Who Cares About Federal Economic Statistics?

Editor’s note: This article originally appeared on October 30, 2009 in Dismal Scientist at Moody’s Economy.com under the title: Who Cares About Federal Economic Statistics?

Two years ago, the state of the federal government’s system for gathering and maintaining economic statistics was like The Dismal Scientist—dismal.

In response to former President George Bush’s threat to veto what he viewed as excessive federal spending, Congress cut the budgets of a number of agencies, including three whose responsibilities include gathering the nation’s vital economic statistics: the Census Bureau, the Bureau of Labor Statistics, and the Bureau of Economic Analysis.

These agencies didn’t yield particularly large savings. Congress snipped some $35 million from the BLS and the BEA and delayed about $60 million for the Census Bureau, all to help cover a $22 billion gap in a $3 trillion appropriations bill.

Statistical values

Congress could order these cuts because most of its members don’t fully comprehend the value that the $1 billion federal economic statistical system provides to the nation. Data collected by government agencies guide fiscal and monetary policy in a $14 trillion economy and inform the investment decisions of 6 million U.S. firms. Without timely and accurate statistics, thousands of private and public agencies would be unable to successfully carry out efforts in transportation, education, healthcare, affordable housing, and economic, community and workforce development.

Congress’ actions pushed the federal statistical system, which had not been faring well for some time, into dire straits. Because of the budget cuts, in 2008, the Census Bureau couldn’t add new quarterly and annual data it planned to collect on the nation’s finance, insurance and real estate industries. As a result, the BEA now lacks optimal data for measuring economic activity in those areas that were at the heart of the recent recession.

The cuts also left the BLS unable to update its consumer price index survey with information from the 2000 census. The CPI’s housing sample thus continues to be based on data from the 1990 census, which has created a visible, growing age bias in the sample.

There were other consequences of Congress’ budget cuts as well.

• The BEA said it would stop publishing industry subsector estimates for metro and county GDP and earnings. Researchers therefore would be unable, for instance, to track Detroit’s dependence on auto manufacturing.
• The BEA also halted the collection and publication of detailed state foreign direct investment data, despite protests from state economic development agencies, which depend on that information to help attract overseas firms.
• The Census Bureau cut back and postponed the 2008 “dress rehearsal” for the 2010 census, leaving it unable to test its procedures and technologies for the upcoming decennial count.
• The BLS killed current employment statistics for 65 small metro areas.
• The Department of Housing and Urban Development discontinued the residential finance survey, which gathered data on mortgage debt, and had dramatically cut back the number of metro areas covered by the American Housing Survey, which examines local housing conditions and markets.

With only a $2 million appropriation, the Census Bureau’s innovative Local Employment Dynamics program barely dodged elimination and was relegated to pilot status. Starved for funds, the program could not realize its potential to use existing administrative records to track worker movement over time and space—knowledge that could transform how we understand the workings of our national and regional economies.

Properly funded, LED could show patterns of labor turnover by worker characteristics such as age, sex, race, industry and occupation, down to the community level. It could trace how workers fare after layoffs, shedding light on such questions as the fate of construction workers after housing markets collapsed. And the program could provide a demographically rich picture of where people live in relation to where they work, a valuable tool for business site location and transportation planning.

Turnabout

One year and a new administration later, the situation is markedly different.

In the budget for fiscal 2009, Congress gave the Census Bureau the $8 million it needed to collect data from finance, insurance and real estate firms quarterly and annually, rather than once every five years. The BLS also received sufficient funds to update the CPI housing sample and restore current employment statistics for the 65 small metro areas cut under the last Bush budget.

The American Recovery and Reinvestment Act gave the Census Bureau $1 billion to improve operations and beef up its “get out the count” efforts, to ensure a more accurate 2010 census. The Census Bureau also initiated a new Business Dynamics Statistics program that allows researchers to examine longitudinal trends in business development.

For fiscal 2010, the House and Senate are on the verge of agreeing to fund several important economic data improvements. Assuming budget passage, these positive developments are in store:

• The BEA will improve its estimates of service industry contributions to GDP using the newly expanded Census Bureau services industry data.
• The BEA will restore detail to regional NAICS for GDP and earnings and also create a new interarea price index to produce real measures of state and local area product and income.
• The Census Bureau’s LED program will become permanent, with increased funding for significant enhancements. Among these would be a “job-to-job flows tool” allowing analysts to see how different categories of workers move among industries and regions over time.
• The Department of Housing and Urban Development will receive a substantial increase in funding for improved housing statistics. In particular, HUD will expand the American Housing Survey and begin capturing residential finance data in a more timely way.
• If the House agrees to Senate funding levels, the BEA would reverse cuts to the state FDI data program and the Economic Development Administration would create an industry clusters research and information center, including detailed cluster mapping.

Apart from this year’s appropriations process, agencies are pursuing several efforts to improve the statistical system. The BEA is exploring the creation of measures of economic well-being and sustainability. The Census Bureau has proposed collecting data needed to create a modern poverty measure as well as a new program to improve its longitudinal business datasets. The Census Bureau also is looking at ways to expand the sample size of the American Community Survey (to increase its reliability) and to improve its annual population estimates program. In addition, an administration-supported effort is under way to improve the accuracy, comparability and usefulness of economic statistics through a series of “data synchronization” activities across statistical agencies.

Looking ahead, however, the political and budget winds blow in different directions. The good news is that senior Obama administration officials understand the importance of reliable and extensive economic statistics. OMB Director Peter Orszag has written about using statistics to drive sound policy, for example.

Time to speak up

The bad news is that for the foreseeable future, Washington will be under enormous pressure to reduce spending. It is quite possible that the administration’s understanding of the need for good statistics will not be reflected in the president’s upcoming budget proposals—unless data users weigh in. And if the past is a guide, Congress may well make cuts because it doesn’t naturally comprehend the nearly infinite economic return on a relatively small taxpayer investment in data collection and product development.

To a great extent, then, the future health of the federal statistical system is in the hands of data users. Historically, consumers of government statistics have had little direct contact with federal decision makers. Congress and the bureaucracy take data users’ concerns seriously when they hear them, but they rarely do. If the federal statistics system is to sustain recent improvements and survive upcoming budget battles, data users need to make their case. Silence in this instance is the opposite of golden.