Around the Institute

Peter Orszag and John Bridgeland Discuss “Moneyball for Government”

January 27, 2015  • Institute Contributor

Above, watch the full conversation with Peter Orszag and John Bridgeland.

Peter Orszag, vice chairman at Citigroup and the former director of the Office of Management and Budget and the Congressional Budget Office, and John Bridgeland, CEO of Civic Enterprises, co-chair of the Aspen Institute Franklin Project and the former director of the White House Domestic Policy Council, sat down to discuss “Moneyball for Government,” a recent book which Orszag edited and to which Bridgeland contributed.

The term “Moneyball for Government” has its origin in baseball. “Playing moneyball” meant relying on data — not scouts’ intuition — to select players who could provide specific outcomes at the lowest cost. Such an evidence-based system could eventually be used to compare government programs like team managers compare second basemen. But first we need to start gathering evidence. In “Moneyball for Government,” government leaders and thinkers make the case to do just that.

This book was inspired by Orszag’s and Bridgeland’s time in government, which Orszag described as “flying blind in an era when we don’t need to be.” Orszag recalled a 2013 Government Accountability Office survey where only 37 percent of federal government managers had had any evaluation of any program they were involved in over the previous five years.

In talking with Washington Post Columnist Michael Gerson, Bridgeland described “Moneyball for Government” as a “means to an end: to boost social mobility and economic opportunity for young people and families and vulnerable populations.” And to reach that end, three steps are necessary across government:

  1. Build an evidence base
  2. Move toward policies backed by that evidence and away from those without supporting evidence
  3. Direct funding accordingly

That last step is the hard one because, as Orszag put it, “no one likes being told their program doesn’t work.”

The panelists argued that such a system would meet the aspirations of both parties. Orszag said it would give Democrats a chance to “make the case to the public that government works,” while Bridgeland said it would allow Republicans to press for a “limited, effective government.”

To build the foundation for such a system, Orszag and Bridgeland advocate “taking one percent of existing departments’ and agencies’ budgets and having that invested in collecting and developing sophisticated evaluation, so that you know — as a taxpayer, and a policymaker, and a stakeholder, and a program developer — how the other 99 percent of that funding actually is being spent,” Bridgeland said. “Marry that with chief evaluation officers, and you really could systemically transform what government funds,” he added.

Although evaluating the success of government programs isn’t a new idea, a combination of factors is making “Moneyball for Government” not only easier, but also more necessary. “It is easier to obtain data and process it than ever before,” Orszag said. “We live in an area in which Google and others are constantly running randomized controlled trials.” In other words, that one percent of funding can go further than it used to.

Additionally, “we will be under tighter fiscal constraints than any other time in modern history,” Orszag said. Accordingly, by focusing on programs that demonstrate evidence-backed results, the government can do more with less.

Orszag also said there is growing pressure from organizations outside government for using this sort of measurement. And Bridgeland added a final factor, saying, “The evidence base [has become] better than over the last 30 to 40 years in a lot of areas, and there’s a lot more to cull from the research in terms of what is effective and what could be effective.”

But a “Moneyball for Government” system faces some obstacles. To begin with, such a system would have to overcome how people think about testing. As Orszag said, “There’s always going to be tension between the gold standard of evidence — the randomized controlled trial — and anything that’s not randomized.” He worries about those who would only consider that gold standard evidence. Although randomized trials are the only way to demonstrate causality, such tests are also very expensive, and they can take a long time. Instead, Orszag advocates embracing a spectrum of evidence and learning to weigh different forms of quality evidence.

Additionally, “Moneyball for Government” would have to change how politicians make decisions about funding. Bridgeland cited a study that found Congressional members were more responsive to party leaders and special interest groups than evidence.

Bridgeland and Orszag speculate about a future where a broad evidence base could be used to create a “Moneyball” index to rate programs and persuade politicians by combining advocacy and data. Such indices could also be used to hold accountable politicians that vote for programs not backed by evidence.

This system could also change how funding is given and rescinded. For example, prizes for results could be offered instead of grants for programs. And instead of permanently eliminating less effective programs, they could be suspended until they provide more data regarding effectiveness.

Momentum is building for “Moneyball for Government.” Orszag and Bridgeland’s ideas are finding successes at the municipal, state, and department level. There is even pending federal legislation to take steps toward creating the mechanisms for evidence-based policymaking. Perhaps it’s only a matter of time before budgets are built more like baseball teams.