In Dire Straits: The Urgent Need to Improve Economic Statistics

Andrew Reamer
Andrew Reamer Former Brookings Expert

March 4, 2009

America’s Challenge

Current, accurate, detailed federal economic statistics are essential to the proper functioning and growth of the U.S. economy. Federal, state, and local government policy-makers and millions of businesses rely on economic statistics to assess conditions, guide investments, and evaluate results. Insufficient, incorrect, or outdated data can result in mistakes costly to the economic vitality and competitiveness of the nation and its regions, cities, and neighborhoods. The availability of good economic data is particularly important in this time of economic recession and uncertainty.

Unfortunately, the economic statistical system is visibly deteriorating before our eyes. Existing economic data series have been eliminated, reduced in detail and accuracy, unadjusted to account for new benchmark data and new methods, or delayed. Consider these examples, particularly in light of current economic conditions:

  • The Bureau of Labor Statistics (BLS) Consumer Price Index (CPI) still relies on the 1990 Census to construct its geographic area sample and housing sample.
  • The Bureau of Economic Analysis (BEA) will no longer publish metro and county estimates of GDP and earnings by detailed industry. So, for instance, the Detroit metro area will lose the ability to track auto manufacturing’s contribution to the region’s economy.
  • BLS no longer collects jobs data for 65 smaller metropolitan areas.
  • The Department of Housing and Urban Development (HUD) is discontinuing the Residential Finance Survey, which gathers data on mortgage debt, and has dramatically cut back the American Housing Survey, which examines current housing conditions and markets.
  • The Federal Reserve Board of Governors eliminated its Survey of Small Business Finances (SSBF).
  • The most recent detailed research and development data from the National Science Foundation (NSF) is for 2004.
  • The Federal Highway Administration’s National Household Travel Survey (NHTS), which guides hundreds of billions in government transportation spending, was delayed for two years and no longer covers long-distance travel.

Moreover, our economic statistical system remains oriented to the mid-20th century manufacturing-based, oligopoly-dominated economy. It does not fully capture the major structural changes of the past several decades, including the enormous growth of services industries; large-scale technological innovation; geographic mobility, increased trade, and global competition; the rise of entrepreneurship; ongoing corporate restructuring (e.g., mergers, acquisitions, spin-offs, outsourcing, failures); and the extensive transformation of regional economies.

In addition, the statistical system is not taking full advantage of remarkable advances in information technology and statistical methods (e.g., synthetic data). These advances offer agencies once undreamed-of opportunities to create data series that reveal the workings of the nation’s economy in breadth, depth, and detail, e.g., tracking business openings and closings, worker hires and fires, and worker flows from job to job by industry and location.

Limitations of Existing Federal Policy
The primary reason for the deteriorating economic statistical system is inadequate funding. A secondary reason is the absence of a strong set of relationships between statistical agencies and data users other than macroeconomic policy-makers.

At an annual cost of less than $1.3 billion to guide the workings of a $14 trillion economy and the geographic distribution of over $500 billion in federal funds, the economic statistical system is one of the federal government’s most cost-effective activities. Essentially, the cost of the system is extraordinarily low and the return on investment is almost infinite. Despite this, federal economic statistical agencies have been subject to significant budget shortfalls, with appropriations flat or declining for almost every one.

As a result, the nation has achieved trivial savings in exchange for inadequate data at a time when good data are sorely needed to guide economic recovery (and in the context of unplanned appropriations of over $1 trillion for recovery purposes). Consider the following:

  • BLS has not updated the CPI using the 2000 Census for want of $10 million;
  • BLS must “borrow” $10 million from the Census Bureau to stave off dramatically cutting the sample size of the Current Population Survey, the primary source for employment and unemployment figures;
  • the Census Bureau lacks the $8 million needed to provide full coverage of the nation’s service industries;
  • BEA is cutting metro and county GDP and earnings detail by industry to save about $1 million;
  • the Federal Reserve saves only $8 million by eliminating the SSBF;
  • the NHTS national survey could not have been conducted had AARP not contributed $500,000 to cover the budget shortfall; and
  • the Census Bureau’s Local Employment Dynamics program, which has the potential to transform our understanding of how the economy works, has limped along for years on a $2 million appropriation when it needs $12 million to achieve its potential.

A number of factors contribute to this troubling situation:

  • Budget decision-makers in parent departments, OMB, and congressional appropriations committees do not recognize the enormous return to the nation on small taxpayer investments in statistics.
  • In the decentralized statistical system, statistical activities are relatively small parts of multi-billion dollar budgets and so easily overlooked. No budget decision-maker has responsibility for the statistical system as a whole.
  • Macroeconomic policymakers have not made it a priority to advocate for adequate funding to improve traditional economic statistics.
  • The absence of a coherent set of federal policies to improve economic structure has resulted in a lack of articulate demand for data on structural elements (i.e., industry capacity and competitiveness, regional economic activity, and the building blocks of innovation, entrepreneurship, workforce development, physical and financial infrastructure, and community stability).
  • Advocates for adequate statistical agency funding are not well informed or organized, largely because of the complexity and opaqueness of the statistical system and the appropriations process.
  • No congressional committee exercises oversight of the economic statistical system as a whole.
  • The OMB office charged with coordinating the entire federal statistical system only has five staff.

The secondary reason for the gap between data supply and need is the lack of strong relationships between the producers and the users of economic statistics. Historically, the three major economic statistical agencies—BLS, BEA, and Census Bureau’s Economic Programs Directorate—have considered their primary customers to be the macroeconomic policy-makers; needs of other data users have been deemed less important. Consequently, agencies have lacked motivation to learn about the needs and priorities of these other data users and how agencies might better serve them. As a result, product development for non-macroeconomic policy uses has been intermittent and ad hoc; existing data for such uses are the first to be cut in times of budget constraint; and data users are not adequately informed to play an effective advocacy role.

A New Federal Approach
To improve the nation’s economic statistical system, the federal government should take the following steps:

  • President Obama should direct the National Economic Council (NEC) to develop and oversee implementation of a plan to upgrade the economic statistical system to support more effective national economic policy and more intelligent public and private sector decision-making.
  • The president’s fiscal 2010 budget request should include sufficient funding to upgrade outmoded information technology and restore deteriorated data series.
  • To inform the fiscal year 2011 budget request, the NEC should commission the National Academies of Science to conduct a six-month reconnaissance of the capacity of the economic statistical system to meet public and private sector data user needs and make priority recommendations for additional improvements.
  • Within its Budget Review section, OMB should create a working group of budget analysts responsible for the various economic statistical agencies to ensure that the president’s budget request meets needs as defined by the NEC.
  • The capacity of the OMB Office of Statistical and Science Policy should be expanded.
  • The White House Office of Legislative Affairs should communicate to the congressional appropriations committees that passage of the president’s budget request for statistical agencies is a priority.
  • To inform product development efforts and resource allocation decisions, each federal economic statistical agency should take steps to enhance relationships and communications with key groups of data users.
  • The Joint Economic Committee should create an economic statistics subcommittee to provide oversight to the economic statistical system.

Throughout his campaign, President Obama called for “smarter government,” one that allocates trillions of dollars on the basis on good information about need and prospective impacts. Quickly restoring the health of the nation’s economic statistical system is the necessary first step in fulfilling the president’s vision.