Chicago Booth Review Podcast How Universal Is the Marshmallow Test?

Oleg Urminsky: People generally want things sooner and want outcomes to be more certain. Are you willing to sacrifice some of that for the potential to have more? We can think of this as like, I can settle and be comfortable by having it right away or having it with high likelihood, or I can be willing to sacrifice a bit of what I'd prefer in order to have more.

Hal Weitzman: The marshmallow test famously gauges children's ability to wait for a bigger reward. A body of research suggests that the happier we are, the more likely we are to be patient. Many of those studies were undertaken in the US and Europe, so how universal is that conclusion? Welcome to the Chicago Booth Review Podcast, where we bring you groundbreaking academic research in a clear and straightforward way. I'm Hal Weitzman, and today I'm talking with Chicago Booth's Oleg Urminsky, whose research examines the connection between emotions and economic decisions around the world.

Oleg Urminsky, welcome back to the Chicago Booth Review Podcast.

Oleg Urminsky: Thank you for having me. It's great to be back.

Hal: We had such fun with you last time. We wanted to have you back to talk about your research on emotions, patience, and economic decision-making. Tell us, what do you mean by patience, and what do you mean by risk-taking in the context of this research? How do we define our terms?

Oleg: We're looking at a particular kind of patience and risk preference that comes up in some economic decisions and not others, which is the situation where there's a trade-off between, in the context of patience, you can imagine the trade-off between having some positive outcome sooner that's smaller, or being patient, waiting longer, and getting a better, larger--

Hal: Classic marshmallow test.

Oleg: Exactly. In the context of risk, we're again looking at one side of the risk question, which is you could have something smaller with more certainty, or you could have something larger with less certainty. It's important to qualify this. I'm glad we started with this because when we think of risk-taking, we often think about the entrepreneur, there's a small probability of a very positive outcome, but maybe what has a high probability is actually a negative outcome. Here, we're just looking at upside patience and upside risk, if you will.

Hal: Okay, got it. One of the things that's fascinating about this is you point out a lot of the prior research on emotions and economic decisions or decision-making studied college students in Western countries. We have our lab here at Chicago Booth, we have one downtown. Why is that insufficient for drawing conclusions about human behavior? Does it just tell us something about what college students do in US universities? Is there anything we can generalize or is that just not applicable?

Oleg: I think that the issue is that we just don't know when the research only looks at it in this way. The reason for it is largely one of convenience. If I'm a psychologist studying the role of emotions in decision-making, which is such a fundamental question, the most feasible thing to do is to go to the people I have around me who are willing to be research participants, which are typically the undergrads at the school where I teach. There's a lot of research on this question, but it's largely been done in this way. When it wasn't done with college students, then it tended to be done with US survey participants, again, because that's what's most feasible.

I don't think that there was necessarily a strong stance that the field took that emotions and how they impact decisions are so universal that we are confident in this generalization, but it's more the combination of this seems, compared to political attitudes or how religion interacts with decision-making or things like that are so obviously culturally-dependent, this seemed basic enough that it was probably fine to study in this way.

Hal: I see. Certainly, the way it gets translated into the media, the way that people will hear about it on TV is not going to have the subtlety that you're talking about, it will be a general statement about humanity. Is that fair?

Oleg: Exactly. We don't say, "University of Chicago undergrads, when they're in a good mood, make these kinds of decisions."

Hal: 100 people said this.

Oleg: People.

Hal: Right. That makes sense. That's fascinating because I'm guessing that quite a lot of behavioral science research falls into that, has been categorized as such, maybe not by the researchers, but it's been translated like that, which suggests that this is a bigger issue than just the paper that you're talking about.

Oleg: Absolutely. I should be clear that this is something that I think has been a fundamental change in psychology and behavioral science more broadly in the last 20 years or so is this growing awareness and appreciation that it can be quite difficult to predict a priori which things are just human nature and are surprisingly consistent across countries, cultures, societies, and which ones differ. There's been lots of research along these lines.

Hal: Sure. In this paper, you're focusing on what you call "incidental emotions." What are incidental emotions?

Oleg: The idea of incidental emotion is the emotional state we come into a new situation or decision with, as opposed to how we feel about that situation or decision.

Hal: It's something in the background?

Oleg: Exactly.

Hal: All right. Because you're not looking at laboratory-type experiments that students do in their lunch hour and get paid $5 or whatever, you're looking at the Gallup World Poll and Global Preferences Survey, and that measures things like emotions, economic preferences, patience, risk-taking. Just tell us a little bit about how it measures those things.

Oleg: One disadvantage of this kind of research is that when the undergrad comes into the lab, we can take our time. We can spend half an hour, an hour if we need to, and measure things very precisely, measure things in multiple ways to see if they're consistent. When we're using this kind of data that was collected to answer lots of different questions and conducted around the world, efficiency was at a premium, so the measure of emotions is not the ideal measure.

The measure of emotions is they get a checklist of a list of different emotional states and just check off all the emotions that you experienced in the past day. You might not still be feeling that emotion. That might be something that happened yesterday and you're not feeling it. It's not the literal incidental emotion in that moment as much as just the broader emotional or mood state that you have been in. That's on the emotion side.

On the economic preferences side, there was a team of researchers who developed the measures, and they actually went to the lab first and looked at both measuring this in terms of actual decisions, so make decisions about small amounts of money, do you want to get $2 tomorrow or $3 in a week kind of questions. Then, also, ask them self-report questions, like, "Do you find it difficult to be patient? What are your preferences?" They developed a combination of these questions that they found to be most predictive of actual choices.

Hal: I see. If you're measuring incidental emotions, if it's not too stupid a question, and people are saying, "I had these emotions in the past week." How do we know what's incidental and how do we know whatever the opposite is? Central?

Oleg: Right.

Hal: Guiding?

Oleg: That is a core limitation in this research, which is each person just takes a survey once. We don't observe changes in people's state over time. That means when we're saying, "We're comparing two people, one said they were happy, one said they were sad, what does that mean?" We're assuming they're reporting that accurately to us, and that the happy person actually was happier and the sad person actually was sadder.

What we can't distinguish between is, these are people with the same baseline level of emotion but one just happened to have good things happen yesterday and the other one had negative experiences yesterday, versus, these are the kind of people that, in general, one tends to have more happy days and more sad days. There is some other research that was done with different surveys that were conducted over time that finds somewhat complementary results that can pinpoint that part of it a bit more.

What we do in this research is we try to address it by tackling the major confounds you might be worried about, like maybe people who have more money, their lives are easier, they're less likely to be stressed and angry and sad, and people with less money have tougher lives. Could this really be about their income instead of their emotions? We control for that in the analysis, but those controls only go as far as the measures we have available to us.

Hal: I see. Got it. You may not have depth, but you do have depth.

Oleg: Exactly.

Hal: You studied 77,000 people. That's great. This survey gives you this reach. 74 different countries. There are patterns there. You've found that positive self-reported emotions, with all the caveats that we just talked about, generally predict two kinds of behaviors. One is a greater willingness to wait for the delayed rewards, that's the patience, the marshmallow that we talked about. The second one is a greater willingness to take favorable risks. Now, those might seem to be contradictory, waiting patiently versus taking risks. What's going on? How do you explain that?

Oleg: The common theme, because we're looking at an upside risk, people generally want things sooner and want outcomes to be more certain. Are you willing to sacrifice some of that for the potential to have more? We can think of this as like, I can settle and be comfortable by having it right away or having it with high likelihood, or I can be willing to sacrifice a bit of what I'd prefer in order to have more.

Hal: They're not contradictory, they're complementary.

Oleg: Exactly.

Hal: All right. Which emotions from this study were most strongly identified with patience, and which with willing to take risk?

Oleg: Happiness and enjoyment, which was very similar, were most strongly associated with willingness to be more patient and take more of these upside risks. Pain was most strongly associated with not being willing to take additional risks and not being willing to be patient.

Hal: That makes sense.

Oleg: That dimension you can think of in terms of just when life is good and my current needs are met, then it's easier to say, "Let's invest for tomorrow. Let's put some effort into things that might not work out, but if they did work out it would be great, and to think in that way. When things are bad right now, what that emotional state tells us is, what are ways that are available to us to improve our current situation? These results are very in line with this theory of valence as a cue to action.

Hal: Just explain that. Valence as a cue to action, meaning?

Oleg: Meaning that when things are good, you don't have to in the moment. That means there isn't a problem to attend to now. Instead, I have the luxury of focusing on the future or focusing on ways to achieve unnecessary extra things. When I'm in a negative emotional state, that's a cue that there's some problem in my environment that I need to deal with. The valence doesn't explain everything, and so there are some interesting deviations from that.

We see this for anger and stress, but the results are strongest for stress. Particularly for stress, when people report more stress, that's a negative emotion. If it was just about valence, we would say, "Okay, they should not be patient and not be willing to take risks. Not be willing to wait and not be willing to take risks." Instead, we see the stress people, like the happy people, which is a contradiction, are both more willing to wait and more willing to take risks.

The intuition there is that the cue that emotion provides to us to guide our behavior is not just about good, bad, but it's been discussed in literature in terms of approach avoid. If I'm sad or in pain, I just want to curl up in a ball on the couch and not deal with anything. If I'm angry or stressed instead, I have more of this approach motivation where it's like something is wrong and I'm going to do something about it. You can think of patience and risk as also being sacrificing immediate uncertain reward for the long-term benefit as something that is taking action to improve my circumstances in the long run.

This is not something we necessarily anticipated a priori because anger and stress, you could think of in either way. You could either think of it as like, "I'm not going to put up with this anymore. I need to change my life," which is in line with patience and taking risks, or you could also think of it as like, "I can't deal with anything right now because I'm so angry and stressed." To us, it was an open question which direction it would go.

Hal: If you're enjoying this podcast, there's another University of Chicago Podcast Network show that you should check out. It's called Entitled, and it's about human rights. Co-hosted by lawyers and law professors Claudia Flores Tom Ginsburg, Entitled explores the stories around why rights matter and what's the matter with rights.

Oleg Urminsky, in the first half we talked about your research about patience, and emotions, and economic decision-making. We talked about the contrast between, just in terms of the study, the deep, small study in the lab and the wide, maybe is more shallow-- that's fair-- study that you got by looking at these surveys, but were the results broadly the same? Did your study confirm the prior studies or append them?

Oleg: One part of this research and actually the starting point that took the most time was to go through the prior literature and accumulate all the tests that different researchers had done. What emerged from that is these ideas had been tested many different times in many different ways. Some of the more conclusive-seeming results got a lot of attention and were cited in textbooks and became part of what we thought we knew about this phenomenon. Then when you pool all the data together, it's actually remarkably inconclusive. Part of that is because these are quite small studies with small numbers of people and it's just a very noisy phenomenon. Those studies didn't really have the precision.

To the degree that the prior literature pointed in a particular direction, our results were largely in line. The main difference is just by having so much more data than had been available before, we can answer these questions much more precisely.

Hal: The stress thing you talked about in the first half, for example, that stress and happiness actually weirdly lead us to the same direction, which is we're more prepared to take risks, is that a new finding or has that not really been tested before?

Oleg: It had been tested, but it was inconclusive. There were some studies that suggested stress had positive effects, some that said negative effects, some that found no effect. We just didn't know.

Hal: Because you're doing this massive global study, you can look at these megatrends. You identify two really big factors that predict how strongly your emotions influence economic decisions across countries. One is economic development and one is individualism. Just walk us through what you found there.

Oleg: To us, this was in some ways the most exciting part of identifying that we could do this with this dataset. As we talked about at the beginning, most of the studies that had been done before were in Western economies, often with college students. The question of how replicable are these findings across cultures and across populations was a very open question. The first thing we saw is just that we saw lots of variation in the estimated relationships between the emotions, particularly between valence and these economic preferences.

Just the fact that it varies a lot is something that we think is new and interesting and should affect how we think about these things. I think there was this view that maybe emotion is so basic and fundamental that it doesn't matter where you test it, you'll always find the same thing. We don't find that. Then that led to the second question of like, "Okay, so it differs, but why?" Here, even with the enormous amount of data we have, the data starts to get pretty thin because we were looking at country-level measures. We went from 77,000 people to 70-some countries, which is a much smaller dataset.

We have two measures, individualism and economic development. It turns out they're very highly correlated across at the country level, and so we couldn't distinguish between them. What we can think of this as a composite, that there are some countries in the world, the United States is one of them, that is different from the typical country in terms of being more individualistic and more economically developed. Then there are countries at the opposite end of the extreme that tend to be more collectivist and have less developed economies at this point in history.

The role of emotions or at least the relationship between emotions and economic preferences differs. In particular, in more individualistic and more developed countries, we see a stronger relationship between emotion and economic preferences. That's what we can say with the data. In terms of interpreting that, that's open to many different interpretations. We're now looking at country-level correlations, confounding many things.

One interpretation that I would offer as a hypothesis rather than a conclusion is that in individualistic countries, with more developed economies, we have the luxury of making economic decisions in line with our incidental emotions. Whereas in countries that are more collectivist, that where compliance with social norms is more important and where there's stronger economic pressures, in less developed economies, you have to get the economic decisions you make right. You don't have the luxury of making the wrong decision just because of the mood you happen to be in at the time. I think this is a plausible interpretation.

Hal: There's another. I feel like the bankruptcy laws change our emotions.

Oleg: Exactly. Right.

Hal: We're more likely to be happier to take a risk, more comfortable taking a risk because of the way that the whole structure is set up. There's a lot of things interacting there. When you say individualism versus collectivism, is that self-reported? How do we measure that in the survey?

Oleg: We used two different measures and found similar results, which gives us some confidence because none of the measures are that great. We used measures, country-level averages from the World Values Survey. There's also this very standard measure that people use that we also used that sounds crazy when you describe it, which is a very early researcher who is interested in cross-cultural psychology worked with IBM to survey their employees around the world. That is one of the standard measures, like how IBM employees differed across countries in the 1980s is how we think about these country-level differences. Subsequent data has validated that that predicts pretty well.

Hal: It's not just differences between countries, it's also differences within countries, right?

Oleg: Right.

Hal: You found these big regional variations. Tell us a little bit about that.

Oleg: Exactly. We looked at differences across countries. Within countries, we looked across regions. We also found quite a bit of variation. There, we were not able to explain that variation as well. We looked at both region-level proxies for economic development and individualism as well as some not-great person-level measures. There, the results were pretty inconclusive. We have stronger evidence at the country level.

That's something that I think would be really interesting to try to do more research on because there are these questions about, is it just a lack of ability to measure it accurately enough that led us not to detect these patterns at the region and person level, or is it really that the norm of integrating, using emotions as an input into your decision-making? That those are widely shared norms that differ across countries, even if your personal circumstances don't align with the circumstances that created the norm in the first place.

Hal: We talked at the beginning about the marshmallow I mentioned, the marshmallow test. It made me think of it, the classic thing where you ask a kid, can they be patient and wait for two marshmallows? That also, it seems to me, has some ideological or cultural underpinning. It may be that in America or in Europe, if you wait, you really will get two marshmallows, whereas if you're in a poor country, it may be that you won't get any marshmallows. That the nature of risk-taking, we think of risk-taking, as you say, as a good thing, but it may be risk-taking is actually not a good idea if you're in a country that's beset by war or a lot of economic uncertainty.

Oleg: Absolutely.

Hal: There's a lot going on there, which means that you might be happier and more likely to take a risk, but that might not have been necessarily a good idea everywhere.

Oleg: Yes. I think that's a challenge for any research like this, that when we see these differences, it does tell us not to jump to assumptions that things are the same, but it can be difficult to tell what the source of the difference is. Is the difference at the psychological functioning level, how people incorporate the emotion they're feeling into the decision, or is it also, in part, that what the decision represents or what the emotion represents is different? Is someone saying, "I was sad yesterday," different in the US than in another country? We did some analysis to look for strictly linguistic differences.

We did the analysis only using Spanish-speaking countries to make sure that it wasn't at the level of how the word was translated, and we rule that out, but at this more fundamental level of what is the experience of being sad in one culture versus another, that is baked into the phenomenon itself.

Hal: Right, remains, but more research on that.

Oleg: Exactly.

Hal: Okay. We'll have you back when you've done that. All right. Oleg Urminsky-

Oleg: Thank you very much.

Hal: -thank you so much for coming back on the Chicago Booth Review Podcast.

Oleg: My pleasure.

Hal: That's it for this episode of the Chicago Booth Review Podcast, part of the University of Chicago Podcast Network. For more research, analysis, and insights, visit our website at chicagobooth.edu/review. When you're there, sign up for our weekly newsletter so you never miss the latest in business-focused academic research. This episode was produced by Josh Stunkel. If you enjoyed it, please subscribe, and please do leave us a five-star review. Until next time, I'm Hal Weitzman. Thanks for listening.