Are agencies using data to improve performance?

Federal agencies still struggle to wrangle their own data as they look to improve how they do business, according to a new survey.

A report from the nonprofit Partnership for Public Service and consultant Grant Thornton found that many workers charged with refining their agencies’ performance “had not begun using data to drive decisions effectively.” According to the research, only 41.7 percent of the so-called performance improvement officers surveyed were using data to measure the factors their agencies needed to determine success. That’s a nudge above the 38.8 percent in 2013 who said the same.

“Several performance staff said they have too much data and not nearly enough information,” authors wrote. They added, “Agency officials described databases full of numbers that were not helpful in determining if the agency was accomplishing its mission.”

Grant Thornton’s Robert Shea, who has previously served as the Office of Management and Budget’s associate director for administration and government performance, said IT staffers have the potential to help agencies become more efficient.

Focusing on “the need to depict information in a way that facilitates decision-making is a real service that the IT community can offer performance improvement officers,” he told FedScoop.

The paper examines where agencies stand five years after Congress passed the Government Performance and Results Modernization Act, an effort to help agencies hone their most important functions and use data to make management decisions. As part of the law, agencies have to set priorities, conduct strategic planning, and select a chief operating officer and performance improvement officer to oversee those activities.

For the report, authors focused their research on the departments of Agriculture, Commerce, Homeland Security, Housing and Urban Development, Labor, and Treasury. They talked to nearly 100 staffers who are working on improving performance at the agencies’ subcomponent levels. About 60 of those participants also answered a set of written questions

The report includes several anecdotes showing where agencies’ data collection could improve. At the HUD’s Office of Multifamily Housing Programs, for example, some local offices collect data differently, which causes problems when the department aggregate’s data sets.

“The standardization of reporting would probably help and would be a very cost-effective means of getting all the data down,” one HUD performance staff member said, according to the report.

One USDA representative said in the report that the agency needed staff with advanced analytics skills who could pull insights from data.

“We have management analysts and program analysts by series and choice of grades, but are they truly doing the analytic work? How do we redefine that and work with OPM so that we do have, in fact, the skill sets that are needed?” she said.

In the report, Partnership recommends that agencies:

• Standardize data collection across regions or offices to make it easier to aggregate data across subcomponents.
• Eliminate outdated data collection requirements where possible.
• Reach out to other organizations that collect data that could inform or complement the information currently available to agency staff.

The report also offers recommendations in other areas, like building meaningful relationships, recruiting, connecting program activities to agency priorities and showing return on investment.

In May, the Office of Management and Budget and the Treasury Department released standards and a guidance to help agencies adhere to the Digital Accountability and Transparency Act, or DATA Act. The law requires agencies make financial, budget, payment, grant and contract data more transparent and accessible. Shea said the DATA Act has the potential to help agencies make themselves more efficient.

“I think there will be a good number of data elements available that weren’t formally available,” he said. “And hopefully there will be a lot of focus on that. And the quality of the data will be improved and, when married with the performance data, will give you real insights into what’s working and what’s not.”