THE AVENUE APRIL 19, 2010
In Washington, it’s the season for many things—spring flowers, baseball, political speech (always in season), and House and Senate appropriations subcommittees delving into the minutiae of the president’s proposed $3.7 trillion budget for FY2011.
Scattered among the nooks and crannies of this massive document are the plans for the multiple agencies in the nation’s decentralized statistical system. And within these plans are disparate items that suggest that, slowly, the federal government is taking steps to improve our ability to grasp the demographic, economic, and social dimensions of the nation’s regions and communities.
The remarkable thing about our statistical system is that, when you realize it’s tracking about 310 million people and a $14 trillion economy, it’s really inexpensive, a few billion a year, on average. With the exception of the 2010 Census process, the proposed improvements that I’m about to describe are ridiculously cheap, well under the cost of one F-35 fighter jet.
In FY2011, the biggest statistical deal will be the Census Bureau’s providing the nation with a string of the tallies of population for use in apportionment, legislative redistricting, and other public and private functions. The Census Bureau also will be evaluating the accuracy of the 2010 Census, responding to state and local government concerns about the count, and, yes, planning for the 2020 Census. Cost: $477 million (or three F-35s).
The Census Bureau also proposes to add two important new wrinkles to its data offerings. First, it wants to boost the annual sample size of the American Community Survey, the detailed assessment of population characteristics, from 2.9 million to 3.5 million housing units. While the nation has been growing, the ACS sample size hasn’t. If the sample size isn’t increased, the reliability of the ACS neighborhood estimates, out for the first time this fall, will deteriorate. Strong ACS reliability is crucial for the smart investment of trillions of public and private funds in cities. Cost: $44.2 million.
The Census Bureau also proposes to develop a new Supplemental Poverty Measure.
The current official poverty measure is based on 1950s household economics, excludes federal assistance, and doesn’t account for geographic cost-of-living differences. While the current measure will still be used to determine federal program eligibility, the new measure would give a realistic sense of poverty in various communities, based on what people spend for housing, childcare, and other essentials and reflecting the cost differences between, say, Topeka and Los Angeles. Cost: $5 million, plus $2.5 million for work by the Bureau of Labor Statistics to improve data on household consumption.
The Bureau of Labor Statistics and the Bureau of Economic Analysis are proposing to add new economic data, which, while not as flashy as the ACS or poverty measures, will be very useful for regional analysts.
- BLS wants to boost the sample size for its Occupational Employment Statistics so it can replace a three-year average with a one-year estimate. This would allow analysts to build a time series of changes in occupational structure, important for workforce development. Cost: $5 million.
- BEA proposes to use a new state cost-of-living index to produce real state personal consumption expenditure data. While the well-known BLS Consumer Price Index measures change over time, the BEA index would measure differences over space, highly useful for all sorts of regional analyses. In light of BEA’s research to date, our hope is that a metro cost-of-living index is not far behind. Cost: some sliver of a $3.9 million initiative for better household statistics.
- BEA also plans to restore detailed state-level Foreign Direct Investment Statistics it terminated because of an FY2008 budget cut. These FDI data are crucial to state efforts to attract foreign business. Cost: some portion of $3.3 million for improved FDI data.
In addition to these various statistical goodies, the White House is proposing a series of regional data improvements in specific policy domains:
- The Department of Housing and Urban Development wants to restore the American Housing Survey to something closer to its former glory, boosting the number of metros surveyed every two years from seven to 30. Cost: $2.9 million.
- The National Center for Education Statistics also is looking for funds to provide technical assistance to states for data systems that track students through the educational system and into the workforce. Cost: $6.7 million.
- To complement this NCES effort, the Employment and Training Administration is asking for increased funding of its Workforce Data Quality Initiative, to help states integrate education and workforce data. Cost: $1.3 million.
- The National Center for Health Statistics wants to increase the sample size of the National Health Interview Survey to allow for state and community estimates of health insurance coverage for the largest states and metropolitan areas. Cost: $8 million.
- The Bureau of Justice Statistics wants to redesign the National Crime Victimization Survey, which would include research on the development of sub-national crime estimates. Cost: some portion of $15 million.
- The Energy Information Administration is asking for funds to increase the number of states covered by the Commercial Buildings Energy Consumption Survey, data useful in sustainable development initiatives. Cost: some portion of $15.8 million.
While it’s sometimes difficult to see, the nation and its regions run on statistics—they’re how we frame problems and solutions. OMB Director Peter Orszag understands this, which in large part is why we are seeing a flowering of low-cost, high-impact data initiatives as well as gardens and trees this spring in Washington. Let’s hope that Congress allows these efforts to bear fruit.