How I got to watch (and help a little bit) in the creation of a new discipline
Learning the Power of Incentives
I think I first understood the power of incentives when I was in my twenties and working as an unlikely salesperson for Chicago public radio. It was my job to find sponsorship from small businesses, local theater groups, etc.When you’d hear on the radio: “Support for this program is provided by… Century Tile and Carpet”–that was thanks to me. I didn’t want the job at first. Before then, I had been producing radio shows and that job was thrilling. I was just out of college and I got to decide what went on the radio–the topics, the guests.
But that job went away and the only thing available at the station was selling ads. As one friend said, it felt like going from fighter pilot to the guy who shovels scrambled-eggs in the mess tent;.
On the other hand, I quickly learned that at least in this case) background actors could make a lot of money. As producer, I was paid $36,000 a year in 1994. As an ad sales guy, I had no fixed salary. My pay was, I was told, “incentive-based.” The more money I brought in, the more money I would make. That first year I made almost $100,000, which, in 1995 was a lot of money for me. It was enough to pay off all the debt I had incurred while making very little as a producer.
I had another drive at the time, one far more powerful than money. I now know to call it a “non-financial incentive.” I hated my boss. I felt sure he had maneuvered to transfer me to ad sales as a punishment for supporting his rival at the station. I needed to succeed just to show my boss that he couldn’t hurt me.
That first year, my salary shocked my boss as much as it had surprised me. He never thought I’d become one of the highest-paid people at the station. So he designed a new incentive-based pay scheme. My commission on existing clients would fall dramatically. I would only get a big payday if I could bring in new clients who spent more than $50,000 a year on advertisements. At the time, we had only one contract that big and it seemed unlikely I would get many other ones. So my boss (I think) was sure I wouldn’t make as much money.
But, like so many salespeople, I became an obsessive expert at my incentive scheme. Since I got very little money from existing clients, I focused all my attention on getting new $50,000-contract clients. I would give prospects the most generous deals I could possibly make–all sorts of free advertising like getting them mentioned constantly during our on-air fundraisers. It worked. I made a lot of money, again.
There was one odd quirk in my incentive structure. To get the largest possible commission, I needed to have the full contract run within the station’s fiscal year, July 1 to June 31. This created a perverse incentive. Once we got to October, I stopped selling. I didn’t want to start a contract that would run from November to November because I’d get less money. So, I would show up at work late, then go for a long lunch alone (hinting I was meeting a big client) and go home early. Around April, I’d start to work again in earnest, hoping to close as many contracts as I could in early July to get that full fiscal year bonus.
It was a powerful lesson in the ways incentives work. In my first year–before my incentives changed–I was a hard worker, caring deeply about my clients. Then, my commission changed and I followed the incentives. I neglected my existing clients and brought in less money for public radio and more money for myself. If you asked me, then, what I was up to, I think I would have justified it by saying my boss was a jerk. (I should mention that, a decade or so later, he and I got over all of this and became friends.).)
That experience of seeing, first hand, how incentives can fundamentally change my actions, my values, my way of seeing my own role in the world, was profound. Later, as a business journalist, I saw incentives–perverse ones and good ones–everywhere. I remember learning about studies of physicians that showed they gave measurably different kinds of treatment to patients when their financial incentives changed. In Iraq, I saw innocent men get arrested because a neighbor was getting paid to report people as terrorists.
I fell in love with the study of economics in no small part because it is the study of incentives. It makes clear that the way people act is not always in line with their stated values but is almost always in line with their incentives.
Getting Serious About Shaping Incentives
About ten months ago, I got an email from Jana Gallus, an economist and professor at UCLA Anderson School of Management, asking me to chat about her new initiative on Incentive Design. Over a thrilling phone call, I learned she was putting together a unique conference on incentive design, looking for new ways to work across disciplines.
I had never considered the challenge she identified: can we not just study incentives–as if incentives are some fixed entity–but can we actually get better at shaping them to achieve better outcomes.
Jana had been wondering about this for some time. But she had a new urgency because of an email she had received a few months earlier.
At first she believed this email, purported to have come from Sytse “Sid” Sijbrandij (billionaire CEO of Gitlab), was actually a scam. It praised Gallus’s work, mentioning by name a new class she had designed. Then it suggested Gallus click on a few links to learn about Sijbrandij’s goal of using her ideas—and those of others—to transform how people understand incentives. The email had a big promise—Sijbrandij would fund the development of Gallus’s work into a new field of study: Incentive Design.
It was all so perfectly appealing that Gallus felt certain some clever scammer had programmed an AI bot to scour academic journals, identify targets, and send emails promising to fund their research. No doubt, there would be some sort of “application fee” and a series of grand promises. Gallus immediately shared the email with UCLA’s IT department so they could warn others.
She never imagined that this email was, in fact, real; that Sid Sibrandij, having made his fortune, had begun researching ways to improve the world. This research had led him to what he saw as a huge hole in scholarship: the study of incentive design from an interdisciplinary perspective that could be applied to pressing social problems. And so he wrote Gallus that very-much-not-a-scam email. It would take some real effort on Sijbrandij’s part to persuade Gallus that his offer was a serious one, and that the two should collaborate.
What Is Incentive Design? And Why Could It Be So Transformative?
Think of areas of society that need improvement. Too many people are sick or poor or unhoused or uneducated. Communities have polluted water or children who are hungry. Typically, change in those areas is yoked to financial incentives and disincentives. Governments or charities pay money to achieve an outcome; they fund schools or pay for food or housing. Governments outlaw and regulate negative behavior, typically by fining those who, say, pollute rivers.
Gallus’s and Sijbrandij’s premise was that any outcome—positive or negative—is not just the result of financial incentives or disincentives. There are a huge range of non-financial incentives that powerfully shape how people and systems behave. Some might be incentivized by fame or glory or friendship. Others might be motivated by a sense of belonging, purpose, or moral duty. Some might respond to social pressure or the desire for status within their community. Others might be driven by curiosity, the challenge of solving a problem, or the satisfaction of mastery.
To understand and improve the functioning of systems––how good behavior is encouraged and bad behavior discouraged––we need to adopt a more holistic perspective on incentive structures. One that looks at financial as well as non-financial incentives. Achieving that perspective naturally requires bringing different disciplines together.
Assembling the Team
Gallus and Sijbrandij began to dream of a new discipline that would study such incentive structures, pinpoint misguided incentives, and develop tools that could transform those incentive structures to create more socially beneficial outcomes.
So, who to involve from other disciplines? Economists are, of course, obsessed with incentives. But economists tend to be most interested in financial incentives for individuals. Gallus understood that people from other disciplines could provide crucial insight, as she had experienced in her own interdisciplinary collaborations. Psychologists, like economists, tend to look at individuals but focus on deeper, non-financial drives. Sociologists study systems and have different analytical tools to uncover the ways systems drive group human behavior. Anthropologists help us understand issues like social status, culture, and the long-term evolution of cooperation and competition. Legal scholars provide perspective on how legal frameworks create and enforce incentives. Computer scientists bring in expertise designing algorithms and systems that optimize incentives for desired behaviors. Then, too, are the non-academics who are on the front lines of activism, analysis, and news reporting.
Gallus and Sijbrandij decided to create a conference that would bring together leading thinkers from these and other disciplines to identify areas of common ground and find the kind of new insights that come when people with different training look at the same problem and unpack it using their varying perspectives. The goal: to drive meaningful social change by identifying big issues that have not yet been fully studied.
As Gallus began talking with academics across many disciplines, she soon saw some challenges to collaboration. For a start: words. Economists use the word “incentive,” Psychologists are likely to talk about “drives,” and sociologists work with “social norms” or “institutions.” Anthropologists often speak of “culture”, “norms”, and “reciprocity,” while education experts tend to refer to “motivation.” All these words speak of the same—or, at least, quite similar—basic concept. They all address what people want. But the different vocabulary makes it hard for, say, a sociologist to pick up and make sense of an economist’s journal article. Another challenge: many academics simply don’t want to engage with other disciplines. Sociologists and economists, in particular, can be quite dismissive of each others’ techniques and core concepts (one economist told Gallus that if she uttered the word “incentives,” sociologists will “vomit and run away.”)
Another challenge to cross-disciplinary collaboration is the incentives that academics face to publish in the most elite journals in their own field. Sociologists don’t get much credit (and can even get some shade) when they publish in economics journals and vice versa. There are few multidisciplinary journals that bestow a more generalized benefit on lots of disciplines.
Finding the right participants took months. Gallus brought on Sandy Campbell, who was finishing a PhD program in Management of Organizations at UC-Berkeley (Campbell is now Gallus’s postdoctoral research fellow). This is also when she reached out to me. After many years at NPR’s Planet Money and as economics writer for The New York Times Magazine and The New Yorker, I became quite friendly with a lot of academics (and quite unfriendly with a few).
Gallus set some ambitious goals for the conference:
- It would be small enough so that everyone would be able to talk with everyone else, hear others’ perspectives and ideas, and discuss them as a group.
- It would be large enough to have a broad range of backgrounds and opinions.
- People would be respectful and open to conversation with those who see the world quite differently.
- People would not be so polarized that little forward-momentum would be achieved.
- People would come ‘unarmed’––they would not be doing presentations with slides about past work; instead, the time together would be focused on finding new ideas, new avenues of discovery.
- Lastly, everything would focus on how to achieve major societal impact. No small tweaks.
In putting the conference together, we would spend hours talking about this scholar or that one. They all needed to be smart and creative thinkers with something to add. But we’d also dig into their personalities. Was this one open to challenging thoughts from others? If we invited this other one, would it tilt the balance and alienate some third group of people? It was like we were planning the perfect dinner party. We needed people who were good on their own but also able to develop ideas in groups. We needed a strong cross-section of thinkers, but we also wanted doers: activists, policy-makers, and others who are involved in seeking better outcomes in housing, healthcare, and education.
Gathering in Chicago
It Could Have Been a Disaster.
We gathered on a windy (natch) Thursday at a hotel in Chicago. One by one, folks walked into a long, narrow conference room with a set of tables. As each new person walked in, you could feel a sort of nervous excitement. “I’m so-and-so, I study such-and-such, and… I’m excited to see what happens.”
Have you ever been to an academic conference? I have been to many (though nowhere near as many as these academics) and, if I am brutally honest, they are pretty boring, especially for an outsider. Typically, the themes and topics were decided long ago by some committee. Conference speakers summarize papers or, worse, just read them out loud in a monotone. A typical panel is just three to five people reiterating what they already wrote in a paper. Spontaneous, surprising new ideas are rare. For someone not in the discipline, the discussions are often completely incomprehensible, using technical jargon to argue over fine points that nobody outside of a few academic departments understands or cares about.
This one was completely different. There were no prepared papers; no printed documents to read from… We were broken up into groups, given a broad topic area to think about, and then we had to … talk. Debate. Argue. Agree. Disagree. Come up with a few shared conclusions.
Gallus was excited and nervous before it all began. What if people didn’t engage? What if the open-ended sessions were… disastrous: just bored people looking at their phones or experts yelling at each other or talking past one another?
Within minutes, though, I could see she had nothing to worry about. Group members began lively, exciting, delightful discussions. Economists and psychologists and evolutionary anthropologists bounced ideas against each other, coming up with new formulations.
Then we all came together, shared each group’s findings, and talked and argued and discussed some more. It was lively. It was surprising. It was sometimes confusing and other times brought major new ideas to light.
Over the four days, the sessions moved from broad discussion to focused strategy geared towards identifying major research questions and possible solutions.
Throughlines
It became clear that there were a handful of ideas that were relevant to each of our areas of discussion–health, education, and housing. These became the throughlines that the group wanted to explore in greater depth.
Misaligned Incentives
One of the most powerful areas of discussion was around misaligned incentives. These can be financial or non-financial, deliberately designed or accidentally emergent, and they occur when –– an incentive for one kind of activity inadvertently hurts a stakeholder group, organization, or society at large. I think of the commission structure I had as a public radio salesperson. My bosses wanted to incentivize me to make more money for the station, so they created a contract that actually incentivized me to barely work at all for half the year. I no doubt brought in much less money than I would have with a different incentive structure. The contract was poorly aligned with their actual goal.
You can see this everywhere.
At the conference, we discussed the strange fact that doctors—through the American Medical Association and other institutions—are able to control the number of physicians practicing in the U.S. It’s in the interest of nearly every citizen to have far more doctors. But the group that has control over the supply—doctors, themselves—has an incentive to limit it (the fewer the doctors, the more they can charge for services).
I thought of an old friend of mine who had become a heroin addict. He was treated at a methadone clinic that was supposed to wean him off drugs. Instead, the clinic did things like give him extra doses of methadone for the weekend, knowing that it was common for addicts to sell the extra and buy heroin with the proceeds. Then he’d fail his drug test and be given even more methadone. Why weren’t they doing more to help wean him off the drug? Simple, he said: They made their money by having a lot of patients taking a lot of methadone. When addicts recovered, they lost money.
Misaligned incentives were even right there, in the conference. Academics are incentivized to write peer-reviewed journal articles. That is how they get tenure or get the respect of their peers. The journal articles that offer the best rewards are, typically, focused on fairly narrow, technical areas that are of interest within a discipline. There is little incentive, within the academy, to do research that will lead to powerful outcomes in the real world. There’s another misalignment, too. At a conference with other academics, many want to look smart, so they talk about things they already know well and they focus on topics they are deeply familiar with. This doesn’t help build radical new ideas. (Similarly, if someone else says something smart, there’s an incentive to trash it, which can make you seem even smarter.)
I think even at an event like the conference, there’s a misalignment. You want to look smart. How do you do that? Talk about what you already know, get the conference to turn to your area of expertise. Will that trigger radical new ideas? Nope. But it’ll make you look good. (Same for trashing others’ ideas, by the way. You look smarter if you’re critical)
There’s nothing inherently wrong with narrow technical academic papers or with well-understood solutions. But they do leave a lot unexamined. The frontier solutions—the new ideas that could transform outcomes—could be radical ones that are perceived as too unsophisticated by journal editors, who want to see theory, math, and modeling. And the ideas might be too controversial for key stakeholder groups, such as funders. Perhaps, the group thought, that would be a key area of focus: those solutions that are largely ignored by the current incentive structure.
A stronger understanding of how incentives come to be, how they interplay, and how they can be shaped in any one of these domains would, doubtless, assist in the others—as well as countless areas of societal concern.
Research to Impact
At one point, a researcher drew a simple table:
| Interesting to practitioners | Not interesting to practitioners | |
| Interesting to academics | Where a lot of the research is, now. | |
| Not interesting to academics | Where the research should be. |
He pointed out that there are a lot of incentives to focus research on areas that are interesting to academics and also to activists. These would be research areas that are provocative and cutting-edge enough to get interest in journals and help academics get tenure or other plaudits. At the same time, these would be areas that activists find impactful and easy to fund because they are attractive to foundations and individual donors.
There are a bunch of areas that might interest academics but activists would find too esoteric or abstract—and areas that activists find important but academics see as trivial.
Paradoxically, the area that needs the most attention might be the space that neither group is attracted to. The truly break-through new ideas are likely to lie in that most-ignored quadrant. It also—and here’s a good incentive—is the area that is least crowded with existing work, where researchers and practitioners can carve new pathways.
Public Communication–There’s No Impact Without It
My own passion during the conference was to focus on how these ideas could be presented to the public, to activists, to politicians, and others, with clarity and force. I’ve seen so many smart ideas that never jump from the academy or from some nonprofit organization to wider adoption because the findings are not presented in a clear, compelling, impactful way.
That is the work I’ve tried to do throughout my career and I argued that powerful public communication must be at the heart of however this incentive design field develops. If it truly seeks to change minds, reshape how people understand social outcomes, transform academia and activism as well as influencing politics and regulation and healthcare and education, it needs to more deeply understand how these different stakeholders consume and process information. Most people share the incentive of wanting better outcomes in health, education, and housing, but they find the systems surrounding them to be so confusing and opaque, they have no idea how they can convey their incentives to those who make decisions. There is a world of actors, writers, producers, and others who could be harnessed to communicate crucial ideas with humor, poignancy, joy.
Are Housing, Healthcare, and Education All One Thing?
Usually research and activism focuses on one area, such as housing, healthcare, and education. But in our conversations during the conference, it became clear that these really are far too intertwined to be neatly separated. People in subpar housing or no housing at all suffer disproportionate health impacts and struggle in school. People who face debilitating illness or expensive healthcare can easily lose access to housing. Our system uses education as a way to serve food and offer nursing services to many under-housed, hungry, and under-treated young people. But the way we fund and regulate each of these sectors creates incentives to not see them through a unified lens.
One conference doesn’t solve the world’s problems. But this one conference made us all feel that our path is a good one. We are onto something here. We are (only just!) beginning to understand how our systems work together and how we can make them work better.
It was exciting to feel that we are, just maybe, at the start of a new approach, a new way of thinking about seemingly unsolvable social ills. If we can identify and disentangle the incentives that have led us to our current state, and if we build a set of tools that allow us to shift incentives in a more positive direction, we might just be able to make transformative changes that have, so far, felt stubbornly impossible.
