many U.S. law schools were already actively engaged in very significant initiatives designed to prepare law students to provide value-adding counsel to entrepreneurs, and called on more to get on board with that movement.3 The current landscape reflects continuation of programs that existed at the time of that report as well as several additions, as many U.S. law schools have answered that call in diverse ways, including through enhanced use of technology in skills development,4 new formal degrees or concentrations,5 institution or expansion of interdisciplinary collaborations,6 specialty areas, centers, and institutes for law and entrepreneurship programs,7 and the creation of post-graduate incubator and residency programs.8

While the core task of training more entrepreneurially minded “can do” lawyers is an important one, it is not the only contribution law schools can and should make to the production of entrepreneurship-conducive ecosystems. Specifically, law schools also are well-positioned to be key collaborators in designing and implementing such ecosystems by engaging with like-minded institutions, including, among others, city governments, to: (i) provide education on legal issues and provide or arrange for direct pro bono legal services to larger numbers of innovators and entrepreneurs of initially modest financial means through entrepreneurship and intellectual property clinics or pro bono panels of volunteer lawyers the law schools help organize; (ii) develop technology-based means of eliminating unnecessary obstacles and accelerating the path to successful, job-creating startup businesses (whether or not tech-based); and (iii) encouraging more faculty to pursue research agendas, with components of collaboration with researchers from other disciplines and institutions, that produces policy-oriented publications that propose meaningful reforms to outdated or unduly cumbersome laws, regulations, and legal processes that affect innovators and entrepreneurs.

This paper will present examples of collaborations in each of those three areas drawn from personal experiences at the University of Missouri-Kansas City (UMKC) School of Law, as well as from several other law schools in the United States and Europe. Many of those examples are directly tied to support from the Kauffman Foundation or are responses to calls for action made by the Foundation. The initial draft of this paper was submitted in connection with participation in the Foundation’s June 2015 New Entrepreneurial Growth Conference(NEG Conference) to pose such discussion questions as: How and why should more law schools commit to being regularly and actively engaged with other entrepreneurial ecosystem promoters to establish fertile environments for the birth, development, and growth of startup ventures pursuing for-profit, social, or civic missions?; how can we get those other players to embrace law schools as partners in building those environments?; and what areas and types of policy-oriented research should we challenge law schools to undertake to help bring about significant reforms to spur innovation and economic growth? Accordingly, the following discussion reflects, among other things, feedback from other NEG Conference participants.

I. Education and Legal Services for Startup Entrepreneurs

Law schools are, of course, primarily educational institutions. Many extend their teaching mission to include interdisciplinary interactions through which they help familiarize student entrepreneurs in other academic units, or involved with cross-campus entrepreneurship centers or institutes, with legal matters commonly encountered by startup ventures, and provide tips on efficiently working with lawyers. Some law schools, often through connections with incubators or accelerators, offer education on legal issues and processes to community entrepreneurs who are not college/university students. And at least one, the Northwestern University School of Law, has cast a wider net in terms of acquainting budding entrepreneurs with startup legal issues through a “massive open online course.”9

All of those teaching activities are useful to the entrepreneurs who access them. However, giving a basic grounding and “heads up” warnings to first-time travelers through the morass of laws and regulations confronting a startup venture only goes so far. Those entrepreneurs also need legal counsel to represent them in an attorney-client relationship and give them legal advice applicable to their particular facts and circumstances. Unfortunately, many of them cannot reasonably afford lawyers in the early stages of planning and forming their business enterprise—stages during which critical mistakes can be made, including mistakes that can erode their intellectual property value, trigger unanticipated tax liabilities or other obligations, or jeopardize their ability to timely raise capital in compliance with securities laws.

U.S. law schools have a long history of impactful response to the need for the delivery of direct legal services to innovators and entrepreneurs of modest financial means.10 The American Bar Association website lists 205 ABA-approved law schools.11 As reflected on the Kauffman Foundation’s Entrepreneurship Law website, well over half of them have transactional clinics that provide legal services to entrepreneurs, some with multiple pertinent clinics.12 They vary in type and scope of work—some are expressly entrepreneurship clinics; some focus on intellectual property; some concentrate on community economic development; some assist only nonprofit organizations; some accept students (at least students from outside of the law school) as clients, and others do not; but many regularly serve for-profit, nonprofit and hybrid startups. In general, they involve law students working with and under the supervision of faculty who are licensed attorneys or private practitioners serving in adjunct faculty or similar capacities. Many of these clinics provide pro bono legal services to their clients, while some charge relatively modest fees for “skin in the game” purposes.

The U.S. transactional clinics have associated themselves in a formidable network initiated by some pioneers, most notably Professor Emeritus Thomas Morsch at Northwestern,13 coalesced and grown around the Entrepreneurship Lawwebsite and related listserv, and further solidified by the organization of the Transactional Clinics Group within the Clinical Section of the Association of American Law Schools. The U.S. network held its 14th Annual Transactional Clinical Conference in April 2015. Significantly, included in the program was the initiation of a plan to create what Jonathan Askin of the Brooklyn Law School has called a “network of networks” by working with iLINC, a recently emerging and expanding network of clinics at European law schools that provide legal services to technology startups.14 Building the international network will serve to both enhance sharing of best practices (including, among other things, effective uses of technology in training lawyers and in delivering legal services) and facilitate collaborations among clinics in addressing legal issues and policies arising in multiple jurisdictions.

The “network of networks” of law school transactional/entrepreneurship clinics can be quite powerful in serving many entrepreneurs, and as laboratories for prototyping and refining technology-based tools for efficiently delivering legal services, as well  as sources of data to inform policy research, as discussed further below. Yet, it has two significant limitations that many of us involved with these clinics have been exploring and are seeking to overcome. First, most law school clinics are located in urban areas, but innovators and entrepreneurs may also reside in rural areas. . Second, the clinics’ capacity to provide timely legal services to large numbers of clients often is hampered by one or more of (i) the limited scope of the services they offer, (ii) the fact that law students (or with the European clinics, recent law school graduates) are being trained “on the job” and cannot be expected to work as quickly as veteran lawyers, and (iii) the reality that the number of entrepreneurs who need startup legal services is more than a clinic run by one or a few faculty members with generally small student enrollment can handle.15

The rural area need has been a subject of discussion among interested faculty at various conferences and inspired the inclusion of questions regarding technology-assisted “remote representation” in a survey of U.S. transactional clinics.16 A notable example of action to address the gap in service to rural-area startup businesses is the work in the Central Valley of California farming region being done by the New Business Counseling Practicum, directed by William Kell at the University of California- Berkeley (Boalt Hall) School of Law, occurring in collaboration with other organizations.17

As for addressing the more general issue of law school entrepreneurship clinics’ limited capacity to take on clients, one possible strategy is to leverage the clinic infrastructure by recruiting qualified volunteer lawyers into a pro bono panel organized by the clinic. For example, at the UMKC School of Law, we encountered a capacity challenge while collaborating with a student ventures facilitation program operated by colleagues in the cross-campus Regnier Institute for Entrepreneurship & Innovation.18 The volume of student ventures going through that program was simply too much for our UMKC Entrepreneurial Legal Services Clinic to accommodate. So, in 2011, we reached out to the local bar and initially asked approximately twenty-six lawyers if they would be willing to consider providing pro bono startup legal services to one or two student ventures per year. Rather quickly, about 90 percent of them said yes.

Currently, including the number of additional lawyers the “point people” on the volunteer panel brought in from their firms, we have mobilized approximately thirty lawyers to assist student entrepreneurs who are not in a position to reasonably pay startup legal fees. Within the panel are lawyers with experience in intellectual property, business organizations, employment, securities, tax, immigration, and other areas of law. Our clinic takes applications for the pro bono legal services program, does an initial assessment of financial need and types of legal services required, and then, unless the applicant prefers an engagement with the clinic, provides the names of usually three panel attorneys from different firms that the applicant may want to interview for a possible pro bono startup legal services engagement. Lawyers on the panel and their law firms have assisted many student entrepreneurs.

The high positive response rate we received from community lawyers asked to join the volunteer panel is noteworthy. Transactional lawyers commonly have a strong desire to do pro bono work for low-income or modest-means clients, but not many get organized opportunities to provide pro bono legal services that are directly in the lawyer’s usual practice areas. Coupled with a particular desire to help student innovators and student entrepreneurs, this made our program very attractive to the lawyers we sought to recruit, and most have stayed with it for several years now. There is good reason to believe that transactional lawyers across the United States would be similarly inclined to devote some significant time and talent to providing pro bono legal services to student startup ventures. Many transactional/entrepreneurship clinics refer matters outside their scope of work to attorneys in their communities, but usually in a rather informal way.19 Adopting a more formal “panel” structure to help get transactional lawyers to large numbers of student entrepreneurs in need on a strictly pro bono basis should be a viable approach in many locations across the country. Leadership of the United States Association for Small Business and Entrepreneurship (USASBE), a leading organization of entrepreneurship educators from multiple academic disciplines, has indicated strong interest in promoting this initiative on a national scale.20

Another notable recent development is the United States Patent and Trademark Office’s desire  to develop panels of patent lawyers to assist inventors across the country on a pro bono basis.21 Many inventors, whether students or not, find the costs of paying market legal fees for patent prosecution prohibitive, which can lead to their inventions never getting to market or to the inventor having to give up a very large “piece of the deal” to obtain outsider investment in the patent prosecution. Several cities and regions have embraced the pro bono patent lawyers’ initiative.22 Others seeking to create well-functioning entrepreneurial ecosystems should consider joining in, as well, as part of their overall strategy.

II. “Law Tech” and Entrepreneurship

Using technology to teach law is not particularly new,23 though its evolution and current uptick in emphasis are subjects of an article in the  Journal of Legal Education.24 However, awareness of the tremendous potential benefits society could reap by utilizing technology in creative ways to make law more accessible to non-lawyers and to make the delivery of legal services by lawyers (or perhaps without lawyers) more efficient, affordable, and available is a relatively recent development that is rapidly expanding. Again, the Kauffman Foundation has played a significant role in this surge through its support of, among other intersections of law and technology, the Reinvent Law Lab developed at the Michigan State University College of Law.25

Within the last approximately eighteen  months, the Kauffman Foundation made grants to the UMKC School of Law that allowed us in July 2014 and April 2015 to bring together in Kansas City two passionate groups we believed should learn more about what the other was doing and identify and pursue opportunities for collaboration. The groups were (i) advocates of addressing the plight of the unconscionable numbers of unrepresented or under-represented individuals and small businesses in the United States who cannot afford legal counsel through traditional means, and (ii) “law technologists” who are breaking new ground in seeking to improve access to law and to the delivery of legal services through such efforts as the development and testing of sophisticated new tech tools and the institution of inclusive, interdisciplinary, and forward-looking “legal hackathons” and related networks designed to, as described in a June 2015 ABA Journal article, “create an app, service, policy proposal, or other work product that can address a real-world problem.”26

The two Law Schools, Technology, and Access to Justice conferences had “all stars” from both groups in attendance—too many to name here—and produced a wide range of project ideas and outgrowths that now are being actively pursued. Some of the projects involve policy initiatives discussed in Section III below. For purposes of this paper, I will present in some detail in this section just one: a collaboration between a city committed to building an entrepreneurship-friendly ecosystem and a law school-led interdisciplinary and inter-institution course that bears very little resemblance to a traditional law school course. This type of collaboration has attracted at least a few schools, is replicable, and can and should become a widely used approach to civic entrepreneurship, elimination of impediments to startup ventures, and facilitation of innovation, economic development, and growth at the local level.

The Law, Technology, and Public Policy course we first offered at UMKC in fall 2014, have offered every semester since, , and plan to offer every semester going forward, originated as a direct result of the July 2014 Law Schools, Technology, and Access to Justice conference. Initial credit goes to Dazza Greenwood of the MIT Media Lab,27 who approached us shortly after that conference and asked if UMKC would collaborate in a projects-based course in that nature of a semester-long hackathon, which would involve the Brooklyn Law School and perhaps other institutions, as well,. Three of us at the UMKC School of Law—Dean Ellen Suni, Michael Robak (a law librarian and our Chief Technology Officer), and I—readily said we were on board. We worked with Greenwood and others to, in essence, “hack” together a course in a matter of a few weeks. Among the three of us at UMKC, we have since co-taught in the course in various combinations.

Over the four semesters completed so far, the course has involved UMKC faculty and students from not just the law school, but also from the School of Computing and Engineering and the Public Administration and Entrepreneurial Real Estate programs in the Henry W. Bloch School of Management. Other regular course participants have been various officials from City of Kansas City, Missouri, government (KCMO), including, among others, its Chief Innovation Officer and an assistant city manager; leaders of the Kansas City chapter of the Code for America Brigade;28 and other Kansas City community activists. In addition, the course has had interactions with Greenwood and The MIT Media Lab; individuals from City of Boston government; Jonathan Askin and students in his Brooklyn Law Incubator & Policy Clinic course; and various “legal hackers.”

The course revolves around student teams working on specific projects, has more team meetings than “all hands” class sessions, and has included a number of interesting experiences led by city representatives (including “charrettes” and “card sorts”). The faculty members are learning as we go along and are, in many ways, more like partners with the students than instructors. As Dean Suni has observed, this is a decidedly new and rewarding experience for the faculty, as well as for the students. Law school courses typically involve instructors with exceptional domain knowledge and skills in a given area of law sharing such knowledge and skills with the students. In this projects-based course, the faculty come in with a need to learn the vocabulary and problem-solving approaches of coders and other technologists, of city officials, and of other participants and a need to learn how to work well with all involved and to communicate with each other in efficient and productive ways. There are certainly significant elements of “law” involved, but solving the problems posed by the projects is a multi-disciplinary team proposition in terms of both knowledge and skill sets—and everyone benefits from learning more about the value each participant brings to the table and in developing proficiency in organizing those diverse contributions to achieve the project objectives.

The key to all of this is having projects with potential for expeditious positive impact. Examples have included: work on the development of tools to streamline permitting, licensing, and rezoning processes and make them much more accessible to individuals and small businesses, as well as to large, established companies; ways to use OAuth 2.0 and Open ID Connect29 to allow efficient data access, but with appropriate terms of use/service and security protocols; exploration of city “living lab” initiatives, again with attention to appropriately balancing privacy and security with reaping the benefits of data collection for enhanced, beneficial urban planning purposes; and work on a legal services facilitation tool, which might be embraced by city small business development offices to help founders—especially first-time founders—of a startup venture of any type or size understand and negotiate their internal rights and obligations and make efficient use of legal counsel, and to help legal counsel, including general practitioners, perform more effectively.

By-products of the course have included, among others: students and faculty learning to be architects of problem-solving tech-based tools and to communicate their designs to coders in formats that work well for the coders; students refining their presentation skills by demonstrating their work in “prototype jams” (a term coined by Greenwood) in which they collect critical feedback for use in the next iteration of their tools;30 student internships in city government; spin-out research projects on municipal regulation of sharing economy innovations (e.g., Uber, Lyft, and Airbnb); and the assembly of a long list of potential future projects with which the city would like to work with our students and faculty. The city, the university, local coders, and other community members clearly are working together through this law school-led collaboration to build a constantly improving environment for innovation and entrepreneurship, and the positive effects are immediate.

We certainly are not the only academic institution engaging with city government in such technology-based problem-solving pursuits.31 The point is that all law schools should be considering involvement in civic entrepreneurship as a “win-win” for their students and faculty, their local communities, and the larger net of communities around the country and in other countries, which ultimately can benefit from these kinds of energizing and impactful collaborations.

Before the NEG Conference, participants received two survey questions regarding predictions on things technology may create or destroy over the next twenty-five years. Based largely on experiences with the Law, Technology, and Public Policy course and familiarity with tech-based initiatives at other law schools described above, I had posited in one of my responses that increasingly effective uses of technology would facilitate entrepreneurial growth by making legal processes and the delivery of legal services more efficient and affordable. In a panel discussion early on the first day of the conference, that proposition was challenged. In the very interesting and helpful discussions that ensued from that challenge, I offered reasons why I continue to believe in what I said in the survey response—certainly with respect to legal services and legal processes in the United States, but, I think, with respect to other countries, as well. I will offer here a summary of my reasoning at the conference and some additional support based on reflections since then.

First, it seemed that the panelist challenging the proposition was expressing skepticism that law firms would harness technology in ways that will reduce the cost of legal services. Upon further discussion, it appeared that the focus was on large law firms. I pointed out that, in the United States, approximately 70 percent of the lawyers in private practice are in firms of ten or fewer lawyers, and nearly half of the lawyers in private practice in the United States are in one or two-lawyer firms.32 The reality is that many startup businesses (from “main street” shops to more complex tech-based ventures) will go to solo/small law firms for legal services because they typically tend to have lower overhead and lower hourly rates than “mega firms” do, and because they are in the entrepreneurs’ communities. However, while there are, of course, small law firms that focus on business practice for small and large enterprises of varying complexity, often the small firms approached for startup legal services, in both urban and rural areas, will have general practices that may not involve a steady diet of transactional work with intellectual property, tax, securities, and other laws that startups regularly encounter. Projects many of us are working on involve uses of technology to provide tools for continuing legal education, fact-gathering, and other startup advice to help general practitioners enhance their competency to represent emerging businesses with a much less expensive (in hourly rate terms) “learning curve” than they would have without such tools.

Importantly, using technology to render the delivery of legal services more efficient is not confined to just solo/small law firms. In Dentons’ recent announcement of its NextLaw Labs initiative, the  mega-firm’s Global Chairman, Joe Andrew, was quoted as saying, “We don't just want to be the world’s largest law firm, we want to transform the legal profession, drive innovation in legal services, and bring clients what they want with better quality, more responsiveness, and greater value.”33 That approach does not mean the firm will become less profitable. If done well, it could mean a higher volume of clients and increased firm profits, but at reduced cost to each client—including entrepreneurs.

Further supporting my belief that law firms increasingly will embrace technology as a means to make legal services more readily available, more affordably, to more people, are the energy and ideas witnessed at the two above-referenced Law Schools, Technology, and Access to Justice conferences. Among the action items coming out of those events is the development of global prototyping and vetting mechanisms34 to take advantage of law school-initiated tech-based tools for training lawyers and delivering legal services, and utilizing law school clinics nd post-graduate  incubators as test beds for such tools—as well as sources of data on thousands of startups or attempted startups per year that researchers of entrepreneurial growth trends may well want to tap into. In addition, it was clear that access-to-justice advocates are prepared to push for changes in “unauthorized practice of law” restrictions in state laws and court rules to allow more extensive use of technology by non-lawyers in law processes, with appropriate safeguards. This comports well with the goal of considering and addressing occupational licensing barriers to entrepreneurial growth discussed in several of the draft papers submitted to the NEG Conference.35

Finally, with regard to the proposition that over the next twenty-five years technology will facilitate entrepreneurial growth by making legal processes and the delivery of legal services more efficient and affordable, I am encouraged by both the “legal hackathons” movement—much of which involves volunteer, open-source solutions to challenges at intersections of law and public policy—and several cities’ interest in streamlining their permitting and licensing processes through tech-based solutions. Our experiences with Kansas City, Missouri, through the Law, Technology, and Public Policy course at UMKC demonstrate the city’s sincere desire and deployment of resources to create (i) better inter-department communications, data sharing, and data usage, and (ii) user-friendly electronic portals for stakeholders engaging with such processes, as well as better access to underlying ordinances and understandable explanations of associated requirements, with much less need to pay professional intermediaries to help steer them though those processes. Similar initiatives are ongoing in many other cities, as well.36

As discussed during the NEG Conference, several of the papers submitted to the conference cited the complexity of laws as a significant barrier to entrepreneurial growth.37 Simplification of federal, state, and local laws is certainly a worthy goal. Unfortunately, as innovators innovate and laws try to keep pace, it seems likely that laws themselves will continue to get more, rather than less, complicated. The above-described experiences at the municipal level suggest that a more realistic goal, which at least some at the NEG Conference seemed to embrace, is to use technology in ways designed collaboratively by regulators and the regulated to simplify the navigation of legal processes. Law schools can and should be facilitators of such collaborations, in many types of legal processes and at all levels of government. In working with Kansas City, Missouri, in that type of role, we have seen that, in addition to assisting with tools to simplify legal processes, there are, of course, opportunities to provide input on eliminating “dead letter” laws, reforming existing laws, and creating new ones to achieve well-balanced regulation of evolving economies and new types of enterprises. The potential role of law schools in producing related policy-oriented research is the subject of the final section of this paper.

III. Law Schools and Entrepreneurship-Advancing Policy Research and Writing

Entrepreneurship is an interdisciplinary proposition, and successful entrepreneurship is usually dependent on well-functioning interdisciplinary teams. Research on policies that may need to be changed or established to promote innovation and entrepreneurial growth will often, perhaps always, involve laws and regulations. Law school faculty are accustomed to casting a critical eye on laws and regulatory policies. But how many law professors have substantial experience as entrepreneurs or as hands-on advisors to founders of startup ventures? The answer is that, on average, not many members of the “regular” (sometimes called “doctrinal”) faculty do. Faculty who work in transactional/entrepreneurship clinics, of course, have more of that experience—but they are not always required or encouraged to publish; though, thankfully, that is beginning to change.38

It follows that, to help them identify and truly understand the implications of policies that may need fixing to promote entrepreneurship in desirable ways, law faculty and law students—who are quite capable of making valuable contributions in writing about laws and regulations and proposing statutory or regulatory language—would benefit from involvement in interdisciplinary teams working with entrepreneurs in planning their ventures. While coming to the academy from a pretty lengthy career as a business and tax attorney provided me with more than the average amount of background in advising business clients, I personally have benefitted substantially from interdisciplinary involvement at UMKC. Working alongside faculty from the School Management and the School of Computing and Engineering in interdisciplinary courses and programs centered on the planning of startup ventures, and, in particular, student ventures, has led me to author or co-author publications concerning university intellectual property policies and accidental partnerships;39 immigration law inhibiting students studying in the United States from being active founders of startup businesses while in the country;40 the need for a new hybrid form of business organization for social enterprises;41 and a work-in-process on difficulties in student ventures raising capital in compliance with securities laws.42

Those are just a few examples of policy topics at intersections of law and entrepreneurship. Several others appear in the Kauffman Task Force on Entrepreneurial Growth paper entitled A License to Grow: Ending State, Local and Some Federal Barriers to Innovation and Growth in Key Sectors of the U.S. Economy.43More recent additions to the list include determining appropriate levels of regulation of the “sharing economy” and issues of efficiency versus privacy surrounding evolving ways of collecting, transmitting, and using data—both topics at this April’ssecond Law Schools, Technology, and Access to Justice Conference, and both subject areas in in which faculty and students at the UMKC School of Law have been collaborating with KCMO on policy-oriented research. Although faculty and students at law schools can bring a lot to the table when it comes to policy research, they shouldn’t dine alone as much as it seems they tend to do. While there are plenty of exceptions, law faculty aren’t generally adept at empirical research and, like many of their students, may have gone to law school because they considered themselves “not a numbers person.” They need to work with regulators, economists, and others who can help ensure that their “legal analysis” is in touch with the actual contexts, behaviors and external circumstances in which legal issues are implicated.

At the NEG Conference I highlighted in particular the work of our faculty and students with KCMO and various stakeholders on regulation of sharing economy issues and suggested that a consortium of cities and law schools in their regions might be formed to share data and approaches to developing such regulations, and perhaps fashion related templates to be considered by other municipalities, tailoring them, of course, to their specific circumstances and needs.44 This seemed to resonate well with some participants at the NEG Conference, who supported the notion that law schools can deliver special credibility to such a consortium, as they can apply their legal research and analytical and “devil’s advocate” capabilities without significant political pressures on them. The type of cities/law schools collaboration proposed can tackle not only regulation of the sharing economy, but also address permitting and licensing laws and processes and other challenges. As a consequence of the NEG Conference, several individuals and organizations  are now actively engaged in  growing a  consortium of several cities and law schools launched at an October 30, 2015 event organized by the UMKC and Brooklyn law schools and Dazza Greenwood, and involving  collaborations with many other academic disciplines, organizations and community activists.. Again, this is one of many ways law schools can facilitate entrepreneurial growth by participating as key members of interdisciplinary and inter-institution teams.


For any kind of collaboration to be truly effective, the participants must respect and appropriately value each other’s knowledge, skills, and contributions, communicate well with one another, and function as a team. Law schools, in general, do not have a distinguished record of modifying how they define and conduct their missions to change with the times and keep pace with innovations that speed up how other disciplines and industries approach, adapt to, and even create innovation.45 Moreover, within university settings, law schools have something of a reputation for being a “silo of silos.” However, as discussed in this paper, many law schools and their faculty and students have in the 21st Century demonstrated that they can be innovative and entrepreneurial and can work well in interdisciplinary teams, adding substantial value to collaborative initiatives. That theme is spreading across more and more law schools. Legal education may be a bit late to the party, but let’s forgive that tardiness and take advantage of the contributions to economic growth and addressing social needs that law schools are capable of making. Now is the time to include them in significant ways in building ecosystems conducive to innovation and entrepreneurship.

About the Author
Professor of Law and Director of Entrepreneurship Programs at the University of Missouri-Kansas City School of Law, and Senior Fellow with the UMKC/Regnier Institute for Entrepreneurship & Innovation. A.B., Dartmouth College (1979); J.D. Stanford Law School (1982); LL.M in Taxation, Boston University School of Law (1986). Prior to joining the full-time faculty of the UMKC School of Law in 2001, Professor Luppino practiced law, primarily as a business, tax planning, and transactional lawyer, for a combined approximately nineteen years with law firms in  Massachusetts, Kansas, and Missouri. He has served for several years as the lead editor of the Ewing Marion Kauffman Foundation’s Entrepreneurship Law website and was the principal organizer of the Law & Entrepreneurship Special Interest Group of the United States Association for Small Business and Entrepreneurship (USASBE).


  1. See, generally, Handbook on Law, Innovation and Growth (Robert Litan, ed., Edward Elgar Publishing 2011).
  2. Id. at Ch. 12 (Anthony J. Luppino, The Value of Lawyers as Members of Entrepreneurial Teams), citing, inter alia, George Dent, Business Lawyers as Enterprise Architects, 64 Bus. Lawyer 279 (2009); Ronald J. Gilson, Value Creation by Business Lawyers: Legal Skills and Asset Pricing, 94 Yale L. J. 239 (1984) (coining the term “transaction cost engineer”); Steven L. Schwarcz, Explaining the Value of Transactional Lawyering, 12 Stan. J. L. Bus. & Fin. 486 (2007).
  3. Anthony J. Luppino, Can Do: Training Lawyers to be Effective Counselors to Entrepreneurs ̶ Report to the Ewing Marion Kauffman Foundation, 6 (2008), available at
  4. See, e.g., the LawMeets platforms and resources created by Karl Okamoto of the Drexel University School of Law at (last visited July 31, 2015).
  5. See, e.g., Law and Entrepreneurship LL.M degree offered at the Duke University School of Law described at; Entrepreneurship Law Certificate program offered at the University of Colorado School of Law described at (both last visited July 31, 2015).
  6. See, e.g., the list of Interdisciplinary Programs with Significant Law School Involvement on the Kauffman Foundation’s Entrepreneurship Law website and list of interdisciplinary courses at UMKC on the UMKC School of Law website (last visited July 31, 2015).
  7. See, e.g., CUBE: Center for Urban Business Entrepreneurship at the Brooklyn Law School at; Innovation Practice Institute at the University of Pittsburgh School of Law at; Palmer Center for Entrepreneurship and the Law at the Pepperdine School of Law at; Silicon Flatirons Center for Law, Technology, and Entrepreneurship at the University of Colorado at; the Center for Intellectual Property and Entrepreneurship at the University of Missouri-Columbia School of Law; the Clayton Center for Entrepreneurial Law at the University of Tennessee-Knoxville School of Law at; the John F. Scarpa Center for Law and Entrepreneurship at the Villanova University School of Law; and the UMKC School of Law multi-faceted Law & Entrepreneurship Specialty Area (all last visited July 31, 2015).
  8. See (last visited July 31, 2015), for the American Bar Association’s directory with descriptions of more than forty incubator or residency programs, many established by U.S. law schools, and others by bar associations or other organizations.
  9. See Esther Barron, The Future of Legal Education? Northwestern Law’s First MOOC: Law and the Entrepreneur (last visited July 31, 2015).
  10. See, e.g., Susan R. Jones and Jacqueline Lainez, Enriching the Law School Curriculum: The Rise of Transactional Legal Clinics in U.S. Law Schools, 43 Wash. U. J. L. & Pol’y. 85 (2014).
  11. See list of ABA-approved law schools (last visited July 31, 2015).
  12. See state-by-state listing and web links (last visited July 31, 2015).
  13. Prof. Morsch founded the entrepreneurship clinic at Northwestern, organized the first Transactional Clinical Conference, and connected the transactional/entrepreneurship clinics network through the LAWBUS listserv, which subsequently was transferred to and became the current EshipLaw listserv maintained by the Kauffman Foundation.
  14. Jonathan Askin, EshipLaw, iLINC, and Building the Network of Networks for Startup Law Clinics. John Cummins, operational leader of iLINC, made a featured presentation at the April conference.
  15. In the United States, the standard appears to be only eight students per faculty member. See March 20, 2013, report, Twenty Years After the MacCrate Report: A Review of the Current State of the Legal Education Continuum and the Challenges Facing the Academy, Bar, and Judiciary, by Committee on Professional Education Continuum, ABA Section on Legal Education and Admissions to the Bar, at note 31 and accompanying text.
  16. The 2013 survey, circulated by Megan Carpenter, William Kell, and Anthony Luppino (and on file with author) also posed questions about clinic collaborations with outside attorneys and representation of student entrepreneurs. The survey results were discussed at the 2014 annual conferences of USASBE and the U.S. Transactional Clinical network.
  17. Notes from conversations with and presentations by William Kell, on file with author.
  18. See description of “E-Scholars” program (last visited July 31, 2015).
  19. As reflected in the survey referenced supra note 17.
  20. For a description of USASBE, see (last visited July 31, 2015). In 2014, the then-president of USASBE, Anthony Mendes, spoke on behalf of this initiative at the 13th Annual Transactional Clinical Conference, and its 2015 president, Alex DeNoble, also has expressed enthusiasm for it to the author.
  21. See United States Patent and Trademark Office, U.S. Dept. of Congress, Pro Bono (Free Legal Representation) at (last visited July 31, 2015).
  22. See Jennifer McDowell, An Update on USPTO’s Patent Pro Bono Initiative.
  23. See CALI (Computer Assisted Legal Instruction), formed in 1982: (last visited July 31, 2015).
  24. Michelle Pistone, Law Schools and Technology: Where We Are and Where We Are Heading, 64 J. Legal Ed. 586 (2015).
  25. See (last visited July 31, 2015).
  26. See, e.g., Jason Krause, “Hackathon movement may jump start a legal career,” ABA Journal (June 1, 2015).
  27. Who is also one of the driving forces in civic entrepreneurship and the legal and civic “hacking” movement—see the ABA Journal article in which he is quoted, cited in note 27 supra.
  28. See visited July 31, 2015).
  29. See (last visited July 31, 2015).
  30. The April 2015 prototype jam was intentionally held on the weekend immediately preceding the second Law Schools, Technology, and Access to Justice Conference so that many of the law technologists coming to that conference could participate. This led to one of the resulting conference initiatives being the proposed formation of a global prototyping platform/laboratory involving faculty and students from many law schools, working with technology companies to produce more and better tools to improve the delivery of legal services to entrepreneurs. A small group, including, among others, representatives of the above-described United States and iLINC (Europe), are working on a corresponding proposal to share with other interested law technology advocates. The need for and timeliness of this initiative have been underscored by one of the world’s largest law firms’ May 19, 2015, announcement that it has formed a subsidiary to facilitate the creation of law practice technologies generally. See (last visited July 31, 2015).
  31. See, e.g., in addition to the Brooklyn Law School and the MIT Media Lab, the approach to “Solving Public Problems with Technology” embodied in The GovLab Academy described at (last visited July 31, 2015) (with key collaborators from MIT and NYU working on “civic technology” projects); and the “Technology Law & Public Policy Clinic” offered at the University of Washington School of Law described at (last visited December 11, 2015).
  32. See Anthony Luppino, Incubating Solo and Small Law Firm Entrepreneurs, Ohio St. Entrepreneurial Bus. L. Journal (2014) at note 1 and accompanying text (citing, inter alia, Am. Bar Ass’n., Lawyer Demographics, (last visited Dec. 29, 2013) (citing the 1985, 1994, 2004, and 2012 editions of The Lawyer Statistical Report, American Bar Foundation).
  33. See announcement cited supra, note 31.
  34. See note 31 supra.
  35. See, e.g., discussion of occupational licensing in the following draft papers submitted to the NEG Conference: John Haltiwanger, Top Ten Signs of Declining Business Dynamism and Entrepreneurship in the U.S. at 9; Brink Lindsay, Accelerating Growth by Curbing Regressive Regulation, at 3: Aparna Mathur, Policy Ideas To Spur Entrepreneurship in the Unites States, at 13; Elliot Schwartz, Kauffman Foundation—Policy and Practice Roundtable, at 3–5; Steven Teles, The Political Economy of Rent Seeking and Inequality, at 6–7 (all on file with author).
  36. See, e.g., references in note 32 supra. See also, Michael R. Bloomberg, Why I’m Betting On Cities And Data (HuffPost’s The Blog, April 20, 2015).
  37. See, e.g., discussion of legal/regulatory complexities in the following draft papers submitted to the NEG Conference: Aparna Mathur, Policy Ideas To Spur Entrepreneurship in the Unites States, at 4–13; Scott Winship, Modernizing Federal Fiscal Policy for Economic Growth, at 3–5 (both on file with author).
  38. As discussed at the 14th Annual Transactional Clinical Conference this April.
  39. Anthony Luppino, Fixing a Hole: Eliminating Ownership Uncertainties to Facilitate University-Generated Innovation, 78 UMKC. L. Rev. 367 (2009).
  40. Anthony Luppino, John Norton and Malika Simmons, Reforming Immigration Law to Allow More Foreign Student Entrepreneurs to Launch Job Creating Ventures in the United States (August 9, 2012), Kauffman Paper.
  41. John Tyler, Evan Absher, Kathleen Garman, and Tony Luppino, Producing Better Mileage: Advancing the Design and Usefulness of Hybrid Vehicles for Social Business Ventures, 33 Quinnipiac L. Rev. 367 (2015).
  42. Adrienne Haynes, Tony Luppino, and Malika Simmons, Following the Crowdfunding: The Need for Better Securities Law Compliant Capital Raising Options for Student Entrepreneurs {working draft on file with author).
  43. Principals drafters Robert Litan and Larry Ribstein (January, 2012) , available at (discussing such potential areas for reform as the rules on authorized/unauthorized practice of law; health care; pharmaceuticals (drug approval and liability); K-12 education; financing of growth ventures; and occupational licensing). On the subject of who can lawfully provide legal services, see Debra Cassens Weiss, Seven people pass test to become nation’s first legal technicians, ABA Journal (June 2, 2015).
  44. This suggestion seems to be consistent with calls for civic action to facilitate entrepreneurial growth reflected in some of the draft papers submitted to the NEG Conference, including: Aaron Renn, Rethinking the Success Strategy of America’s Cities; Tokumbo Shobowale, Facilitating a Shift towards Entrepreneurial Activity in New York City. See also the following draft paper submitted to the NEG Conference: Martin Kenney and John Zysman, Choosing a Future in the Era of Digital Platforms: The Implications and Consequences of the Platform Economy, at 14–15 (discussing the challenge of fashioning appropriate rules to govern such “platform economy” enterprises as Uber, Lyft, and Airbnb).
  45. See, e.g., criticisms of the traditional law school curriculum in William M. Sullivan et al., Educating Lawyers: Preparation for the Profession of Law (Carnegie Foundation for the Advancement of Teaching Report, 2007).

How Can We Do Better?

Lessons from HIV/AIDS

By Amar Bhidé Thomas Schmidheiny Professor at the The Fletcher School Tufts University
Amar Bhidé
Amar Bhidé


The relentless development of new products, practices, and ideas has transformed everyday life to a degree we scarcely could have imagined a decade ago. This transformation spans entirely new artifacts, such as scarily smart mobile phones; how we perform old tasks, such as checking into airplanes and paying tolls and taxes; and even the revival of defunct practices, such as pedaling around cities on rented bicycles. We take for granted that things will continue getting better, faster, and cheaper, with hulking sport utility vehicles providing the fuel efficiency of the small cars of yesteryear.

Recent medical advances have been equally dramatic. Human immunodeficiency virus/acquired immune deficiency syndrome (HIV/AIDS) was tamed in a matter of decades after its sudden outbreak in the 1980s; cholera, leprosy, the plague, tuberculosis, and polio took millennia to cure or control. Gleevec and related drugs have nearly doubled the five-year survival rate for people with chronic myeloid leukemia. Harvoni, a recently approved drug, offers a cure for most hepatitis C patients. Minimally invasive surgery has revolutionized knee replacements, and cataracts are removed and lenses replaced without hospital stays.

But everyday experience suggests—and health statistics corroborate—that advances in healthcare have not had the same transformative impact on the common person’s life as, for example, information technology. Total deaths from cancer actually have increased, although survival rates for many cancers have gone up, and age-adjusted deaths from cancer have declined. Life expectancy is increasing at a snail’s pace. HIV/AIDS and hepatitis C apart, many new drugs target diseases that afflict relatively few patients. Obvious applications of information technology have not been used to improve the delivery of care: We can renew driver’s licenses online, but we can’t make medical appointments in the same way. And while technology (and technology-enabled supply chains) drives down the cost of clothes, computers, and other goods, thus holding down the overall inflation rate, health-care costs continue to rise in spite of significant public and private investments in medical research and development.

Several plausible explanations can be offered about why progress has been slow over the past few decades. The low-hanging fruit of diseases caused by a single bacterium or virus already has been plucked. The great afflictions of our time, such as cancer and aging disorders—whose causes are murky—are more intractable. Great advances in basic scientific knowledge that have been made by mapping the human genome, for instance, cannot be expected to produce breakthrough treatments immediately. Social norms about safety and privacy impede medical innovation to an exceptional degree.

This essay focuses on how limiting pluralism has slowed medical innovation. In other fields, I argue, innovation has become highly democratized and decentralized, and this multiplayer system has fostered a high degree of dynamism. In contrast, Western medicine has long relied on elite researchers with extensive, virtually identical training. Modern approaches to funding and regulating medical research have reinforced this age-old exclusionary tradition and the inherent conservatism of oligarchic governance.

HIV/AIDS stands out as an exception, I argue, in the speed and manner in which the disease was contained. Special circumstances spurred an unusually multifaceted, multiplayer effort. The outbreak of the disease was sudden and ubiquitous, so the impetus for a quick response was strong, with no entrenched paradigm standing in the way of try-anything solutions. Perhaps more importantly, patients included many well-educated and well-placed individuals who forced regulators and other procedure-bound organizations to deviate from their routines. Such potent patient coalitions are unlikely to reestablish.

Yet exceptions also can serve as beacons, potentially making the success against HIV/AIDS more than just a one-off. Some of the practices and attitudes it catalyzed—such as the willingness of the FDA to approve drug “cocktails,” and the skeptical assertiveness of patients —might persist. These and other trends could make medical innovation more open and fast-paced.

I will discuss:

  • The nature and role of multiplayer innovation.
  • The degree to which advances against HIV/AIDS fit the multiplayer pattern.
  • Barriers to ongoing multiplayer medical innovation that make HIV/AIDS an outlier.
  • Trends and policy changes that could make innovation in medicine and health care more inclusive.  

Inclusive, Multiplayer Innovation

Before the industrial revolution, highly talented, ambitious individuals of undistinguished lineage could shine serving God or their sovereigns as priests, soldiers, or colonizers of distant lands. The first and second industrial revolutions allowed exceptionally creative and enterprising individuals, who lacked pedigree or formal qualifications, the opportunity to accumulate wealth and power by serving consumers. However, while the revolutions were meritocratic, they were also highly selective. Geniuses such as Henry Ford had ample scope to revolutionize transportation and build empires. But popular theories of Scientific Management and Taylorism sought to reduce rank-and-file employees to automatons or clerks. At the Ford Motor Company, assembly line workers were well paid, but they also were worked hard and told what to do by a small cadre of industrial engineers.

Eventually, however, big business gave up on extreme forms of centralization. It was demotivating to tell workers exactly what to do—paying high wages did not buy Ford great loyalty.  And as Hayek would have predicted, it was wasteful: Workers who had knowledge of specific circumstances weren’t empowered to make adjustments or undertake new initiatives. Therefore, organizations adopted what Peters and Waterman call “loose-tight” controls, giving frontline workers scope to exercise creativity.

Innovation in general became a broad-based multiplayer game during the twentieth century (Bhidé 2008). The personal computer or the Internet, for instance, does not have a solitary Alexander Graham Bell. Innumerable entrepreneurs, executives of large companies, standard-setting institutions, scientists, programmers, designers, and even investment bankers, lawyers, and politicians revolutionized computing, communication, and commerce. While some players and their contributions are more visible and celebrated than others, the accretion of small advances is as crucial as the conceptual or scientific breakthroughs. Rosenberg, (1982, p. 62), for instance, highlights the importance of the “slow and often invisible accretion of individually small improvements in innovations” that are ignored because of “a preoccupation with what is technologically spectacular.”

The interlacing of small and large advances across different innovators and across time isn’t centrally directed or planned. Although some players participate through organized teams and hierarchies, coordination between teams and hierarchies is ad hoc, with interdisciplinary advances emerging without prior agreement or conscious design. Similarly, although individuals and organizations do plan ahead, because they have limited knowledge of what others are doing or of what will work, the development of new products and technologies involves considerable improvisation, and trial and error.  

While entirely indigenous innovation always has been rare, international collaboration and rivalry now play an unprecedented role in developing the science, technologies, design principles, and business concepts undergirding new products and processes. And because ideas now travel so quickly and easily—and intense competition forces producers to cede more than ninety percent of the value of innovations to their consumers, according to Nordhaus’s (2005) estimates—where innovations are widely and effectively used matters more than where the underlying ideas originate.

Widespread, effective use is not automatic, however—the democratization of what I have called “venturesome consumption” also now plays a critical role. Unlike rich hobbyists who bought early automobiles, millions of the not-so-well-to-do scoop up products, such as the Apple iPad, from the get-go. But buying a new product involves a leap of faith: We cannot know in advance whether it will be worth the price. Similarly, using new products effectively often requires resourceful effort. Modern artifacts are rich in features and are complex. Few products, iPads and iPods included, “just work” out of the box; we have to learn about their quirks and nonobvious attributes. Affordable products also are standardized for mass production, and have to be hacked and tweaked by consumers to suit their individual needs. 

Greater inclusivity in the twentieth century improved our capacity to develop, make, sell, and effectively use new products and services, even if pessimists say that innovation peaked with the second industrial revolution (see Appendix A).[1] Inventions between 1850 and 1900 include the monorail, telephone, microphone, cash register, phonograph, incandescent lamp, electric train, steam turbine, gasoline engine, and streetcar, as well as dynamite, movies, motorcycles, linotype printing, automobiles, refrigerators, concrete and steel construction, pneumatic tires, aspirin, and x-rays. These may well overshadow inventions credited to the entire twentieth century. But in the late nineteenth and early twentieth centuries, extraordinary products and technologies typically were developed by a few exceptional individuals, and sold to a few wealthy buyers. Alexander Graham Bell invented the telephone with one assistant. Automobile pioneers were one- or two-man shows—Karl Benz and Gottlieb Daimler in Germany, Armand Peugeot in France, and the Duryea brothers of Springfield, Massachusetts. But small outfits couldn't develop products for mass consumption. Early automobiles were expensive contraptions that couldn’t be used for day-to-day transportation because they broke down frequently, and lacked a supporting network of service stations and paved roads. One or two brilliant inventors couldn’t solve these problems on their own. Rapid, widespread adoption also was hampered by a paucity of venturesome consumption.

Multiplayer innovation continues to deliver the goods in the twenty-first century, though not all is hunky dory. In my 2015 paper, I argued that the decline in promising, informally financed businesses is an alarming portend for inclusive innovation. And as we will see later in this essay, the problem of limited inclusivity is especially acute in health care, a sector that accounts for a large and growing portion of the economy.

Rolling Back HIV/AIDS: A multiplayer success[2]

The story of how HIV/AIDS rapidly flared across the globe and then was contained is a tale of our times. Unlike diseases of earlier pandemics, HIV/AIDS does not itself kill. Rather, HIV infections induce AIDS, making patients vulnerable to other life-threatening diseases.

The disease isn’t age-old. It is thought to have jumped from apes to humans in the 1920s. After the first recorded fatalities in the early 1980s in North America and Europe, infections and deaths grew at fearsome rates. The pandemic wasn’t airborne, waterborne, or vector borne. Rather, the virus spread through bodily fluids transferred in distinctively twentieth-century ways. The retreat of colonialism brought Haitian doctors to Central Africa, who then carried the infection west. Artifacts from the twentieth century—notably syringes, blood banks, and blood products—and growing drug use, anonymous sex, and international travel after the 1970s helped infect diverse groups, including heroin addicts, gays, bisexuals, and hemophiliacs.

But in contrast to older epidemics that took centuries or even millennia to control, the HIV/AIDS epidemic was contained, at least in the West, in just a few decades. In the mid-1990s, the number one cause of death for individuals ages twenty-five to forty-four in the United States was HIV/AIDS-related illnesses. Now, although there is no cure, drugs have become so effective that GlaxoSmithKline foresees that, in about a decade, its AIDS treatment unit, now Glaxo’s most profitable business unit, may no longer have a purpose. According to Glaxo’s chief strategy office, “The industry has done a fantastic job of taking the fear of the late ’80s, and the death sentence, to one tablet a day.” Glaxo’s and Gilead Science’s drugs leave little room for improvement, short of a cure (Staley 2015).

The advances that rolled back HIV/AIDS had a lot in common with the multiplayer game whose features I have summarized. They drew—to a nearly unprecedented degree in medicine—on the contributions of a diverse cast. These included researchers working in government-funded and pharmaceutical-company laboratories, physicians working in hospitals, public health officials, providers of private capital and research grants, and community organizations. The venturesome role of at-risk individuals and patients in mobilizing a multifaceted, multiplayer rollback of HIV/AIDS was pivotal and unprecedented. They were determined, articulate, resourceful, well-educated, and affluent. They formed advocacy groups, and recruited Hollywood stars to lobby for funding research and treatment. When “cocktails” showed therapeutic promise—but the FDA refused to approve their use in the U.S.—advocates discovered a legal loophole for importing cocktails through “buyers clubs.” Eventually, they persuaded the Food and Drug Administration (FDA) to drop its traditional opposition to cocktails.

The multiplayer character of the effort against HIV/AIDS was evident from the very outset when astute clinical observation, prior advances in immunology, and technologies that enabled rapid sharing of information helped public health officials recognize a pattern across a relatively small number of disparate and geographically dispersed cases. The pattern provided the basis for naming and categorizing the disease before much had been learned about its underlying pathologies.

Similarly, even as a consensus about its name emerged, many actors undertook multifaceted efforts that drew on different knowledge and capabilities to control transmission, test for infections and treat patients.  

Initiatives to control transmission focused on modifying behavior and practices, rather than on scientific breakthroughs or technological innovations. Transmission through unprotected sex was controlled by education about the risks, by new rules (such as the closing of bath houses), and by condom-distribution programs. Infection through contaminated syringes was attacked by instituting procedures to protect doctors and nurses from accidental needle sticks, and by distributing clean needles to heroin addicts. Transmission through transfusions of contaminated blood was controlled first by screening donors and later by treating the blood.

Researchers developed tests for detecting HIV infections by building on the paradigms, knowledge, and techniques of virology. Testing was crucial because of the long lag between infection and the appearance of clinical symptoms. Early detection facilitated the control of transmission (since extra precautions could be taken with individuals who tested positive) and increased the effectiveness of treatments (since treatment could be administered before the virus had seriously compromised the patient’s immune system).

Researchers who developed treatments followed a different approach than researchers who developed tests. They did not rely on a scientific paradigm, but simply tried drugs that had shown efficacy in treating other viral infections and boosting immune systems in an ad hoc, “see what helps” way. Development of treatments also was subject to more stringent regulation. In turn, patient groups pressured regulators to modify existing rules and standards.

Progress on all three fronts was accretive, proceeding through many incremental advances informed by novel discoveries and concepts, as well as by large and small disappointments. The first efforts to control transmission through contaminated blood, for instance, were crude: Blood from anyone in a group considered to be at risk was simply rejected. Similarly, early tests could not detect early-stage infections or show how far the infection had progressed. And AZT, the first effective drug to treat HIV/AIDS, had serious side effects, often damaging the liver and causing anemia; patients also quickly stopped responding to AZT treatments.

These advances, including the breakthroughs, also were accretive in not being sui generis: They built on scientific knowledge, instruments, technologies, practices, and organizations that preceded the outbreak of HIV/AIDS.  

As is common in contemporary multiplayer innovation, participants were interconnected but not tightly coupled within and across their specializations. Sometimes, they consciously agreed to collaborate; in other instances, they drew on ideas and artifacts developed by strangers; and in yet other cases, they engaged in head-on competition.

International collaboration and rivalry were likewise salient in the HIV/AIDS effort. French scientists first identified the AIDS-inducing virus—debunking a prior hypothesis advanced by the American, Robert Gallo—while using a chemical agent developed by the American scientist. Reliable, economically viable tests for the virus were developed by competing multinationals. AZT had been first developed in Detroit to treat leukemia, shelved after it failed to deliver hoped-for results, shown to have antiviral properties by German scientists, and ultimately turned into an anti-AIDS drug by a UK-headquartered pharmaceutical company.

The rhetoric of national competitiveness notwithstanding, who made breakthrough discoveries or pioneered critical innovations and where was less important to the public good than local, on-the-ground effectiveness in exploiting the therapies, tests, and practices derived from high-level scientific or technological breakthroughs. Countries in Europe, or states and provinces in North America where no significant contributions to HIV/AIDS research were made, derived tremendous benefits. In contrast, African countries that lacked the capacity to use advances against the disease realized far fewer benefits.

The campaign against HIV/AIDS deviated from the typical multiplayer pattern in one important respect: The participation of private businesses was almost entirely through large, public companies, and a few new and growing businesses that could raise significant amounts of funding from professional venture capitalists or public markets. Informally financed and self-financed ventures did not play the role they often play in nonmedical innovation, in terms of conducting pioneering experiments and diffusing new technologies.

Contributions of Informally Financed Entrepreneurs

Historically, improvised informally financed ventures have conducted pioneering experiments, especially in fields where the payoff was small or murky. For instance, Ed Roberts bootstrapped the launch of the first personal computer in 1975, and Paul Allen and Bill Gates quickly provided its first high-level computer language (BASIC), also without outside funding.

Improvised entrepreneurs willing to pursue small opportunities also can play a significant role in the diffusion of new technologies. After the success of IBM’s PC legitimized personal computers, a swarm of self-financed entrepreneurs accelerated the computer’s deployment. This second wave of mostly uncelebrated entrepreneurs encouraged wavering customers to take a chance on the new technology, helped tweak mass-produced hardware and software to suit individual needs, and developed complementary products and extensions. In other words, informally financed entrepreneurs facilitated customers’ willingness to adopt new technology.

In my earlier work (Bhidé 2008), I had suggested that U.S. Food and Drug Administration (FDA) rules limit multiplayer innovation found in other sectors. Here, I will make a broader argument that, before there was an FDA, longstanding traditions dating back to Hippocrates discouraged inclusive innovation in medicine. The FDA and funding agencies such as the National Institutes of Health (NIH) then reinforced these traditions.

Barriers to Multi-player Medical Innovation: Why HIV/AIDS is an Outlier[3]

A. Standardized medical training and knowledge. Building on a tradition initiated by Hippocrates nearly 2,500 years ago, the practice of medicine in Europe (and subsequently in America) has over time been restricted to credentialed physicians who undergo lengthy training[4] Although the training entails much more than book learning, in Europe it long has been based in universities with instruction by professors who are themselves physicians. This practice dates back to a school in Palermo, founded primarily for the study of medicine, that is considered to be the first university of any sort in Europe (Nuland 2008 p. 70),

Restricting the practice of medicine to physicians who received university educations had (and continues to have) advantages. It helped establish a baseline of competence and create a common body of knowledge and beliefs among physicians. In turn, this helped provide consistent treatment of patients. University-based professors screened new ideas and helped preserve accepted ones into a canon that would be passed on to succeeding generations. Medical faculties and textbooks also helped disseminate standardized knowledge across geographies. For instance, William Harvey, who discovered the circulation of blood, studied medicine in Padua, Italy, after receiving his bachelor of arts in Cambridge, England. In other fields that relied on masters to train apprentices, knowledge was less codified, and communities that could build on the ideas of past generations were not as cohesive.   

There was also a downside to the collectivized canon. As Thomas Kuhn has argued, once a scientific community accepts a paradigm, its foundational ideas and assumptions are not open to question. The medical paradigm was similarly resilient. Galen, a second century Greek physician, provided a foundation that was almost as unshakable as it was flawed (See box: Galen’s Anatomy).

Galen’s Anatomy (and Humorology)

Galen’s Hippocratic predecessors had only a surface knowledge of anatomy because of “cultural prohibitions against dissecting human bodies,” but they were not overly bothered by their ignorance (Bynum 2008 p, 11). Galen, who saw himself as “completing” the framework of Hippocrates, included a detailed anatomical account in his texts, but derived his schema entirely from animal dissections—he never saw a human dissection.  Worse, even his animal dissections were guided by his religious, creationist imagination.

Galen’s anatomy became deeply entrenched withal. It even survived the introduction of human dissection into medical training because low-status barber/surgeons dissected bodies while professors in elevated chairs recited from Latin Galenic texts to bored students, never looking at the exposed cadavers below (Nuland 2008 p. 72). Ultimately, it was in the sixteenth century when Andreas Vesalius, a Brussels-born professor of anatomy at Padua who had studied medicine in Paris and did his own dissections, overthrew the Galenic paradigm. In 1543, Vesalius published the sumptuously illustrated De humani corporis fabrica (On the fabric of the human body), an unvarnished attack on Galen. Galen, Vesalius declared, had been deceived by his dissections of monkeys. 

The reductive theory that disease was caused by an imbalance or corruption of the four humors (yellow bile, black bile, water, and blood) survived even longer. The humoral doctrine was “at the heart of Hippocratic physiology and pathology,” although it “was not contained in all the Hippocratic treatises.” The Hippocratic ignorance of anatomy likely encouraged reliance on the simple humoral model, since its “operative elements” were “the bodily fluids” (Bynum, 2008, p 11). Galen codified the Hippocratic doctrine of the humors, creatively tying it to a complicated—and erroneous—account of physiology and anatomy. Galen’s interpretation “gave humoral medicine such prestige that it dominated medicine until the eighteenth century” (Bynum, 2008, p. 10, 15).

Standardized medical training also likely contributed to the persistence of treatments and resistance to new therapies. The Hippocratics, for instance, believed in vix medicatrix naturae, the healing power of nature, favoring treatments that would help a patient’s body do its natural work. These included bloodletting, since symptoms such as local inflammations or flushes of fever were regarded as evidence that the body had too much of the blood humor. The practice of bloodletting continued to be a “mainstay of therapeutics until the mid-nineteenth century, and physicians abandoned it only gradually and reluctantly” (Bynum, 2008, p. 13).

Restricting who could practice also restricted who could innovate—or at least whose innovations would be included in the canon. The history of medicine is almost entirely a history of new ideas and techniques developed and advanced by physicians. The title of Sherwin Nuland’s book, Doctors: The Illustrated History of Medical Pioneers, is telling. Individuals who were not physicians faced two obstacles—lack of exposure to patients and treatments, and lack of credibility (See box: Exposure and Credibility).

Exposure and Credibility

French physician Rene Laennec’s invention of the stethoscope, a mechanically simple but medically revolutionary device, illustrates the importance of exposure to patients that only practicing “insiders” could have. Before Laennec’s 1816 invention, physicians placed their ears on patients’ chests to diagnose conditions such as heart murmurs. Laennec’s invention apparently was accidental: Because he was uncomfortable putting his ear directly on the chest of a plump young woman, he used a tightly rolled notebook and found he could hear the sounds more clearly. He then crafted a stethoscope out of a hollow tube with a bell at one end and a diaphragm at the other. Over the next three years, Laennec created the vocabulary still used today to describe breath sounds, and correlated several heart and lung diseases with their auditory patterns. Someone who wasn’t a physician would not have stumbled into Laennec’s discovery, or then been able to categorize body sounds that made the stethoscope a ubiquitous diagnostic tool. 

The discoveries and ideas of non-physicians also could not easily secure the credibility and audience to be accepted into the canon. Leonardo da Vinci preceded Vesalius in his dissections and anatomical drawings; but, although da Vinci’s drawings are famous now, they had virtually no influence on physicians in his time. Ambrose Paré, the son of a cabinetmaker who could not afford a proper university education, instead apprenticed as a barber/surgeon and served in French military campaigns. Paré’s military service secured him an outstanding reputation, and his books eventually transformed surgery. Yet they were disdained by University of Paris professors because Paré wrote in French, not Latin.

Nineteenth century French scientist Louis Pasteur risked prosecution for conducting the first human trial of the rabies vaccine on a nine-year-old boy who had been mauled by a rabid dog. Pasteur did not hold the syringe, and the head of the pediatric clinic at Paris Children's Hospital was present. But because Pasteur wasn’t a licensed physician, his supervision of the vaccination was illegal. As it happens, because the boy was cured, Pasteur was spared prosecution and hailed as a hero.

In the twentieth century, even after the role and training of nurses was professionalized, their lower status in the medical hierarchy severely circumscribed their role in medical innovation. For instance, Dame Cecily Saunders—after securing a degree in Philosophy, Politics, and Economics from Oxford—qualified as a nurse. In the late 1950s, when working as a volunteer nurse, she proposed a specialized hospice. Told that her ideas would never be accepted unless she earned a medical degree, Saunders did just that. Eventually, in 1967, she opened the pioneering St. Christopher’s Hospice and, after another 20 years of trying, persuaded the Royal College of Physicians to recognize palliative care as a medical specialty.

Restrictions on who could practice and innovate were not foreordained. Physicians across much of Asia traditionally were not university trained. Even in Europe, builders of complicated artifacts such as bridges, cathedrals, ships, and aqueducts were not required to have formal university educations. Individuals acquired the necessary knowledge and skills through observation, self-study and formal or informal apprenticeships. Their ability to undertake technically challenging tasks was assessed not by their diplomas, but by their record of past projects, the quality of their mentors and patrons, and, in some cases, by their membership of a guild. Just as more individuals could practice outside medicine, more individuals, sometimes complete outsiders with spotty educations, could innovate (See box: Autodidacts).


As is well known, many of the storied names in the information technology revolution—including Bill Gates, Steve Jobs, and Mark Zuckerberg—have been college dropouts.  They continue a long tradition, outside medicine, of pioneering contributions by the self-taught. John Harrison (1693-1776), for instance, followed his father into carpentry and taught himself to repair and build wood clocks in his spare time. When gridiron pendulum clocks appeared, Harrison’s inventions, such as the nearly frictionless grasshopper escapement, improved their performance. In the 1720s, another clockmaker, Henry Sully, invented a marine clock to help navigators determine longitude, but it worked only in calm weather. Harrison then spent more than four decades creating chronometers that kept accurate time in rough seas, eventually winning a prize from the British parliament after the intervention of the king. Notably, Harrison not only lacked much formal education, he also had difficulty communicating his ideas cogently.[5]   

Outside medicine, unschooled innovators whose contributions spoke for themselves flourished in the nineteenth century. George Stephenson, considered the “father of railways” in Britain, was illiterate until age 18, and learned to read and write in night school.  He became an engine wright in a coal mine after repairing a pumping engine. In 1815, the year before Rene Laennec’s stethoscope, Stephenson invented a mining safety lamp. Then, after studying the workings of a locomotive used to haul coal, Stephenson constructed his own locomotive in a workshop behind his home. 

Thomas Edison, who had just three months of formal schooling, provides another example. Edison’s first inventions derived from a brief stint as a telegraph operator, but he proceeded to rack up more than a thousand patents for a wide range of inventions ranging from electric lighting, to power generation, to sound recording, to motion pictures. Certainly, autodidacts did invent some medical artifacts (such as Benjamin Franklin’s bifocals) but, as a rule, virtually all medical pioneers were trained, practicing physicians steeped in the prevailing medical paradigm.

B. Patient/physician relationships. Traditions governing the relationship between physicians and patients also have discouraged innovations from outside the field, and restricted venturesome consumption in medicine.

The holistic Hippocratic concept of medicine placed individual patients at the center of diagnosis and treatment, whereas a rival school of the time, the Cnidian, focused on the disease rather than the patient. Accordingly, “the Hippocratic doctor needed to know his patient thoroughly: What his social, economic, and familial circumstances were, how he lived, what he usually ate and drank, whether he had travelled or not, whether he was a slave or free, and what his tendencies to disease were” (Bynum, 2008, p. 7).  

Later, even though holism waned (because advances in medical knowledge emphasized diagnosis and treatment of disease) physicians continued to idealize the treatment of patients as unique individuals. Practices pioneered in Parisian hospitals in the nineteenth century made the examination of patients more personal and comprehensive. Physicians began to routinely inspect patients for furriness of tongues and coloration of eyeballs, palpitate to feel lumps or enlarged organs, percuss chests and abdomens and auscultate with stethoscopes. Systemizing the traditional patient-by-patient approach preserved the idealized practice of medicine as artisanal, even as the knowledge and training of the “art” was codified and standardized. As Abraham Verghese of Stanford University’s School of Medicine puts it, “A very important, I would say ministerial, function of being a physician is to be attentive, is to be present, is to listen to that story, is to locate the symptoms on that person of that patient, not on some screen, not on some lab result, but on them.”1 Mass inoculations, mass screenings and the like emerged as specialized “public health” exceptions to the normal practice of medicine.

Much of capitalist enterprise after the nineteenth century has, however, pursued the large profit possibilities of mass production for mass markets. The limited economies of scale in an artisanal system likely curtailed commercial interest in medical advances. Thus, while innovations of the first and second industrial revolutions transformed activity upon activity—plowing fields, slaughtering hogs, freezing and canning produce, preparing breakfasts, making shoes and clothes, lighting streets, transporting goods, traveling for work and pleasure, and performing household chores—few innovators targeted health care. In many lists of great inventions in the second half of the nineteenth century, for instance, only x-rays and aspirin are overtly medical. Of course, the development of electricity, chemical processes and so on did have medical implications. My suggestion is simply that the artisanal approach in medicine (and fields such as law and education) dampened interest in developing and harnessing new technologies for medical care. Matthew Josephson’s (1934) Robber Barons lists magnates from many industries—real estate, steel, copper, railroads, shipping, hotels, coal, sugar, tobacco, and even furs and barbed wire. None of the barons made their fortunes in medicine or medical technologies, or in professional fields such as education, accounting, and law, which preserved artisanal practices.

Another salient feature of the traditional relationship between physicians and patients has been the passivity of the latter, which is expected by the former. The responsibility for diagnosis and treatment has been with physicians who were thought to have the necessary training and experience. The first lines of Hippocrates’s Aphorismi observed, “Life is short, and art long, opportunity fleeting, experience perilous, and decision difficult.” How could lay-patients who had not devoted their lives to learning and practicing the medical art make difficult decisions? Medical professionals, therefore, severely discourage any choosing, mixing-and-matching, hacking, and tweaking that help bridge mass-produced goods on the one hand, and our personal circumstances and preferences on the other. The good patient is compliant, not venturesome. Even the diagnostic reliance on symptoms reported by patients has been progressively diluted in favor of objective tests, typically performed on rather than by patients.  

C. Church and State. “By about 1900,” writes Pickstone (1996, p. 306), “most Western states were supervising and subsidizing the education of doctors, underwriting the policing of medical misconduct, and protecting the regular professional against false claims to qualifications.” Well before the 1900s, however, the state—and church—had a strong influence on the development and use of medical knowledge.

Theological concerns, Nuland suggests, are among those that helped sustain Galen’s claims about human anatomy. Christianity before the Renaissance had “exerted an inhibiting influence” on the study of the human body, writes Nuland. Christian doctrines “diminished the importance of man’s corporeal being as compared to his soul,” and religious authorities were “quite satisfied with the teleological precepts of Galen.” Although Galen’s pagan creator was “quite different from the Judeo-Christian God, the church and synagogue were united in the belief that the Galenic construct accorded with their dogma far better than did any obtrusive efforts of objective research” (Nuland, 2008, p. 71).

The overlap of theology and healing exposed incautious innovators to sanctions imposed by religious authorities. The Spanish physician and theologian—and contemporary of Vesalius—Michael Servetus had corrected Galen’s erroneous account of how blood passes between the heart and lungs. However, Servatus published his thesis (which preceded Harvey’s more general theory of the circulatory system) in a religious anti-trinitarian tract, Christianismi Restitutio (The Restoration of Christianity). Servetus was denounced as a heretic, and in 1553, Calvinists burned him and his book at the stake (with green wood to prolong the agony).[6]

On the positive side, religious orders started hospitals—which later became hubs for the development and dissemination of medical knowledge—to offer refuge (“hospitality”) to pilgrims and the needy. Hospitals, like monasteries and nunneries, also sometimes included infirmaries to tend to the infirm and sick. Hospitals assumed an exclusive medical role in the Middle Ages, many specializing in caring for leprosy and plague patients. Religious support continued as hospitals served patients rather than pilgrims, and the names of storied European hospitals such as Hotel Dieu in Paris, St. Bartholomew's Hospital in London, and Santa Maria Nuova in Florence reflect their religious origins.

The state, to the extent it was distinct from the church, had a secular interest in medicine. Infectious diseases, especially the plague, could disrupt the public order and infect ruling elites. Wealth and power did not protect against illness. Appointments as court physicians were coveted and could provide valuable platforms for advancing medical ideas. The Greek Galen, for instance, served the Roman Emperor Marcus Aurelius, whose patronage protected Galen from the “vendettas of his rivals” and gave him the freedom to pursue his investigations with “plenty of help in the preparation of his manuscripts” (Nuland, 2008, p. 50). Treating wounded soldiers was of vital interest to the state, and also was a pathway to influence for some innovators and a spur to their innovations (see box: Ministering to the Military).

Ministering to the Military

Serving in military campaigns was a path to influence for some pioneering physicians. Before Aurelius appointed Galen his court physician, the emperor had asked Galen to join his campaign against the hordes of the Marcomanni. Similarly, while barber/surgeons had low status in the medical hierarchy, they were considered indispensable on the battlefield. Paré’s service to French forces in several campaigns of the Italian Wars and the Wars of Religion—which started before his birth in 1510 and ended after his death in 1590—secured the barber/surgeon royal patronage and protection: King Charles IX personally intervened to spare Paré in the St. Bartholomew’s Day massacre of Paré’s fellow Protestant Huguenots.

Changing military technologies also spurred medical innovation. “Each new war demands yet further improvements in medical care,” writes Nuland (2008, p. 96), because it brings “more efficient methods of destruction. The injuries are more complicated, requiring increasingly sophisticated knowledge of the body in order to treat them.” European armies began using small guns and gunpowder in the fourteenth century, and artillery was first extensively used in the Italian Wars. Doctors of the time mistakenly believed that hot oil was necessary to “detoxify” gunshot wounds, providing Paré the opportunity to discover—after his supply of hot oil ran out—that bandages with soothing balms were more effective

While state patronage might have given a boost to some innovators, the fact that it was restricted to credentialed physicians likely reinforced barriers to pluralistic innovation. Long, expensive educations (including learning Greek and Latin) typically restricted entry to the relatively wellborn. The pioneering anatomist Vesalius was “fifth in his family’s line of distinguished medical men, all of whom had been either scholars or physicians to royalty” (Nuland, 2008, p. 75). Restricted entry also helped elevate incomes and brought “gentlemanly status” to physicians who could “read Latin and dispute the niceties of Galen” (Bynum, 2008, p. 27). Physicians drawn from this milieu who could successfully compete for court appointments would not be expected to be the scrappiest or most innovative of their peers.2

The French Revolution, which initially tried to make medical practice more inclusive, had the opposite effect: It triggered a revolution in “hospital medicine” that systematized treatment and clinical research, reinforcing the specialized nature of medical training and innovation (see box: A Tale of Two Revolutions).

A Tale of Two Revolutions

The French Revolution first “swept away” the “institutions of medicine—physicians, surgeons, hospitals, the old academies, and faculties”—and in the “heady” early 1790s, “it seemed best for everyone to be his or her own doctor” (Bynum, 2008, p. 45). The Faculte de Medicine of Paris that had controlled teaching since the Middle Ages was dissolved, leaving the training of doctors to individual enterprise, apparently in keeping with the philosophy of the new Republic. But the Revolutionary government’s army needed trained doctors, too, and in 1794, it reopened three medical schools (in Paris, Strasbourg, and Montpellier) primarily to train physicians for the military. 

The National Convention established public competitions (concours) to fill the chaired and adjunct professorships it created in the schools. It decided, in a break with tradition, to pay faculty salaries (and allow them an unrestricted private practice) so that professors did not have to depend on tuitions paid by students. The National Convention also appointed a commission to report on how the reopened schools should function (Bynum. 2008, p. 45).

The commission recommended intensely practical training (the student ought to “read little, see much, do much”) studying hospitalized patients and their corpses. The forty-eight hospitals in Paris overflowing with destitute patients were expected to provide ample opportunity. The report also urged students to be trained both in medicine and surgery, since they would be expected to provide services to the military. Adoption of the recommendations had unexpected, far-reaching consequences. “Teaching hospitals” became centers for systematic research, not just hands-on training. Faculty members could observe and treat more patients in hospitals than they could through house calls.  And hospital patients who “were mostly the poor and uneducated, and therefore powerless to have much of a say in the way they were treated,” (Bynum, 2008, p. 47) served as unproblematic subjects for research. 

Including surgery in the curriculum elevated the standing of surgeons and broadened the nature of medical research. According to Bynum, whereas physicians had been concerned with dysfunctions of the body as a whole (such as might arise through an imbalance of the humors), surgeons dealt with “solid” local problems (such as broken bones and abscesses). The surgical propensity to focus on local solids led to an interest in lesions and pathological changes in organs induced by disease. The opportunity to study many patients with similar symptoms led to pathologico-clinico research.

The clinical side involved recording detailed patient histories that included symptoms reported by patients, as well as the objective examination (inspection, palpitation, percussion, and auscultation) pioneered in the French hospital system. Similar signs and symptoms provided a basis for categorizing diseases.

The pathological side involved associating diseases with pathological changes in organs (the lesions) discovered through postmortem examinations. Hospitals provided a natural venue for these postmortems since their autopsy rooms were as full as their beds. Correlating lesions and symptoms helped differentiate diseases rather than cure them or even identify their causes. It did, however, provide a foundation for later advances such as the germ theory of disease that did lead more directly to cures.

Thus, while the hospital system catalyzed by the French Revolution broadened medicine to include surgical knowledge and sensibilities, it also increased the scope and importance of specialized training and research. Reforms intended to reduce the role of lectures and textbooks in medical education actually increased the scope and reliability of codified textbook knowledge. Hospital medicine also then set the stage for further specialization through laboratory research pioneered by Pasteur and the German physician Robert Koch.

While faculty appointments in teaching hospitals were more meritocratic than court appointments, the role of patrons and pedigree did not disappear. The stethoscope inventor, Laennec, for instance, had been a top student, done important research, edited a medical journal, and developed a successful private practice. Yet because Laennec was a devout Catholic and Royalist and had no powerful sponsors, he could not secure an invitation to enter a concour for a faculty position. Laennec could only secure a hospital-and-teaching appointment after Napoleon fell and the monarchy was restored —and even then it took a personal connection. The appointment had a nearly immediate payoff: Within weeks, Laennec’s hospital rounds led to the invention of the stethoscope.

Contemporary rules

In medicine as in so many other fields, the larger resources commanded by the state in the twentieth century and the broadening of its regulatory reach significantly expanded the influence of the state in shaping innovation. The expansion in the U.S. was pronounced. As the Federal Government increased the share of private incomes it secured through taxes, especially following the Sputnik scare, the Federal Government became a major source of funding for basic and applied research.

Federal agencies and foundations—including the National Science Foundation, National Institutes of Health, Department of Energy, and Defense Advanced Research Projects Agency —developed a structured process to evaluate grant applications. Unlike private philanthropists, taxpayer-funded agencies had to avoid the perception of caprice or bias. Peer review of thoroughly documented grant applications by other well-established researchers became the norm. The process, in turn, favored projects that addressed problems that could be deduced from the prevailing paradigm (“normal science,” in Kuhn’s terms) or sought solutions along lines suggested by such paradigms. Projects based on inchoate or out-of-the-box hunches, and projects whose steps could not be specified in advance, usually were not funded. Similarly, despite blind peer review, credentialed insiders with impressive curriculum vitae and training in preparing proposals had advantages. Plus, because universities received a share of grant funds as overhead, they favored faculty (for promotion and honors) who undertook larger, more expensive projects.

The research and development—and new business development—projects of large, professionally managed corporations, which also became a major feature of innovation in the twentieth century, had a similar bias. Objective, independent scrutiny by bosses and committees could help reassure diffused shareholders that their funds were being judiciously used, just as peer review of grant proposals might comfort taxpayers. Here, too, the approval process favored projects whose risks and returns could be objectively investigated, whose execution could be systematically planned, and that were proposed by credible advocates. The fixed overhead of review and ongoing oversight encouraged large corporations to undertake a few large projects rather than many small ones.

In many fields, systematic big-ticket research financed by governments and large corporations did not displace smaller scale and more ad hoc efforts. Autodidacts and freelancers continued to invent and innovate in homes and garages, even in technical fields such as light polarization and instant photography (pioneered by Edwin Land, who dropped out of college to further his technological interests), Xerography (refined in Chester Carlson’s kitchen), and personal computing. As previously discussed, synergies between organized and ad hoc initiatives made innovation more heterogeneous and multiplayer. New cryptographic techniques today are developed both by researchers working at behemoths such as Google and by offbeat individual programmers such as Moxie Marlinspike, whose simple encryption programs are used by the likes of Facebook (Yadron, 2005).  

Heterogeneity was not universal, however. In some industries, such as nuclear energy and aircraft manufacture, significant government research, funding, and procurement—in conjunction with high industry concentration sometimes maintained by regulatory barriers to entry—gave innovation a uniform character. Innovative projects had objectively verifiable prospects (low “Knightian” uncertainty), and large capital requirements. They were carefully planned and undertaken by experienced, credentialed personnel. This was also the pattern in medicine. Increased government funding and regulation in the twentieth century expanded the prior role of the state in promoting innovation by credentialed specialists working in research institutions, rather than in homes and garages.

The National Institutes of Health (NIH) is an important case in point. It evolved from traditional government concerns with the care of military personnel and with epidemics, and it originated in the Marine Hospital Service (MHS) first charged with providing medical care to serving and retired Navy personnel. When the U.S. Congress asked the MHS to investigate epidemics, such as cholera and yellow fever, it established a lab to study bacteria. In 1930, that lab was designated as the NIH.

In 1967, the NIH created a division to fund research on noninfectious diseases, notably cancer, strokes, and heart disease, which had historically been of lesser interest to governments than infectious diseases. (The NIH already had started supporting cancer research in the 1920s through a partnership with Harvard Medical School and, in 1937, had taken over the previously independent National Cancer Institute.)

After President Nixon declared “war on cancer,” Congress passed the National Cancer Act of 1971, greatly increasing funding for the NCI’s (and thus the NIH’s) budget and responsibilities. In the 1980s, the resources of the NCI were used in the campaign against HIV/AIDS (discussed earlier). In the 1990s, the NIH increased its emphasis on basic genetic research not tied to a specific disease, joining with international partners to launch the Human Genome Project. Overall, the budgets of the NIH increased more than 500 fold3 in the latter half of the twentieth century.

The NIH now undertakes research in twenty-seven of its own institutes and centers (such as the NCI) that employ more than 1,000 principal investigators and more than 4,000 postdoctoral fellows, making it the largest biomedical entity in the world. The NIH also disburses funds, amounting to four-fifths of its total budget, to researchers at universities, medical schools, and research institutions such as the Mayo Clinic.4 There is dispute about whether the NIH favors prestigious researchers and organizations. The NIH, which likely is sensitive to the need to maintain support in Congress, points out on its website that it funds research at more than 2,500 institutions spread across every state in the union. But however broadly the funds might be disbursed, it is a virtual certainty that nearly all of its grantees have doctoral degrees and institutional affiliations. In medical research, freelance Moxie Marlinspikes do not apply for or receive government funds. Even researchers from prestigious institutions can be shut out if they challenge the prevailing paradigm (see box: Swimming Against the Tide).

Swimming Against the Tide

Obstacles faced by researchers who want to study the body’s immune response to cancer illustrate the difficulty of securing funding for research that questions the prevailing paradigm.

In the early 1890s, Dr. William Coley, a prominent New York surgeon, noticed that cancer patients who contracted acute bacterial infections experienced spontaneous remissions. Acting on the hunch that the infections had induced the remissions, Dr. Cooley audaciously injected bacteria into a patient with an inoperable tumor to induce a “virulent infection.”  When the patient recovered completely, Dr. Coley developed a bacterial mixture, known as “Coley’s mixed bacterial toxins,” for treating cancer patients.

But Coley’s approach of stimulating the body’s immunological response was overshadowed by radiology and chemotherapy, and his work was forgotten. After Coley’s death, his daughter, Helen Coley Nauts, found records of her father’s “toxin treatment” as she was going through his papers. For the next twelve years, Mrs. Nauts, a housewife with no medical training, “taught herself oncology, immunology, and record keeping,” tracked down 896 patients who had been treated with Coley toxins, and published findings showing the beneficial effects.  Nauts also secured a $2,000 grant from Nelson Rockefeller to start the Cancer Research Institute (CRI) in 1953.

In 1971, the CRI recruited Dr. Lloyd Old, a physician/researcher, as its medical director, and started a fellowship program to “attract outstanding young scientists to immunology.”5  According to Don Gogel, who has served on CRI’s board since 1981, the fellowships stimulated “basic research that provides the foundation of today’s immunotherapies. The researchers we funded now form the core of the new wave of cancer immunology leaders. But for my first 20 years on the CRI board, I saw a sustained lockout of immunotherapies from the mainstream of funding and research blessed by NIH. The prevailing paradigm of chemotherapy and radiation treatment was favored.  We stayed the course largely because of the conviction of Dr. Lloyd Old, who also served as chair of the Department of Immunology at Memorial Sloan Kettering.”

Lone-wolf innovators acting on hunches or challenging the prevailing views of their peers are not altogether absent in the medical sphere. Dr. Charles Kelman, an ophthalmologist in private practice in New York, invented the cyroprobe, an instrument to freeze and extract cataracts, in 1962. The following year, he developed freezing techniques to repair retinal detachments.[7] Most of Robin Warren’s and Barry Marshall’s paradigm-defying work on how bacterial infections cause duodenal ulcers was done by the two Australian physicians—“after hours or at home,” according to Warren (2006)—without a research grant. When Marshall presented his findings in October 1982, the response was “mixed.” The “standard teaching,” according to Warren (2006), was that “nothing grows in the stomach.” 

After Marshall was denied renewal of his hospital contract in Perth, he resumed research at a hospital in Fremantle in the face of continuing skepticism. He writes that “most of my work was rejected for publication, and even accepted papers were significantly delayed. I was met with constant criticism that my conclusions were premature and not well supported. When the work was presented, my results were disputed and disbelieved, not on the basis of science, but because they simply could not be true. I was told that the bacteria were either contaminants or harmless commensals” (Marshall 2006).

In 1950, Ewing Marion Kauffman, who was working as a salesman for a pharmaceutical company, started Marion Laboratories in his basement. The business—selling calcium supplements made from ground oyster shells—produced revenues of $36,000 in its first year and a net profit of $1,000. When Merrell Dow bought the company in 1989, Marion Laboratories had grown to become a diversified health care giant with nearly $1 billion in sales.

Where these “unfunded” exceptions make their mark is noteworthy. They usually succeed in areas that are not of primary interest to the NIH or grant applicants from mainstream research establishments. Thus, while developers of new surgical techniques or diagnostic equipment (or business models, as in the case of Ewing Kauffman) will face greater difficulty in securing grants than researchers doing cutting-edge genetic research, they also are less handicapped by their inability to secure grants. Similarly, development of treatments for conditions such as cataracts or ulcers—which people have learned to live with, or have resigned themselves to available treatments—offer more opportunities to unfunded innovators than diseases such as cancer, which the NIH prioritizes.

Ample opportunities to innovate outside the NIH’s and the mainstream research community’s interests should, in principle, produce considerable freelance innovation by outsiders. But outsiders face disadvantages beyond their inability to secure research grants. One disadvantage, as previously discussed, is the restriction of medical practice to trained physicians. Like Laennec and other preindustrial revolution innovators, Drs. Charles Kelman, Robin Warren, and Barry Marshall developed their novel ideas in the course of caring for patients. Just as importantly, FDA rules pose formidable barriers to frugal development that innovators can undertake in, for example, cryptography or smartphone applications.

The FDA became a formidable force in the twentieth century[8]; before that, there were few federal laws regulating the production and sale of food or pharmaceuticals. The agency’s foundational legislation was the Food and Drug Act of 1907, passed to control the adulteration and “misbranding” that had increased along with high-volume production and interstate sales.  The U.S. Department of Agriculture’s Bureau of Chemistry, which was tasked with enforcing the act, mounted an aggressive campaign but found its authority checked by the courts. In 1911, for instance, the Supreme Court ruled that the 1906 act did not cover false claims of therapeutic efficacy. Congress responded by expanding the definition of “misbranding” to include “false and fraudulent claims,” but the courts again limited enforcement by setting high standards of proof for fraudulent intent.

In 1938, Congress passed the Food, Drug, and Cosmetic Act after an elixir formulated with a toxic solvent had claimed more than 100 lives. The 1938 legislation gave the FDA (as the Bureau of Chemistry had by then been renamed) sweeping powers. The law did not just mandate premarket review of the safety of all new drugs (which could have prevented the elixir tragedy), it also allowed the FDA to ban false therapeutic claims without proving fraudulent intent, authorized it to inspect manufacturing facilities, and brought medical devices (defined broadly enough to cover toothbrushes) under the FDA’s authority. Courts after the New Deal were also more helpful to the FDA as it enforced its new powers against drug manufacturers that made unsubstantiated claims. For instance, a 1950 court decision (in Alberty Food Products Co. v. United States) held that omitting the intended use of a drug from its label did not provide a safe harbor against “false therapeutic claims.” However, the FDA could not yet prevent the introduction of ineffective drugs; it could only force recalls.

A 1962 amendment to the 1938 act authorized the FDA to put the onus of providing “substantial evidence” of efficacy on developers before they could market a new drug or device.[9] The act defined substantial evidence as comprising adequate and well-controlled investigations, including clinical investigations, by experts qualified by scientific training and experience to evaluate effectiveness. The FDA (1998, p. 2) further took the position that, in passing the 1962 legislation, Congress had “intended to require at least two adequate and well-controlled studies, each convincing on its own, to establish effectiveness.” Moreover, in 1962, the “prevailing efficacy study model” had been “a single institution, single investigator, relatively small trial with relatively loose blinding procedures, and little attention to prospective study design and identification of outcomes and analyses.” Over time, the FDA required efficacy studies to be “multicentered, with clear, prospectively determined clinical and statistical analytic criteria” (FDA, 1998, p. 12).  

There was an obvious downside to tougher efficacy requirements. As the 1998 FDA document acknowledges, “the demonstration of effectiveness represents a major component of drug development time and cost. The amount and nature of the evidence needed can, therefore, be an important determinant of when and whether new therapies become available to the public.” Congress ameliorated the problem of the delays faced by developers of new drugs by passing the Hatch-Waxman Act of 1984. This legislation extended patent protection given to new drugs, tying the extension to the time required to secure FDA approval. But, while the extension increased the incentive for pharmaceutical companies to continue developing new drugs, it did little to speed up the process. It also did nothing to reduce the costs of regulatory compliance and thus the prices consumers had to pay.

High costs of efficacy requirements also limit innovation to established companies and the relatively few new businesses that can secure funding from professional venture capitalists. According to DiMasi, Hansen, and Grabowski (2002, p. 162), the mean cost (in 2002 dollars) incurred during Phase I trials (which provide the first screen for safety) for drugs approved in the 1990s was $15.2 million; the Phase II cost (in which new drugs are tested for safety and efficacy on as many as  a few hundred patients) was $23.5 million; and the Phase III cost (which involves large-scale randomized and blinded testing of thousands of patients) was $86.3 million. These sums are outside the scope of informally financed new businesses. For instance, most founders of companies on Inc.’s list of the 500 fastest-growing companies in the U.S. (which exemplify exceptionally successful, informally financed ventures) start their businesses with less than $20,000. Unsurprisingly, very few “Inc. 500” companies develop products that require FDA approval.6

FDA and intellectual property rules also discourage continuous iterative innovation, even by well-capitalized enterprises (including venture capital-backed companies), which is the hallmark of innovation outside the biomedical sector. As I have previously argued (Bhidé, 2008), the FDA has developed the sensibilities of researchers in the natural sciences (who try to discover universal, parsimonious laws) rather than that of engineers or technologists (who try to develop artifacts that solve specific problems and are often complex).7 In pharmaceutical regulation, “science-mindedness” has apparently engendered strong preference for simple, single-molecule drugs whose therapeutic effects on a specific indication can be more easily isolated than with cocktails or mixtures.[10] The scientific orientation of the FDA also is reflected in rules requiring well-specified, controlled experiments to test the efficacy and safety of molecules (and medical devices).

This regulatory posture limits the scope for “try it, fix it” innovation. Once a compound has been submitted for approval, and the FDA has approved a testing protocol, the developer is in the same position as someone conducting a science experiment: It is what it is. A skilled chef can make up for a missing ingredient by modifying the recipe, but if a compound is discovered to have an unexpected toxicity, the FDA’s single-molecule approach makes it impossible to effect compensatory adjustments by adding something to offset the toxicity. In fact, because FDA-approved protocols (like good science experiments) are more or less cast in stone, the developer cannot adjust the dosage or other aspects of how the drug is administered to patients once a trial is under way. Flexibility afforded to developers of devices, especially ones implanted or attached to the body, to make changes during trials is only modestly greater.[11]

Developers of new surgical techniques do not seem to be significantly constrained in making alterations, however. There are, in fact, no explicit federal regulations governing innovative surgery, save general Department of Health and Human Services Institutional Review Board (IRB) guidelines covering research on human subjects. According to a survey by Reitsma and Moreno (2002), surgeons experimenting with new techniques rarely seek IRB review, and few are even familiar with how research on human subjects is defined.            

The rules discourage ongoing development after a new product has made it to market, as well.  In pharmaceutical products, the costs of establishing safety and efficacy represent an obvious barrier in developing new applications or versions. In addition, according to one expert, pharmaceutical companies that discover an approved drug produces better results in a different dosage, or has utility in treating a different condition, are reluctant to go through FDA trials (for the different dosage or indication) because they are afraid the trials might produce data that will lead the FDA to reexamine the drug already being sold. At the same time, strong intellectual property rights limit the competitive pressures that induce companies like Intel to introduce advances on regular “tick-tock” cycles (improving the manufacturing process on the “ticks” and the architecture of chips on the “tocks”).8 As long as the patent hasn’t expired, pharmaceutical companies typically focus on ways to increase sales (by persuading more doctors to prescribe or insurance companies to reimburse) rather than on ongoing improvements.

Barriers to incremental advances are somewhat lower in devices than in drugs, although the intellectual property incentives are similar. The FDA reduces the evidentiary burden for incremental advances; it clears devices after reviewing a premarket notification (known as the 510 (k)) showing that the device “is substantially equivalent to a device that already is legally marketed for the same use.” New devices, in contrast, must be “approved” rather than “cleared”9 Therefore, ongoing incremental improvements may be a more routine feature in medical devices, as Gelijns, Rosenberg, Nathan, and Dawkins (1995) find. The true extent of incrementalism is hard to pin down, however, because device developers have an incentive to seek FDA clearance rather than go through the onerous “approval” process for new products. Regardless, even in devices, there is no equivalent of the weekly or monthly updates that companies such as Microsoft can issue (without any regulatory approval). With both drugs and devices, regulatory (and intellectual property) rules also preclude third parties from adding value by making small changes or developing complements.[12]

Additionally, the FDA isn’t a friend of venturesome consumption. As mentioned, consumers make leaps of faith in deciding whether new, nonmedical products will be worth the price and risk. They often mix and match or hack standardized, mass-produced products to suit their idiosyncratic needs. But the FDA has been mandated to make choices about safety and effectiveness on everyone’s behalf. And as a practical matter, the FDA’s evaluations pertain to standardized interventions. The agency deems something to be safe and effective when “used as directed” where the “as directed” matches the precisely specified conditions under which preapproval trials were conducted. As in the traditional physician/patient relationship, the FDA’s approach favors patients who comply with the “use as directed” injunction rather than making choices of their own.

One manifestation of the tension between venturesome consumption and the FDA’s mandate and modus operandi is in the area of at-home and direct-to-consumer testing. The FDA treats all home-use testing equipment—and tests sold directly to the consumer—as medical devices. Therefore, manufacturers of testing equipment (or providers of tests) have to establish that the tests are safe, reliable, properly labeled (with warnings prescribed by the FDA), and satisfy the agency’s good manufacturing practice requirement as specified in its quality systems regulations. In addition, the FDA also “requires the results to be conveyed in a way that consumers can understand and use.” In particular, the FDA requires that “user comprehension” studies “must obtain values of ninety percent or greater user comprehension for each comprehension concept.”

If the agency deems consumers incapable of understanding the results of a test, it requires that the results be channeled through a “licensed practitioner.” The cost of satisfying these requirements can make at-home testing commercially unviable, hindering the ability of venturesome individuals to take charge of tweaking interventions and therapies—an outcome that the FDA perhaps prefers.[13]       

Promoting Multiplayer Innovation

There is considerable concern about the nature and cost of medical innovation. Critics argue that new therapies now target diseases of the few rather than of the many, provide small incremental benefits (adding only a few weeks to the lives of the terminally ill, for instance), or control rather than cure chronic conditions. Cheap, effective treatments are underutilized because of inadequate incentives for their widespread adoption. Meanwhile, spending on medical research—and on healthcare overall—continues to rise even as prices in much of the rest of the economy remain steady or even fall.

Typical remedies seem to focus on increasing the effectiveness of specialized researchers. One such approach is to promote “translational” and “interdisciplinary” research. In 2006, for instance, the NIH created the Clinical and Translational Science Award (CTSA) program, which had expanded to about 60 academic medical institutions in the U.S. by 2015. Similarly in 2005, the NIH launched an Interdisciplinary Research (IR) program to “change academic research culture such that interdisciplinary approaches and team science spanning various biomedical and behavioral specialties are encouraged and rewarded.” The program’s components included interdisciplinary research consortia, training programs, a “Multiple Principal Investigator (Multi-PI) Policy,” and the fostering of new “interdisciplinary Technology and Methods.”10

The history of innovation inside and outside medicine suggests, however, that the relationship between knowledge developed through basic research, typically undertaken without regard to its practical use, and the practical application of such knowledge is difficult to predict and control. In some instances, researchers have been able to systematically and successfully apply basic research; in other cases, applications have been discovered serendipitously (the use of the transistor principle in transistor radios) or after frustrating lags (as in the effort, started in the 1980s, to apply knowledge of nano-molecules, which is only now bearing fruit). In yet other cases, practical knowledge and inventions have led the development of scientific knowledge. As L.J. Henderson observed, “Until 1850, the steam engine did more for science than science did for the steam engine.”

In medicine, too, clinical practice often has preceded scientific understanding (Nelson et al, 2011), and long lags between scientific discovery and treatments have been commonplace. Harvey’s revolutionary discovery that blood circulates had virtually no impact on treatments (including the treatment of Harvey’s own patients) for nearly a century. The practical consequences of the many disease-pathology correlations discovered in French hospitals in the nineteenth century took just as long to materialize. Linus Pauling and his colleagues demonstrated in 1949 that sickle-cell disease occurs as a result of an abnormality in the hemoglobin molecule. This was a milestone in the history of molecular biology. Yet the disease remains incurable. How much an organized effort can accelerate the translation of findings in genomics research into treatments—or whether a systematic process can be developed to shorten lags between medical science and medical treatment—remains to be seen.

Similarly, while many important medical and nonmedical innovations have resulted from the cross-pollination or integration of ideas across fields, the process often has been serendipitous. As often as not, innovators have borrowed ideas from other domains without any formal collaboration. For instance, Kelman was inspired to develop photo-emulsification (to remove cataracts after pulverizing them with ultrasound) as he was having his teeth cleaned by a dentist. There are certainly examples of successful, structured collaborations, which are integral to modern design-thinking approaches to organized innovation. Some of the important advances in cataract treatments that followed Kelman’s serendipitous insight resulted from purposeful multidisciplinary effort.  But where and how structured multidisciplinary innovation works better than the ad hoc cross-fertilization of ideas remain open questions.

The success of multiplayer innovation outside medicine and in the campaign against HIV/AIDS suggests consideration of a different regulatory approach—that of trying to make medical innovation more pluralistic, decentralized, and continuously accretive. Granted, such a change would run counter to a very long tradition. As we have seen, medicine long has favored specialization and an epistemological monoculture that induces centralization by standardizing knowledge and training. There are, nonetheless, signs of increased openness.

Outside players have increased their roles. As I suggested earlier, medicine was on the periphery of the first and second industrial revolutions, perhaps because physicians and patients could effectively resist changes to artisanal practices. However, many large companies that emerged from the industrial revolutions did later diversify into the medical sphere. Kodak started making x-ray products in 1896 and, after the Second World War, produced equipment and film for tuberculosis screening. General Electric, Phillips, and Siemens developed large medical-equipment businesses. Similarly, after the success of Genentech, venture capitalists who previously had specialized in nonmedical technologies, such as computers and software, started investing in biotechnology, medical-device, and medical-services companies. Companies such as IBM and Google now are seeking to apply their big-data analytics and cloud computing capabilities to health care.

As Nelson et al (2011) observe, many technologies that weren’t developed for medical purposes have nonetheless been incorporated into medical devices. They point out that lasers, which had “no connection with research aimed to understand disease,” became “a central component” of many “effective medical treatments.” Similarly, computerized tomography (CT) scanners “drew heavily on advances in computers and mathematics, ultrasound had its origins in submarine warfare, and magnetic resonance imaging (MRI) had originated in the work of experimental physicists.” We may further note that large industrial companies and venture capital-backed businesses—not traditional medical researchers—drove much of this cross-fertilization.  

Widespread online information-sharing has encouraged venturesome consumers to take many medical matters into their own hands. Online information-sharing started modestly with the formation of Usenet groups at the University of North Carolina and Duke University in 1980. Through the 1980s, Usenet membership was restricted to the few individuals who had access to the Internet and the technical skills necessary to participate in a group. Since then, Internet connectivity has become ubiquitous, and posting or retrieving information has become mundane. In addition, channels for information-sharing have multiplied to include online forums and social networks, such as Facebook and YouTube. Easy information-sharing has, in turn, prompted individuals to investigate and try to solve problems in domains—including medicine—where they have no training or experience. For many, a web search has become a complement to—and in some cases a substitute for—consulting a physician.[14]

The FDA has cautiously changed some of its efficacy requirements. In the mid-1980s, AIDS activists had pressured the FDA to allow several thousand patients access to AZT before it had been approved. In 1987, the FDA formalized the conditions under which patients could get new drugs that were still in trials. Under “treatment” Investigational New Drug (IND) rules, doctors could prescribe the drugs to patients who were not enrolled in a clinical trial only if the patients had advanced life-threatening diseases for which no other treatment was available. The rules also required the drugs’ developers to “diligently” pursue normal trials, and refrain from promoting or otherwise commercializing not-yet-approved drugs. By August 1994, twenty-nine drugs had been granted treatment INDs, of which twenty-four had received normal approval by the end of that year (Flieger, 1995). In the 1990s, the FDA began “priority” reviews for applications that might produce major advances. It also began accepting evidence of proxy effectiveness—for instance, approving drugs that reduce cholesterol on the premise that reducing cholesterol reduces the risk of heart disease.

At the same time, the FDA increased the scope of its regulation to cover new trends in self-diagnosis and self-testing by individual consumers. As mentioned, the FDA now regulates home-testing devices and tests sold directly to consumers. In one celebrated instance, the FDA forced 23andMe to stop marketing its $99 Personal Genome Service, which provided consumers who mailed in a saliva sample more than 200 health reports. The FDA’s 2013 warning letter complained that the tests could “produce false positive or false negative assessments for high-risk indications,” and that patients who did not adequately understand the test results might use them to “self-manage,” possibly even abandoning necessary treatments. (In early 2015, the FDA allowed 23andMe to offer a single test, the Bloom syndrome carrier test, after the company conducted two separate studies to demonstrate the accuracy of the test— one “usability study” to show that consumers could adequately follow instructions about how to submit saliva samples, and another test to show that consumers could understand the results.11)

Similarly in 2013, the FDA declared that apps on smartphones and other such mobile devices would be regulated as medical devices if they were “used as an accessory to a regulated medical device,” or if they “transform[ed] a mobile platform into a regulated medical device.” It encouraged “app developers to contact the FDA—as early as possible—with questions about mobile apps, their level of risk, and whether a premarket application is required.”12

As it happens, consumers aren’t always mistaken in questioning accepted treatments. Radical mastectomy was the standard treatment for breast cancer until quite recently. The few doctors who questioned its effectiveness would have been ignored had it not been for a patient revolt against the drastic surgery. And FDA regulation hasn’t been foolproof in terms of safety or effectiveness. Recent products that had to be recalled after passing safety tests included the anti-inflammatory drug Vioxx and Guidant’s defibrillators and pacemakers.

Overestimates of effectiveness are arguably even more commonplace. The use of antibiotics to treat ulcers (after research by Marshall and Warren) displaced treatments of dubious utility that, nonetheless, had received regulatory approval. Even if completely useless treatments rarely make it through FDA scrutiny, efficacy in actual use often tends to be much lower than reported in earlier randomized, blind trials. This so-called “decline effect,” according to Lehrer (2010), is “extremely widespread” in medicine, affecting therapies such as cardiac stents, antidepressants, Vitamin E, and antipsychotic drugs.        

Some libertarians might see these shortfalls as a reason to shut down the FDA. Believers in the FDA’s current mission—and in the improvability of randomized trials—would see them as reason to provide more funding for safety and efficacy trials. The analysis of this essay suggests consideration of a third retroradical alternative—require the FDA to return to its foundational mission of controlling safety while limiting its role in testing for effectiveness of those interventions (such as vaccines) where lack of effectiveness poses obvious public health risks (or where the lack of efficacy also unambiguously jeopardizes safety). As the costs of satisfying regulatory efficacy requirements are reduced, the intellectual property protections offered to innovators also should be scaled back. In other words, the implicit bargain reflected in the Hatch-Waxman Act—more patent protection to offset longer efficacy trials—could be reversed. Licensing under “fair, reasonable, and nondiscriminatory” terms now required by standard-setting bodies similarly could be adapted for medical patents, especially those originating in publicly funded research.   

Tough safety rules and little to no regulation of efficacy have become the norm outside medicine. Regulators did not subject transformative nonmedical advances—including steam engines, light bulbs, cars, computers, and the World Wide Web, to randomized tests for effectiveness or comprehensibility of operating instructions. It is only in medicine (and increasingly in foreign aid to third-world countries) where “evidence based” is considered synonymous with controlled randomized testing. Evidence of effectiveness or value typically is collected quite differently.  For instance, in September 2014, Microsoft launched its Windows Insider Program to collect continuous feedback about its operating system as it was being developed. By the end of the year, nearly 1.5 million people had signed up. They did not at all comprise a random sample. Nor did Microsoft test features in successive releases in any structured or controlled manner. Similarly, competing products or technologies are subjected to a pluralistic, Darwinian sort of selection that isn’t blind, standardized, or centralized. Rather, in the multiplayer game, many buyers decide whether to take a chance on new offerings, using their own objective and subjective standards.

In some cases, this trial by the many may lock everyone into a poor choice (such as, allegedly, the “Qwerty” keyboard or VHS videotapes). As a rule, however, decentralized consumer choice supports the diversity of innovators and their offerings by protecting innovators against the prejudice or bias of a few expert judges. (Many industry experts, it may be recalled, panned the iPhone when it was first introduced, but the device nonetheless secured a fanatical following). Moreover, unstructured pluralistic testing produces much more data (by millions of testers in the Windows Insider Program, compared to the thousands enrolled in FDA trials). Potentially, to the extent a large number of testers encompasses greater diversity, pluralistic testing better matches products and features with the heterogeneous problems and preferences of users. 

At the same time, the U.S. government, like governments abroad, has significantly expanded its regulatory powers to promote personal and public safety. For instance, the National Highway Traffic Safety Administration’s New Car Assessment Program encourages manufacturers to build safe vehicles. The Federal Aviation Administration has an elaborate process to oversee the design, manufacture, and maintenance of aircraft to ensure that they meet “the highest safety standards.”13 The EPA seeks to control pollution in the air, land, and sea. The Federal Communications Commission screens new computers and other digital devices for “harmful interference” with “police, ambulance, and fire communications” and “air traffic control operations” (Federal Communications Commission, 1996).

In medicine, too, we should expect that an FDA that made safety its primary focus would reduce the incidence of dangerous drugs or devices brought to market.[15] The FDA could devote more money and time to safety. Plus, safety trials that don’t have to be randomized to establish efficacy could enroll more patients.

Meanwhile, scaling back the regulation of efficacy (or making FDA approvals of efficacy, like Good Housekeeping seals, voluntary) promises two important benefits:

  • Sharply reducing the costs of regulatory compliance should foster some of the hectic, frugal innovation that we find in so many other fields.
  • Replacing centrally supervised, randomized trials with more pluralistic evaluations (by medical associations, insurers and other third-party payers, and online consumer communities) should improve the matching of treatments and patients.

Useless treatments might increase with more innovations coming to market. But, as is the case outside medicine, widespread sharing of diverse experiences about actual use might yield more knowledge of what works best and under what circumstances. We could sip a little more of the holy grail of personalized medicine on the cheap simply by allowing more ad hoc user experimentation.

We certainly should not suppress science, disdain biotech and Big Pharma, or replace trained physicians with barefoot, Maoist doctors. But we could be less credulous about imminent research breakthroughs, and offer more scope for nurse practitioners and even completely uncredentialed outsiders to innovate. Placing ever-larger bets on exclusive innovation is a poor remedy for its debilities. Harnessing the enterprise and ingenuity of the many and for the many should be the way ahead.            

Author’s Acknowledgements

I wrote this paper for the 2015 Kauffman Foundation New Entrepreneurial Growth conference. The key section about HIV/AIDS relies heavily on an ongoing project to compile a large collection of case histories about medical innovation. I am very grateful to David Roux, whose most generous gift made the project possible; to Katherine Stebbins McCaffrey, research associate at Harvard Business School who is diligently and thoughtfully writing up the case histories; and to Srikant Datar, my valued collaborator and partner on the project. Srikant also provided valuable comments on this draft. I am solely responsible for any and all factual inaccuracies, questionable inferences and sweeping over-generalizations contained in this essay. 

About the Author
Amar Bhidé is the Thomas Schmidheiny Professor of International Business, member of the Council on Foreign Relations, editor of Capitalism and Society, and a founding member of the Center on Capitalism and Society. He is the author of A Call for Judgment: Sensible Finance for a Dynamic Economy (Oxford, 2010), The Venturesome Economy: How Innovation Sustains Prosperity in a More Connected World, (Princeton, 2008). In addition, he has written numerous articles in the Harvard Business Review, the Wall Street Journal, The New York Times, BusinessWeek, and Forbes. Bhidé was previously the Glaubinger Professor of Business at Columbia University and served on the faculties of Harvard Business School and the University of Chicago’s Graduate School of Business. Bhidé earned a DBA and MBA from Harvard School of Business with High Distinction and a B. Tech from the Indian Institute of Technology.


Amar Bhide, The Demise of U.S. Dynamism Is Vastly Exaggerated – But Not All Is Well, SSRN, (January 26, 2015).

Amar Bhidé, The Origin and Evolution of New Businesses (New York, Oxford University Press, 2000)

Amar Bhidé, The Venturesome Economy: How Innovation Sustains Prosperity in a More Connected World (Princeton, N.J., Princeton University Press, 2008)

William Bynum, The History of Medicine: A Very Short Introduction (Oxford, Oxford University Press, 2008)

Daniel Carpenter, Reputation and Power: Organizational Image and Pharmaceutical Regulation at the FDA (Princeton: Princeton University Press, 2010)

Ken Flieger, FDA Consumer Special Report, FDA Finds New Ways to Speed Treatments to Patients (January 1995) (downloaded July 16, 2015)

Federal Communications Commission, Understanding the FCC Regulations for Computers and Other Digital Devices (OET Bulletin 62, 1996) (downloaded July 17, 2015)

Food and Drug Administration, Guidance for Industry: Providing Clinical Evidence of Effectiveness for Human Drugs and Biological Products (1998) (downloaded July 12, 2015)

Annetine Gelijns, Nathan Rosenberg, Holly V. Dawkins, Sources of Medical Technology: Universities and Industry, Medical Innovation at the Crossroads, v. 5, Institute of Medicine (U.S.). Committee on Technological Innovation in Medicine (Washington, D.C., National Academy Press, 1995), 67.

Robert J. Gordon, “The Demise of U.S. Economic Growth: Restatement, Rebuttal, and Reflections” (NBER Working Paper 19895, February 2014)

Robert J. Gordon, “A New Method of Estimating Potential Real GDP Growth: Implications for the Labor Market and the Debt/GDP Ratio” (NBER Working Paper 20423, August 2014)

Jonah Lehrer, “The Truth Wears Off,” The New Yorker, December 13, 2010, (downloaded July 16, 2015)

Barry J. Marshall, Autobiographical Essay (, 2006) (downloaded July 11, 2015)

Robert Merges, Richard R. Nelson, “On Limiting or Encouraging Rivalry in Technical Progress: The Effect of Patent Scope Decisions,” Journal of Economic Behavior & Organization 25(1):1-24. DOI: 10.1016/0167-2681(94)90083-3 (1994)

Richard R. Nelson, Kristin Buterbaugh, Marcel Perl, Annetine Gelijns, “How Medical Know-How Progresses,” Research Policy, Volume 40, Issue 10 (December 2011) 1339-1344, ISSN 0048-7333,

Neil Osterweil, “Medical Research Spending Doubled Over Past Decade” MedpageToday (September 20, 2005) (downloaded July 10, 2015)

John Pickstone, “Medicine, Society and the State” in Roy Porter (ed.), Cambridge Illustrated History of Medicine (Cambridge, Cambridge University Press, 1996), 304-341.

Nathan Rosenberg, Perspectives on Technology (New York, Cambridge University Press, 1976)

Nathan Rosenberg, Inside the Black Box: Technology and Economics (New York, Cambridge University Press, 1982)

A.M. Reitsma, J.D. Moreno, “Ethical Regulations for Innovative Surgery: The Last Frontier?,” Journal of the American College of Surgeons 194(6), (June 2002) 792-801

Robert Solow, “Building a Science of Economics for the Real World,” prepared statement for House Committee on Science and Technology Subcommittee on Investigations and Oversight, July 20, 2010

Oliver Staley, “Success at Glaxo’s HIV Unit May Mean Having to Call It Quits,” BloombergBusiness (July 9, 2015),

Sherwin Nuland, Doctors: The Biography of Medicine (New York, Alfred A. Knopf, 1988)

Sherwin Nuland, Doctors: The Illustrated History of Medical Pioneers (Black Dog and Leventhal, 2008)

Warren G. Vincenti, What Engineers Know and How They Know It (Johns Hopkins University Press1990)

Robin J. Warren, Autobiographical Essay (, 2006) (downloaded July 11, 2015)

Danny Yadron, “Moxie Marlinspike: The Coder Who Encrypted Your Texts,” Wall Street Journal, July 10, 2015, B1

Appendix A: The Exaggerated Death of Dynamism

Robert Gordon (2014a, 201b) offers a pessimistic assessment of innovation during the past forty years and a discouraging prognosis for what we can expect in the future.

The golden age of innovation drew to a close nearly a century ago, according to Gordon, with the end of the second industrial revolution. The great advances of that revolution sustained continued improvements in productivity growth for another fifty years. But because there have been fewer technological breakthroughs, productivity growth—which Gordon and other economists say is the best possible measure of innovation—has been in a slump since the early 1970s.

I long have argued the opposite (starting with Bhidé, 2000 and more extensively in Bhidé, 2008, chapter XV)—that our innovative capacity improved in the twentieth century. More recently (Bhidé, 2015), I also have used the notion of inclusive innovation to buttress well-known criticisms about the utility of abstracted estimates of productivity. I have argued that the essence of multiplayer innovation—the relentless changing of what, how, and by whom it is produced and priced—is assumed away in models used to estimate productivity.

Consider, for instance, Gordon’s (2014a) dismissal of the consumer surplus that goes uncounted in standard measures of the output that innovations help increase. He writes that “real GDP measures have always missed vast amounts of consumer surplus since the dawn of the first industrial revolution almost three centuries ago,” but provides no evidence for why this vast miss should be constant over time. To the contrary, we should expect that changes in industry concentration, product and capital market fashions (buoyant stock markets encourage pricing for market share), the proportions of tangible goods and intangible services, and the government’s role in providing and purchasing goods and services will change the consumer’s share of the innovation surplus in an irregular, unpredictable way.

Estimates of productivity themselves are as implausible as their underlying assumptions. Robert Solow (2010), who, according to Gordon, was instrumental in establishing total factor productivity (TFP) as the main measure of technological advances, criticizes DGSE models for failing the smell test. The model Solow pioneered to estimate productivity also produces results that would fail many people’s olfactory standards. A model showing that productivity—and, by implication, innovation—was much higher during the Depression than in the Roaring Twenties beggars belief. So, too, do results showing that there has been little improvement in the productivity of the service sector, even after cutthroat competitors have invested trillions of dollars, and have visibly transformed banking, retailing, transportation, and even government-provided services.

The great advances in controlling and treating HIV/AIDS discussed in this paper also underscore the pointlessness of relying on conventional estimates of economic output and productivity growth to measure innovation. Practices that have prevented untold infections are reflected in the national accounts, and thus in productivity estimates, mainly in terms of the value of artifacts—condoms, needles, test kits, and the like—that are used or distributed, or in salaries paid to public health workers. The value of prevention itself, which is not a commodity sold in a market, goes uncounted. The situation with accounting for the value of treatments is a bit better because drugs sold to treat infections—and counted as an economic output—are expensive. Even so, there is a wide, uncounted gap between the value to the patient and the cost of the drugs. And, as drugs became more expensive and effective, we cannot tell whether the increase in their prices—which are not set in a classically competitive market—exceeds their greater value to patients, and thus whether there is any increase in real economic output. Yet who would argue that advances against the disease are anything but significantly, viscerally, real, or that there will be an actual decrease in useful economic output—and a fall in productivity — when the AIDS drugs go off patent and their prices fall?


  1. The thesis of pluralistic, multiplayer innovation is not novel and should not be controversial. Richard Nelson has been emphasizing its importance, and the policy implications thereof, for decades. (See, for instance, Merges and Nelson, 1994). Nathan Rosenberg’s (1976 and 1982) arguments about accretive incremental advances implicitly attribute a pivotal role to pluralistic innovation. Unfortunately, Schumpeter’s stirring rhetoric about great innovators who “found kingdoms” does not. And the “sharp disjunction” Schumpeter posits between “the high level of leadership and creativity involved in the first introduction of a new technique as compared to the mere imitative activity of subsequent adopters” has helped obscure the value of multiplayer innovation.
  2. This section draws heavily on a case history written by Katherine Stebbins McCaffrey for a project on which Srikant Datar and I are collaborating.
  3. The historical material in this section relies mainly on Nuland (2008) and Sherwin (2008). I have taken the liberty, however, of giving their stories and analyses a multiplayer framing.
  4. The restriction, which advanced at different rates in different countries, came relatively late in the U.S. According to Pickstone (1996, p. 305), U.S. medical practitioners in the 1840s did not have to be licensed. They could train, if they chose to, in a “variety of competing medical schools, attached to different brands,” including herbal medicine and homeopathy. Reliance on “credentialed” physicians also varied, with the wealthy and powerful more likely to require formally trained physicians. 
  5. Many discoveries in the nonmedical sciences, such as geology and astronomy, also were made in the seventeenth, eighteenth and nineteenth centuries, by gentlemen of leisure. The inventors’ accomplishments would earn them membership in bodies such as the Royal Society, even though they lacked any university credential.     
  6. The burning of witches also may have unwittingly suppressed the development and use of folk remedies and un-credentialed medical practice (as Elisabeth Barsk pointed out to me).
  7. Kelman did, however, receive a $250,000 grant to develop his breakthrough photoemulsification technology (introduced in 1967) and secured a clinical appointment at New York University.
  8. Carpenter (2010) provides the definitive account of the history of the FDA, its influence on medical innovation, and on the structure of  industries it regulates.
  9. As with the 1938 legislation, the 1962 amendment is widely thought to have been catalyzed by an outcry about drug safety, namely drug defects induced by thalidomide, rather than by a shortfall in efficacy. The only nexus between the expanded role of the FDA in premarket trials and the thalidomide deaths was that they occurred during safety trials whose design the FDA then lacked the authority to regulate. The FDA’s (1998) own account of the 1962 legislation makes no mention of the thalidomide tragedy. Rather, it asserts that “the original impetus for the effectiveness requirement was Congress's growing concern about the misleading and unsupported claims being made by pharmaceutical companies about their drug products coupled with high drug prices.”
  10. Until 2004, a company seeking FDA approval for a therapy based on an herbal extract, for instance, had to identify the single active ingredient that is doing the job—and prove its safety and efficacy. In June 2004, the FDA did, however, issue guidelines that would make it easier to secure approval for botanical drugs that have not been “purified” to a single molecule.
  11. Some people I interviewed for my previous research, whose companies developed medical devices, said they first sought approval in Europe, where regulators were more tolerant of the need for ongoing adjustments.
  12. According to Nelson et al (2011), however, “the cumulative result of a series of more incremental advances in medical knowledge often is a major improvement in ways of treating patients.” This raises the question of how the incremental advances cited by Nelson et al (in treating coronary diseases, cataracts, diabetes, and angioplasty) overcame the regulatory barriers. Did they, for instance, pertain to improvements that are not regulated by the FDA? Did they involve use of “clearance” rather than “approval” rules for medical devices?  
  13. To cite a personal example, my physician told me to take Vitamin D supplements to bring my Vitamin D levels to normal. But because there is no way to know how much additional Vitamin D would do the job, the obvious solution would be to experiment with different amounts and monitor the effects. Unfortunately, there are no home tests that would permit such monitoring, even though Vitamin D deficiency is a common condition. Similarly, I have been unable to find cheap and reliable tests to monitor, and thus help control, cholesterol, blood sugar, and sleep apnea. There is no technical reason why, in this day and age, there should not be such tests, much as there are for pregnancy and HIV infections.
  14. In 2003, for instance, I had a bout of nausea and giddiness. An emergency room doctor performed an “Epley maneuver” on me, which immediately resolved the problem that apparently had been caused by benign paroxysmal positional vertigo. Many years later, when the vertigo reappeared, I looked up a video of the maneuver on YouTube, avoiding another visit to a doctor. An online search also allowed me to figure out why I had periodically suffered from cramps and how to solve the problem—something diligent and competent physicians had been unable to do for more than a decade.  
  15. Maurice Mason, an airline industry veteran (and son and brother of physicians) observes, “Safety had always been used in aviation as the excuse not to deregulate, but deregulation broke the regulatory capture dynamic and, in my view, has significantly helped improve overall safety levels.” 


  1. Interview Transcript, “PBS Newshour” interview, (downloaded July 16, 2015)
  2. Similarly, the state’s interest in controlling plagues or treating wounded soldiers likely reduced the resources available for widespread noninfectious diseases
  3. Congress, which appropriated $28 million for the NIH in 1949, increased that amount more than 500-fold in the next 50 years to $15.6 billion in 1999.  Appropriations doubled again to $30.5 billion in 2009, but have leveled off thereafter
  4. In 2003, NIH grants paid for twenty-eight percent of $94.3 billion spent on biomedical research in the U.S. (Osterweil 2005)
  5. The material in this box and the quotes are from CRI’s “About US,” (downloaded July 26, 2015)
  6. The high costs of satisfying regulatory requirements also arguably encourage developers to piggyback off NIH-funded research. This incentive also tends to narrow the scope of development to areas favored by the NIH, and also favors individuals (particularly NIH grantees) and organizations plugged into NIH research.
  7. Vincenti’s 1990 book, What Engineers Know and How they Know It, provides an excellent analysis and persuasive examples of the differences.
  8. Intel’s description of its tick-tock process.
  9. “What Does It Mean When FDA ‘Clears’ or ‘Approves’ a Medical Device?”, (downloaded  July 23, 2015)
  10. Summarized from an overview, (downloaded July 15, 2015)
  11. FDA press release (February 19, 2015) July 16, 2015)
  12. See (downloaded July 16, 2015)
  13. See (downloaded on July 17, 2015.)

Session Summary


New Entrepreneurial Growth Conference: Prioritizing

There are many potential ways to renew entrepreneurial growth, so how do we go about prioritizing and deciding what are the best immediate and most effective steps? Conference participants discussed this question and raised several interesting points about different barriers to—or facilitators of—entrepreneurship.

Reform the political environment. One participant pointed out that economic stagnancy is closely related to political stagnancy. Many of the resources for entrepreneurship and the decisions that affect the supply of entrepreneurship (from occupational licensing to patents to land use) are determined by politics. These decisions, however, are unduly influenced by wealth and corporate lobbying. As funding is currently such an important part of any election, the wealthy are the gatekeepers who determine who runs for office—and this population does not always share the same priorities as the rest of the country.

One participant suggested that there needs to be an effort to convince the wealthy to think more about the long-term health of their entire communities. Furthermore, we need changes that would enable those who run for office to be more independent of wealth, allow sitting politicians to depend less heavily on lobbyists, and encourage more good people to go into government. Lowering barriers to entry in politics would give us greater political dynamism and make it easier for policy entrepreneurs to voice their messages. Reducing congressional leadership power and increasing committee power could also increase political dynamism, which will yield greater economic dynamism.

Expand the scale and scope of the safety net. Some participants argued that a stronger safety net is important for entrepreneurial activity, as well as for combating inequality. They contended that we need a system that provides universal basic income, gives people a level of stability, and is independent from institutional employment. Health care insurance, in particular, needs to be decoupled from full-time work, and there need to be opportunities for coverage for those outside of the Medicaid window. The biggest constraint that entrepreneurs face, participants said, is access to insurance and a safety net. Entrepreneurship requires a leap, and we need to offer people a net if we want them to jump. If we reduce the risk inherent in starting firms, more people will take on the challenge.

In addition to giving new entrepreneurs the safety net they need for their own families, this effort would give new firms greater access to talented employees. Potential employees, especially those with children, are reluctant to leave stable corporations to go to new, small firms if they need health care and other benefits. As a result, new firms can’t compete for talent with larger firms that have these structures in place. Private companies are effectively administering the welfare state. It would be more efficient and economical to have a public safety net that could take advantage of economies of scale for benefits and then pay lower wages to employees.

Several other participants agreed that a stronger, more efficient safety net is important for a wide range of reasons, including mitigating inequality and increasing the mobility across firms that benefits entrepreneurs. They questioned, however, its role in increasing entrepreneurship, arguing that people who start new businesses would not be influenced by a stronger safety net. A very large proportion of Inc. 500 entrepreneurs, one contributor explained, have neither equity capital nor debt capital at the start. He suggested that there is a natural selection process that produces only a small group of people who will take the plunge with no debt, no equity, and no safety net. It is this process that yields the high-quality entrepreneur. One contributor pointed out that there is little entrepreneurship in Sweden, where the safety net is strong, although another participant disagreed, noting that the expansion of the safety net in France has resulted in more entrepreneurship. In the United States, a contributor suggested, the creation of S-CHIP in the 1990s may have played a role in increasing entrepreneurship for those who became eligible for the new health care benefit.

Improve access to affordable legal services. A discussion regarding increasing legal complexity during the first session was revisited here. The complexity of legal issues, participants said, are not necessarily proportionate to the size of the deal, with the result that smaller entrepreneurs’ legal costs are often as high as those of larger companies. These high legal costs serve as a barrier to entrepreneurship, participants said, and disparity in access to high-quality legal services affects the probability of new companies’ success. Some participants believe that there may be even greater inequality in access to these services in the future, as law becomes a more concentrated industry. Participants suggested the following potential solutions to this problem:

  • Engage law schools’ clinics for pro bono services. One participant emphasized that law schools can be productive collaborators in encouraging entrepreneurship, offering pro bono legal services through their clinics to entrepreneurs who cannot afford to pay. These services may benefit more entrepreneurs starting main street businesses than those developing high-growth firms.
  • Change professional licensing restrictions in law and medicine. Allowing legal technicians to do some litigation work would make costs more affordable.
  • Simplify navigation of the law. Rather than simplifying the law, which may be impossible, one participant suggested that we develop new platforms that deliver legal services in ways that make lawyers more cost-effective. Technology, they said, can be used to make the law more accessible and reduce costs somewhat.

Develop legal best practices to share between cities. One participant noted that there would be a benefit to having a nationalized set of best practices in local regulations. For example, cities could start with a consumer protection law that already is vetted by legal experts and tailor it for their specific community so that every city isn’t starting from scratch. Law schools could play an important role in developing these templates for balanced regulation, as they lack the bias that taints developers or other interested parties. While all cities are different and there inevitably will be some variation in these policies, the templates could be adjusted for specific locations.