hen the Republicans took over Congress in 1994, federal statistical and research activities seemed headed for trouble.
Freshmen in the House complained that Washington-sponsored research duplicated what was already available from private sources. The prevailing view in the GOP seemed to agree with the famous Disraeli maxim: "There are three kinds of lies: lies, damned lies, and statistics."
At the same time, however, official Washington was just digesting a new law, the Government Performance and Results Act. One of its key goals was to "improve congressional decision-making by providing more objective information on achieving statutory objectives, and on the relative effectiveness and efficiency of federal programs and spending." The new law suggested the need for more collection of information by the agencies, not less.
That leaves many agencies in a quandary. Pushed to do more with less, their managers need to know more about the workings and impacts of their operations. At the same time, budgets are tight, and research, particularly seemingly esoteric policy research on boring topics, is often seen as an easy mark for cutting. Cuts, after all, would play to the stereotypical Republican views on the subject.
Because of these conflicting currents, the movement to slash federal statistical spending has failed to gain steam. A few small programs have been sharply cut, but there's been no rush to downsize, and many research programs survive and flourish despite the climate of austerity, as agencies and congressional allies have found the keys to saving them.
Faces of Research
Policy research is a bit like pornography: you know it when you see it, but it is difficult to define. If one thinks of the Results Act as asking the question, "How are federal programs and policies affecting or likely to affect outcomes in the larger society?" policy research can be thought of as the means agencies use to seek answers.
The federal statistical agencies are the best-known providers of such answers. And chief among them are the Census Bureau and the Bureau of Labor Statistics. Their work-tracking trends in wealth and poverty, employment and unemployment, inflation or deflation, race and ethnicity and more-influences both the operation of existing programs and the formulation of new ones.
Between them, these two agencies spent $875 million in fiscal 1997, about a third of the $2.75 billion spent governmentwide on statistical research. (The Office Management and Budget has identified 70 federal agencies which spend at least $500,000 a year on statistical research.)
These statistics-gathering activities are the tip of an iceberg that includes a host of policy research activities scattered throughout government, often deeply buried within particular agencies.
Although not always easy to characterize, such activities are usually focused on moving beyond the numbers and into interpretation. They are aimed at development of ideas regarding the effectiveness of existing programs, the potential of new programs and the understanding of real-world phenomena and ways in which those phenomena might be changed through government policies.
The scope of policy research is so big as to defy definition. Yet many of the research programs are at the center of both federal managers' responsibilities and the daily lives of U.S. citizens. They include, for example:
- The Agency for Health Care Policy and Research in the Public Health Service, which funds research into health outcomes and rural health, conducts the National Medical Expenditure Survey, provides technical assistance to states involved in health care reform and creates practice guidelines for use by various health practitioners.
- A joint effort by the Social Security Administration and the Office of the Assistant Secretary for Planning and Evaluation at the Department of Health and Human Services to examine the causes of the growth in adult disability caseloads.
- The Compliance Research and Planning program of the IRS, which is designed to examine the reasons and remedies for tax evasion, which costs the U.S. Treasury more than $100 billion a year.
- The Advanced Measurement Initiative of the Environmental Protection Agency, which seeks advanced monitoring tools for enforcement of environmental regulations.
- The National Advanced Driving Simulator program of the Transportation Department, which helps predict driver behavior to improve safety policies.
- The Intelligent Transportation Systems, also within Transportation, which examines uses of technology to reduce traffic congestion in major urban areas.
The federal budget includes no category for such research, but a tabulation of selected items with budget classifications indicating such activities (and reported by the General Accounting Office) suggest that it cost roughly $4.8 billion in fiscal 1996.
When the Republicans took over Congress in 1994, there were early indications that policy research and data collection might be in trouble. Budget resolutions suggested cuts, and some came true. The Congressional Office of Technology Assessment, with an annual budget of $19 million, bit the dust. The Bureau of Mines, whose budget of almost $60 million funded a lot of research, also was abolished, as were smaller agencies including the Advisory Commission on Intergovernmental Relations and the Administrative Conference of the United States.
But the budget-cutting movement failed to gain momentum. The outlook for policy research now is not as grim as three years ago. Nor are the partisan lines so brightly drawn. And several patterns on funding of federal research have emerged.
The political football factor. When research and data collection efforts become political issues, nonpartisan agencies are in a position to intercept a pass and score a game-winning touchdown. That's because government data shops are seen as reliable and apolitical and thus to be trusted with sensitive jobs.
A case in point this year was the politically charged move to revise the Consumer Price Index. The bipartisan commission appointed to look at the matter argued that the CPI was overstated and that large budget savings in federal benefit programs could result from a change. But neither the commission nor Congress had the expertise to make any revisions in methodology, and so Congress in the end gave more money to the BLS to improve its own CPI measures.
On the other hand, agencies with perceived political agendas can find their research efforts quashed. Thus, the IRS compliance initiative has been stalled by members of Congress concerned with keeping the agency's investigative powers at bay. The agency has alienated a few taxpayers by subjecting them to intensive audits designed to collect information for taxpayer profiles, and they complained to Congress.
Even more controversial was the Interior Department's U.S. Biological Survey. Interior Secretary Bruce Babbitt billed the survey as an overdue attempt to collect essential information on species diversity. But the effort became seen as a vehicle for overzealous enforcement of the Endangered Species Act at the expense of private property rights. So the $200 million-plus research program was radically cut and banished from Babbitt's office to a new bureaucratic home at the U.S. Geological Survey.
Environmental research is OK, if packaged innocuously. Republicans in Congress made environmental research an early target, following through on a threat to cut EPA's budget. But as time passed, an environmental focus in research emerged as a plus in the view both of the administration and Congress, so long as private property rights did not seem directly threatened and the research encompassed regulatory flexibility, technology and performance. Thus clean cars, air pollution research and regulatory reengineering have essentially held their own and even the President's Climate Change Action Plan is continuing, despite congressional critics' early attacks on the motivation behind such research.
Health care and transportation remain winners. The National Institutes of Health, the Transportation Department's University Transportation Centers and the Volpe Transportation Research Center in Cambridge, Mass., to name a few examples, have more than held their own in an otherwise bleak budgetary picture. (Republicans in Congress actually increased the administration's fiscal 1996 budget for NIH.)
The success of these two categories of research is attributable in part to the power of their constituencies-health care recipients, doctors, consulting engineers and urban commuters, to name a few. They've also been helped by the close association between policy research in health and transportation and the much larger federal scientific and technological research functions associated with them. This is particularly true in the health care arena, where billions of dollars are spent on technical research. Since scientific endeavors in health care have turned out to be fairly well-insulated, policy research functions associated with them also have been protected.
The pork barrel factor. Some research programs fare well for the age-old reason that particular members of Congress see them as beneficial to their constituencies. Thus high-speed rail research, undertaken by the Federal Rail Administration chiefly in the northeast United States, is a particular favorite of Washington-area representatives such as Sen. Barbara Mikulski, D-Md.
Sometimes, though, policy research programs are seen as easy targets because they lack easily identifiable or organized constituencies. A couple of years ago a number of legislators, both Republicans and Democrats, proposed sharp cuts in the Social and Behavioral Sciences Directorate of the National Science Foundation, where much of NSF's policy research is conducted. One big reason for the attack was that, unlike some other programs at NSF, such as engineering research, social and behavioral inquiries did not seem to have any clear group out in front in support of them, except policy wonks without much lobbying clout. However, once the cuts were proposed, universities and groups interested in particular areas of research, such as the transportation community, made their opposition known and the cuts were averted.
Statistical sampling and modernization are still politically charged issues. The statistical agencies face several unique issues. The most contentious is whether the 2000 census should use statistical sampling to augment the actual enumeration of people. Advocates of sampling note, as did L. Nye Stevens, GAO's director of federal management and workforce issues recently, that "the 1990 census, though it was the most expensive in history, for the first time produced results that were less accurate than those of the preceding census."
Sampling advocates feel that sampling of hard-to-count segments of the population would produce a more accurate count of the population. The debate engages suburban (mostly Republican) congressmen and their inner-city (mostly Democratic) counterparts. The Republicans fear a sampling procedure would shift the count toward Democratic constituencies. Whether this would be the result of a statistical sample is not at all clear, but a rider prohibiting the census from using the procedure was attached to one of the appropriations bills earlier this year, ensuring that the battle would continue.
Modernization of existing data collection routines and definitions is another issue. Most government economic data collection programs came of age during the 1950s and 1960s, when the economy was simpler and when computers were less powerful. Despite dramatic changes in both areas, statistical agencies are still tinkering with basic data designs developed 30 or more years ago. The result, say critics, is a slow deterioration in the accuracy of government economic information. For this reason some claim that although funding levels have been relatively steady, major new investments in systems redesign need to be made.
Concern about data quality has also motivated some to push for consolidating the statistical agencies under one roof. During the 104th Congress, this became a hot topic because of the proposal to close down the Commerce Department, home of the major statistical agencies, including Census Bureau. A bill for creating a statistics super-agency was introduced but, after some serious consideration, has gone nowhere. Given Washington's inertia in changing its institutions, the status quo of federal data collection seems likely to continue.
Demand for Information
All in all, the outlook for federal policy research and data collection programs seems brighter now than in the recent past.
Demand for more and better performance information is growing, albeit slowly. Two years ago, the President's budget contained barely any performance information. This year, almost every agency and major program has at least included one piece of performance-related information in program descriptions. OMB's requirement that agencies start providing such information, and Congress' continuing interest in it, play to one of the main strengths of policy research programs, namely their ability to answer questions about how programs achieve their results and how they might be improved.
Tight budgets, ironically, also may strengthen the research programs' utility. Agencies, after all, need to provide ever-more convincing evidence that their programs are working, or demonstrate why new policies will work in the future. Policy research shops are often best-positioned to provide such information.
At the same time, the fiscal squeeze that sometimes made research activities seem a luxury seems to be easing. Strong economic growth has already lowered the federal deficit to its lowest level since the 1970s.
How long might we expect the favorable climate for research to last? Perhaps that's a question best answered by your friendly neighborhood policy research shop.
Cameron Gordon is an assistant professor at the University of Southern California's Washington Center.
NEXT STORY: GIS Puts Information on the Map