Institute for Research on Innovation and Science

Recently, the American Association for the Advancement of Science (AAAS), an international non-profit organization dedicated to advancing science for the benefit of all people and also the publishers of Science magazine, completed an interview with Kauffman Foundation Research Director E.J. Reedy and Jason Owens-Smith, Professor of Sociology and Organizational Studies at the University of Michigan and Executive Director of the new Institute for Research on Innovation and Science (IRIS). The Kauffman Foundation is a founding donor of IRIS, along with the Alfred P. Sloan Foundation.

As a complement to this discussion in AAAS, E.J. Reedy reached out to Ron Jarmin, Assistant Director of Research and Methodology at the U.S. Census Bureau, which is an important partner in IRIS, to learn more about how the Census is continuing to create new innovative data resources. This post is the first of a multi-part series from that interview, with future updates released on Kauffman’s Growthology blog. You can read part two of the interview here.

Part 1: Institute for Research on Innovation and Science and Modernizing Data Collection and Release

Q (Reedy):  How would you describe the directorate that you oversee at Census?

A (Jarmin): I oversee approximately 200 folks who are in charge of a central component of research on statistical methods, survey methodology, and applied research using Census Bureau data. This also includes the network of federal statistical research data centers around the country.

Q:            And is it in that context that you became involved with the Institute for Research on Innovation and Science (IRIS) initiative?

A:             Probably a little bit of that, but also my professional career as an economist, so some of the IRIS researchers have been colleagues of mine over the years. But there’s a dual interest of mine in this. First, as an applied economist, I’m interested in the process of technological change and entrepreneurship in general. Second, I am trying to lead the Census Bureau to modernize our methods. I think this project can serve as a prototype of what we would like to do in other sectors of the economy going forward.

Q:            So, in terms of modernizing methods, what is it that you are attempting to modernize here?

A:             If you look at what our users say, there are two things that people want more of from the Census Bureau and the other statistical agencies. They want more timely data, and they want data that’s more detailed, whether that be detailed geographically, or detailed for certain economic sectors or demographic groups. Basically, they want data that’s more timely and richer in its detail. Using surveys as the primary mode of data collection limits the ability to be both timely and detailed because you’re obviously not going to be able to produce detailed data and maintain confidentiality if you’re surveying a relatively small part of the population of interest. And then, surveys by definition have a certain lag structure applied to them. So our new Innovation Measurement Initiative, which is how we refer to the Census Bureau portion of the IRIS work, is an attempt to use more administrative records from various organizations, in this case universities, to both speed up the production function of statistical information and to increase the richness of the underlying data we employ to construct our estimates. So we’re getting data closer to real time, and then turning it around quickly on our end so that we can produce statistics for users that are much closer to the reference period that they purport to describe.

Q:            How do you bring together all this data cohesively?

A:             We are linking this administrative data from the universities to data assets that we already have, whether they be administrative or survey-based, and then creating value-added products from that. So, not all of the information has the same timing characteristics, and that’s one of the challenges of this type of exercise. You’re mashing up data from different sources with different timing characteristics, and trying to reconcile those things to produce new data products that make sense to folks and help them understand the world and make better decisions. So that is one of the challenges that we are currently dealing with. I think we’ve been doing things that, in a general level, haven’t really pushed the boundaries of reconciling the different sources of data in a very detailed way. But that’s something we’re going to have to get better at as we go forward.

Q:            What is your favorite insight to date that has come out of the existing prototype research from IRIS?

A:             We’ve been able to describe the production function of scientific research that goes on at universities in a rich way both within the university and then what its downstream impacts on the broader economy look like later on. So that’s done by following firms that are associated with these projects, either as contractors or as vendors. But also by following the workers, the various scientists and students and faculty members that then go on to either take jobs in industry or move to other places, perhaps maybe even starting new firms. We’ve been able to start describing that. I think we’re still at a very nascent stage of this type of work. But I think it shows a lot of promise. A key lynchpin in all this is at the Bureau we do have universe information on firms and workers, so we know that we can usually find these types of entities. When combined with other sources, like universities in this case, whether it be the vendor data or the contractor data, we can find those firms and workers in the universe files here at the Census Bureau.

Q:            When you are looking around at other international or government organizations that are really being leaders in this type of process, who are you looking at or what are some other examples you see that are quite interesting to keep track of?

A:             So, obviously, as a statistical agency, we are often envious of the data infrastructure that some of the Scandinavian countries have or, say, the Dutch have. You know, we’re probably not likely to be able to replicate that kind of infrastructure here in the United States anytime soon. But I think by being able to link things from different sources, we can at least replicate some of those capabilities in our ability to match various data sets. For example, in this case, matching the university data to the Census data. In some ways, we’re trying to follow the lead of some of these other organizations, but I think in other ways we might be treading new ground with the work in the IRIS-IMI project that I don’t think anyone has quite tried to do—the rich linkage from a particular sector of the economy to the statistical infrastructure with this much broad scope of information coming from an organization like a university.

Q:            This is really an example of something that’s almost a public/private partnership in many ways. Is that an accurate kind of description?

A:             Yes, plus many of the universities that we’ve dealt with are public universities, so maybe a public/public/private. But it’s still a partnership between a statistical agency and other segments of the economy to try to produce information that is useful both for the universities, for their purposes, but also useful for more general statistical and economic measurement activities. This is information that the Census Bureau, other organizations, and the public would be interested in. 

You can read part two of the interview here.


comments powered by Disqus
e.j. reedy data maven

E. J. Reedy

As a director in Research and Policy, E.J. Reedy oversees the Ewing Marion Kauffman Foundation’s research initiatives related to education, human capital development, and data.

Since joining the Kauffman Foundation in 2003, Reedy has been significantly involved in the coordination of the Foundation’s entrepreneurship and innovation data-related initiatives, including the Kauffman Firm Survey, for which he served as a principal investigator, and the Foundation’s multi-year series of symposiums on data, as well as many web-related projects and initiatives. He is a globally recognized expert in entrepreneurship and innovation measurement and has consulted for a variety of agencies.