Economics for Inclusive Prosperity

Zero-sum thinking's profound effects on economic policy

Episode Notes

Optimists see the glass as half full. Pessimists see it as half empty. Zero-sum thinkers believe that if someone else’s glass is full, that water was more than likely taken from their glass. Zero-sum thinking is a primary field of study for our guest on this episode of the Economics for Inclusive Prosperity podcast: Economist Stefanie Stantcheva. She’s the Nathanial Ropes Professor of Political Economy at Harvard University and the 2025 recipient of the prestigious John Bates Clark Medal, which is awarded to a young economist who has made a significant contribution to the field. Through her groundbreaking survey research at Harvard's Social Economics Lab, Stantcheva has shown that the zero-sum mindset has profound implications for politics, policy, and polarization and that it’s deeply rooted in our lived experiences, our family histories, and even the struggles of ancestors we’ve never met. She and her research partners have also found that zero-sum thinking cuts across party lines in surprising ways, and that it has strong predictive power in policymaking. In this episode, she talks to EfIP Podcast host Ralph Ranalli about why understanding the implications of zero-sum thinking is crucial to economists trying to design policies that are politically viable. 

Stefanie Stantcheva is the Nathaniel Ropes Professor of Political Economy at Harvard and founder and the director of the Social Economics Lab. She studies the taxation of firms and individuals, as well as how people understand, perceive, and form their attitudes towards economic issues and policies. Her recent work explores people’s attitudes towards taxation, trade, immigration, climate change, and social mobility using large-scale social economics surveys and experiments. Stantcheva was the recipient of the 2025 John Bates Clark Medal, awarded each year to an American economist under the age of forty who is judged to have made the most significant contribution to economic thought and knowledge, and the 2020 Elaine Bennett Research Prize. She has also been named a Sloan Foundation fellow, a Guggenheim fellow, and an Andrew Carnegie Fellow. She received her Ph.D. in Economics from MIT in 2014 and was a junior fellow at the Harvard Society of Fellows 2014-2016 before joining the Harvard Department of Economics in July 2016. She is currently co-editor of the Quarterly Journal of Economics.

Follow Stefanie Stantcheva:

LinkedIn: https://www.linkedin.com/in/stefanie-stantcheva-5543581a3/

Bluesky: https://bsky.app/profile/s-stantcheva.bsky.social

X: https://x.com/S_Stantcheva

Episode Transcription

Preroll: (Music)

(Dani Rodrik) economics has changed and in particular has become much more empirical. It's become less wedded to theoretical preconceptions about, such as, markets will always take care of problems or that the governments cannot ever solve things… 

(Stefanie Stantcheca) What we do find in our survey work is that fairness matters a lot to people… 

(Suresh Naidu) …that inclusive prosperity meant thinking about democracy and its relationship to economic policy and democratic politics as something that we were committed to…

(Atif Mian) …those deeper questions of how we want to organize ourselves 

as a society, so we deliver not just economic growth, but we also deliver the values that we need to aspire to. 

Intro (Ralph Ranalli): Hi. It’s Ralph Ranalli. Optimists see the glass as half full. Pessimists see it as half empty. Zero-sum thinkers believe that if someone else’s glass is full, that water more than likely was taken from their glass. Or to put it another way, if a foreign student enrolls at a U.S. university, they must be taking a spot from an American. Or that a person of color who benefitted from a DEI program took something away from a white person in the process. Or that the riches of the wealthy were invariably gained at the expense of the poor. 

 

Zero-sum thinking is a primary field of study for our guest today here on the Economics for Inclusive Prosperity podcast: Harvard economist Stefanie Stantcheva. She’s the Nathanial Ropes Professor of Political Economy and the 2025 recipient of the prestigious John Bates Clark Medal, which is awarded to a young economist who has made a significant contribution to the field. Through her groundbreaking survey research at Harvard's Social Economics Lab, Stantcheva has shown that the zero-sum mindset has profound implications for politics, policy, polarization, and division and that it’s deeply rooted in our lived experiences, our family histories, and even the struggles of ancestors we’ve never met. 

 

She and her research partners have also found that zero-sum thinking cuts across party lines in surprising ways, and that is has strong predictive power that economists need to understand when they go to design policies that are politically viable. 

 

Let's get started.

 

Ralph Ranalli: So, Stephanie, welcome. It's really nice to have you here on the Economics for Inclusive Prosperity Podcast. I've really been looking forward to our conversation because I think your research on zero-sum mindsets is very timely, very important. But also intriguing, because there seems to be a lot more to it than meets the eye.

 

Now, I think to start, most people have a sense of what zero-sum thinking is in a general way. Which is basically that if you get more of something, it must mean I am getting less. But can you talk about how you define it specifically in your field, which is economics? What is it, and maybe what are contrasting alternative mindsets, in other words, what's the opposite of zero-sum thinking?

 

Stefanie Stantcheva: Yeah, it's great. And I'm also delighted to be here and chat with you. So, zero-sum thinking is exactly as you say—this belief that if you or an individual or a group of individuals wins, it must come at the expense of someone else. So it's this belief that the world is like a fixed pie. If you get a larger slice, I must be getting a smaller slice. And, um, in economics, we think of situations, where many different groups, uh, could be having these, these types of zero-sum interactions.

 

So, in our study which you're referring to, which we did at the social economics lab here at Harvard, we try to ask people about actually many different domains. So, we ask them the extent to which they believe interactions between groups are zero-sum when it comes to the rich versus the poor, when it comes to immigrants versus non-immigrants. You know, do the gain of immigrants come at the expense of non-immigrants? Do they think the gains of some racial or ethnic groups come at the expense of other racial ethnic groups? When two countries trade, do they feel like the gains of one country must mean the losses of another. So, we try to really ask about a lot of different domains. And from there we extract your, your general tendency to think in zero-sum terms. Think of it as a, as an index or sort of summary measure that tells you not in a specific domain particularly, but just in general, you tend to think in zero-sum terms.

 

And this contrasts to a positive sum mindset, where you might believe that, you know, we could all be better off, either through cooperation, or innovation, we can grow the pie. And so there could be, you know, things could be better for everyone involved.

 

Ralph Ranalli: So, say I'm a up-and-coming economist. How do you make the case to me for the utility of this research; that understanding zero-sum thinking is important to my future work or the development of my economic thinking? Why do I care about this?

 

Stefanie Stantcheva: Yeah, this is an excellent question. A good one to ask for students. So in general, you know, we economists tend to not put excessive weight on survey data. On things that, you know, people just say or reveal about themselves, but rather try to look at their actions, try to look at at what they do. And so, what we do at the Social Economics Lab in that sense is quite unusual because we do try to understand how people reason and think about different issues that are, um, very important in their daily lives- about economic issues, about policy issues.

 

And so the zero-sum thinking, which is work that is such in depth survey work, is exactly part of trying to understand how people reason, how they view the world. And it has very important, in this case, policy consequences. So if you're someone who views the world in zero-sum terms, you know, you tend to think that there's essentially some groups that are made worse off because of other groups and you would like the government to ideally step in and do something about this. So it's going to, you know, lead to some really interesting policy views.

 

And before I get to them, I need to tell you something very important, which is that, unlike so many other things today, zero-sum thinking is not a partisan issue. So it's something that is actually quite evenly spread across the political spectrum. So it is not a very clearly Republican issue or very clearly Democrat issue. And this is important because the policy views, I'll tell you now will sound a bit unusual because they're not clearly left-leaning policies or clearly right-leaning policies.

 

So if you're someone who views the world in zero-sum terms, you're going to want more redistribution because you tend to believe that the gains of, of richer people come at the expense of poor people. You'll also support more racial equality policies or gender equality policies, again, with this idea that some groups are better off just because others, you know, have been made worse off. But you'll also be a proponent of anti-immigration policies. So restrictive immigration policies, again, with this idea that the gains of immigrants come at the expense of non-immigrants. So it will lead to this mix of what we might traditionally consider more left-leaning and right-leaning policies. So it is, it is a mindset that that cuts across party lines. And given that it has very strong predictive power for your policy views, it is very important to understand and study.

 

Ralph Ranalli: Right. I want to dive into the politics part in a little, but I did want to take a minute to explore your research methods because they are very, I not sure what the word would be, expansive? You used large data sets. Can you give us a little bit of background on the methodology, how you went about this research and why you have such confidence that it's accurate and on really solid ground?

 

Stefanie Stantcheva: So, our mission at the Social Economics Lab is to really dive into people's minds, and try to understand how they reason, how they think, how they perceive important economic issues, economic policies that shape their daily lives. And we do this with carefully designed large scale surveys. They're usually done online, and these are, you know, not like your traditional say, opinion polls that just ask, you know, very short, very direct questions. But rather it's, it's a deep dive into a whole reasoning process.

 

So to really try to understand the causes of something- people's thinking, people's concerns, people's preferences. And this has really, really, really big value because our preferences, our constraints, our mindsets are going to shape how we view the world, what type of policies we want, but also how policies are going to affect the economy. You know, we're going to be better able to understand do people perceive policies the same way? Are they going to react the way that maybe our models predict or not? So, it is a key input, I think, to complement our more traditional economic data that we have and to inform our models to allow us to make better policy predictions, better policy design.

 

Ralph Ranalli: So, you found that zero-sum thinking can be traced back to the experiences of not just the individuals that you surveyed, right, but also their ancestors. Can you explain that and talk about some of the factors what you found contributed to people formulating their opinions?

 

Stefanie Stantcheva: Yes. So what we find is that zero-sum thinking is very rooted in people's experiences. So I think it's tempting for many people to say: "Oh, zero-sum thinking is just a bias, or just a misperception." But no, it really reflects people's reality. So, in this large scale study, we, asked people about their whole ancestry. So, about the family history of their parents, of their grandparents, where they lived, what occupations they did, how well they did including, you know, the respondent's own life experiences. And so we can, we can look at what influences this mindset, what makes someone be more zero-sum or less zero-sum.

 

And we really focus on three important groups of factors. The first is, is really economics. So it's growth and mobility. And you can see that it plays a big role, first at the macro level. So at the macro level, what you see very strikingly, uh, is that younger generations in the US and actually in other rich countries too, are much more zero-sum than the older generations. And it's a very stark generational pattern. And we show that you can trace it back to the growth and mobility that was happening in the economy when people were growing up. The younger generations in the US and in many rich countries are growing up at times of lower growth, lower mobility relative to the older generations. And the pattern is actually completely flipped in, in some emerging economies or developing countries where actually growth is now higher, mobility is now higher than before, and there we see younger generations being less zero-sum than the older ones.

 

It's also true at the micro level, at the individual sort of family level. So if your family has experienced upward mobility, so if you're doing better than your parents and your parents are doing better than your grandparents, then you today are less likely to think in zero-sum terms. So this growth and mobility matters at the aggregate, at the macro level and also at the individual family level.

 

Ralph Ranalli: I thought that was interesting, the whole cohort effect. You know, I was talking to Danny Roderick and Suresh Naidu in our first episode about the mindset of younger economists—especially economists under the age of 40, and how they are a generation of people who, at least in America, have never grown up in an era where that wasn't getting progressively more unequal. And we talked about how that has shaped their mindsets. So I guess it's true for both economists and non-economist in terms of that cohort effect—what you experience in your surroundings growing up influences your mindset. But I digress, and you were about to give us another factor.

 

Stefanie Stantcheva: Yeah. So, another big factor, very important in US history, is immigration. And that is another experience that is actually related to less zero-sum thinking. And that's both true directly- so if you have immigrant ancestry, you are less likely to think in zero-sum terms. And it sort of fades back in time. Like if you're a first generation immigrant, you're particularly less zero-sum, second generation immigrants, a bit less zero-sum, and so on. And then also indirectly. If you or even your parents, grew up in areas that had a lot of immigrants- which historically in the US have done really well, have typically climbed up the social ladder, done well economically for themselves, and actually also contributed to their local economic communities- you are less likely to be zero-sum. So that is another important part of the U.S. experience that has really shaped this mindset.

 

On the flip side, a very negative experience—again, related to US history—is a history of enslavement. So, if your ancestors were enslaved, and that can come in many forms, of course, you know, slavery in the US South, but also the Holocaust, or indigenous people's reservations, forced labor. So all of these enslavement experiences, which are very, very zero-sum, if they happen to your ancestors, then you today, you know, many generations down the line, are much more likely to think in zero-sum terms. So it is something that is, very tied to the family history, and has these longstanding roots.

 

Ralph Ranalli: Yeah. I saw one of the charts in your paper that showed that black Americans have a strong prevalence of zero-sum thinking. And I think that probably correlates to their collective racial experience.

 

Stefanie Stantcheva: That's right. And I would just add that this history of enslavement, again, also has these indirect effects. So, it is not only for people who directly experienced enslavement in their family, but also people who you know today or, you know, their parents or their grandparents live in counties that had higher shares of enslavement as of 1860. So, far back in time, you know, persistently until today, if you grew up in those places, you are more likely to be zero-sum. So clearly the local, you know, the local economy, the local community, the local environment, is still shaped by this many generations down. And if you grow up there, you are more likely to hold the zero-sum mindset.

 

Ralph Ranalli: Yeah, I think you also found—I don't know what you'd call it, maybe a carpet-bagger effect—where people actually took their attitudes with them as they migrated to different places. And you found that migrants from places where there had been traditional slavery brought those attitudes with them.

 

Stefanie Stantcheva: Yes.

 

Ralph Ranalli: One fascinating finding that I wanted to ask you about was about education levels and zero-sum thinking. First you found that people with low levels of education tend to be very zero-sum. And then that people as they climbed the educational ladder through high school, two-year college, and four-year college became progressively less zero-sum. But then surprisingly it kind of flipped around at the end, where people who did postgraduate and postdoctoral education—PhDs, postdocs—actually were more zero-sum again. Can you explain that? That one was really fascinating to me.

 

Stefanie Stantcheva: Yeah, so we have this basically non-linear pattern between zero-sum thinking and education exactly as you described. Initially, as you start to acquire more education, it seems like it is correlated with being less zero-sum. But then once you get into the very high degrees PhD, MDs, JDs, this is where the, the pattern flips and people are much more likely to be zero-sum.

 

You know, to speak in, in technical economic terms, it could be either selection or treatment effects. By which I mean, it could either be that people who are more zero-sum tend to go into these type of degrees. Or it could be that these perhaps very competitive degrees with limited slots, a lot of competition, make you more zero-sum. So we cannot say which story is true. Could be a bit of both. Could be a mix. Could be one of them only.

 

Ralph Ranalli: So, I have to ask, does your research suggest or infer ways that people can be coaxed to move beyond zero-sum thinking? If zero-sum thinking is in some ways a trap, like how do people escape that trap, or how do you coax them out of it?

 

Stefanie Stantcheva: I, I have thought about this question, and I get it a lot, but I think what our research shows is that this is very rooted in your true history, in your lived experience. So you are not zero-sum just because you know you want to be zero-sum or because it's a bias. The world isn't really zero-sum and you believe it is. No, it is because you have lived through zero-sum experiences. It is because you have grown up in times with less growth, less mobility. So resources truly are more scarce and the world is more zero-sum than in a high growth, you know, high, high mobility world. So it does really reflect people's experiences.

So to me, the, the right question, which is perhaps more complicated, is what type of policies can make the world less, zero-sum. And actually create more—you know, I think inclusive prosperity is a great, is a great word for this—can create more positive sum outcomes.

 

So that is a bit the question. And I think as economists, we try to study such, such policies, right? We try to study which things might cost something, but then yield very large benefits either in the medium run or longer run. Which policy seemed to be much more damaging than that. So, this is what we try to study. And while it's not, it's not easy to give clear, clear and simple answers here. There's at least a few candidate policies on which I think many economists would, would agree. One type of policies is policies that help children. Whether through education or nutrition for children, investments in children, policies that level the playing field for children in general, you know, pay for themselves over the medium or longer run. It is something that, yes, it does cost money in the short run, but it will pay for themselves. It has very high multiplier value, as has been shown by my colleagues, by other researchers.

 

Innovation policy is another, you know, potentially positive sum activity. It creates something new. It. Grows the pie. There's then of course the question how to distribute it in a fair and desirable way, but it has the potential to grow the pie rather than just assign different slices to everybody.

 

On the more perhaps mitigation side, you know, things like policies to mitigate climate change. It's not that it's necessarily growing the pie, but it at least prevents the pie from shrinking. If we, you know, invest some resources to prevent more depletion of resources or the erosion of our world. So that is another direction, which it's possible to think, to prevent perhaps more zero-sum situations from escalating.

 

Ralph Ranalli: Sure, and in fact I was going to ask you about the compatibility of the notion of zero-sum thinking with the efforts of Economics for Inclusive Prosperity, of which you are a co-director as well as being Harvard professor. And you were also awarded the John Bates Clark Medal this year. I wanted to congratulate you for that—that was quite the achievement.

 

Stefanie Stantcheva: Thank you.

 

Ralph Ranalli: But I did want to expand a little bit more on this notion of sustainability, and the relationship between zero sum as a mindset and situations which have elements of being zero sum in the real world. As we’ve said, sustainability is a big goal of Economics for Inclusive Prosperity, to change this mindset of the economics profession around sustainability. But in terms of positive-sum and negative-sum thinking, the reality is that the earth is, after all, a finite resource. A hundred years ago, there were 2 billion people who inhabited it. Now there are 8 billion. So, we've increased the population fourfold and the climate crisis is shrinking the amount of habitable land on the planet. So, in the sustainability context, is there both a positive and a negative side to the notion of zero-sum thinking?

 

Stefanie Stantcheva: So, yeah, this is a tricky, tricky question, but the way I was trying to phrase it is that, as you say, if nothing is done, resources are being depleted. So, we are moving towards a truly more zero-sum world in the sense of it will be harder and harder for all the groups to get the resources that they need. So, climate mitigation policies, you know, policies to either slow down or stop that process are policies to, to preserve resources and to ensure that there is a set of the pie that remains for sufficiently many people. So to me, this is, this is the important question and what type of climate policies can be implemented.

 

And here we actually do have quite some interesting, uh interesting research at the social economics lab. It would probably be too long to talk about it in great depth. But, maybe I can leave you with one key message, which is, if you survey people in 20 countries about their climate change perceptions and what type of policies they want, one of the key factors that that appears as being really predictive of support for climate policy is the extent to which you consider policies to be progressive, equitable. So to not hurt disproportionately lower income or vulnerable households. And of course, people care about their self-interest, so they care about, you know, is this policy going to hurt my household? So in general, people who are, more aligned on cars or on polluting energy are going to be more against climate change policies. So that's very understandable.

 

The effectiveness of the policies also matters. Do they perceive them to be very effective in reducing emissions? Et cetera. But there is that really important third factor, which is, are the policies progressive? And many policies today, whether it's carbon taxes, including with some transfers, or you know, things to regulate specific types of cars, specific types of energy. Many policies today are considered quite regressive by people. So they're considered to really hurt more the lower income people, disproportionately relative to higher income people. And when people see policies as progressive, this is when they're actually really willing to support them.

 

So what we show very strongly in this paper is that today it's not about a lack of concern or worry about climate change. Many people are already very worried and very concerned. So it is not about trying to convince people climate change is a serious problem. In a sense, my feeling from this project is we've already moved past that and we're at the stage of: Well, what exactly can be done against it? And here's where there's a real hurdle because many policies today are considered just not fair enough and the burden is not spread equitably enough. And so that is something I think very important for, you know, policy makers to to think about.

 

Ralph Ranalli: Yes, you mentioned unfairness and that's actually the question I wanted to ask next, which is: How much is zero-sum thinking informed by people's feelings of unfairness? Many interactions are technically positive-sum but still feel unfair to a lot of people because one side extracts all the surplus from that positive sum equation. For example, take productivity gains, right? In theory they should benefit everyone. But in the last five decades, all the benefits of productivity gains—virtually all the benefits—have gone to business owners rather than workers. So to what extent do notion of fairness interact with zero-sum thinking?

 

Stefanie Stantcheva: So, they are related in the sense that, if you perceive the world as zero-sum, you will generally think that it is less fair. So the idea that one group becomes rich at the expense of another will make you consider the success of that group to be achieved less fairly and to want to, you know, do something against it. Which is why it correlates with those types of policy views. Basically, policy views where, you know, if the gains of rich people came at the expense of poor people, if that is your belief, you'll say: "The wealth distribution is less fair because it came at the expense of someone else. And I would like more distribution to, you know, to restore the fairness here. So it is very, very related to this.

 

And I think it's a very important mindset in the sense that it is very predictive of, of such policies and comes on top of other mindsets, which we have studied, which are also very important. But is quite distinct. So just to give you a very clear example: For very long, you know, we have studied the extent to which people consider income inequality or wealth inequality as being fair, depending on whether it comes from effort or your own merit versus whether it comes from just luck or circumstances. And, you know, people who believe income comes mainly from, from effort are much more willing to tolerate inequality. People who think it's mainly luck and circumstances are much more unwilling to tolerate inequality.

 

And what we say here is absolutely, and that is a very important predictor. But there is that second dimension here, which is, and has it been a zero-sum interaction? So, for instance, let's pick Jeff Bezos, just as, as some example. People don't just care about whether Jeff Bezos worked hard to obtain his wealth. You some people will say yes, others will say there is luck. I mean, but that's not the only thing people care about. They also care about: Did Jeff Bezos' wealth come at the expense of others? Maybe small businesses, bookstores, workers, you name it? Or did it contribute to society and to others? Did it lift... was it the tide that lifts all boats? And so both of these concerns will really matter when people are thinking about, do I want to tax Jeff Bezos wealth or not? Both of these will be highly relevant.

 

Ralph Ranalli: So, how do I present my policy proposal in the most positive light in terms of this fairness or unfairness question?

 

Stefanie Stantcheva: So I think, I think the real question here is to try to understand what are people's concerns and preferences and constraints. And ideally, if you are a young economist designing policy solutions, you would like to, to really address those. And for many policy areas, what we do find in our survey work is that fairness matters a lot to people. So, fairness is a key concern.

 

It's something that is sometimes very complex considerations, we spoke about a few. Sometimes it's simpler. But fairness ultimately is something very, very important. So, it'll look different in different cases. For climate change policies, for instance, you'll find, perhaps as we do in our paper, that people would be much more favorable to outright bans on polluting vehicles rather than taxing polluting vehicles. And it might sound puzzling from an efficiency perspective to an economist, and yet you'll see that it's driven by this equity or fairness concern, which is people consider it fair that nobody's allowed to drive such a car. It's sort of across the board, it's fair rather than some people can just pay to pollute and those who can afford it will pay to pollute.

 

So, an interesting, strong fairness view here that, as a policymaker, you can or cannot take into account. But that is there very, very strongly. And there's similar examples in other settings. I think it's very important for young economists to understand in different settings what fairness is and what people mean by that.

 

Ralph Ranalli: Before we wrap up, I wanted to ask you: What's next for you and the lab in this line of inquiry, and what do you hope your research on zero-sum thinking can achieve? And what lines of inquiry do you hope maybe other economists pursue in this area? Because it seems like it just gets more important every day.

 

Stefanie Stantcheva: Yes. Lots of things to do. You know, something that is related that we're looking at right now- also a little bit unusual perhaps for economics, is to look at the role of emotions in policy views. We see that many recent, you know, events like inflation, the decline in, in, in mobility, et cetera, trigger really strong emotions. And these emotions then feed into what policies we want, and they're also amplified by the policy debate. They're amplified on social media. So we document specifically a very strong rise in anger. It is not just an illusion if you feel like things are getting angrier. We can see it in congressional speeches, in campaign speeches, in social media discourse, both from policy makers and from citizens. So, that is an area that I'm very interested in and seeing what can be done there.

 

But overall, you know, for young economists, there is so much work to do on all these things we spoke about that are related to understanding how people think, how people, perceive things. What is fair? What is not fair in different settings? So much work in those directions.

 

Ralph Ranalli: Well, thank you very much. This has been a really enlightening conversation and I'm glad you were here to share your wisdom with us. And good luck with your future research, we'll be waiting eagerly for your upcoming reports.

 

Stefanie Stantcheva: Thank you so much for having me. It was great talking with you.

 

Outro (Ralph Ranalli): Thanks for listening. The Economics for Inclusive Prosperity podcast is based at the Malcolm Wiener Center for Social Policy at the Kennedy School of Government at Harvard UniversityEconomics for Inclusive Prosperityis a network of academic economists from Harvard, Princeton, Columbia, and other leading universities who are committed to an inclusive economy and society. EfIP members are working to transform their field around a new vision of prosperity—a vision that includes traditional economic metrics but also expanded measures of wellbeing including access to health, to democratic participation, and to a livable planet. They’re also highlighting important changes in economics that are already underway, like increasing use of big data and the questioning of old theoretical orthodoxies.

 

Please join us again in two weeks for another new episode featuring University of Oxford economist Max Kasy discussing the political economy of artificial intelligence and the make-or-break battle between the tech titans and the public over who controls what he calls “the means of prediction.”