Editor’s note: This story is one in a series, #PolicyForward, that spotlights how faculty, students and alumni at the Harris School of Public Policy are driving impact for the next generation. Leading up to the May 3 grand opening of Harris’ new home at the Keller Center, these stories will examine three of the most critical issues facing our world: strengthening democracy, fighting poverty and inequality, and combating climate change.

The U.S. economy grew faster under President Harry Truman than under any other postwar president, while it was the most sluggish under George H.W. Bush. Under Mayor Rudolph Giuliani, violent crime fell by more than 56 percent in New York City in the 1990s while it increased 70 percent in Chicago since 2015 under Rahm Emanuel. But how much credit or blame should political leaders receive for those outcomes?

To answer this question, Harris Public Policy scholars Christopher Berry and Anthony Fowler devised an innovative method for statistically testing the effectiveness of leaders. Although many elected leaders take credit for economic growth, employment and crime, the new study downplays the impact of politicians around the world, especially U.S. governors and mayors.

“At a time when public trust in political institutions and leaders has fallen dramatically in the U.S., it is important to understand whether these leaders have a meaningful impact on the well-being of their constituents,” said Christopher Berry, the William J. and Alicia Townsend Friedman Professor, a leading policy scholar renowned for innovative analysis. “Our research bore surprising results.”

To demonstrate the value and accuracy of their method, which they call Randomization Inference for Leader Effects (RIFLE), Berry and Fowler implemented it for world leaders, U.S. governors and U.S. mayors in the areas of economic growth, employment and crime, although the model can work for virtually any area of interest.

Among their findings:

World leaders:

Berry and Fowler found that as a country switches from an average leader to one that is one standard deviation above the mean, they can expect GDP growth to be about 1.5 percentage points higher than normal. Dividing countries into three categories—those that were always democracies, those that were always autocracies, and those that changed their status at least once—they found some evidence of leader effects in autocracies and strong evidence of leader effects in transitional countries. They found little evidence of leader effects in democracies, possibly because democracies more consistently select high-quality leaders, meaning that the differences between leaders are less meaningful. or because good institutions can partly counteract the effects of a bad leader.

U.S. governors:

Berry and Fowler found no evidence that governors matter for economic outcomes in their states. Although governors differ in their abilities or appetites to raise and spend money, including federal aid, those differences don’t translate into differences in state income and employment. Governors do, however, influence property and violent crime rates, but not because of their political affiliation. There is little effect on crime when looking at Democratic versus Republican administrations.


Berry and Fowler examined the effects of mayors in the 100 largest U.S. cities and found little evidence of mayoral effects on income or employment­­. What’s more, they found no evidence of mayoral effect on some of the most important outcomes in a city—the economy, the size of city government and crime rates. One possible explanation is that mayors lack control over governance and service provision within their jurisdictions.

“While previous studies have focused exclusively on aggregate economic outcomes, our results highlight the importance of matching different offices to relevant outcomes when estimating leader effects,” said Fowler, an associate professor at Harris Public Policy. “RIFLE can be applied to virtually any setting with leaders and an objective outcome of interest, so its continued application should improve our understanding of where, when and why leaders matter.”

Another study by Berry and Fowler utilized the RIFLE methodology to determine that coaches have a significant impact on success in both the professional and collegiate ranks, accounting for 20 percent to 30 percent of the variation in team outcomes.