Interview transcript

On behalf of the Nobel e-Museum and Nobel Foundation I have the pleasure of welcoming the winners of the 2003 Bank of Sweden Prize in Economic Sciences in memory of Alfred Nobel. The winners are Professor Robert Engle, New York University and Professor Clive Granger, University of California, San Diego. I am Karl-Gustaf Lőfgren, Professor of economics at Umeå University and next to me is Anne-Sophie Crépin who is an Assistant professor of economics at Stockholm University. I’ll ask Anne-Sophie to start the conversation.

To my knowledge none of you has studied economics as undergraduates and this means that you have a background in another discipline. How did that affect your research when you were doing?

Robert Engle: In fact, when I was an undergraduate, I was a physics major and I had lots of room mates and friends who studied economics and they kept telling me I should take an economics class so I could find out what this was really like. So, my senior year I had a choice between taking a religion class and economics class because I had an extra spot, and I took economics and it was fascinating. That was my first actual experience and that’s probably why I ended up doing the switch. But then when I did my switch from physics to economics I took pretty much a full set of undergraduate classes on a listening basis while I was taking the graduate courses.

Clive Granger: I had just one third of my first undergraduate year in economics which I enjoyed and it was a very non-mathematical approach. I kept trying to translate into my mathematics that I was more used to and I had trouble doing that, but I enjoyed what I did. I always felt that it was both a disadvantage not to know economics and an advantage because economists think about things differently than everybody else, but also when I came to economics proper I came to it in a different way than all my colleagues did and I think that added an extra dimension to the approach. Then in pair with economists together we would then have a wider way of looking at things and so it was both a disadvantage and an advantage.

Has any particular teacher really been important to you? As an undergraduate or graduate?

Clive Granger: Yes, when I was undergraduate my professor was Brian Tew who was a very good macroeconomist in England and he was one of these people who was totally non-analytical but he was appreciative of mathematical methods. He was very /- – -/ believed the future was in a more mathematical, technical approach to economics. At Princeton I worked with Oscar Morgenstern and he was a great man and had very wide interests and again he didn’t teach me any game theory, I learnt some again just being around him and others. But I found that his leadership was just magnificent.

Robert Engle: I guess the teachers that I would like to mention were my couple of teachers at Cornell in graduate school. The very first day I arrived there I met my thesis supervisor whose name is Ta Chung Liu or as you would say in Chinese Liu Ta Chung and he was a wonderful mentor to me. He was interested in econometrics as a tool to do real problems. He was building models of the US economy and the Chinese economy and you could really see that this was a tool that could solve problems. Then two other men that were really very important were Berndt Stigum who is now in Oslo and was really a great inspiration in teaching a lot of the advanced econometrics that I later used, and John Fei who really gave me my introduction to some of the more traditional economic areas.

Some researchers work in teams and some others do more alone work, how does a usual workday look for you?

Robert Engle: I write a surprisingly large number of papers with my students. I like taking a graduate student and just starting on a project at early stages of their career and then slowly it develops and we get a little further and little further and little further and by the time maybe two years has gone by we have a paper, or maybe three years. Then pretty soon maybe there’s a new important idea that they do. That’s one way I do it. But the other way, which is a wonderful way to do it, is to collaborate with colleagues like Clive and we’ve had an awful lot of wonderful times just talking about big problems and how could you solve this and how could we solve that? I’ve seen this thing and what do you make of this? I don’t know what I make of that, what do you make of it? We just kind of go round and round in circles and all of a sudden there’s something new there.

Clive Granger: I second that totally. Occasionally I spend time just by myself playing with a model and just trying different alternative approaches. But I’ve had over 80 collaborators with my publications over the years, so you see I get along well with people and together we produce. When I was at Nottingham recently, I was there just by myself for quite some years and then Paul Newbold arrived and the difference in my ability to produce and to get my ideas was greatly enhanced by two of us, two of us was much better than two single individuals. Then I moved to San Diego and then I had a year there where I was by myself and then Rob came along and suddenly everything blossomed and we could interact and be very productive. Then we have some very great students come along and the whole thing was just built up. So the answer is there’s no single answer, we do both individual work to initially begin a project but then the discussion and interaction and throwing ideas out to people is very nice. One thing we did at San Diego was to have a regular lunchtime meeting every Tuesday of all the econometricians and we always tried to make that meeting and it was a totally un-programmed meeting, we just would go along and see what one talk about. It was a great help to people, sometimes we’d talk about nothing. just sport or something, but other days it was a really interesting meeting and visitors would come along to it and that was I think a helpful thing to do.

Robert Engle: We also did some gossip at these meetings.

Clive Granger: Don’t tell that!

Robert Engle: Who was running around with whose wife or switching universities or what was happening so … a lot of things happened!

According to Nobel you are supposed to have done a discovery to be awarded the prize and I think this time it is obvious that both cointegration and ARCH is really a discovery, but when did you find out or how did you find out that this was really something that could change the way statistics and also financial econometrics could be?

Clive Granger: Certainly the cointegration idea was one of these things that once it became clear that we understood what we’d found and what the implication of it were to other things that it solved many problems. Things that we were being puzzled about beforehand, we were seeing papers which were inconclusive and people didn’t know quite what to do with their data, suddenly we found we found we could solve all those problems. It became clear how to explain all many things. I think almost within days we both realised that this was going to be an important discovery. Exactly how you sell that is another question but I think we had no doubt how important are the question but we were excited by it and we quickly told people about it and other people were excited by it. Word spread quite quickly on that, on this cointegration.

What about ARCH?

Robert Engle: Arch really came about when I was on leave on sabbatical from San Diego at the London School of Economics but it was stimulated by things we had been doing in San Diego beforehand and it was a project I was working on and kind of in the back of my mind. I had a project in the front of my mind that I felt like I had to do first before I could really think about this time-varying volatility question. Then my wonderful student Mark Watson wrote me a letter I guess, couldn’t have been email in those days, saying that he really wanted to figure out the first problem that I felt I had to solve first for his thesis. I thought to myself, hm, if I let him work on that then I can try to work on this other thing that I’m interested in and all of a sudden I had days of empty time ahead of me and I had these ideas that I wanted to build models of time-varying variances and I had a lot of external stimulation from people at the London School of Economics and all of a sudden the idea came, got put together and it was just very exciting. I did feel like that was important too although that is an idea that it seems to me it took a lot more time for the profession to realise it was interesting than the cointegration idea where everybody just jumped on it.

You seem to talk a lot about students and I wonder what would the dream student be for you?

Robert Engle: You know students are very deceptive, it’s not necessarily the ones that do the best in class that are best to work with. What I find is that there is this transition that students go through, that at the beginning you’re explaining things to them. I always like them to have taken my class so that we’re talking a common language. Then at the beginning I keep explaining to them and I always like talking about things I don’t understand, how could we think about it, this way or that way? They have something to say or they don’t have something to say. Then in a few short months all of a sudden they’re explaining things to me or maybe a year or two later: Could you explain that to me again? They’ll tell me again sort of why this is the right answer. That’s the dream student for me.

Clive Granger: I find some students are very compatible, that you talk to them and you’re very comfortable talking to them. They’re just naturally intelligent and they absorb what’s being said and initially they’re taking everything from you. Then slowly your relationship changes and you find that they’re doing more than you asked, they’re bringing back to you more than you expected and they have their own results which you say Wow, I didn’t realise that was going to occur from this problem. That’s great and you’ve established somebody on the road to doing good research and you know they’re going to have good ideas from thereon.

Robert Engle: Another thing to say along the same lines is that when you’re working on something that’s new and actually that’s mostly what we’re doing, you’re headed into the uncharted area so you don’t really know what’s going to happen and some students get stuck a lot and you spend all your time with them trying to get them unstuck. Other people seem to avoid getting stuck, they figure out a way around it or get a good idea sometimes that you sort of think back on, you say Wow, you know we really made a lot of progress, versus in the first case Oh, there are all these roads blocks, road blocks, road blocks, we couldn’t get anywhere. I think that’s actually an important quality that students can have and we have to have ourselves as we develop in these areas.

As I see it, it has been a fantastic develop in econometrics during say the last 20, 30 years, this is not the first prize in econometrics. What would you say, computers, better computers, has that contributed to this or is there something else?

Clive Granger: We discussed this in the past when computers first became rather fast, the data became more plentiful we wondered whether there were going to be lots of fantastic results appearing. New things about the economics were appearing and it hasn’t really quite happened as to the extent we expect. We can now think of quite different ways to approach econometrics, we’re not having the constraints we used to have on computing and often data availability now. Whereas at one time it was an effort to do analysis with 100 pieces of data for a single series, now if you go out with 400 pieces of data and 500 series at the same time and have this enormous paper and how much better that is than the one series done properly is not always so clear. We learn a bit more but not a huge amount more, so I think we haven’t yet learnt how to best summarise and how to best search this extra tool. It’s going to get worse because we start dealing with multi-freight distributions next and even to describe some of these things and to know, we’ll have a lot of results we can’t even look at. It’ll be too complicated for anyone to look at the results, they’ll be on the computer, we can ask the computer question about what the results say to us but they’ll be too complicated, they’re just too numerous. We have to learn how to deal with that situation and that’s exciting, in my opinion that’s something which the physicists have faced for many years, they have multi-dimensions and too many data points. We haven’t quite got there yet.

Robert Engle: In time series one of the things that happened is that the big macro-models were pretty much static models or almost static models. They had very simple kinds of dynamics in them and when people started to look at how well they forecast especially short run forecasting they found out that very simple time series models could out-perform the static which I would call traditional macro econometric models. By thinking about this short-run forecasting problem it gives you a different way of formulating econometric models. Econometric models are formulated very largely in terms of given what we know today what’s tomorrow likely to look like, so you build up these models in what we call recursive fashion. That point of view gives rise to an awful lot of these developments and time series, so it maybe just that change in formulation that’s given rise to a variety of different things.

Doing research is certainly good exercise for the brain, what other kind of exercise do you do?

Robert Engle: You’ve been peeking! We both do a lot of other kinds of exercise but one of the things that I love to do is ice dancing and you’ve probably seen some of it here in this Nobel Week, film footage and I’ve been adult competitive ice dancer for many, many years and it’s a wonderful escape from my economics. My skating friends don’t know that I do economics and my economics friends don’t know that I skate, it’s like you just change your personality and that lets your brain relax and other things happen and then when you come back to work its fresh.

Clive Granger: I’ve nothing as spectacular as this to discuss but every day I try and walk on the beach for half an hour, one of the advantages of San Diego it’s very relaxing on the beach. The sound’s nice and the whole atmosphere’s nice and in the summer the water’s quite warm in San Diego so all 25 years, every day in the summer I’ve surfed, I’ve body surfed, I call it body surfing it’s really not that spectacular but it’s fun and it’s very good exercise.

Robert Engle: You take a boogie board don’t you?

Clive Granger: No, I just use my body, just lie on the surf. It’s fun and I really enjoy it and exercise that you enjoy doing you do, and you have a good time and you come back and you’re fitter from doing it.

Don’t you enjoy art too?

Clive Granger: Yes, I go to art galleries quite a lot, whenever I’m in a city on a conference I always put half a day aside if I can and go to that local art gallery and look, I’m not expert in art I just like looking at art. It’s relaxing.

Do you talk a lot about economics at home or what kind of discussions?

Robert Engle: My wife’s not much of an economist; I don’t think Pat is either.

Clive Granger: Not at all, no. How to spend the household money is one of the problems!

Robert Engle: The closest we come to an economic discussion is talk about how was your day and all this sort of things. Then I’ll say something like You know, I had a good idea today and she’ll say Oh really, oh that’s great. Because a good idea to me is so exciting and she knows it’s exciting and she appreciates that, and it feels like a real accomplishment to her as well as to me.

Clive Granger: I’m told off for working too long and too hard and I go away in my study and do my own thing but that’s all, we don’t discuss economics at all.

But you have children both of you and are they economists or academics?

Clive Granger: My daughter is on the edge of an academic career. She’s a science writer and she’s thinking about working in the university doing administrative type academic jobs. My son’s a programmer or developer of animation programmes, so he’s not academic at all but quite successful.

Robert Engle: My daughter took a class at Princeton from Helen Blinder and was very interested in it and I got my anticipation up a little bit but she decided that wasn’t what she wanted to do. She’s just finished her PhD in cognitive psychology at UCLA and I think wants to be a professor so I feel like there’s some carry over there even though it’s not economics. My son took an economics class at Williams and after he was done he said, You know Dad I really learned something important in this class – that I’m never going to be an economist! So, he’s more on the creative side and he wants to be an actor and has been doing a lot of high-level photography and pursuing a more creative career.

I think this interview is coming to an end so on behalf of Nobel Foundation and the Nobel e-Museum I would like to thank you very much for giving us this interview. Thank you.

Clive Granger: Pleasure.

Robert Engle: Thank you.

Did you find any typos in this text? We would appreciate your assistance in identifying any errors and to let us know. Thank you for taking the time to report the errors by sending us an e-mail.