The Sentencing Matters Substack Interview: Economist and Executive Vice President of Criminal Justice at Arnold Ventures, Jennifer Doleac
How data holds the key to safer communities at lower cost
Here in Washington, it’s a rare policy pitch that doesn’t claim to be “evidence based.” But those claims often don’t hold up to even the smallest bit of scrutiny. Behavioral psychologists like Richard Thaler, Daniel Kahneman, and Robert Shiller have shown that for most of us, public problems we face beget policy conclusions first, driven by ideology, politics, and an unconscious mental shortcut to get us to an answer. Only then, once we have that gut response, comes the search for information and data that align with our preexisting beliefs.
But Jennifer Doleac is a reminder that there is still reason to believe in genuinely thoughtful and rigorous policymaking driven by research and data rather than ideology and politics; and still people like her who practice it. Doleac is an economist and Executive Vice President of Criminal Justice at Arnold Ventures. She spent over a decade as an economics professor, and fortunately for all of us, her research has been in large measure focused on how to lower crime at lower human and fiscal cost. In her recent book, The Science of Second Chances: A Revolution in Criminal Justice, Doleac shows how coming to a crime problem with an open mind and a commitment to testing different hypotheses and approaches to address it can lead to remarkable results and effective and less expensive public policy. She follows the data, even when it leads to uncomfortable conclusions. If you haven’t seen it, here is a review of Doleac’s book we published a few weeks ago on our Substack —
These days, you can spot someone with integrity working on public policy rather easily: it’s the person who both the left and right respect, and someone who both the left and right are angry with. Doleac is one of these people. She recently announced she will be leaving Arnold Ventures this summer. I spoke with her in late June, and we discussed her decision and much more.
The interview has been edited for length and clarity.
- - -
Jonathan: Thanks so much for speaking with me today. Could we start with just a little bit about you and where you grew up, what that time of life was like, and where you went to college?
Jennifer: Yeah, so I grew up in Massachusetts, about 40 minutes outside of Boston, in the suburbs. I felt like I had a pretty normal childhood. I have one sister, and she also lives in Texas now, somehow. I went to Williams College.
Jonathan: Why Williams?
Jennifer: It’s a small school in the middle of nowhere. My hunch is a lot of people have this experience where you do a bunch of college shopping, and some place just feels like it fits. You can’t quite articulate what it is. I’ve come to think of the college selection process as making your first big choice about your future. It doesn’t really matter what the choice is in the end; it’s just the fact that you’ve made it. I did a tour of Williams, and I just kind of fell in love. I really enjoyed my time there.
Jonathan: Did you study economics at Williams, and why economics?
Jennifer: Yeah, economics and math. I went in knowing I wanted to major in math. I’d always loved math growing up. What I did not realize at the time was that math in college tends to be a whole lot more letters than numbers. I loved math with numbers. It becomes more theoretical the further you go, and I actually really love the more applied, quantitative stuff. I took a couple of computer science classes, and took some econ classes, and wound up following econ farther and farther, and that’s what I really fell in love with in the end.
Williams had these tutorial classes where it’s two or three students and the professor, and you basically read a paper every week and write responses to it. I took a class with the then-college president, who was an economics professor, and we read a bunch of economics studies about higher education and the impact of different types of policies, including some of his papers.
It was the first time I realized that professors were researchers; like that’s a job you can have, to study this stuff.
I wound up talking to him about how one gets a job like that, and he’s like, well, you need a PhD in economics. I was like, okay, how do I do that? I wound up getting a job as a research assistant at Brookings in D.C. for a couple years after college. I then looked around and realized everyone who had a job I really admired and thought was really cool had a PhD in economics. So, I realized I probably needed to go back to school and do that.
Jonathan: You studied for your PhD at Stanford. What was that like, and how did you end up focusing on criminal justice issues?
Jennifer: I went in not expecting to study crime or criminal justice. It really hadn’t crossed my mind. I knew I loved economics. I figured I would study the economics of education. That was the world I really knew. But the practice of actually being an econ PhD student is you’re hunting around for a natural experiment, or an opportunity to run your own experiment. You’re looking for data that you can use to actually measure the impact of some policy or program that is interesting and can tell you something about human behavior.
At some point, I read a newspaper article about law enforcement DNA databases. The point of the article was that every state has its own rules about which groups of offenders are required to provide DNA to the database, and that means you can have states that are right next to each other where someone convicted of robbery goes in the database in one state and the next state over they don’t. The article suggested that was crazy and random and bad policy. I read it as an econ PhD student thinking it might be bad policy, but it’s great for research. It was a beautiful, natural experiment. So, I wound up looking at the effects of DNA databases on crime and criminal behavior.
Jonathan: And that fascination with crime continued?
Jennifer: I kept studying crime over and over again, because there are so many interesting questions. I think at this point, the most useful and most innovative work in this space is coming out of economics departments.
I gave a book talk recently, and a lawyer asked me “how do you think about normative values and the role of people’s rights when you’re advising what to do and what not to do?” I looked at him and I was like, I don’t think about that. That is not my role. That is your role as a lawyer. An economist’s approach is very consequentialist. What’s the impact of this on the ground. Then, we can discuss as a community whether we want that in light of all the various trade-offs we might be facing. It’s just a different lens to look at these problems.
Jonathan: When I was in college, I read a book by Richard Posner. It was on law and economics. Posner argued in favor of the use of the death penalty because of its efficient use of society’s resources. It seemed off to me that he wasn’t grappling with any moral issues.
Jennifer: What I have found most useful is to name the moral issues; name the elephant in the room and recognize that the policymaker has to consider lots of things when making a decision.
The piece I can help you with is this one piece. You might also be thinking about these other pieces. Take sentencing, for example? Why do we put people in prison?
When I’m in a room with policymakers, I point out that there are four broad reasons we lock people up. One is deterrence. We hope doing so will keep them from offending again. One is incapacitation, physically removing you from the community. The third is rehabilitation; it’s an opportunity to intervene. And the fourth is retribution; you’ve done something bad and you deserve to suffer for it. As an economist, I can tell you a lot of useful things about the first three of those. Those are all empirical questions. Do you get deterrence? Do you get incapacitation? Do you get rehabilitation? The fourth is a moral concern, and each person in the room might have different preferences about how much retribution or mercy someone deserves.
As an economist, I think it is fine that we all have different preferences, and data cannot tell you whether you’re right and I’m wrong, or vice versa. Everyone might come to different conclusions about how long a sentence should be. But we all would prefer to have less crime for less money. Whatever you think about retribution, you probably also care about what the impact on public safety is and so let’s nail down what that is.
Jonathan: While we’re talking about sentencing, do you have any thoughts about sentencing guidelines and whether they’re a good idea or a bad idea?
Jennifer: I don’t know that I have strong opinions about them as a concept. I think we generally swing back and forth as a society between how much discretion to give different criminal justice actors. Sentencing guidelines are one way we operationalize that. I will say that sentencing guidelines are great for research. They set up nice, natural experiments where people who are pretty similar, but some got one point higher on the guidelines and suddenly they get more years in prison.
Jonathan: Can we talk a little bit about Arnold Ventures (AV)? From the outside, it looked like you were thriving as a tenured economics professor before going to Arnold. What happened?
Jennifer: I got into the research game because I viewed it as the highest impact way I could help make policy better. It turned out being a junior faculty member was a really great place to build your expertise. You have all this freedom. But what I found in academia was that the more senior you become, the harder it is to do that externally-facing work, because the university is structured in a way that once you get tenure, it’s like, congratulations, you are now an expert, so your job is to serve on university committees and teach all the classes so that your junior colleagues can become experts.
I felt I was an expert now; I want to use this for something. It was getting more and more difficult to do that. I went to a couple conferences with some publicly facing academics. I asked them all, why are you still at a university? They each had compelling reasons for themselves that didn’t apply to my situation. So I knew it was time to leave. I then posted something on Twitter about how I’d realized academia was no longer the right fit for me. I think two days later, John Arnold reached out and said we’re looking for someone to lead our criminal justice program, any chance you’d be open to a conversation? The stars aligned. So, I took this crazy leap. Everyone thought I was crazy. I thought I was a little bit crazy. But I was ready for the new adventure.
Jonathan: Can you talk a little bit about AV, its role, and the role of other philanthropies in furthering best practices in criminal justice. I’m especially interested in your thoughts about how philanthropy and DOJ — and in particular the Office of Justice Programs — live side by side in promoting research and best practices?
Jennifer: I honestly had not thought much about this before taking this job. But now, three years later, I have a lot of thoughts about the role of philanthropy. What I knew about AV going in was that it was focused on evidence-based policy. It was originally founded by Laura and John Arnold as the Laura and John Arnold Foundation, and it funded only academic research to measure impacts of different interventions in the real world. Criminal justice was one of its earliest areas of focus.
Several years ago, they added policy and advocacy components and became Arnold Ventures. And the reason they did that is they looked around and realized we funded all this amazing research, and none of it has turned into policy and impact on the ground. They added policy and advocacy arms to become the lobby for evidence. My role was to focus on modern research standards to find better solutions to public safety problems, trying to double down on strong causal research, measuring, testing things on the ground.
The way I’ve come to think about the role of philanthropy is almost like the seed funder on the business side. Philanthropy is most adept at taking risks; trying lots of new things, pivoting quickly. And just as in the finance world, as things become more proven, once we have a sense of which of these thousand crazy ideas actually work, then you can think about bigger investments over time from a broader set of philanthropies, perhaps, or maybe government. Like an IPO when a company goes public, we hand over this idea to government and have government fund it at scale.
Jonathan: Can you talk a little bit more about what you call the risky investments?
Jennifer: John Arnold talks a lot about a two-by-two matrix of funding opportunities. On one axis, you’ve got impact, low or high impact, and on the other axis, you’ve got probability of success, low or high probability of success. In the first quadrant, you’ve got low probability of low impact. Nobody wants to fund that. The opposite quadrant — high probability, high impact — everyone wants to fund those things. But those are unicorns; they don’t come along very often. And so that leaves two quadrants left that philanthropies can go after. One is the high probability, low impact. Because there is a tendency to want to rack up quick wins, there’s an overabundance of philanthropic dollars going to the high probability, low-impact stuff.
Where there’s an undersupply of investment is the low probability, high impact ideas; the moonshots. John was always encouraging us to try to look for the moonshots. It can be really hard to tell up front which those are. So, we have been trying lots of stuff with a failing-fast mentality. One of the projects we’ve been developing over the last couple of years is looking at the effects of GLP-1s on substance use disorder and ultimately time in jail. There’s lots of anecdotal evidence that GLP-1s seem to have this impact; they reduce cravings for drugs and alcohol, but we don’t have the rigorous evidence yet. There are some ongoing clinical trials, but the question of the impact on the ground is quite different. There are lots of reasons this might not work. But if it does, it would be a game-changer in the public safety space. Finding projects like that has been really fun and rewarding.
Jonathan: There’s some recent research that seems to show that young people today are committing crime at a far lower rate than earlier generations. We now have the lowest homicide rate in a generation, and there are some researchers who are predicting that we’re not done with this crime decrease. It may be coming from our obsession with cell phones or something else. What do you think of this research?
Jennifer: After the most recent uptick in crime we had around COVID, we just went back to the previous downward trend in crime that is longstanding. So, a bettet question really is why did crime go up during COVID, and why did that uptick last so long? This makes me think the decline is due to something other than just the phones. There are lots of compelling reasons that young people are not out doing risky stuff as much as earlier. My hunch is it’s a whole bunch of stuff, including some of the crime control policies we put into effect, but also, we took lead out of gasoline and out of paint. That’s going to have bigger and bigger effects over time as lead continues to dissipate. There’s just all of these things; life has been getting better, and we’ve been doing lots of things that create progress.
I think all of the investments we’ve been making in society to try to give people more economic opportunity and make them healthier can have public safety payoffs that we just see accumulate over generations. That doesn’t mean crime isn’t still a problem. It’s certainly less of a problem than it used to be, and that’s wonderful. But there are still many communities where crime is THE problem, and is THE biggest concern of parents who are trying to raise children and want to make sure that they live into adulthood, much less be able to thrive in other ways.
I think one of the ways in which our public safety conversations kind of went off the rails in the early 2020s is people lost sight of how costly crime is. We were focused on the harms of the system, and those are important. But the system exists for a reason, which is to reduce crime, because crime is so costly. So, we have to think about both of those things.
Jonathan: You announced recently that you will soon be leaving Arnold Ventures. What’s coming next for you?
Jennifer: I’ve been at Arnold Ventures for three years; it’s been an amazing three years, I’ve learned so much. I’ve been really inspired by the momentum we’ve been able to build. I have been surprised — very pleasantly surprised — by how much appetite there is on the ground, especially among state lawmakers and state practitioners, for evidence-based policy. They just need help. They want to know what the evidence is; they want that evidence to guide policy. Arnold Ventures is basically the only major funder in this space. Federal government funding has been cut. But it’s crazy that for such an important problem, there’s only one major funder that is investing in finding new solutions, measuring the impacts of those solutions, and scaling them.
My goal in the coming years is to try to bring other donors into this space and make this pitch for an ROI-focused public safety agenda. It is different from previous conversations and previous pitches around the criminal justice system which have been more focused on social justice — which is wonderful. But it’s not what my motivation is here. I think public safety is really important, and we need to be smart about measuring ROI, figuring out what the most high-impact solutions are, and scaling them. That’s what I’ve been doing at AV. We have a moment here where there are a lot of people out there with a lot of money and I think a lot of opportunity to try to pull some of them into this space and use some of that money to make our communities safer. So, I’m going to try to do that.
Jonathan: We’ll be watching very closely. Last question. If you could change or institute two criminal justice policies that you think would bring more public safety and more justice to communities, what would those two things be?
Jennifer: Well, one would be to move a really big share of our prison budgets to investments in police investigations; to focus more on solving crimes faster, rather than making sentences longer. We know from tons of evidence that that is the best way to deter and prevent crime, and we, at this point, have dramatically under-invested in figuring out how to increase the probability of getting caught. So, let’s solve more crimes faster.
The other one would be, if I had a magic wand, I would embed some sort of experiment or quasi-experiment into every federal grant program that trickles down money to all the state law enforcement agencies. I think it’s just a huge, missed opportunity that the federal government has all of these COPS grants, for instance, all of these amazing funding programs that fund local and state law enforcement and we’re not learning from them as fast as we could be. We’re lucky if we learn what works and what doesn’t. But we could be systematic and have a more structured way of embedding experiments into those different funding streams, so we have more coin flips built in that tell us what new technology or what new program worked and what didn’t.
Jonathan: Thank you so much, Jennifer, for speaking with me today. It’s been fascinating to learn about your work. We’ll be watching closely for your new venture. Thank you for your service to the country
Jennifer: Awesome. This was fun. Thanks.
Great interview!
Jennifer Doleac (Research Karen) repeatedly distinguishes between empirical and moral questions. Economists, in her telling, can tell us what works; lawyers and philosophers can tell us what is just. It sounds intellectually modest, but the distinction is not nearly as clean as she suggests. In criminal justice, deciding what to measure is itself a moral decision.
Why measure recidivism rather than dignity? Why prioritize cost-effectiveness over democratic legitimacy? Why define success as reduced crime rather than stronger communities, civic participation, or repairing historical injustice? Every research agenda makes judgments about what matters before the first dataset is assembled. There is no value-free criminal justice research.
That matters because Doleac’s framework privileges institutional forms of knowledge over lived experience. Natural experiments, causal inference, and randomized evaluations are valuable, but justice is not simply a laboratory. Someone who has lived inside a prison understands humiliation, coercion, informal power, and human adaptation in ways no administrative dataset can fully capture. A family separated by incarceration possesses knowledge about punishment that may never appear in a recidivism statistic.
The question is not whether quantitative research matters. It does. The question is who gets to produce knowledge and whose knowledge counts. Formerly incarcerated people should not simply be research subjects whose experiences are converted into data for experts to interpret. They should be recognized as producers of knowledge themselves. The hierarchy in which economists produce knowledge while impacted communities provide the raw material is not neutral. It is a political choice.
The same problem applies to evidence. Doleac seems to believe better evidence can discipline politics and move policymakers toward better decisions. Of course it can. But evidence never speaks for itself. The same study can support expanded policing, investment in housing, behavioral-health treatment, or less incarceration depending on what one believes government should accomplish. Evidence can tell us whether an intervention produces an outcome; it cannot tell us which outcomes a democratic society should value most. Evidence informs decisions; values ultimately determine them.
This becomes particularly important when Doleac invokes return on investment. ROI can be useful when governments and philanthropies allocate scarce resources. But it becomes dangerous when it becomes the dominant language for thinking about justice. Democratic societies protect rights even when doing so is inefficient because those rights express values that transcend economic calculation. We do not defend freedom of speech because it saves money. We do not measure due process by how efficiently it processes defendants. Justice is a constitutional and moral commitment before it is an economic calculation.
Perhaps the most significant omission in this framework, however, is power. Evidence-based policymaking often assumes policymakers lack information and therefore need better research. But American history suggests that ignorance is frequently not the problem. We have known for decades about the harms of solitary confinement, the unequal consequences of mandatory minimums, the criminalization of poverty through cash bail, and the destabilizing effects of mass incarceration.
The obstacle has often been less a shortage of evidence than a shortage of political courage. Evidence does not automatically defeat entrenched interests or redistribute power. Institutions can ignore compelling research when acting on it threatens those who benefit from the status quo. In criminal justice, power, not ignorance, is often the obstacle.
That is also why Doleac’s description of philanthropy as venture capital for public policy deserves scrutiny. The analogy suggests experimentation and innovation, but it raises a democratic question: What happens when private foundations and unelected billionaires become major agenda setters for public policy?
Foundations can determine which questions receive attention, which interventions are tested, which organizations receive resources, and which ideas enter the policy conversation. Even with the best intentions, private wealth can exercise public power without the accountability that accompanies democratic authority. The question is not simply whether Arnold Ventures funds good research. It is whether private wealth should possess so much influence over the direction of public justice in the first place.
Doleac’s distinction between moral and empirical questions also breaks down when she discusses punishment. She treats retribution as primarily moral while deterrence, incapacitation, and rehabilitation can be examined empirically. But the way a society understands punishment shapes empirical outcomes. Punishment that communicates accountability and legitimacy may have very different consequences from punishment that communicates exclusion and humiliation. The moral and empirical are not separate worlds; they constantly interact.
This is where institutional legitimacy matters. People do not obey laws simply because they fear punishment. They also comply because they believe institutions are legitimate and that they have been treated fairly. Legitimacy may be harder to quantify than recidivism or incarceration rates, but that does not make it less consequential. It may be one of the most important ingredients of public safety.
And this points to the larger limitation of Doleac’s framework: it leaves too little room for transformation. The interview approaches criminal justice primarily as a system to be made more effective, efficient, and evidence-based. But another tradition asks a more fundamental question: Why have we assigned prisons and jails the burden of addressing problems created by failures in education, housing, employment, behavioral health, family stability, and economic opportunity?
The irony is that Doleac is arguing against ideology while advancing an ideology of her own: technocratic consequentialism, the belief that rigorous empirical measurement should occupy the center of policymaking and that interventions should be judged primarily by demonstrable outcomes.
There is much to admire in that approach. It challenges dogma, demands accountability, and asks policymakers to distinguish what we hope will work from what actually works. Criminal justice has suffered enormously from policies based on fear, intuition, and assumptions that were never tested.
But empirical rigor becomes a problem when it is treated as a substitute for democratic judgment rather than a tool for improving it.
Data can tell us what works. They cannot tell us what is worth working toward.
That question belongs not only to economists, lawyers, and policymakers, but also to philosophers, historians, victims, families, communities, and people who have survived the criminal legal system. The strongest public safety policy does not choose between evidence and values. It brings them into conversation: empirical evidence should discipline our assumptions, while democratic values determine the ends toward which we work.
That is not an argument against evidence. It is an argument against pretending that evidence can answer questions that only democracy can legitimately decide.