Memo adds detail to president's innovation agenda

What: OMB memorandum on the administration's Evidence and Innovation Agenda

Why: In a White House speech July 8, President Obama promised an "aggressive" second term management agenda designed to capitalize on the power of technology to improve the efficiency of government operations, optimize policy outcomes and innovate government services. A memo to agency chiefs from OMB Director Sylvia Burwell and other top policy staff released July 26 goes a little deeper into explaining what Obama has in mind.

Agencies are directed to develop "evidence-based practices," that draw on data and experimentation to drive policy, with the goal of driving resources toward programs and policies that work efficiently and effectively, and away from programs that do not. In September, OMB will kick off a series of five workshops designed to acquaint senior agency personnel with how to apply experimental, data-driven approaches in their own budgeting.

The language mirrors a proposed $14 million Data-Driven Innovation effort in Obama's 2014 budget that would give the U.S. CIO's office a mandate to look for savings and improvements across government. When the proposal was unveiled in April, CIO Steven VanRoekel said, "We'll work with agencies to scale out what this evidence-based analysis and agenda can drive for the whole of government."

The memo identifies a few early efforts in this direction. The Treasury Department plans to test whether people with delinquent government debt can be nudged into paying by using a more personalized approach, simpler language, and a reminder that the recipient is in a small minority of those with outstanding debt. A joint data-matching program between the departments of Health and Human Services and Housing and Urban Development is looking to find out more about the health of senior citizens living in federally subsidized housing. OMB is also asking agencies to develop more innovative funding methods such as "pay-for-success" grants, in which programs initiated by non-profits and the private sector receive government funding after they are shown to have achieved results. Finally, agencies are being asked to share effective research and policy approaches in government-wide repositories.

Verbatim: "Proposals should utilize randomized controlled trials or careful quasi­experimental techniques to measure the effect of interventions on important policy outcomes. We particularly welcome proposals that draw on behavioral insights to improve results and lower costs in direct operations."