Lawrence Katz, the Elisabeth Allison Professor of Economics, has been named a 2025 Citation Laureate, an annual award that recognizes influential researchers considered likely to win the Nobel Prize in their field.

The pioneering labor economist, who joined Harvard’s faculty in 1986, has produced decades of highly cited findings on wages, inequality, and technological change. According to London-based Clarivate, the company that publishes the annual list of awardees, Katz now totals nearly 26,000 citations across 72 academic publications.

“It’s very nice to see the research recognized,” said Katz, who was honored with Massachusetts Institute of Technology Professor David H. Autor, M.A. ’94, Ph.D. ’99. “It’s particularly nice that it’s shared with my longtime collaborator and former student. We’ve done a lot of work together.”

A total of 22 Citation Laureates were announced this year, with Clarivate touting its program as a shortlist of worthy recipients for the world’s top scientific distinction. Since 2002, 83 Citation Laureates have gone on to win Nobel Prizes.

Katz, who is also the editor of The Quarterly Journal of Economics, discussed everything from AI to the power of connecting across socioeconomic class in this conversation, which has been edited for clarity and length.

---

Let’s start by revisiting “The Race Between Education and Technology” (2008), co-authored with your wife, the economic historian and Nobel laureate Claudia Goldin. The book feels so relevant today.

We were building on work by the first Nobel Prize winner in economics, Jan Tinbergen, who showed that improvements in productivity and new knowledge tend to increase the demand for highly educated workers. Claudia and I documented rising inequality during much of the 19th century. The U.S. had shifted from what you might call the artisanal shop, with a lot of tacit knowledge learned by experience, to mass production. That tended to erode the value of individual craft workers’ skills. But it increased the demand for bookkeepers, managers, engineers, and skilled production workers.

“We showed that inequality narrowed during first half of the 20th century, when education kept pace with technological change.”

Then the U.S. started expanding access to education in the early 20th century through a grassroots high school movement. Access to high school allowed many more people to shift from agriculture to industry, from operative and labor positions to clerical and managerial jobs with very high economic returns on their education. We showed that inequality narrowed during first half of the 20th century, when education kept pace with technological change.

And then in the mid-20th century, as we shifted to the Information Age, the returns on a college education started rising. For a while, the U.S. did a pretty good job of expanding access to higher education. Think of state universities built in Florida or California. But that slowed down with funding cutbacks in the 1980s and ’90s.

How did economic inequality become an interest?

I grew up in Los Angeles, where I was involved as a K-12 student with some school integration programs. And my mother was a psychologist in some of our district’s less affluent schools. I just ran into so many talented people who didn’t have the same opportunities.

Trying to understand the lifelong impact of neighborhoods and schools has been a focus from the start. Going to UC Berkeley, and just being in the Bay Area, inspired me as an undergraduate to work on land-use regulation and residential segregation in California. I studied what today is called NIMBYism. If the voters in every area try to maximize their own property values without considering access for people who aren’t in their jurisdictions, the broader polity won’t build sufficient housing and ends up with highly segregated outcomes by socioeconomic status.

But by imposing some restrictions on individual communities that want to ban new multifamily housing, a state such as Massachusetts or California might end up with less segregation and huge benefits — both for people who are less advantaged and for greater diversity of interactions for more advantaged kids.

How did the economic forces of the 1980s, when you were a student, further drive your interests?

Back then, the growing gaps between the more and the less educated were just very clear. I could see it by looking at numbers from the Current Population Survey. I saw something I’ve called the “fractal” nature of rising inequality. It was rising between education groups, within education groups. It was rising across regions, with richer cities getting even richer.

So I started working on frameworks to explain what was happening. My colleague Richard Freeman, Ph.D. ’69, and I eventually came up something we called “the supply-demand-institutions framework” (1994). We wanted to understand how much was due to changing technology, how much was due to a slowdown in education, and how much was the decline of unions or the stagnant federal minimum wage. We found that about three-quarters was due to slowdowns in access to education while technology accelerated.

Another approach, which I developed with Kevin Murphy, assumed inequality was driven by supply and demand in an effort to decompose the effects of technology, trade, immigration, and education (1992). We started our project when I visited the University of Chicago in 1989. At the time, computing was not what it is today. Analyzing each single year — each month, for that matter — was a major undertaking. Putting together 25 years of microdata from the Current Population Survey, from 1963 to ’87, was no trivial matter.

You served from 1993 to ’94 as the U.S. Department of Labor’s first chief economist. How did that experience shape your thinking?

It reinforced the importance of my early work with Kevin Murphy, Richard Freeman, and Claudia Goldin. Having historical narratives, and knowing the facts, was incredibly important to understanding what policies to pursue.

I was also struck by the importance of transparent, clear evidence from randomized experiments and natural experiments, as seen in medicine and many sciences. I helped develop the Moving to Opportunity housing mobility demonstration while I was there because we wanted to do a randomized experiment to truly test the impacts of providing low-income households the opportunity to move to higher-opportunity areas.

I later co-founded, with my former student Amy Finkelstein ’95, J-PAL North America to help other social science researchers do randomized control trials — true experimental work — on the causal effects of different policies.

We were building on work that started in labor economics in the 1970s. MCRC, originally the Manpower Demonstration Research Corporation, did the first randomized control trial of a job-training program. What they showed is that, without the experiment, you got misleading estimates because of what’s known as the selection problem. That is, the types of people who enter training programs are not random. A true counterfactual is needed.

You’re known in the field as an early adopter of big data. Can you say more about its power?

If an LLM were trained on only 1,000 lines of text, it would not be very useful. The fact that it’s trained on trillions makes it very good at answering your questions. Because you need to find the digital twin of that person entering a job-training program. If you can get a rich history of their earnings, their background, and what labor markets they’re in, it’s not a true experiment but you may come close to approximating an experiment. And you can validate your findings against existing experimental estimates.

Accessing big, administrative data is something my collaborator, colleague, and former student Raj Chetty ’00, Ph.D. ’23, has really helped develop at Opportunity Insights. Over the next year, while I’m on sabbatical, Raj and I will be revisiting the question of how to combine experimental and non-experimental data to test different programs. We’re particularly focused on workforce training initiatives.

“Integrating neighborhoods has big positive gains for less-advantaged kids without negative effects on economic outcomes of the more advantaged.”

You mentioned Moving to Opportunity earlier. Two of your former students recently called it “one of the most important social science field experiments in history.” How do you think of the initiative and its legacy today?

It helped a bunch of younger scholars see the power of a large-scale randomized control trial.

Our original motivation was evidence that childhood environments really seem to matter for life outcomes. In the short run, we could see all the adult outcomes. When the families moved to a part of their city with greater opportunity, there wasn’t much of an economic return for the parents. They still didn’t have the training and education to take advantage of it.

But I eventually worked with Raj and Nathaniel Hendren to follow the kids over the long run (2016). And we found that the program really changed the trajectory for children. Growing up in a higher-resourced environment had large positive effects — 30 to 40 percent earnings effects.

Raj and Nathan’s later work repeated these findings at the national level (2018). We also find that integrating neighborhoods has big positive gains for less-advantaged kids without negative effects on economic outcomes of the more advantaged.

You’re also known for frequent appearances in the acknowledgement sections of economics papers and books. Can you speak to the place of teaching and mentorship in your career?

It’s central. Being in a place like Harvard is so amazing because we get to work with so many great undergraduates and graduate students. It’s a joy to learn from them. But you can also have a much bigger impact on science and policy by influencing their trajectory.