Raj Chetty has reshaped our understanding of social mobility in the United States.
The Harvard University economist’s research, often featured in The New York Times, has renewed scrutiny of America’s identity as “the land of opportunity.” In 2015, Chetty and Nathaniel Hendren, a frequent co-author, posted a detailed analysis of every county in the U.S. sorted by how likely their poor children are to move up the socioeconomic ladder. In 2018, Chetty, Hendren, and two other co-authors showed that Black men consistently earn less than their white peers from families with the same incomes, once they become adults. Chetty’s work has been used as a rallying cry to urge municipalities to do more to help their poorest residents reach the middle class.
Chetty has also dramatically recast how colleges are measured. Among his landmark findings: Many of the most selective institutions enroll more students in the top 1 percent of family income than in the bottom 60 percent. In recent months, he and his colleagues have revealed the extent to which many selective colleges put their thumbs on the scale for students from very rich families, as well as how strongly SAT scores correlate with family wealth. Yale University said it would integrate Chetty’s research into its admissions processes.
Economics is a notoriously white and male field, and Chetty has been credited with helping to change that. Born in India, his family moved to the U.S. when he was 9. He earned his economics Ph.D. from Harvard at age 23, won a MacArthur Fellowship in 2012, and was elected to the National Academy of Sciences in 2018. After stints as a faculty member at the University of California at Berkeley, Harvard, and Stanford University, he returned to Harvard in 2018, where he and two colleagues founded a research group called Opportunity Insights. He also started an introductory economics course that’s been lauded for attracting more diverse students to economics. Opportunity Insights has garnered tens of millions of dollars in grants from the Bill & Melinda Gates Foundation, the Chan Zuckerberg Initiative, and others.
But inside the lab, Chetty and his colleagues have not always practiced what their research preaches, several former employees say. When hiring for their prestigious “pre-doctoral fellowship” program, for instance, the lab uses a rubric that explicitly favors students from the very colleges that its own research has called out for reinforcing elitist systems. Opportunity Insights didn’t have its first Black pre-doc until 2021. Seven former employees who spoke to The Chronicle about their experiences were bothered by what they saw as contradictions between the lab’s practices and its stated values.
After landing the fellowship, some employees said they were also disturbed to find a culture of overwork that left them fried but feeling forced to impress in order to secure a letter of recommendation to a top Ph.D. program. For some employees, it took a toll on their health. Harvard even reviewed the lab following claims of unsustainable working hours.
Chetty and the research group’s other leaders, John N. Friedman, an economist at Brown University, and Hendren, an economist at the Massachusetts Institute of Technology, defended the practice of privileging graduates from a select group of well-resourced colleges. They said they value the economics training those students received.
“If you’ve been at this handful of colleges we’re talking about, you have been taught probably by recent Nobel laureates, or students of Nobel laureates, about methods they developed in the last five years,” Chetty said. “We can feel that very clearly when we’re doing interviews or talking with students after we’ve hired them here.”
Chetty acknowledged the lab had room to improve. But he argued that it was much more diverse than other, similar pre-doc programs, and than the field’s doctoral programs. “We’re proud of that,” he said. “We view that as basically part of our mission. We’re trying to advance equity in opportunity and through our research.”
Interviews with dozens of former Opportunity Insights employees, several of whom would only speak anonymously out of fear of damaging their career prospects, painted a nuanced picture of what it was like to work there. At best, the trailblazing lab is making a real effort to be true to its stated values while constrained by larger forces of homogeneity and elitism within economics. At worst, its leaders have been content to let the lessons of their research go unrealized within their own organization.
Over the past decade, big data has revolutionized economics. The mountains of numbers have meant two things. First, students need more specific, cutting-edge research skills to succeed in graduate school. Second, researchers need more people to help analyze the reams of data that now go into a study.
Enter the pre-doctoral fellow, or “pre-doc.” Inspired by the example of labs in the sciences, Chetty and his collaborators were among the first economists to hire increasingly large cohorts of recent college grads to be the foot soldiers in their intensive research. Chetty envisioned researchers at different stages of their careers spending years solving big problems under one roof, the way a biologist might cure a disease.
Some economists hoped the advent of pre-doctoral fellowships would also help diversify the profession. The field is striking for its lack of representation. A paper published this fall in the Journal of Economic Perspectives reported that 65 percent of U.S.-born economics-Ph.D. recipients have a parent with a graduate degree. That was the greatest share among 14 broad Ph.D. fields — like humanities, biological sciences, and business — the researchers analyzed. “Economics is one of the least socioeconomically diverse fields,” the paper’s authors wrote.
That paper also found that among U.S.-born doctorate holders in those 14 fields, economics Ph.D.s are the least likely to be underrepresented racial and ethnic minorities, the least likely to be first-generation college students, and among the least likely to be women. (Math, engineering, and computer science have fewer female doctorates).
In theory, pre-doc positions could help to level the playing field. They give recent graduates a chance to try out life as a researcher before committing to a six-year Ph.D. This opportunity is especially important for people who didn’t attend a college with robust research programs. Pre-docs are seen as a good alternative to master’s degrees because they’re paid — at roughly $45,000 to $65,000 per year, according to one survey. (The pre-docs at Opportunity Insights make $65,000, Chetty said.)
Known for its prestige and mission, Opportunity Insights is a sought-after destination for prospective pre-docs. The group receives some 500 applications a year for the roughly half-dozen spots that open up annually. To sort through the stacks of materials, the lab created a rubric that assigns points based on different measures. Current pre-docs use the rubric to help make a first cut in the applicant pool for the successive class.
In a 2021 copy of the rubric obtained by The Chronicle, “caliber of university” was worth two points, out of nine. The other criteria were research experience, worth four points; grades, worth two points; and a bonus point that could go to applicants “who truly stand out,” have a unique skill or perspective, or go “above and beyond.” Opportunity Insights’ leaders confirmed that the college an applicant graduated from is still worth two points out of nine on the rubric, but declined to share the latest version.
A college’s caliber is determined by a list the lab leaders maintain. One version of the list obtained by The Chronicle catalogs the 75th-percentile math SAT scores at more than 1,200 colleges. The rubric stipulates that the full two points go to alumni of colleges where the math SAT scores of top admitted students were 790 or higher; 1.5 points for scores of 750 to 789; and one point for 700 to 749.
Based on these copies of the rubric and calibrating list, applicants from the California Institute of Technology, Carnegie Mellon University, Swarthmore College, and 32 other colleges, got the full two points. Graduates of the University of California at Irvine, Rutgers University at New Brunswick, Kenyon College, and 68 other colleges got one point. Public universities that had lower SAT scores, but which were their state’s flagship, as well as “top” historically Black colleges, got a half point. All other HBCUs, along with about 1,000 other colleges — the vast majority of the institutions listed — got zero points. (It’s not clear which HBCUs were considered “top.”)
Sarah Oppenheimer, executive director of Opportunity Insights, wrote in an email that the lab has since more than doubled the number of colleges that score two points on the rubric, but she declined to share the current list. “We do not want to inadvertently discourage applicants who may not be from these top-scoring institutions from applying,” she wrote.
Several former employees saw the rubric as conflicting with the findings of Opportunity Insights’ own research. Data from the lab’s latest college-admissions paper show just how tightly students’ SAT scores correlate to their families’ incomes. That paper also showed that at “Ivy Plus” colleges — defined as the eight Ivy League institutions, plus Duke University, MIT, Stanford, and the University of Chicago — family income has an outsize influence on admissions. Even among students who have the same SAT scores, students from families in the top 1 percent of American incomes are more than twice as likely to get into an Ivy-Plus college as students from middle-class families.
“We conclude that highly selective private colleges currently amplify the persistence of privilege across generations,” reads the paper’s abstract. All of the Ivy-Plus colleges received two points in the calibrating list The Chronicle obtained. In the list the lab currently uses, Oppenheimer wrote, many state flagships are also worth two points, including the Universities of Florida, Illinois, and Maryland. “Hence, the rubric does not differentiate between applicants from the University of Florida or Harvard,” she wrote. “We hope this gives some insight into why the final pool of pre-docs at OI consists of students from such a broad group of colleges.”
In interviews, the lab’s leaders said they saw benefits to hiring graduates of high-scoring colleges. The pre-docs often learn from one another, Chetty said, so it’s helpful to have at least some who come in highly trained. “If you’re trying to do the best research,” he said, “the reality is, I think it is valuable to hire students from top colleges.”
As for why the lab defines “top colleges” using their students’ math SAT scores, Friedman acknowledged it’s not a perfect measure, but, he said: “We’re trying to be consistent about how we do it, in a way that isn’t just us saying, ‘Oh, we like that school. We don’t like that school.’” Even as Opportunity Insights’ research has shown a close correlation between SAT score and family income, it’s also found that among students of the same race or level of family wealth, higher scores are associated with better incomes and jobs later in life.
The lab’s leaders also emphasized that applicants’ undergrad institutions play only a small, or no, role in their hiring. The rubric is one factor in determining who advances to the next stage in hiring, which is a test of applicants’ data-analysis skills. The final stage is two interviews, focused on the research applicants have done.
Chetty said he was aware that favoring prestigious colleges could “cut against having a diverse pool.” To account for that, Chetty said the lab also keeps an eye out for applicants “from those backgrounds that are less widely represented” and advances most of them, regardless of how the rubric scores them.
Underrepresented perspectives in the lab might mean skills, such as computer programming, or being from parts of the country or world that are unusual for Opportunity Insights, Chetty said. A large share of underrepresented-racial-minority applicants get the coding test and an interview, he said, adding that pre-docs working on admitting the next class weren’t told about this aspect of the hiring. He said he didn’t want to “create the social dynamics of: ‘This person was admitted through one process, and then this person got some other consideration.’”
The hiring process has yielded a relatively diverse group of pre-docs, the lab leaders say: Since 2018, about 15 percent of their pre-docs have been Hispanic or Black, and 44 percent are women or nonbinary. For comparison, just 4 percent of doctoral students in economics are Hispanic, Black, American Indian, or Pacific Islander, according to the latest available data from the National Science Foundation. Thirty-six percent of current economics Ph.D. students are women.
But the lab’s pre-doc alumni overwhelmingly hail from colleges of high “caliber,” as the rubric defines it. Of 76 former pre-docs for whom The Chronicle could identify an undergraduate institution — out of 81 pre-docs in the lab’s history — only one non-international student attended a college that received zero points in the 2021 copy of the rubric and calibrating list.
The problem isn’t the rubric, the lab’s leaders argue. It’s that less than 2 percent of people who apply to Opportunity Insights come from zero-scoring colleges.
Any efforts the lab made to specifically recruit Black pre-docs yielded no results until recently. By 2020, after nationwide protests sparked by a police officer’s murder of George Floyd, the lab still had never hired a Black pre-doc. Some staffers saw the situation as conflicting with the moral thrust of the lab’s research.
At least two classes of Opportunity Insights pre-docs advocated for changes to how pre-docs are hired, as did Maddie Marino, Chetty’s assistant in 2019 and 2020. Several pre-docs in the 2019-21 class compiled data for Chetty on who applied and got into the lab, showing which undergraduate programs were not well represented among applicants. After Chetty reviewed the data, Marino sent emails from his address to economics professors at those programs, to encourage applications from their top students who would soon be graduating. At least one pre-doc was hired as a result of this new outreach, Marino said.
In 2020, a new group of pre-docs saw problems with the way they were hired. This group pushed to widen the rubric even further — for example, advocating for points for applicants in the honors college of any public university. Former employees say the lab leaders made some changes — graduates of honors colleges do get a bonus — and emphasized that the pre-docs should flag applicants whose rubric score might not reflect their abilities, and who deserved a second look.
In the spring of 2020, Kai Matheson, a pre-doc at the time, worked with other staff members to organize an antiracism discussion. Later in the year the lab held an antiracism training for employees. After the training, staff members took an anonymous survey. Three responses pushed Opportunity Insights to hire more Black staffers and researchers. A fourth response suggested reviewing the pre-doc hiring process, to make sure it was equitable and effective at finding talent.
Some staff members from that period told The Chronicle they weren’t satisfied with the lab leaders’ explanation that the problem was the pipeline.
“Relative to the representation in the pipeline of applicants,” Chetty told The Chronicle, “I think we are, even with those low numbers, at or above the share you’d expect.” Since 2018, the proportion of applicants who identified as Black or African American has been 3.7 percent, according to numbers from Harvard’s hiring portal.
Opportunity Insights’ first Black pre-doc began the fellowship in 2021.
Opportunity Insights is not the only pre-doc program that has found it challenging to hire diverse researchers. To try to remove barriers to careers in research, in 2020, some economists started PREDOC.org, a group dedicated to expanding the talent pool for pre-doctoral programs, said Pietro Veronesi, one of the founders and a professor at the University of Chicago’s Booth School of Business. The Opportunity Insights economists have been active in the organization. PREDOC.org has a free summer class, modeled on Chetty’s at Harvard, meant to showcase the important work economists conduct. Chetty, Friedman, and Hendren have all lectured for the course. In a recent cohort, the students — of whom nearly half were underrepresented minorities and a third received Pell Grants — got a stipend to participate. (Veronesi said he could not share demographic information about Booth’s pre-doc hires.)
Pre-doc programs are an alternative path into economics, but could they actually hinder diversity in the field? Some economists worry that a pre-doc has become just another credential for already qualified Ph.D. candidates.
PREDOC.org recently hosted a webinar for students thinking of taking the plunge. During the panel, Ilyana Kuziemko, a professor of economics at Princeton University, answered the question: What do professors look for in a pre-doc application?
“We’re looking for somebody who, 10 years ago, would have gone straight to Ph.D., who looks really talented and ambitious and prepared,” she said, while acknowledging that her own research group was part of the problem. “Just to be completely blunt, I think that is what these more competitive pre-doc programs are looking for.”
Adding another bar to clear on the way to a professorship may disproportionately reward certain students. A study by Kevin A. Bryan, an associate professor at the University of Toronto, showed that between 2013 and 2018, pre-docs increasingly appeared on the CVs of the most sought-after economists coming out of Ph.D. programs. Those who are willing to do research for a couple years before entering a doctoral program, rather than get a higher-paying job, Bryan said, are often children of academics or others with advanced degrees. “Anything that delays the time before you have a good job is bad for diversity,” he said.
In addition, professors who run pre-doc programs are incentivized to wring as much labor from pre-docs as they can. Pre-docs are skilled workers motivated to put in long hours, solely on a professor’s projects. Unlike graduate students, they’re not pursuing their own theses.
“You’re giving a huge amount of control over your life to one person,” Bryan said.
Once pre-docs were hired at Opportunity Insights, the lab leaders expected them to work hard, as they themselves do. That could mean 12-plus-hour days, long nights, and Slack messages with fresh assignments at late hours as deadlines approached — before major presentations, for example, or New York Times stories. (Lab leaders say 12-hour days and required late-night assignments are very rare.) Within the lab, people called those periods “pushes.”
The intense workload reached a new level early in the pandemic, some former pre-docs said. When offices like the Internal Revenue Service and Census Bureau closed, Opportunity Insights lost access to the data sets it was using for its research. At the time, the lab had about 18 or 19 pre-docs, its largest group ever, according to Chetty. The pre-docs and the lab’s leaders pivoted to a new project: a vast tracker of how the U.S. economy was reacting to shutdowns.
Moderate crunches before publication deadlines are a normal part of being an economics researcher, economists within and outside of Opportunity Insights say. One pre-doc who was critical of the workload nevertheless understood the long hours in part as the price the lab pays for its influence. Chetty said the data that emerged from the shutdown tracker, showing that wealthier Americans didn’t spend their Covid stimulus checks, led to income caps for later stimulus bills, and more money available for low-income households. “The work the pre-docs did then had actually enormous impact,” he said.
But most pre-docs who spoke with The Chronicle and worked on this push found it to be extraordinary. The lab turned into a “zombie apocalypse,” one former pre-doc said. Instead of a push of a few days or a week, which was typical before a deadline, the pre-docs estimated that the Covid-tracker push lasted at least a month. Lab leaders were divided over how long the push was, exactly.
Pre-docs endured the long hours in pursuit of the most precious credential of all: a letter of recommendation from Chetty to a top economics-doctoral program. For the fellows, that is the coin of the realm and a chief reason they become a pre-doc. “Our compensation is prestige,” said a former pre-doc. “It’s rec letters.” Opportunity Insights advertises itself as an avenue to prestigious locations, listing on its website where pre-doc alumni land after their two years at the lab — MIT, Northwestern, Stanford, Harvard.
“The implicit assumption is that the pre-doc who’s spending all their time at their desk is the one that’s going to have the better letter of rec,” said one former staff member. Another former employee who worked on personnel issues said the pre-docs were intensely focused on getting into a top-five program, and considered it a letdown if they did not.
One former pre-doc pointed to a study to help explain his peers’ preoccupation with prestige. It found that in the 96 economics departments that U.S. News & World Report ranks, 60 percent of faculty members had trained at a top-15 doctoral program. Fifteen percent had trained at MIT or Harvard alone. As one economics professor who is not involved with Opportunity Insights, Trevon Logan of Ohio State University, put it: “Economics has many problems. The biggest problem is not racism, sexism, or homophobia. The biggest problem in economics is elitism.”
Logan’s overall impression of Opportunity Insights, however, was that it appeared more diverse than other pre-doc programs. “I know Raj has been committed to that,” he said. “He has been purposeful about trying to diversify that applicant pool as much as possible.”
In the spring of 2020, the hours at Opportunity Insights got so bad that Harvard’s Office for Faculty Affairs began a review. A Harvard administrator contacted the pre-docs on their non-Harvard email addresses. The administrator, a program officer for Title IX and Professional Conduct in the Office for Faculty Affairs and FAS Human Resources, explained in a message that she was worried about “the level of silencing folks are experiencing and fear of retaliation.”
“I feel bad if somebody felt that way,” Chetty said during his interview with The Chronicle. “I certainly cannot think of an instance where I or others have said anything that would suggest we would retaliate. I have never withheld a letter or something like that.”
After interviewing pre-docs and other staff members, Harvard administrators attributed the long hours at the start of the pandemic to “a competitive culture that had emerged in the lab particularly while remote and more difficult to monitor,” said Kirsten Colton, assistant dean of administration in Harvard’s Division of Social Science, of which Opportunity Insights is a part. Although some former employees understood the review to be an investigation, Harvard didn’t consider it one, Colton wrote in an email, because no evidence emerged of misconduct that violates Harvard policy.
There was no report as a result of the inquiry, but in October 2020, Chetty, Friedman, and Hendren held a meeting to explain its results to the pre-docs. From then on, Opportunity Insights would hire fewer of them. Fewer pre-docs would mean the aspiring Ph.D. students would get more face time with the lab leaders. They would get the mentoring and strong recommendations they were there for.
For some former employees, however, that time period left lasting scars.
Both Marino and Matheson said the stress of working in a place that they believed conflicted with their values contributed to severe mental-health crises. For Matheson, the long hours also contributed.
These and other pre-docs’ setbacks were not easy for those who witnessed them, either. At least three former employees who said they had fair or good experiences at the lab tempered their enthusiasm because they’d watched their colleagues struggle.
“These concerns contradict what we have heard directly from our staff,” Chetty, Friedman, Hendren, and Oppenheimer wrote in an email to The Chronicle, saying they encourage their staff to connect with Harvard resources when they need support and educate them about those resources. “We strongly prioritize mental health among predocs and the full OI team.”
Marino and Matheson each left Opportunity Insights in under two years.
Since 2020, Opportunity Insights has hired about six pre-docs per cohort, so there’s a total of roughly 12 at any given time, Chetty said. Each pre-doc is paired with an economist, sometimes including collaborators elsewhere at Harvard or at another university, so that everyone has a mentor with whom they work closely and who can write them a letter of recommendation for graduate school. The Chronicle spoke with four pre-docs hired after this change, who had been assigned to mentors outside of the lab. All were satisfied with their advisers, and one mentioned working with Chetty alongside her mentor, and getting both to write her letters.
“I just think the program is in great shape and is better now,” Chetty said. “Even aside from the pre-doc experience, it allows us to do better work and serve our mission better.”
Harvard, too, was satisfied with the changes Opportunity Insights made. “We believe that OI is a leading example of a positive and diverse workplace culture within the division thanks to dedicated leadership making incremental improvements over the past several years,” Colton wrote. “Indeed, OI as an organization strives to live its mission.”
Many pre-docs relished their time in the lab. Oppenheimer, the executive director of Opportunity Insights, put The Chronicle in touch with 13 former pre-docs who worked for Chetty and his colleagues between 2014 and 2023. Eleven of the 13 spoke on the condition of anonymity. Five pre-docs The Chronicle reached independently also said they had very good experiences.
These former pre-docs said they learned a lot, made great friends, and got into top-ranked graduate programs. At least seven disagreed with the idea that the pre-docs were competitive with one another. Some said that yes, they worked long hours, but the pushes were a chance to work closely with Chetty, Friedman, and Hendren, and to watch some of economics’ best minds solve problems in real time. Plus, pushes were balanced by more relaxed weeks, they said. “It was a really great experience,” said Isabel Almazan, who finished her two-year fellowship in 2023. “I still can’t believe I was actually a pre-doc at OI.”
In defending their group, the lab leaders cited experiences like these. But they also acknowledged their pre-doc program doesn’t yet fully embody the lab’s mission of spreading opportunity more widely. In discussing how Opportunity Insights hires, Chetty frequently came back to the field’s limitations — its homogeneous pipeline and the underlying inequities in American life. “We’re not perfect,” Chetty said. “We’re trying to do this ourselves and we’re living in the constraints of the system as they are.”
