Does Economic Growth Help the Middle Class and the Poor? Russ Roberts Speaks at Hoover's Inaugural PolicyEd Talk

Tuesday, March 6, 2018
Hauck Auditorium, Hoover Institution, Stanford University

When looking back at our parents’ economy, we see both bad and good. We would love to go back to 1970 and buy a single-family home in Palo Alto for $35,000, but we’re glad we get to have the cell phones, quality of medical care, and range of job opportunities available in 2018. American economic life is clearly different than it was in 1970, and the overall economy is significantly larger. But are Americans, particularly the middle class and the poor, better or worse off as a result of those changes?

In the new PolicyEd series The Numbers Game, Russ Roberts, the John and Jean De Nault Research Fellow at Hoover, takes a look at that question. Responding to the dominant story of economic stagnation among the poor and middle class, he offers viewers a nuanced picture of what economic growth has meant for individuals across all levels of the income distribution.

“The truth is, it’s complicated,” he explains in the series. Conclusions about the poor and the middle class depend on how the data is treated and interpreted. Tweaks like swapping the price index used to adjust for inflation, choosing whether to include the elderly, or correcting for changes in family structure can dramatically alter measured trends and implications.

Russ weighs different approaches and arrives at relatively encouraging estimates of income growth across the board. But the more important point, which he emphasizes throughout the series, is how difficult it is to draw conclusions at all. Approaching economic data with an open mind, whether that data deals with income growth or anything else, produces a much more measured perspective than the ones we typically hear on TV or in political debates.

Extended Cut: Russ Speaks at Hoover’s Inaugural PolicyEd Talk

On February 18th, 2018, Russ spoke about The Numbers Game as part of the new Hoover event series PolicyEd Talks. Content featured at is concise by design, but the new series lets Hoover fellows explain the big-picture ideas behind their work. More than 230 students and other guests attended the talk, which was one of the first public events held in Hauck Auditorium in Hoover’s new David and Joan Traitel Building. Russ screened a “director’s cut” version of The Numbers Game, pausing throughout the two episodes released so far (more are in production) to expand on the economic concepts, methodology, and counterintuitive findings mentioned in the series. What follows are some lessons from his talk:

1. We are own biggest sources of misinformation. Russ began by comparing discussions about how economic growth has affected the middle class and poor to the story of Romeo and Juliet—we already know the ending. When interpreting information about the economy, we’re biased by our preexisting conclusions. He reinforced his point with a quote from physicist Richard Feynman featured in The Numbers Game: “You must not fool yourself—and you are the easiest person to fool.”

2. Outliers can seriously distort average income. Russ shared an extreme example: Indianapolis Colts starting quarterback Andrew Luck earned a degree from Stanford in architectural design in 2012, but having signed in 2016 what was then the most lucrative contract in NFL history, he skews the average salary among his graduating classmates. Returning to the video, Russ explained that counting only production and nonsupervisory workers when calculating average hourly earnings helps remove data on the rich and super rich that might make the average less representative.

3. Watch out for shifts in sample composition. Changes in the composition of a sample can significantly affect median income figures. The marriage rate, for instance, has declined among less-educated Americans while staying relatively stable among high earners. Having more single-earner, lower-income households at the bottom of the distribution reduces median income for the sample without representing any decline in individual income.

4. To measure progress, follow real individuals. Snapshots of sections of the economy over time don’t capture individual progress. Russ shared a 2016 paper co-written by Capital in the Twenty-First Century author Thomas Piketty that finds incomes for the bottom fifty percent of earners have stagnated since 1980, while the top one percent of earners have seen their incomes triple. Russ showed that following actual people within those percentiles gives a much more hopeful result. In one longitudinal study of earnings across generations, ninety-three percent of children born in the bottom quintile earned more than their parents, compared to just seventy percent of those born in the top quintile. Another found that children born in the bottom quintile earned twice as much on average as their parents, while the average income of those born in the top saw no change at all.

5. We shouldn’t be so sure. Russ concluded his talk by restating the principal theme of The Numbers Game: interpreting data about income and economic growth is complicated. “I hope that economists making dramatic statements about these things can be a little more nuanced and be a little less dramatic going forward,” he said. “We’ll see.”

More PolicyEd Talks at Hoover and Online

Education is the heart of the PolicyEd program, and PolicyEd Talks put Hoover fellows in a classroom-style environment where they at their best. Fellows can connect with guests face-to-face, convey their excitement about ideas, and in Russ’s case, keep the discussion going well into the night.

This and other PolicyEd Talks will be available on the web as part of a new video series, and PolicyEd Talk events are free and open to the public. Stay updated on future video releases and events by following PolicyEd on Facebook or registering at