Nathaniel Burke is a Ph.D. candidate at the Sam M. Walton College of Business. He received his BA in economics from Manhattan College, then earned his MS in resource and applied economics from the University of Alaska Fairbanks. After serving in the Army, Nathaniel developed an interest in identity and diversity as it relates to economics.
His research interests include behavioral economics, experimental economics, micro-development, education, and social networks and identity. Nathaniel was a Mercatus Center Frédéric Bastiat Fellow during the 2018-2019 academic year.
Episode Transcript
[music]
00:08 Matt Waller: Hi, I'm Matt Waller, Dean of the Sam M. Walton College of Business, welcome to Be Epic, the podcast where we explore excellence, professionalism, innovation and collegiality and what those values mean in business, education and your life today. I have with me today Nathaniel Burke who is a PhD student in Economics here in the Walton College, at the University of Arkansas, he is also a Bastiat fellow at the Mercatus Center at George Mason University. And I know that Nathaniel you study Behavioral Economics, and experimental economics and a number of other things, like micro-development and I know the Mercatus Center has a history of experimental and Behavioral Economics.
01:01 Nathaniel Burke: Yeah so George Mason as a whole has a pretty strong history with experimental economics. Vernon Smith, who's kind of regard to as the father of our sub-fields, he taught there. And actually if you go to George Mason, there was a funny error that was made where they accidentally renamed the garage in his name when they were supposed to rename an entire building in his name. So if you show up, it still says, yeah, it still says the Vernon Smith garage and it's really supposed to be the building but yeah, so there's a really strong cohort there of behavioral economists, they are still doing really good work that I get to, I have the pleasure of working with whenever I get to travel out there.
01:36 Matt Waller: Well, that's great. What is the difference between Behavioral Economics and experimental economics?
01:42 Nathaniel Burke: That's actually a really important question. And a lot of people kind of lump them together. Experimental Economics is really a tool box that economists use to collect data, right, so when we say someone is experimentalist we really mean that they use experiments to test their models and test hypotheses. Behavioral economics is the study of how individual behaviors impact rational decision-making. So you don't necessarily need to do an experiment to test behavior, you can do that using empirical data that already exists. I've done human capital projects where I used IRS data, for example, to see how likely someone is to move based on where their parents are from which is still behavioral study but we didn't use an experiment for that one.
02:24 Matt Waller: So which one are you more interested in?
02:27 Nathaniel Burke: I say that I'm a behavioral economist and I use experimental methods and when I'm in the realm of behavioral economics I will classify myself as applied micro-problems, so I do a lot of education, labor, things like that.
02:42 Matt Waller: How did you get interested in it?
02:44 Nathaniel Burke: Originally, I thought I was into Environmental Economics, my master's degree is actually in resource and Applied Economics. I did that at the University of Alaska and I knew that experiments seemed kind of interesting to me and what really pushed me over the edge is Cary Deck who used to work here. He was a visiting chair at the University Of Alaska Fairbanks. And so while I was up there, Cary and I got to meet, I also got to meet a few other prominent behavioral economist that use experiments and they kind of nudge me over the edge and say, if these are the problems that you're interested in looking at then, maybe you should consider doing experimental methods and go to university, that would be able to foster that.
03:22 Matt Waller: So you got your Master's degree from the University of Alaska Fairbanks.
03:27 Nathaniel Burke: Yes.
03:28 Matt Waller: And in Economics?
03:29 Nathaniel Burke: Yeah. Resource and Applied Economics, very similar to a general e-con masters. The difference was we focused on applied issues, a lot of them were specialized to Alaska so I got to do a really cool project where I tested how much Alaskans were willing to pay for reindeer meat from Alaska and it was my first introduction to seeing the economics of identity and how much someone's identity could actually impact their decision-making process when they're spending real money. And from that study we're submitting it to a journal soon, but kind of a preview, that study we actually found there was over $3 per pound premium for Alaskans to buy meat from Alaska.
04:06 Matt Waller: Because of their identity as Alaskans?
04:09 Nathaniel Burke: Exactly, because their only other option for reindeers is obviously Canada, just based on geography, but they were willing to pay up to $3 extra just so they can be supporting the Alaskan economy versus importing it from Canada.
04:21 Matt Waller: How did you figure that out?
04:24 Nathaniel Burke: So we did something called an adaptive choice based conjoint survey which is really just a fancy way of saying, it was a smart survey. We figure out what people's baseline preferences are. We kinda ask them, "Hey, if you got to pick any type of reindeer meat organic, non-organic, what cut? How lean? Where it's sourced from, what would you choose?". And then we used that information and we started offering them different baskets, different combinations of attributes and we start putting price tags on them to see which ones that they would be willing to spend money on and from that we can use this kind of updating model and start to figure out what are the cut-off points for individual people for how much they're willing to pay for each different attribute.
05:03 Matt Waller: So is that using part of revealed preferences, concepts?
05:07 Nathaniel Burke: Yeah, so it's a different approach to revealed preferences. We're using something called Bayesian updating and so through the process we find out not just how much people are willing to pay for certain goods, we're also finding out what things people are not willing to buy at all. So throughout the process, we found that a lot of Alaskans, they prefer organic meat. A lot of Alaskans would never buy something that was not organic certified, which is a really interesting problem because Alaska doesn't have a processing center, that's organic certified by the USDA. So these are a lot of tools that have been used in marketing a lot and now they're kind of making their way into mainstream economics, finally.
05:46 Matt Waller: I'm curious 'cause I'm familiar with its use in marketing, how would an economist approach that kind of question differently than, say, a marketing?
06:00 Nathaniel Burke: So, I think one the biggest things that is different is the way we kind of model the overall problem. So the first run, when I did that project in Alaska, we did it as a hypothetical survey and we sent it out to 100 plus Alaskans in the Fairbanks area, which doesn't sound like a lot, but in Fairbanks, Alaska that's a lot of people to get responses from. [chuckle] So, we set this kind of a baseline, and then what we do next is, now we're gonna set up an experiment where people earn a certain amount of money and then they spend that money in a virtual grocery store. So, not only do we have this baseline where people got to take the smart survey, but now we're gonna take people and we're gonna actually throw them into a virtual grocery store with money they had to earn from us and see what their actual willingness to pay is with money that they earned. And that's kind of the big jump that we make from a traditional marketing study to an economic study, where now we're actually looking at how people are allocating their resources.
06:54 Matt Waller: So, how did they earn the money from you all?
[chuckle]
06:58 Nathaniel Burke: So, there's a series of standard economic games that they can play, where they might have to solve a certain number of problems, they might have to count zeros and ones in the table, they might have to do and make investment decisions where they're playing in a simulated market and there's buyers and there's sellers and we start off everyone with $5 or $10 and they go in and they have to bid on certain items, and they can use those assets and kind of sell them in an experimental marketplace, so it mimics what happens actually in the real life market.
07:25 Matt Waller: Interesting. So, were there differences in your findings between the hypothetical group and the group that had to spend money they earned?
07:34 Nathaniel Burke: Yeah, so, I think this is kind of a general thing where people will always overstate in the hypothetical realm. So, in ours, we just find that people are going to spend less money in reality, but even when we put people through this earned money type setup, they're still spending over $3 a pound, whereas in the hypothetical they're spending maybe like $3.60, but there's still a significant amount of money per pound for reindeer products, especially when you consider that a lot of common meats that someone would buy, say in Arkansas, a lot of them aren't even priced at over $4-$5 a pound at the base level, we're adding on top of that.
08:06 Matt Waller: Yeah, that is remarkable. I wouldn't have guessed that.
08:08 Nathaniel Burke: Yeah.
08:09 Matt Waller: Did you look at people that were also not from Alaska?
08:13 Nathaniel Burke: Yeah, so it's hard to find non-Alaskans in Alaska. In northwest Arkansas, if you want to run a study like this, it's really easy. A lot of people aren't from here. You have a huge student population from Texas, you have a lot of executives working for these massive companies that are coming from all over the world. But in Alaska, most people are Alaskan. But in the project, we did control for how long someone had been in Alaska, and most of those are gonna come from students who came from the Pacific Northwest and came up to Alaska for college, or military members. There's a base in Fairbanks that I was stationed at.
08:46 Matt Waller: 'Cause you were in the military?
08:47 Nathaniel Burke: Yeah, I was in the Army for five years, and that's how I ended up in Alaska. But a lot of the military members are only there for a couple of years, so we're able to control for how long someone's been in Alaska, to kind of get an index of how Alaskan they really are.
09:00 Matt Waller: And was that significant?
09:03 Nathaniel Burke: It was moderately significant. It wasn't extremely significant. There was a break point. If someone had been in Alaska for at least four years, it was very significant. Using it as kind of this continuous variable, it wasn't very significant, but if we broke it up four years and we say, "Hey, someone who has been in Alaska for more than four years is Alaskan at their identity," then we get significance.
09:25 Matt Waller: Interesting. So, I can see how that kind of study could be quite helpful for companies when they're trying to figure out how to price products and so forth. A lot of times economists use analytical models to create hypotheses, right?
09:43 Nathaniel Burke: Kind of. So, in experimental, we stood very close to the scientific method. In order for experiments to be valid, you don't wanna do what's called data mining, you don't want to just sift through the data and find a hypothesis in the data, 'cause that's very dangerous.
10:00 Matt Waller: 'Cause you can have randomness that you find.
10:02 Nathaniel Burke: Right. And there's a lot of confirmation bias that comes up there. Our typical method is to... We come up with a hypothesis, we build the model that we think is gonna work, and then we build the experiment around that model so that we were only varying one or two things at a time, and we can actually test the specific variables in the model to see how they work out.
10:18 Matt Waller: And that's one of the problems with data science today. People are using "big data" and data science, and they're running and looking for correlations, and you can find correlations that are not really... There's no causation.
10:37 Nathaniel Burke: Right. And I think that's one of the main benefits of experimental economics, is that we can control what we are varying. When I go into the lab, everyone's doing exact same thing except for one thing that's different or two things are different, and I can observe just those differences and see how different randomly formed groups behave when there's like one thing different between them.
11:04 Matt Waller: So, I know that experimental economics and behavioral economics, I know they're different, you explained that, but they're still relatively new compared to the history of Economics, at least from what I'm aware of.
11:17 Nathaniel Burke: Yes, so, it takes... I think that's true for all academic disciplines. It takes time for the new idea to become the mainstream or to even be accepted. Experimental Economics had issues in the beginning, where people were questioning the power because the sample sizes were small. When you're so used to having a field undertaking federal level data or international data with hundreds and thousands of observations, and now I'm saying that I have a significant result and I ran it with only a couple hundred people in the lab, at first there was a lot of skepticism with that. I think as the methods become more refined, people become more accepting. And now we're seeing... Where before experimental methods were really for micro-economists, now we're seeing in macro-economics that behavioral and experimental methods are starting to leak their way into that as well.
12:02 Matt Waller: So, do you feel like economic theory is advancing because of this?
12:08 Nathaniel Burke: I think economic theory is becoming more nuanced because of this. We originally started off with this idea of the principal agent. We can use this one person to represent the economy. And no one really questioned it for a while, and now people are starting to question it, like, "Is this really realistic?" And when you do more behavioral models, and when you start using experimental methods to test these behavioral models, now you can start teasing out the differences between people, you can say, "Okay, there's different types of people, there's different educational levels, there's different socio-economic backgrounds."
12:39 Matt Waller: Different probability distribution.
12:40 Nathaniel Burke: Exactly. I think we're trying to refine a lot of things.
12:43 Matt Waller: Well, I remember when I was in graduate school, I think it was about 30 years ago, that I took a class in graduate school in Economics, an advanced Micro class and I remember the principal agent model. And the question is, what's the optimal incentive for the agent? And the model that they came up with was quite general. But it showed that, because you can't observe the effort of the agent you should have the optimal policy involved. And tell me if I'm wrong about this, a combination of a salary and a commission kind of...
13:27 Nathaniel Burke: Right, yeah.
13:27 Matt Waller: Where they're exposed to the risk.
13:29 Nathaniel Burke: Yes.
13:29 Matt Waller: Under a probabilistic outcome. So in other words, if the agent tries hard, the agent will increase the probability of favorable states of nature, but there's still uncertainty.
13:44 Nathaniel Burke: Right. And I think that's really relevant to today's labor market. A lot of people want salaried positions.
13:49 Matt Waller: Yeah.
13:49 Nathaniel Burke: 'Cause there's a lot of security behind a salaried position. But the firm knows that when they incentivize the employee, the employee is going to work harder. If your paycheck is based on your performance, then you're going to optimize your performance on an individual level. I think, we're starting to realize that different people react differently to those incentive structures.
14:07 Matt Waller: Yeah, and I remember even back then, 30 years ago, there were some papers starting to come out showing that if you varied some of the assumptions... One paper I remember saying that suppose the agent can invest in a way that increases the probability of favorable states of nature. An example would be advertising.
14:32 Nathaniel Burke: Right.
14:33 Matt Waller: Then the general solution to the principal agent problem doesn't hold if you have a probability distribution that isn't symmetrical. Things like that that are kind of odd.
14:47 Nathaniel Burke: Yeah, so there's a joke that goes around. And basically, like there's an economist, a physicist, and a mathematician are stranded on desert island. And they're trying to figure out what to do, and the physicist comes up with some fancy physics solution for how to create fire. The mathematician is trying to work out angles and everything else. And the economist is just like, "Why don't we just assume we have a boat?"
[laughter]
15:08 Matt Waller: Assume we have a boat.
15:10 Nathaniel Burke: Yeah. And I think economists have gotten... We've taken a lot of heat in the past decades for over-assuming to simplify our models. And now with the kind of rise of behavioral and experimental methods and popularity, now we don't need to rely on those assumptions so much anymore.
15:29 Matt Waller: Yeah.
15:30 Nathaniel Burke: Now we could say, "Well, let's get some data for this, and let's actually figure out what's going on on a more idiosyncratic or even a group level."
15:38 Matt Waller: What made you decide to go into the PhD program?
15:41 Nathaniel Burke: Originally, I did not think I was gonna get a PhD. The funny story is when I was graduating from undergrad, I went to Manhattan College in New York City, and I was an economics major. And I actually had a job offer in Europe, where they were going to fund my master's degree, which would eventually lead to a PhD. And I kinda had a moment where I freaked out, and said I wasn't ready to sit behind a desk. And I couldn't think of myself having a real job, and that's when I decided to enlist in the army. And then after a while in the army, I got hurt pretty bad, and I couldn't go in the field anymore, so I decided to do my master's. And I was like, "Okay, I'll do the master's, I'll get a good job out of this, and then I'll move on." And I applied for a couple of jobs, I did get a couple of job offers. And really it was my master's thesis advisor, we were doing my thesis together, and he says, he's like, "You know, you can do a PhD and you can keep doing work like this." For me, I'm a first-gen college student from Brooklyn, so none of my family had gone to college before me. I had no clue what I was doing when I went to undergrad, let alone during my master's. Everything for me was just kind of a, "This sounds like a good idea, let's try this out." And I like to say, I've been kind of falling forward. And so I didn't believe him at first, I didn't think... I was like, "No, I'm not the kind of person that does a PhD."
16:53 Nathaniel Burke: When I met Cary Deck, he had a conversation with me, he talked about, "You know, the PhD is really what you make of it. If you are interested in studying people, then that's the way to do it." And so I told my advisor, I'll at least apply for five PhD programs, and I'll apply for five jobs. And I got into three PhD programs, I got two job offers. And I just kinda sat down one day, and I decided, "You know what? I am good at this research thing, let me see if I can take it a little bit further." And I think the big thing that really pushed me over the edge was realizing that I'd be able to help people on a more general scale, if I do the PhD. Because I can do meaningful research that'll actually help people, in addition to getting my chance in the classroom. And here at the Walton College, I get the opportunity to teach hundreds of students every semester. So I get to directly impact people both on a research and on a teaching level.
17:47 Matt Waller: Let's... That's neat. It really goes along with the vision of the Walton College, because the vision of the Walton College, is to be thought leaders and catalysts for transforming lives. And it sounds like both of those really appealed to you.
18:01 Nathaniel Burke: One of the great things about the Economics Department here is that it's a relatively small department given the size of our university, but we have a very, very strong research faculty. Which means that I get to impact change on a much more direct level, and I get a lot more control over what kind of projects I'm interested in. Whereas, I have some friends who went to larger departments, and they kinda just get buried in the department hierarchy. And they end up having to hop onto a project with their advisor, hop on a project with someone else in their committee. I've been able to get support to actually do things that are directly impacting the Arkansas Community, and are actually leading to my research agenda as well.
18:36 Matt Waller: And so what are some of the current things that you're working on?
18:41 Nathaniel Burke: Yeah, so my biggest project that I'm working on... And this is actually the project proposal that I submitted to George Mason for Humane Studies Fellowship. I recently piloted a free college mentorship program for Arkansas high school students. So I'm running it through the ACT Academy on campus. And the way it works is that students come here for ACT Academy Prep Course, they're here for a week. And while they're here, they're matched with a mentor. They live with the mentor for the week in the dorms, and the mentor is kind of there to answer questions, kind of do a little community group building, kind of raise the esprit de corps, if you will, of the group. And then after the ACT academy is over, and they go back to their hometown, wherever in the state they're from, the mentors then work for me for the entire fall semester helping these kids get ready for college.
19:30 Nathaniel Burke: So that can involve things like helping them fill out the FAFSA, helping them apply for different colleges, helping them make a college list, because the kids that we're targeting... While we do have a pretty balanced representation of students upper class, lower class, White, Black, Hispanic, the kids that really need the help are the first-gen students and the kids coming from the lower socio-economic backgrounds. And so a lot of these kids they don't have anyone in their network that's gone to college, they don't have any parents that have gone to college. So that's where the mentor steps in. So while, many people may look at that project and say, "Okay, that sounds really nice. You're doing a good thing for the community. How is that economics?"
20:07 Nathaniel Burke: While they're here, I actually take them down to our behavior lab, explain all about the economics field and what we do here at Walton, and they go through a bunch of different surveys and some experimental games I mentioned to you before, and we learn empirically what kind of risk preferences these kids have, what kind of time preferences these kids have. And then throughout the experiment, we are randomizing whether or not their mentor is someone of their own racial and ethnic background, and we're also randomizing whether or not the students will get the same general college information that everyone else gets, or if they're gonna get the more specialized information that targets their specific backgrounds. And through this we can learn about how students are reacting to information. There's this assumption that if the information's freely available, people will be able to freely receive it.
20:57 Nathaniel Burke: My hypothesis though is that when you have, for example, a lower-income black male student from a town where no one goes to college, and he gets a brochure from the University of Arkansas main campus, he's gonna have a hard time identifying with the information of brochure, so he's not gonna internalize it. However, if that information is more specialized to him, if the person handing him the information is someone from his community or someone that at least looks like him or talks like him, and the information talks about scholarships, like NAACP scholarships, for example, the Dr. Martin Luther King Foundation, if it talks about the community up here that understands his background, then he's even more likely to accept that information and actually try to enroll in the University of Arkansas versus just saying, "You know, that's not for me, this is for someone that has more money or comes from a college family."
21:47 Matt Waller: So this is similar to your identity research on the reindeer meat in some ways.
21:51 Nathaniel Burke: Yes. Yeah, that's the common theme through all my research is measuring how much someone's identity's impacting their decision-making. Just now we're just starting to get into how someone is going to invest in their own human capital based on their identity.
22:07 Matt Waller: Do you have any findings yet, or are you still in the midst of it?
22:10 Nathaniel Burke: So I just finished a pilot. The pilot, we had 77 students, which was pretty good for a first run, and had 17 mentors working for me who were all fantastic. We found, especially for bi-racial students that they have the strongest identity poll. It's most important to bi-racial students that they have someone that they can identify with on a cultural level. And me myself, I grew up in a mixed ethnicity, mixed race family, I can understand how important that's gonna be, because when I was growing up, I was too black to be white, but I was trying to get an education, I was kind of a nerd, I read a lot, and so, ethno-linguistically, I didn't fit in with my neighborhood. And so I had this consistent identity crisis growing up. And it wasn't until later in life when I started getting mentors that understood what I went through that I started realizing that I could go in this path and it's okay if I don't fit in with everybody.
23:02 Nathaniel Burke: So we did find strong significance for bi-racial students, and I also found that for minority students, on average, the minority students that got information that was relevant to them s have higher aspirations as far as where they applied to school, but even more curiously, students that were minorities, their parents ended up having a higher belief in the student's ability to go to a high-ranking school, if the student had a mentor of the same race as the student. And so this was something I didn't expect to happen. And so, when I adjust the study for the full run of the field experiment I'll pay more attention to, but when we kind of brainstorm, they realized, well the mentors are going out and picking up the students and their hometown, or the parents are dropping them off on campus, depending how close they are to campus. And the parents are meeting the mentors. They're meeting the mentors, they're shaking their hands, they're seeing who the child's mentor is.
23:56 Nathaniel Burke: And I realize a lot of these parents didn't go to college, and they have a lot of concerns about their student going to college, especially when you have a institution where if I'm black, my son or daughter is black, and they're going to school that's only 4% black, I might be concerned about them being able to fit in culturally. But if I meet a mentor from that school and say, "Okay, there's a mentor that looks like my son or daughter, has the same background, from the same area, and they're doing fine," it made the parents feel a lot more confident in the student's ability to succeed.
24:24 Matt Waller: What course do you teach here in the Walton College?
24:27 Nathaniel Burke: This year I'm teaching Principles of Microeconomics. Last year, I taught basic economics and now I'm teaching Principles of Micro, but I put my own spin on it. And so, in addition to teaching all the core stuff that's in the course catalogue, we also learn about the economics of different people. So what I do is I pull from a different culture that exists in the classroom organically, and we talk about issues such as the economics of vices, we talk about economics of discrimination, we talk about economic growth in urban communities and rural communities. And I try to make it more relevant not just to the students in the classroom, but also help them branch out to learn about people that they might not encounter in the University of Arkansas, but they definitely will encounter in the workforce. So it's been a lot of fun though. It's been really interesting to engage with the students here.
25:12 Matt Waller: Well, Nathaniel, thank you so much for taking the time to visit with me about your research and your program.
25:18 Nathaniel Burke: No problem, really glad to come and talk to you. It's really good to get a chance to talk about what I'm doing outside of the kind of the walls of Walton.
[music]
25:27 Matt Waller: Thanks for listening to today's episode of the Be EPIC podcast from the Walton College. You can find us on Google, SoundCloud, iTunes, or look for us wherever you find your podcast. Be sure to subscribe and rate us. You can find current and past episodes by searching "BeEPIC" podcast, one word, that's B-E-E-P-I-C podcast. And now, be epic.
