The Obama administration is reporting progress on its ongoing effort to prod agencies to increase their use of data analytics in making “evidence-based” evaluations of programs—even to the point of attracting interest from Republicans in the highly polarized Congress.
Kathy Stack, adviser for evidence-based innovation at the Office of Management and Budget, on Wednesday summarized the multi-agency effort to embed “rigorous evidence in combination with collaborative decision making that is critical to getting change” for more effective federal grant-making. She spoke at a Washington forum for agency representatives co-sponsored by Johns Hopkins University, which is offering new graduate programs in data analytics, and REI Systems, which contracts with the government to operate Data.gov, Performance.gov and USAspending.gov.
Fleshing out details on agency responses to the push for inexpensive data from a wider variety of sources laid out in a July 2013 OMB memorandum, Stack said there is interest on Capitol Hill in funding agency data analytics offices, particularly among the Labor, Health and Human Services and Education appropriations panels.
In what may come as a surprise to some, she said OMB had spotted possible common ground on evidence-based program reviews in the white paper on reducing poverty and expanding opportunity released in July by conservative House Budget Committee Chairman Paul Ryan, R-Wis. It contained a sentence advocating “results-driven research,” and called for “a commission to examine the best ways to encourage rigorous analysis of our safety-net programs. Specifically, the commission would consider the implementation of a new Clearinghouse for Program and Survey Data to enhance research capabilities and help us policymakers design more effective programs.”
On Aug. 26, with little fanfare, an OMB team joined academics and nonprofit specialists in a meeting with the House Budget staffers organized by Robert Shea, a principal with Grant Thornton and George W. Bush-era OMB veteran who now chairs the board of the National Academy for Public Administration. The discussions, Shea told Government Executive, centered on how to tap the “treasure trove” of agency data that the law permits to reduce the cost of program evaluations without violating privacy protections. “The combination of President Obama’s evidence agenda and the apparent acknowledge of its importance by Chairman Ryan show that this is clearly an area where we can find common ground no matter what happens in the next election.”
In her talk, Stack said “good data on what works is not overhead,” but integral to rigorous program evaluation. “There’s a lot of confusion across government, with lots of collective data, but if agencies don’t think of in within a framework, it gets confusing.”
The best model currently in government, Stack said, is at the Labor Department, which since 2010 has had a chief evaluation officer with a 10-person staff who coordinates a “data-driven learning agenda to drive cultural change,” in which all assistant secretaries participate. The Commerce Department, she noted, on July 14 announced its first-ever data officer, which is also a model.
Her brief history of the data movement characterized implementation of the 1993 Government Performance and Results Act as having “gotten off to a bumpy start.” The subsequent Bush administration’s grade-based Program Assessment Rating Tool “got deeper into the weeds, but became incredibly adversarial as agencies became defensive,” said Stack, who worked at OMB under both administrations.
So the Obama team stepped in determined to bring in data enthusiasts and experts, she said, to provide evidenced-based tools to show “what works” in good grant-making, in part by leveraging partnerships within agencies, among agencies and with existing third-party data from academia, the nonprofit world and private business.
Examples of the greatest agency impact, she said, are the Health and Human Services Department’s use of data to determine best practices in home health care delivery, and the Education Department’s simplification of the student financial aid forms. Programs actually terminated by Congress following data-driven negative evaluations included the Education Department’s Even Start program for students from low-income families and a program to provide counseling to children of prisoners, she said.
The General Services Administration and the White House Office of Science and Technology Policy have assembled data experts who are available to other agencies. Some have advised the United Kingdom on use of data from behavioral sciences in the area of tax compliance, Stack said.
“Last fall OMB had workshops across agencies and within agencies,” Stack added, “and we made sure they came in teams—someone from policy, from statistics, from data.” In many agencies there is resistance to creative use of multi-sourced data heard from the “mission support officers,” such as the general counsels and budget officers, who “mostly love standard operating procedure,” she said.
The path forward, Stack said, “does not involve setting up a big new organization, but having agencies come to OMB to beef up their analytics unit by tapping from within. It means finding “the sweet spot” among the data seekers, the problem solvers and the agency powers, who are the enablers.
(Image via wrangler / Shutterstock.com)