Reports out this morning from the official monitoring the Troubled Asset Relied Program for the U.S. Congress highlight the difficulty of tracking the success of this program because of a lack of data.  These reports reminded me of a recent presentation I sat through from Statistics Canada discussing two of their programs for monitoring lending to small businesses in Canada.  A bit more background on the two projects as I understand them.

• Survey on Financing of Small and Medium Enterprises.  Completed every three years, with most recent data available from 2007, this survey gets detailed business financing data from establishments.  You can read much more on the details of sampling, etc.
• Survey of Suppliers of Business Financing.  Completed on roughly the same cycle but gathering data from banks instead of businesses, this survey gets aggregate information on bank portfolios.  Read more on the details.

What I found most interesting about hearing Statistics Canada talk about both collection efforts back-to-back was the informal comments that were given about the merits of collecting data from banks rather than businesses.  As I remember, the comments basically alluded to the difficult time that Statistics Canada has of getting banks to provide data for their survey, even though the survey is mandatory under Canadian statutes.  On the survey of businesses, they have a much easier time of getting the data, relatively, and have felt they are able to do more detailed manipulations with the data, providing richer outputs, and helping that program to have more stable funding support. 

What I read into these stories is that banks hold a great deal of power in deciding who gets what information, even when the world is watching closely or statutes compel them to participate.