"Auctions, Privatization, Turkey, Math and Hünkârbeğendi with Professor Isa Hafalir"

In this episode, our guest is Professor of Economics at UTS Business Isa Hafalir. We talked about auction design, Turkish economy, school choice, math Olympiads and why is it hard find good Turkish food outside of Turkey.

Behzod: Hello everyone! Today’s guest is Isa Hafalir. He is a professor of Economics at the University of Technology, Sydney. Before that, he worked for long time at Carnegie Mellon. His PhD is from Pennsylvania State University. Before Penn State he had been a grad student at California Institute of Technology. Welcome to the show!

Isa Hafalir: Thank you!

Behzod: I’ll start right away. You’re an auction theorist, you are a micro theorist. That’s why I want to talk a little bit about micro theory. I am from Uzbekistan and this show will be watched mainly by people in Uzbekistan. We are talking about auctions a lot in a public domain. Why? Because the government wants to sell its assets. To give you a little bit of context everything in Uzbekistan was owned by the government during the Soviet times. Even trivial things like coffee shops or wedding venues, barbershops. Everything was owned. Then government starts selling it little bit by little bit. Still vast majority of assets is owned by the government. Right now, 2020, they want to start selling the government property. My first question will be how should we think about auction theory and auctions when it comes to selling government property?

Isa Hafalir: Auctions are excellent tools for selling or for buying because as long as you have easy entry and you don’t have very tin markets in terms of whoever is going to buy these assets the markets will reveal the tool value of these assets because you don’t have to do a lot of job on your own in terms of understanding what the value of this is. The serious buyers will do the regulations and they will take it to account what the other buyers will think about the value, they will optimally choose and then you will get the best possible value, the fair value. There are a lot of theoretical work showing that if you have… This may be too theoretical but there are papers that show that if you want to sell an item and underspoken some assumptions the best possible auction of selling this item would be to run a standard auction the first price or a second price option with some suitably chosen reserve price. Of course, for this reserve price you would need to know something about potential values of the bidders, the buyers but even the effect of the reserve price is not that huge. Auctions are theoretically proven to be the best mechanisms to sell when you are trying to maximize the revenue. They also have similar properties. If you want to maximize the social welfare, which is the efficiency, the total value generated. Both theoretically and practically proven to be good mechanisms to use but of course under some assumptions. 

Behzod: I see. Some countries even in the west like the UK or Switzerland. You are probably familiar with some empirical work that shows saying the UK when they were selling the communication like 5G and stuff they were pretty successful. They make like 30 billion dollars out of it. And when they did It in Switzerland they made like 30 million dollars, thousand times a difference. There is a huge difference between how you design auctions. What the Switzerland missed? What the FCC in the United States when they tried to sell 5G networks they talked about collusion, they talked about the firms  not behaving optimally in the models. What can go wrong in auctions?

Isa Hafalir: A lot of things can go wrong. The experts in this area some of the names are Paul Clumporth in the UK and Ken Binmore in the UK, in the US Paul Milgrom, Born Sausebell, Peter Cramton, they are all theorists who had basically consulted the FCC in the US. Ken Binmore had done a lot of work in Europe. They had written about what are the successes and failures in different spectrum auctions. Their writings 10-15 years ago were mostly about 2G, 3G. These are not simple tasks. When you sell spectrum this is not like one spectrum that you sell for the whole country. There are the countries divided into regions and then in those regions you have some dense areas some non-dense areas. There are some potential entrance, big firms that can provide for the whole country. There are some smaller firms that would like to be regional and provide this service to small regions. Big players, small players have different incentives, different capabilities. I think the bottom line is the thickness and the thinness of the market. If you do not have enough players attracted. Let’s say you are selling 5 licenses and there are only 5 bidders. It will be a disaster. But if you have 6-7, who are serious this going to be ok, this going to be quite good. If you have even more, this going to be better. There are also issues about collusion. They don’t have to provide paper trails about collusion but they can implicit collusion.

Behzod: When they show the price…

Isa Hafalir: There are very nice examples. There are signals during the auction. These auctions take place a couple of days. There are some rules about how much you can increase. Sometimes the rule says the bid size has to increase by at least 10 percent and they choose the number exactly value increases 10 percent it becomes the value of other asset. They signal that you go by that asset instead of this asset and leave me alone here. A lot of things can go wrong and the bottom line these researchers have said is the following. Devil is in details. If you try to take one auction for a month that has been done very successfully in one country and try to adapt it exactly the same way in another country it maybe a disaster. Small difference that may have a huge effect. For now, this is what I can say.

Behzod: If you are designing auction or a group of you are designing auctions they think about few goals. Social welfare goal, revenue maximizing goal and then there is a goal that can be neither social efficient nor revenue maximizing. If we are selling government owned to company we want to keep the number of employers constant, adding some additional constraints onto the price. In that kind of world when you are filling the market naturally as not a lot of people want to sign up because there are so many requirements. How you do you want to tell to the public that selling an auction in a socially maximizing way will translate into firing most of the people. How do you have to think about that?

Isa Hafalir: When you change the objectives, it may give different results. Standard objectives are usually either revenue maximizing or welfare maximizing in this context. Welfare maximization here makes sense. But even the revenue maximization would need to take into account some other issues like. Other than firing people you would need to make sure there is no monopoly in the market. There is already a big player who is very efficient in providing the service. This firm has a lot of incentives to buy at a very high price and overbid the others and you can generate more revenue this way. It looks like sufficient because this firm has already a lot of towers and so on. But at the same time if this sector is not regulated thenthis firm will become the monopoly and the customers will bear the consequences because they can put very high marginal profits. Like some countries have some constraints to make sure there is enough players. They say that you cannot buy more than this number of licenses. Even for developed countries who are trying to do this very professionally there are different ways. The firms can overcome this. There are some cases they want to make sure after they buy their license no other firm can buy license. Resale was banned. What happened is in one year rather than buying this license some other firm actually bought the firm who bought the license. According to the rules, this was ok and this went to the courts and I don’t remember the ultimate decision but what I am trying to say is that this is a very complicated problem. There maybe a lot of constraints. Maybe I am not answering your exact question about issues about people getting fired but lots of issues have to be tackled. There is a recent auction that has been designed and implemented it’s called the incentive auction. In the past they actually had sold licenses to the TV companies but TV companies do not need this spectrum anymore and mobile companies they needed. At the same time they bought this anews spectrum from the TV companies, rebundled it in an optimal way and then sold it to the mobile companies. The US government. There are mass of constraints going on and they use auction theorists and apprentice research people to make sure this is all done in an optimal way.

Isa Hafalir: Socially maximizing auction is simple. It is when you adapt all the utilities of all the players. The whole country basically.

Behzod: Like consumers…

Isa Hafalir: Yeah.

Behzod: So revenue maximizing is when the government is selling the assets.

Isa Hafalir: Yes. And they try to maximize the money they get.

Behzod: Here is the question. Is revenue maximization a goal? The government wants to maximize amount of money they get for the asset. And socially maximizing. Are they always the same or do they contradict? If they contradict under what circumstances?

Isa Hafalir: They are not always the same. Even in the very simple set up with symmetric bidders what we call independent private values. The values of the bidders are not correlated. They are private, they just care about their own valuation. Even in this case the social welfare maximizing will be requiring you not to set the reserve price. Revenue maximizing will require you to set the reserve price. The things are more complicated if you have common values. My value for this asset depends not only on my own valuation but I also cared about your valuation. In this cased there are differences between simultaneous auction formats. The bids are given at the same time in close envelopes versus sequential auctions which can be like ascending price or descending price auctions. Under some assumptions some auction formats are shown to be better than the others under some assumptions easier but if you want to say something about practical worth it is a difficult question. It is a question that a group of experts should come together and think about how these markets are working what are the important properties here how these bidders will behave. In the standard game theory, we always assume that the bidders are perfectly rational but sometimes they may not be that rational. Lots of issues. In developed countries like in the US and Europe what happens is basically when you are designing the auction government is having experts working for how this will work out. The companies also hire experts to determine the optimal strategy for them and the things are quite ok. But when you are in the country in which you are not quite sure what’s going on, what has been done under the back of the table there are issues these are even more difficult.

Behzod: Let me ask you a historical question about auctions in a former Soviet Union. When Soviet Union collapsed, they started selling government property sometimes with the bad design, sometimes with the horrible design, something in between. Usually was between bad and really terrible. One of the ways they sold this and the public kind of bought into. Some Russian economists and economists in Soviet Union they call it beauty pageants meaning basically the government bureaucrat sit down they look through the bidders and say this guy will get it.

Isa Hafalir: Form of corruption.

Behzod: Yeah, this is a form of corruption basically. But for a lot of public when you are giving it to somebody you know and they make promises, invest thing, 10 millions. I can give a recent example. In Uzbekistan one of the refrigerator factories was sold to a bidder for a zero price but with promise of investing 50 million or something. When you talk to the bureaucrat and say why are you guys doing this they say if we sell it fair and square they will buy it for amount of money not zero but first we don’t know what they going to do with it whether they will keep it a refrigerator factory, second we don’t know whether they going to fire people, and third we don’t know how much investments that would make. How would you tackle this?

Isa Hafalir: Basically what’s going on here is that environment is as simple as everybody is. This is a procurement environment. There is something I need to do and I need someone to do this and the bidders are defraciated between the abilities of doing so. They are not all providing the same plan. This one is saying I am doing this and this one is saying I am doing this and there’s a difference between them. I think beauty pageant are used.

Behzod: Are they good at all? Can we create circumstances when they are good?

Isa Hafalir: It could be good if it’s benevolent. If you can make sure they are really doing for public interest rather their own interest. But it’s impossible to check.

Behzod: Theoretically. Sorry for interrupting you. Let’s say there is a beauty pageant, super benevolent. Like angel came from sky and is selling it. What I was thinking is the results of the beauty pageant at best would be equal to the results of good auction. Benevolent person could do better than an ideal auction.

Isa Hafalir: It depends. The problem of angel coming is that angel should tell transparently in advance, what selection criteria it is going to use to the agents. Angel also tells them he is basically going to use that auction formats this beauty pageant will become an auction format and is going to generate the same outcome. But most of the time what happens is that it’s not clear what they are looking for. It would be fair and efficient working properly. Things should be clearly stated in a transparent way. So if I tell here is the criteria. You will all come up with different projects. Some of them will say I will give you zero dollars but I will invest 50 million and someone will tell you I will give you 10 million but I will only invest 30 million. If you can have a clear and transparent way of ranking this, these bidders can basically compete with each other. If they don’t know what the selection criteria then the competition will be not tough. You want to make sure there’s competition. The competition will do the work.

Behzod: To do like recap you are saying this. The limits of the beauty pageant goes to the auction. The first welfare theory in which like the best allocation is really close to allocating the resources. Let’s talk about Turkey. You are from Turkey. Lets talk about Turkish economy. I am from Uzbekistan, you are from Turkey. When Uzbekistan got independence Turkey was the first country to formally establish diplomatic relationship with our country. It was very important at the beginning.  Relationship soured for a while and now they are pretty good. But the idea why I am asking about Turkey our kind of elite and leadership Turkish economic model and way of thinking about the economy was the main idea for first 6 years. And now it’s coming back. When Erdogan visited Uzbekistan our president said Turkish economy already went through these steps and we want them to teach us those steps that’s why I am saying Turkish economy is related and relevant for our listeners. And right now as an observer who is not in Turkey or in Uzbekistan I can see the Turkish economy slowing down. There are a lot fundamental institutional problems, political risks are going up. Generally, are you pessimistic or optimistic about Turkish economy next 10 years, not 50 years, 10 years?

Isa Hafalir: 10 years it depends on what’s going to happen in political sphere. First of all I am not an expert in macroeconomics. My understanding mainly based on reading experts in this area Berunu Jemulu…

Behzod: What about Yavas?

Isa Hafalir: Yes. Yu Kainak. AKP in the…

Behzod: AKP is the main Turkish party.

Isa Hafalir: In the first and some of their second term – first 5-6 years. They actually had done quite well in every aspect including the economy. Ali Babacan who is now becoming an opposition leader was one of the very important people when this was happening. Turkish economy is not really producing new technology. They are adapting technology and selling to less developed countries.

Behzod: Like ours.

Isa Hafalir: They heavily depend on the foreign investments. Even when there was crisis there was enough trust in the Turkish economy. Turkish economy had not done too bad. There was still money coming from outside. This has changed a lot. All the unfortunate injustices done by the party. They are becoming more and more autocratic. The progress has been very negative in the last 10 years. It doesn’t look like it’s going to change because they don’t listen.

Behzod: Let’s talk about the first part. They got something right at the beginning. AKP came to power, they got something right. I was a kid. I went to Turkey first time in my life in 2003 or 2004 when I went few times. In 2014 or 2013 I was in Istanbul as a foreigner and I observed how different it was in 2004 or 2003 it was a completely different country than it was in 2014. Even I could sense it. They got something right. What they got right?

Isa Hafalir: I think what they got right was for the people who knew what they are doing in power Babacan and Mechinshek and Erdonbashi governor of the Turkish Central Bank. They didn’t produce new technology but they started the trust of Turkish economy. There is some experts, people produced. For a long time what they try to do is they try to keep Turkish economy alive by pumping money and giving the opportunities for the…

Behzod: Foreign investors…

Isa Hafalir: Not for the foreign investors. I am talking about all the constructions. People are impressed with all of the new bridges and the tunnels and the buildings. And even in the small towns in Turkey there are lots of apartments. This actually had money moving in the economy. But it’s not producing movable value.

Behzod: I see. You are saying construction is not a productive sector in economy. There is like zero economists who would say no. Let me tell you a story right now. In Uzbekistan we have a construction boom. A lot of constructions happened by Turkish construction people. Our government is financing it. Now we are pumping money to construction. It was a successfully of Turkey. We want to repeat that and construct all these bridges, hotels, apartment buildings and stuff. We are going to hit the same wall.

Isa Hafalir: You are going to hit the same wall. Its not the sustainable goal. You will be trapped. It will help but there is an upper limit that can get you there. Turkish economy is struggling and AKP’s policies, paints are already falling out wall. They are playing with the numbers a little bit in terms of the inflation.

Behzod: Do they play with the numbers?

Isa Hafalir: Some theorists say yes.

Behzod: I am from the country that cooked up inflation…

Isa Hafalir: It’s not too bad but manipulative. If they give this job to a neutral third party I am sure they will find some problems. I am not saying they are making up numbers but they are changing the methodology to calculate this that now doesn’t make sense. Property should be independent. The Turkish Central Bank hasn’t been independent for a long time. That’s a big problem.

Behzod: Are you sure it hasn’t been independent from 2002 or something till like 2013? Was it more independent?

Isa Hafalir: There is different independence. I am not saying they are just

Behzod: Completely part of the government…

Isa Hafalir: Yeah. They are struggling with the government. The struggle is getting harder and harder.

Behzod: What I thought though when they were independent is that if you fire somebody it reveals that institution was independent. You can’t buy two things simultaneously. If you believe the Central Bank was part of the government then it has to fall to whatever the government was telling. And the fact that they fired they want fully complying… Logically thinking. I am not familiar how the Central Bank in Turkey works but as an observer, oh the guy was fired. Probably that means…

Isa Hafalir: How about this theory. This guy wants to implement A and the government wants him to implement B. And he is ok with implementing B and he continues implementing B. But he says C is crazy. But the government says implement C. Then he says I cannot implement C and then he is fired.

Behzod: Correct. Still there is independence, right?

Isa Hafalir: There is independence but there is big influence. They are under big pressure.

Behzod: I wanted to ask about you research. The thing I find quite interesting was that. I think that theorists have a very good toolkit and when they try to solve problems. Paper where economists are some kind like engineers. You are trying to look through many aspects of human life. One part that kind of struck me you briefly discuss quotas and affirmative action. What is your general view of the affirmative action, was it a mistake or not?

Isa Hafalir: Affirmative action? It’s not a mistake. Basically affirmative action is defined as for type of people it can be gender it can be race it can be ethnicity it can be religion. There are people unfortunately who were discriminated in the past and we should give them more chances to make it fair. It is debatable subject. Some people say there is discrimination, disadvantage because of these policies. I hear what they say but I think that due to historical discrimination you need to make sure the underrepresented groups should be given more priority in schools in labor markets everywhere. I think that affirmative action policies are good policies.

Behzod: Do you know this paper by … and his co-authors they send out resumes of black pope and white people and black people don’t get a lot of call back. And some people in a blogger sphere or whatever argue that this is mainly because of statistical discrimination and affirmative action. If you are holding constantly a degree, a black guy has a Harvard degree and a white guy has a Harvard degree. And then they are like oh because he was from specific background. He is not as successful and she is not as successful. What do you think of that? Hurt people whom it should help basically?

Isa Hafalir: The alternative in which these guys won’t even have Harvard in their CV they have only Penn Sate in their CV.  Then what will happen is that do you think labor market will respond in a way this Harvard’s white person is worse. I am not saying the reason for less callback for the same CV or the African-American black name is due to the fact they are white supremacists who hates black people. Things like that could be in their minds. What they had done is wrong and we need to do even more affirmative action. The market will respond in a way making this even not effective.

Behzod: Let me ask you a different question. In Uzbekistan they want to have quarters on woman politicians. They want have 40 percent of the parliament to be woman. There will be within woman elections. A lot of woman who have high government jobs and stuff some of them and some men raise concerns about other woman’s fate. Right now if you are woman and if you are in a leadership position everybody understands that you are there for your talents, abilities, hard work and all sort of that. Now imagine rooms full of men will become half women, 40 percent, I am not sure what was the quota. Then people who get in without the quota would be inside the quota. Would people treat them seriously?

Isa Hafalir: Those are valid issues, not just issues specific to Uzbekistan. For instance, in India there is a very rigid cast system. In universities have implemented these quotas for different casts. They have backwards cast which is the lowest cast and if you are this cast members even in medical schools you can get into the medical school with quite low score and people are complaining that I don’t want to be treated by a doctor who is not capable and has become a doctor just because he was from a backward cast. This is a public debate issue. I cannot say that they are wrong or they are right. These ratios should be carefully chosen. These ratios should not be exaggerated too much. Even for incapable woman, for Uzbekistan case be in the power positions. There are lots of ceilings and glass ceilings for minorities in terms of race or ethnicity or religion and also gender like woman. Woman should be given more opportunities. How much more? It’s a tough question. I agree that giving too much may result in unforeseen other problems. But I actually fully support this kind of initiative but in 40% rights ratio.

Behzod: You are saying adding like epsilon more.

Isa Hafalir: Not epsilon but something like…

Behzod: What is the epsilon?

Isa Hafalir: There is an optimal way in which you…

Behzod: What you are saying it is a maximisable function.

Isa Hafalir: I think so. That maximsation is not zero. They should be given. That’s my view.

Behzod: You wrote about school choice. What was the reason of writing about school choice? What’s your contribution to school choice?

Isa Hafalir: Before that let me start with a joke that I like telling. There are so many Turkish economists who have…

Behzod: Have school choice.

Isa Hafalir: matching theory and school choice. One of the biggest names here is Tayfun Sonmez. My first paper was a matching paper then I moved from Caltech to Penn State I started working with my advisor Vijay Krishna who was an auction theorist I started working auctions and mechanism design and  for a long time I wrote papers about auctions and mechanism design. Then we hired my old friend Yanmas who got his PhD from Stanford and we got together and we started working on this matching problem. When I completed this paper I told my advisor I wrote a matching paper. He told me good, now you can renew your Turkish passport. We wrote two papers with Bumin and Mohammed. We were actually teammates in 97 Turkish IMO team.

Behzod: IMO, really? What was your medal?

Isa Hafalir: I get silver. Oh, no sorry. I got bronze.

Behzod: Did you solve the sixth question or third question? Could you solve any of these?

Isa Hafalir: 97?

Behzod: No, you know there are three question a day. There is a third question usually geometry and sixth question.

Isa Hafalir: I was good at geometry. I don’t remember my scores exactly but I was one score below a silver medal.

Behzod: So you got the highest bronze?

Isa Hafalir: Yes.

Behzod: You went only once to IMO?

Isa Hafalir: Yes, my friend Bomin went 4 times and he got one gold and two silver.

Behzod: From this answer, I have three questions I wanted to ask. First, the IMO question. Why some countries do so much better in IMO than others. Like Iran is really good. Turkey is not so good. Not even as good as say Kazakhstan.

Isa Hafalir: Like all ex Russian countries.

Behzod: What do you think about this? That’s the first question I ask. Let me just start. Why do you think some countries do much better I IMO than others? And is there anything we can learn from the countries’ education system from their IMO performance?

Isa Hafalir: Big countries like China, US, they have good resources. Big pool of students to choose from and some countries like Romania.

Behzod: But it’s traditionally very good. Because IMO started like Eastern-European. Hungary, Romania.

Isa Hafalir: Yeah, they are excellent in math. Turkey…

Behzod: Why Turkey is so bad?

Isa Hafalir: It’s not so bad. There has been years after, they got three gold medals in one year.

Behzod: Singapore is a tiny country but they have gold medals sometimes.

Isa Hafalir: Turkey ought to be better but I disagree, respectfully disagree that it’s bad.

Behzod: I mean for it’s sake. For its population, for its GDP, if you control for a lot of these things I think Turkey still punches below its weight.

Isa Hafalir: It’s all about getting smart students who love math and giving them very good training. Excellent good training. And some countries are historically very good at it. Turkey had been quite good, it has gotten badder, and it may have gotten worse because Turkish education system in the last 5 years.

Behzod: Is it correlated though?

Isa Hafalir: It may not be because this education is not the average education. Me and my friend Booing got interested in this problem because we were reading another paper and we thought this paper is missing something.

Behzod: So you renewed your passport basically.

Isa Hafalir: Basically they were publishing 2013 and 2014 and there were more than 160 citations which is quite good in our theory. For engineering it’s small, 160 is nothing but for our cases…

Behzod: The second question I wanted to ask you is why Turkey has so many good economists. And the third question would be why turkey has so many good theorists. But let me start with the second one. In my opinion, Turkey has so many good economists empirically why. I am thinking for a country of its size its incoming general education level Turkey punches well above its weight in economics profession. Can you tell me why? What do you think is a secret?

Isa Hafalir: Again these all boils down to a few people. That makes a huge effect. There are two guys Semi Quaray and Murat Sertel. Murat Sertel passed away a while ago. He got his PhD from MIT and he was a faculty in US and then he came to Bosphorus University and then Semi Quaray was about to get his PhD in math and then Murat Sertel became one of his PhD committee members. They became very good friends. Semi Quaray decided to switch to economics and he became an economics professor. He was the main person for Turkish IMO team too. Murat Sertel and Semi Quaray get produced so many students. These are people who are academically and personally very nice and very supportive of the students. In the past there has not been a lot Turkish undergraduate students who were able to get into good PhD programs. After … had done very well In Rochester Rochester wanted to have more and more Turkish students. When… done so well in making theory more and more smart Turks said I can do this too.

Behzod: Do you think … was more important than Jemolulu in Turkish economics?

Isa Hafalir: Jemolulu is a completely different level. Haroni Jemolulu he had graduated from Robert college and then he went for his PhD in LSE. His PhD was good enough to produce three PhDs. … in Princton, … in MIT …, the new generation. All their names influenced Murat Sertel, myself influenced by Semi Quaray, many influenced by Typerson. This is a snowball effect but it is decreasing. Nowadays micro theory is not very well.

Behzod: But there are still Turkish economists in every department.

Isa Hafalir: There is but there aren’t new PhD candidates who are coming out from top school’s good schools who are doing well on the market and doing theory. Most of them doing other areas like macro for instance. Macro has become quite good. There are not too many Turks who are doing economics…

Behzod: Generally speaking, what I am saying is look at Germany. Germany has maybe a little bit more population much better education system. Very few famous German economists. This is my theory you can disagree with it. What I think is if economy of the country sucks, they produce good economists. I have good examples. Look at South Korea, Japan.

Isa Hafalir: Big countries like US, China, India, people have a lot of smart people…

Behzod: But even in the countries like Germany. Compare France and Germany. France has much worse economy, much better economists. Look at Italy. Very bad economy, for like 40 years stagnating, very good economists. Spain, even Catalonia has so many good economists.

Isa Hafalir: You can write an economics paper on this.

Behzod: But generally, I can’t believe how weird it is. Germany has a tremendous economy, very good. Very few German economists. South Korea has very good economy in the last 50 years. Very few South Korean famous economists. Singapore. I lived there. It is not big, 5-6 millions, zero good economists.

Isa Hafalir: When you say good economists, you need to explain what you mean exactly like different levels. If we locate Nobel laureates, they are almost all Americans. This has butterfly effect. The small things can change a lot of things. Murat Sertel, Semi Quaray influence smart kids to do PhD in economics can have a big effect.  They will have students, they will send their students.

Behzod: I was thinking like Argentina, very bad economy, very good economists. Chili for example. Chili is much richer than Argentina right now and they have Chicago boys whatever. There are much more Argentinians than Chileans in academy. If you look at Europe or South America or everywhere you will find countries with worse economies having better economists. And Russia lately has bad economy and very good economists. So we talked about stuff in terms of school choice. Now you have this paper with Yavas and one more co-athor about auctions and negotiated sale. We know how auctions are really good but in that paper you basically say those two thing auctions and negotiated sale can be comparable. Can you tell us a little bit about it.

Isa Hafalir: In the US most of the property sales are via negotiated sale. Pool set price, then you come up with the offer, then this is accepted or they just wait for another offer. Except for foreclosure properties. Whereas in Australia where I am now most of the regular house sales are auctions. In Singapore, they both appear quite commonly. Honestly, … and his  student had this paper I am looking at the data for different periods in terms of the seasons and whether the economists doing good or not these prices of the sales for the auctions negotiated sales write a model about it. In that model, what I look at is observed prices. Expected revenue is not the same as the observed prices because sometimes house is not sold. There can be differences. In terms of expected revenue in the model that I consider auction is supposed to be always better than negotiated sale. But in terms of observe prices sometimes negotiated sale may be better. Of course, in the real world auctions and negotiated sale have different properties that is not sceptured in the model. For auctions you need to make sure everybody shows up at the same time. For negotiated sale, maybe you can get larger audience.

Behzod: Do you think auctions are better in terms of dealing with the discrimination and all sorts of other bad things like negotiated sale sometimes may be personal selling a house ‘’hey, this guy is from my high school’’.

Isa Hafalir: In the real world, there is some kind of auction fever. They may be get charred away with extra. If you give them sometime and take a day how much money you will offer to this house maybe they will not offer that much but at the auction time they say ‘’Ok, I will give you 50 thousand dollars more. It’s not a big deal.’’

Behzod: What do you think of behavioral economics in this sense.

Isa Hafalir: Behavioral economics is a very important area. I do not personally work in behavioral economics but…

Behzod: As a theorist what do you think? Can you incorporate some behavioral insights and experiments?

Isa Hafalir: Yes. People do. I have not worked on papers except for one paper. I had loss aversion and risk aversion they are actually seem to be more like rational agents having different utility functions rather than behavioral agents. And there are different behavioral theorists. Some of them are quite close to standard economics and some of them are quite close to psychology. Behavioral economics they play very well on the market right now and for a right reason it is an important topic. More and more people will be doing this because homo economicus it’s important, it’s not obsolete but it is not enough.

Behzod: Let me push back a little bit on this. We all know that Earth is round. But when you are building a building if you are an engineer you assume that Earth is flat. For sake of building it it’s pretty good approximation of the reality, that Earth is flat. So you build your building as if Earth is flat. I think about economic theory this way too. Sure people are pretty rational in many many ways but in terms of designing auctions or stuff like that the approximation that we do what we call homo economicus is pretty good. What do you think? Do you agree?

Isa Hafalir: Long time ago I attended this conference in MU in which... Ken Arrow was there and he gave an example of a plane flying. For a plane to fly air economics is good enough. You don't need to know all the details about electronic systems in the plane to have a good understanding why the plane will fly. He was referring basic economic theory for aerodynamics and rest of the stuff is details similar to what you were saying. There are cases in which economic theory is quite successful especially about individual decision-making. For instance, like retirement, how much people retire. How much they exercise, how much money they put into retire make out, how do they deal with addiction. They are really psychology.

Behzod: So what is the value of theory then in the modern world? Economic theory.

Isa Hafalir: It is very important. Everybody should know. They should build something on top of it but it’s the basis.

Behzod: What do you think undergraduate education right now? It’s very heavily theory focused in most of departments. Do you think it will change to empirics more or not?

Isa Hafalir: I don’t think so.

Behzod: What do you think is the most important contribution of economic theorists to human wellbeing? In the last 50 years. Empiricists might tell you some stuff, like this is correlating with that. But as a theorist I have pretty good answers like kidney exchange or… 

Isa Hafalir: Mathematicians laid foundations for the whole nations sciences. 

Behzod: For the whole world yeah. 

Isa Hafalir: I can tell you examples of areas like market design in which billions of spectrum licenses have been sold and online auctions are everywhere. The … theory. School choice programs are helping kids, another question Typerson …, the organ exchange, the kidney exchange, liver exchange. Actually, in that paper Typhon claims and I believe him that Typhon’s contribution is larger because that paper… 

Behzod: The paper that got the Nobel prize. 

Isa Hafalir: That's exaggerating but. Typhon I think rightly claims he should be the main person. 

Behzod: Does Aura thinks the same way or not? 

Isa Hafalir: I don’t think so. 

Behzod: Cause it’s rare. Do you know Newey-West estimator in econometrics? We had econometrics teacher, the west the guy is in this concept. Jack Porbor was teaching the class and he said Newey he knew West and Newey told him about West saying that estimator most of the work was done by West, by other author. When he asked Wes,t he said most of the work was done by Newey. Jack Porbor said it was the first time in my entire career when I heard somebody claiming less when it actually is. Anyway, let me ask you question about Indian food. Is Indian food better in Australia or in the US?

Isa Hafalir: I think the US. 

Behzod: Why, what is your theory? 

Isa Hafalir: I was not able to go to Indian neighborhood in Australia where the Indian food is good but if you go to a random place in Australia, it's not good. Like Japanese food, Thai food, Korean food, Chinese food all Eastern countries foods they are much better in Australia than in the US. 

Behzod: What's your theory about food, why some foods are replicable? I think the Turkish food is not very replicable. Turkish food in Australia and America is pretty bad.  

Isa Hafalir: Turks don't process their food a lot. They need good quality of meat. We don't add a lot of spices. If you have good quality meat you just grill it and it's good. And you do not always have it here.

Behzod: The quality of ingredients is very important in some cuisines than in others.

Isa Hafalir: Turkish cuisine is closer to Japanese cuisine. Japanese cuisine and Turkish cuisine are simple; they don't mix up a lot of stuff. You just add some ground beef and when you make …

Behzod: What about Hünkârbeğendi? 

Isa Hafalir: It’s complicated. It’s the dish that I can cook well. 

Behzod: Me too but why is it so bad in restaurants even in Turkey?  

Isa Hafalir: It's kind of home food. You use good butter… 

Behzod: But hear me out. I never paid money for a good …. Good … either I cooked or somebody cooked. In Uzbekistan I ate, in Turkey I ate in US I ate. Never I ate good …. What other foods are always bad like this?  

Isa Hafalir: Iscander is always bad. Even in Turkey. 

Behzod: I not talking about levels. There are foods that are usually bad. Like for me … is usually bad. Are there foods that are usually good in Turkish cuisine? 

Isa Hafalir: In Turkey if you go to regular places it is not going to be bad.  

Behzod: So here is  the part of the show when I ask short questions and short answers. Towards the end of the show. You have to choose. Ali Babacan or Shamshak?  

Isa Hafalir: Babacan. 

Behzod: Elif Shafak or Orhan Pamuk? 

Isa Hafalir: Elif Shafak.

Behzod: Do you hate both? 

Isa Hafalir: Yeah I had problems with both. I like Elif Shafak’s writing more than Orhan Pamuk so. The thing is that in the last novel she claimed to steal from some other thing that I didin’t know. I am not sure so.. But I like her style more. 

Behzod: What's your favourite Pamuk title? 

Isa Hafalir: Snow. 

Behzod: Ok. Medal East Technical University or Bosphorus University?  

Isa Hafalir: Bosphorus. 

Behzod: Why? 

Isa Hafalir: Nicer view.  

Behzod: Caltech or MIT? 

Isa Hafalir:  MIT. 

Behzod:  Why? That is really surprising.

Isa Hafalir: There are different levels. If you look at MIT economics, department it’s better than Caltech department. In terms of undergraduate education, they are both good. Caltech undergraduates are quite smart but at the same time, they are quite kimberly. 

Behzod:  Von Neumann or John Nash?

Isa Hafalir: Von Neumann, no question. He is a genius. 

Behzod:  Hal Varian or Paul Milgrom? 

Isa Hafalir: Paul Milgrom because of my personal. 

Behzod: Thanks a lot for your time. I really appreciate our talk.

Isa Hafalir: Thank you.