Data for an Evolving Economic and Financial System

Karen Dynan
Karen Dynan
Karen Dynan Professor of the Practice of Economics - Harvard University, Nonresident Senior Fellow - Peterson Institute for International Economics

February 20, 2010

Over the past several years, the U.S. economy has experienced a devastating financial crisis and the most severe recession since the Great Depression. Specific developments in credit markets and the macro-economy are often cited as precipitating the crisis, but important longer term changes in the economic and financial system laid the groundwork for these developments. The morning session of the recent conference “Measuring Innovation and Change During Turbulent Economic Times,” jointly sponsored by the Brookings Institution and The Heritage Foundation, focused on the implications of an evolving economic and financial system for our nation’s data needs. Policymakers, staff from statistical agencies, and data users from the public and private sectors discussed changes to our data infrastructure that would enhance our ability to anticipate and respond to today’s crises and cyclical downturns, as well as to create an environment that fosters long-term economic growth.

Perspectives from policymakers

The morning featured remarks by Rebecca Blank, Undersecretary for Economic Affairs at the Commerce Department, and Alan Krueger, Assistant Secretary for Economic Policy at the Treasury Department. Blank laid out an agenda for federal statistical agencies that included improving information about innovation as well as exploring how we might capture alternative aspects of well-being such as environmental and health concerns. Assistant Secretary Krueger focused primarily on the inadequacies of our current measurement practices revealed by recent events, characterizing the crisis as providing a “data stress test.”

Krueger noted the decline in the relevance of some data traditionally used by policymakers because of changes in the economy and (especially) the financial sector. For example, most of our high-frequency measures of credit extension are based on outstanding bank loans. However, in a world with large securitization markets and where up to half of all loans are made by nonbank financial institutions, outstanding bank loans are a misleading gauge of the strength of credit markets.

He also emphasized how economic policymaking is hindered by the low frequency and long publication lags associated with key household finance and spending data. The most complete source of aggregate information on household financial flows, the U.S. flow of funds accounts, is published only quarterly and with a lag of more than 2 months. The problem is even worse for household-level data, which are critical to monitoring and understanding the effects of asset price fluctuations, job loss, and counter-cyclical policy measures. Comprehensive household-level data on consumer expenditures for 2009 are not likely to be available until late 2010. The Survey of Consumer Finances, the best source of information on Americans’ balance sheets, is released just once every 3 years, with a publication lag of more than a year.

Krueger was optimistic about the potential for improving data. He pointed out that some important gaps in financial data may be filled when financial reform is enacted. For example, the bill that passed the House last year called for a Financial Services Oversight Council that would issue semiannual reports describing the state of financial markets and detailing the size, scope, scale, concentration, activities, and inter-connectedness of the 50 largest financial institutions in the United States. Other provisions of the bill would facilitate the collection of information on financial derivative products and hedge fund activities. Among other data reforms, the federal government could build “‘rapid response’ data gathering capacity that could be tailored to answer specific, one-shot questions” to address the lack of timely data for real-time policy analysis. In addition, the federal government could make private data more useful by setting standards that would certify private series as being of sufficiently high quality to be used for policy analysis.