ladies and gentlemen it's my uh um job to try and persuade you um together with my colleague Dr harun Chang of the truth of this proposition that there is uh too much math and too little history in economics as it's currently taught and conceived and well let's start off with an obvious question why do people uh study economics um basically to understand how economies work work um and um there are two different ways um of understanding that I mean there's maths and there history and they stand at opposite poles um in terms of epistemology let's take math first um first two things that maths tries to do um claims to do first of all it's concerned with propositions that are necessarily true that is it's a branch of logic um but secondly math is concerned with propositions that can be proved and economists um really endorse maths in the second um in the second um uh sense that is for generating provable propositions but because the social system to which they apply maths um is so complex too complex to obtain reliable proofs they're tempted by the first or what I would call the platonic view of the subject and throughout the history of Economics one can find economists uh praising maths as the pathway to rigorous and elegant laws Ain as one famous um uh Economist said akin to the laws of celestial mechanics and this powerful um Temptation platonic Temptation was very well summarized ized by Nobel Prize winner Paul Krugman um when he wrote this after the crash of 2008 2009 the economics profession went astray because economists as a group mistook Beauty clad in impressive sounding mathematics for truth economists fell in love with the old idealized vision of an economy in which rational individuals interact in perfect markets this time gussied up with f fancy mathematics it's an old Vision gussied up with fancy maths and this drift towards this Insidious drift towards platonism um explains something um of I mean it's particularly you can find it in the work of Bob Lucas um explains something which a number of people have noticed that and that is the strong nor formative thrust of mathematical economics now an economic an economic model is not like that of like a model of an airplane that is reality in a miniaturized form it approaches Reality by way of assumptions an assumption is said to be chosen for its purchase on reality um but in fact um more usually assumptions are are selected because of their convenience typically to make possible the solution of an equation by excluding extraneous elements but these acts of exclusion are choices one's always got to remember that which Define the problem of Interest or direct the inquiry along certain paths that is they reflect the values of The Economist soon as you get into an assumption based um uh mathematical modeling you are actually inserting the problems you think are important and also inserting your view on how they ought to be discussed and excluding other views on how they ought to be discussed and this exclusionism together with the virtual impossibility of proving anything um gives mathematical econ economics a powerful incentive to make the real world more like their models than to make the models more like the real world so there's one very very important bias in mathematical economics what then is the role of history in economics I would argue that the role of history is that of a reality check propositions from Financial Theory such as risks are correctly priced on average could not survive any actual knowledge of of um Financial a market the way financial markets work or of economic history the branches of History history especially suited for reality checks are economic history the history of economic thought and political and social history all are excluded or minimized in the standard economics curriculum they have been largely pushed out by mathematical economics the general reason for the exclusion of history is quite Insidious it arises from the belief that everything to be learned from history has already been incorporated into the latest textbooks and this ludicrous belief has taken a heavy toll on the study of Economics thought it rises I think from what I call physics Envy some an Affliction which to which many economists are prone economics wants to be a hard science if economics is like a natural science the lessons of history have already been Incorporated in the latest models and no useful purpose is thereby served by recalling the history of mistakes um to quote JB say famous for something known as say's law um and this was already true in the early 19th century he wrote what useful purpose can be served by the study of of absurd opinions and doctrines that have long since been exploded and deserve to be uh it is mere useless pedantry to attempt to revive them the more perfect a science becomes the shorter becomes its history our duty with regard to errors is not to revive them but simply to forget them so much for the history of economic thought and contemporary courses in economics have taken say se's dictum to heart and made it their business to exclude past errors and that's why you don't actually have any study really of economic thought in most economic curriculum today because they would be the study of mistakes and we've transcended those mistakes haven't we that's why we never have any crashes now but the point the point of studying economic his history of economic thought is precisely to realize that the important debates in economics have never been resolved and continue to Vex us today is the quantity theory of money true is it true that state spending is inherently wasteful economics have always economists have always been divided on such issues and the different contemporary schools are heirs to these disputes to sanit I them by delivering a mainstream textbook account or opinion is to rob the subject of its richness diversity and and an ability to illuminate reality knowledge of the history of economic thought can help us avoid hubris or absolute confidence in our own favored school of thought be it NE neoclassical or or post canian and so economists shouldn't aim for a comprehensive Universal and Timeless Theory but must try to keep up with an understand a continually changing system studying history makes it clear that the truth of economic propositions depends on the context in which they are applicable for example can anyone doubt that the Keynesian Revolution arose out of the circumstances of the Great Depression can anyone doubt that fredman's monetarism was a reaction to the inflation of the 1970s to take a less obvious example and this I think is controversial um the marginalism the marginalist revolution of the 1870s um was part of an attempt to refute Marx's labor theory of value um the notion that the factors of production are paid their marginal products was a very useful affirmation of the value of the businessman to society economists try to get some historical hinan into their understanding of of of of the economy but by copious use of statistics and econometrics but as Robert solo rightly points out in a very an article very well worth reading all narrowly econ IC activity is embedded in a web of social institutions Customs beliefs and attitudes concrete outcomes are undoubtedly affected by these background factors some of which change slowly and gradually others erratically as soon as a Time series gets long enough to offer hope of discriminating among different hypotheses the likelihood that they remain stationary dwindles away and the noise level gets correspondingly high under these circumstances a little cleverness and persistence can get you almost any result you want finally history shows us how economic ideas reflect power structures can one reasonably doubt that Financial Theory reflects the interests of the banking system in freeing itself from State regulation so let me end with a final thought progress in economics is not from the particular to the general I think that's a big mistake we don't start with partial and incomplete theories which evolve through a process of maths to ever more General theories which contain progressively more of the truth about economic life that's the model of a natural SCI that is not how economics works it's never been true of economics I would like to Champion the view of horses for courses for choosing those bits of Economics appropriate to different problems and situations without any presumption that one approach is superior to others and for example development economics requires relatively little maths because the real problem has to do with government and governance on the other hand if you want to forecast the conditions of supply and demand for a particular product or a particular sector of the economy then maths is very helpful and useful and in indeed essential so there's a sort of M there is there is a case for mathematical microeconomics in fact it's very important but but having said that the banishing of history from standard economics curricula is a symptom of a loss of pluralism and Imagination in the discipline it's become complacent it's become archaic in many many of its intellectual structures students now leave universities thinking there's only one way to do economic the way they've been taught in years 1 two and three and this lack of diversity leads to intellectual arrogance economists treat anyone who doesn't do economics in their way as s somehow inferior they may be politely condescending but the condescension is undoubtedly a and I expect the conversation to be polite this [Applause] [Music] evening awareness of the past should make our discipline less arrogant and more useful thank [Applause] [Music] you thank you very much Lord Keli I now call upon Professor celi who will not be in the least bit condescending I'm sure [Applause] [Music] hello all right so I don't know if there is too much math in economics uh maybe there is too much may this too little I think as a starting point I think maybe the most useful thing I can do is to try to tell you what math is for me as an economist and how I use it because I use it and and why I use it so what do I do um well it usually starts or actually not usually always starts with observations okay always are on the lookout for facts that are perhaps a little puzzling a little difficult to understand hopefully important to understand stuff happens and I think economists are here if there are if stuff that like to do with economics are here to explain why it happened okay if fact actually economics in many in many ways is is history uh because it's about looking at something that happened maybe it's just an event that has been described by by other historians or or by even journalists or maybe it's in the data and there is some data that tells you look there's a fact out there and then as many historians do you you look at this thing and try to interpret it understand it to tell a narrative to tell a story for what happened in that particular case and it almost never starts with math uh almost never it starts with an intuition it start with a hunch it start with some kind of story you tell yourself in your mind and you tell it yourself with words okay you tell it with yourself with words you say hm this is an interesting fact that just happened there or maybe happened 200 years ago for that matter it doesn't matter um how can I explain it maybe you say well here is here is a possible story a there was fact a that fact a caused fact B fact B then caused fact C you tell yourself a narrative okay now what happens when you tell yourself narrative so this is doing this is bilding a theory okay now you building a theory doesn't mean to use math okay you know you can you can you can build verbal theor sociologies have theories historians have theories every time you explain something that's a theory you're making a theory but one thing that I think any Theory should have whether it's verbally expressed or whether expressed with math it should be logically consistent you should hang together you should not miss things you should not not draw the the wrong conclusions you should not let be led astray so why is ma useful for me because what I do is I first tell myself the story in words and then I retell the story IET tell this try to retell the same story in math and why is that useful because sometimes you know if you really want to know why this useful it's useful because I'm not smart enough okay I'm not smart enough to just ver think about all the things I might be missing in my argument but if I miss something the masth will tell me the math will ring a bell and says look you're not getting what you were expecting so you're missing your story was too simplistic you missed something okay let me try to give a maybe a flavor this is a bit risky but let me try to give a little flavor of how that might work okay now this is a bit difficult because I think many of you I don't know how not I don't know how many of you are are Eon majors are or or have taken uh my course for example in the first year uh so um so I need to come up with something that is uh accessible to people who haven't taken even first year economic so let's it's going to be something very simple let's say that uh well Steve here my colleague Steve is a labor Economist he he worries about what happens to labor Supply when wages go up change okay will workers want to work more when the wages go goes up uh than than they were before the wage went up so if I approach this question in a very verbal simple way I think I'm going to come up with something very clever which is this wages went up when worker should be willing to work more okay the the reward for working is up so they should be working more and I might be tempted to run with it you know I might be tempted to say well that makes sense I'm going to run with it now if I force myself to also write the same story in math I will immediately realize that I'm missing a big chunk of the story which is as wages are higher of course as you probably all all those you who are in economics already know as wages are higher people also can afford to consume more Leisure and so actually the effect of wages increases on labor Supply could be an increase or or a but I think you can see how I might have too excited by my wages goes up so label Supply goes up and stop there if I force myself to retell the story in math I'm going to catch that mistake so that's one that's one that's one way I use math I use it as a thinking Tool uh I use as a thinking tool I use it to catch mistakes in my in my reasoning now let me be very clear I and this is where the point about uh you you almost never start with the math is important if I just found that my simple intuition was wrong just because the math said so and then I went to my colleagues and say well the math says is this I get kicked out of the I get kicked out of the room I need to go back and understand again in words what the mass is telling me okay why was I wrong what did they miss and if it's just the ma telling that and there is no intuition there is no reasoning there is no way of rational Iz in verbally what you have found with the math then that's not good good economics okay so it's it's a it's a double narrative there are two narrative reinforcing each other there's a verbal narrative and there is a mathematical narrative the mathematical narrative keeps you on your toes but the the verbal narrative ultimately was is most important which is why when we teach economics I mean those of you who I took my course they you know I don't I don't use any math I believe that I can teach most what's important in economics without any math but I wouldn't necessarily have believed it if I hadn't checked the reasoning with math before but then once I've checked it I can rely on the verbal version of it and just and just go with it so that's one way I use math as a thinking tool now there is another uh way of using math which I think is more and more important especially in my field with much microeconomics which I think is becoming more and more important because uh it works very well with computers and that is precisely to do what my one of the opponent one of proponents of the proposition just told you that you cannot do with math which is to deal with complexity I think it's completely the contrary I think math is a wonderful tool that allows us to deal with things that even people smarter than me cannot deal with because they're too complex okay so think about microeconomics microeconomics there is some Shock that occurs you know maybe the government is doing austerity the central banks is changing interest rates and that affects the behavior of literally hundreds and hundreds of thousands of different agents consumers firms other branches of governments foreign agents and the way these people respond to this to the changes in the economy will then trigger farther responses from the other guys so there are then second round responses the responses firms affect the responses of consumer the respons consumers affect the respons of government the respons of government affect responses of firms I I I don't know anyone who's smart enough to to maybe K was uh but uh since then I don't think there have been many people smart enough to kind of trace this incredibly complex web of reactions and counter reactions but if you have if if you think you have reasonable and this is a big if but that's what we're talking about here we're talking about math if you if you think you have a reasonable description of how agents process information and act on that information then you can use a mathematical formulation representation of this put in a computer and have some um create again A Narrative of the respones to the economy so that's the other of the two uses of mass so this is why mass is important to me this this is why math is a critical tool in my toolkit it's not the only one it's not the main one uh first of all as they say there is always thinking through it uh but but um without the math I would be like a you know a plumber who has been taken away one tool one of the many tools uh that he uses um dayto day let me make a couple of um s of kind of impromtu responses to to what we just heard uh um the perhaps the the biggest uh um misconception that I think there is about um economics and it's not so much about math and actually let me let me first make a more General point I think there is a lot of confusion between math and other potential shortcomings of Economics okay and if this debate was about some other potential shortcomings of Economics I might probably be on the side of those who criticize economics um so for example and I think already came up very often the discussion of whether there is too much much in economics gets mixed up with a discussion about the whether the assumptions of Economics are right or wrong are realistic or unrealistic should they over they should change over time if that was a debate I would be much more sympathetic for example I think there are huge problems in economics about the assumptions we make for example the the assumptions about extreme forms of rationality that agents have or the extreme forms about the information information capacity of a agent have but that has nothing to do with math you can have the wrong assumptions and be completely verbal and you still get junk at the end okay and you can have the wrong assumptions and be mathematical and get junk at the end you can have the right assumptions and be verbal and get it right you can have the right assumptions and be mathematic it right it's very important to me that we separate the debate about whether the Assumption economics are right I think many cases they're not and the assump and the discussion of the methodology of mathematics which is completely separate two more points another huge huge misconception about economics this idea which I don't know where it's coming from that there is this just one overarching model and that we are all slave to this one model and we try to fit everything into that model it's just not true it's just not true Krugman by the way I don't need to share krugman's opinion about economics I don't think Krugman has been in a seminar room in for many many years uh uh um it's just not true uh again go back to my initial examp ex You observe something and you tell yourself a story okay you're not going to try to squeeze this in into some you know big brother kind of platonic model uh that is out there no you're going to try to write something that uh is essential and that try to get at that particular question and so I completely agree with roseli we there isn't you should have a a different approach the model the story that you tell is going to be targeted uh um tailor to the particular questions you asking and that's what we do all the time history I have very little to say about that um I think there is an enormous amount of history in economics there is an enormous amount let me just give you one very small piece of information but it's it's not uh dispositive but I think give you a sense uh in the economics Department we give every year actually no not every we started this year so but from now on it going to be every year a prize a prize a prize to the two best PhD thesis okay to the two best PhD thesis over that year so we have thesis on the great on the Great Recession of course we have thesis on development in villages in Bangladesh today we have thesis on labor markets today we have all sorts of stuff focusing on the present the two winners of this year prize wrot thesis on respectively one wrot a thesis about the consequences of the Napoleonic blockade on the development of the cotton industry in France this person won the thesis for the best PhD Pride for the best PhD in economics Department this year the other winner won a prize for the consequences of the integration of the telegraph between Europe and North America at the end of the 19th century these are the two people we the econom economists the history blind economists have given prizes to this year um in the department of course we knew that there was this debate coming up uh so uh um we are we're dealing with data all the time we are constantly looking at data and I think that's one way of making history of looking at history uh the data tells us histories and and our job is to try to understand understand them so I think we are historians more than anything [Applause] else than very much Professor I now call upon Professor harun Jang to present for the proposition [Applause] please right um thank you everyone for coming thank you the organizers for organizing this uh wonderful event uh yeah let me start with this uh you know according to Dr Sheldon Cooper the greatest scientist of our time well after Steven Hawking according to him at least uh from Big Bang Theory in unraveling the mysteries of the universe we need three things Math Science and History it's uh there yeah note that there is no economics there [Applause] [Music] [Applause] yeah well of course there are people who think history is a rather boring thing studying which is rather useless and in this movie based on Alan Ben brilliant play The History Boys there's a character called Raj is uh this boy uh played by Young Russell toi who is asked by his uh history teacher to Define history and he gives a most succinct answer that many people might agree with and is just one thing after another Well indeed uh these days uh many Economist uh oh the what happened here slid show where is the slideshow of course at these days uh many economists that consider the study of history to be a rather low great intellectual Pursuit on a power with train sporting but watching and collecting Pokemon cars no seriously I mean that you know it's that beginning to slowly change but let's face it I there is this purpose intellectual hierarchy in our profession in which uh the less oh this is not moving uh no scroll scroll okay so let's leave it at that yeah now this is a PO intellectual hierarchy in which uh the less what you do is related to the real world somehow the intellectually most Superior you are so if you're the cleverest you do mathematical modeling of something that has very little to do with the real world like I don't know touring machine no don't get me wrong I mean these things have to be done you know I'm not one of those people who deny that these things are useful but somehow these people are thought to be superior to other people and if you're not that clever you to well become econometrician but then never running regressions but the economic theoretician so ideally you should have a PhD in statistical theory if you're not that good then you do macro if you're not even that good you become a development Economist like I am and if you are not even that good then you become economic historian you know and if you are running around surving Factory managers trying to understand how firms organize their production you know there must be something wrong with you no this kind of intellectual hierarchy unfortunately uh exist uh in other fields you know many of you have watched The Big Bang Theory yeah there's a clear hierarchy theoretical physicist yeah experimental physicist well there Sheldon and Leonard yeah and then Howard yeah he only went to MIT and did engineering yeah so it's like the scum yeah so I'm not saying that economies are unique in this but somehow we have develop this uh uh thinking that uh somehow abstract uh theories mathematical modeling that's useful but I would argue that economists need to study history they need to understand history and they need to draw lessons from history and I can think of at least five reasons for that well first very obvious but important Point history is where the president has come from so without understanding the past well we can understand the present well yeah is a very obvious proposition but people often forget that yeah second what people know about the history or rather what they think they know about history affects the present and by implication the future by informing people's decisions including how they for example build their models when their economies so the despite the pretend that economists do not need to know anything about history in order to make the right policy recommendations a lot of actually policy recommendations are backed up by historical examples so free trade Economist have the biggest the the guns in the Arsenal yeah Britain and the United States became all economic superpower because they practice free trade you well there's nothing like spectacular real life cases successful otherwise in persuading the others of the wisdom of your policy recommendations you so if you want to recommend anti-inflationary policy you talk about B Germany and Zimbabwe yeah instead of say South Korea in the 1960s and70s where inflation rate was nearly 20% but the economy grew fantastically so that people actually use this selective examples and you know unfortunately that this is a very widespread I cannot go into that but basically I have written quite a few books where know yeah this is a commercial break uh where I show that uh this St shouldn't be here but I put it there by mistake no no that has come later you [Applause] know I written these things to show that uh this uh sort of uh official versions of uh British or the US History are very Incorrect and if they knew that Britain was the most protectionist country in the world in the late 18th early 19th century if they knew that the US was the most protected economy for 120 years since the 1830s you know the promoters of free trade whether they are economic theorist or policy makers may not have such conviction in their policy themselves and would have hard a time in persuading the others thirdly I think that the history is are very useful in letting us question some taken for granted assumptions you know once again I if I tell you this many of you would say yeah I mean I knew that but you know this was never integrated into your kind of Economics education so I mean I'm I'm still going to put it to you you know lots of things that cannot be bought and sold today you know human beings Labor Service of children government offices used to be perfectly marketable on the other hand there are things that could not be sold 100 years ago even 50 years ago but there they are that so Center left and right today you know a lot of countries didn't have patterns until the late 19th century Financial derivatives are became what they are today only because in 1982 two of the financial regulatory that bodies in the United States agreed that you can actually finan derivative contract based on unreal things like inflation rate or weather or things like that yeah things that cannot be delivered yeah so that created a whole new set of financial products yeah and once you know this things that you will realize that the boundary of the free market into which political logic shouldn't penetrate is actually not drawn by some Timeless law of science you it's drawn by political and eth IC decisions another example you know if you learned that the advanced capitalist economies grew the fastest in history between the 1950s and the 70s the so-call Golden Age of capitalism when there are a lot of and increasing regulations and high taxes you will the question today is accepted wisdom that the best way to generate grow is to cut taxes and red tapes you you know the United States in the 1940s and 50s had two communist presidents maybe you haven't realized that yeah they were called Harry Truman and Dwight eisener yeah under them the top income tax rate was 92% despite that the US economic grate at 2.5% during the golden age in the last three decades of neoliberal experiment it has grown at 1.5% you know these are only Ina per cap terms yeah so once you know that you will think I mean the the the differently of all this underlying assumptions fourth history is useful in making you see the limits of economic theory you know as I like to say life is often Stranger Than Fiction and history has given us so many successful economic experiences at all levels Nations companies individuals that cannot be explained very well by any single economic theory I mean Robert already talked about nine different schools of many different schools of economics and in that book uh the users guide I try to explain that there are at least nine oh God this is useless yeah at least nine different schools of economics and they all have their strength and weaknesses and it's because they are all they were all develop in order to solve particular problems they are based on different political assumptions assumptions of a human nature Assumption of ethics and so on and they are good at different things so that to illustrate that I give you the Singapore problem you know when you read about Singapore in I mean the the newspapers or or standard books on globalization will only hear about this free trade policy and his welcoming attitude towards foreign investors which it has but you will never be told that 90% of land in Singapore is owned by the government 85% of housing is provided by government owned housing Corporation they have a for savings scheme and they that that a staggering 22% of GDP is produced by uh State own Enterprises including the famous Singapore Airlines so I put it to my gradate students give me one economic theory doesn't matter what it is neoclassical Kian institutionalist tarian one economic theory that can single-handedly explain Singapore and I dare say that there's nothing like that yeah so you need to know that these different theories different historical experiences you know even just to look at the data I think uh we need to learn a lot about history you know let me give you a beautiful example I mean all this openness and growth regression yeah so you measure some our Count's openness by looking at say that effective rates of protection or some indicator of openness but unless you actually know something about the countries that you put in this regression you can make a big mistake for example in South Korea between the 1960s and 1980s there was a very rigorous foreign exchange rationing by the government which made tariff rates almost meaningless yeah the Tariff rate could be anything the government that can refuse that to sell you foreign exchanges because the government owned all the banks yeah so if you know something about Korea you would be immediately skeptical about using tariff figures to measure openness yeah I mean I said that because I happen to know a lot of Korea but what about other countries yeah you know if you that look at other countries there are 50 70 countries where tariff race are not that meaningless meaningful so that even just to that use data in intelligent way you need to know uh some details about that history and last but not least we need to uh look at history because we have the moral duty to minimize live experiments with people no this this is a serious Point yeah I mean Economist have killed actually a lot of people Yeah Yeah from Central planning uh in the Soviet Union the Big Bang transition back to capitalism austerity policies following the Great Depression structure adjustment programs in the developing world I mean that they have ruined many people's lives yeah I of course uh learning history doesn't make you repeat the mistakes yeah uh doesn't make you avoid repeating those mistakes but it reduces his chances so we need to Lear history out of moral duty yeah so that's the summary but uh let me spend last minute or two to talk about the more practical side of the argument because you know a lot of economists say yeah I mean the history is useful but then when it comes to actually assigning hours to different parts of the curriculum they say well however there's so much math to teach let's forget history let's drop it yeah some of them even say well you know history people can't teach themselves yeah they can read books yeah but we have to teach them math yeah well I think this is quite problematic because unless you put your money to your words so to speak uh you are not really being serious about it you know if you say that yeah I mean I recognize that history is important but we have no time to teach history then you are just giving lip service so that if you really think history is serious you have to teach this yeah teach this in a systematic way yeah because that history you know some we have developed this idea that if something's are written in words it's easier to understand no it's not like that I mean history requires a particular type of you uh the the reasoning and understanding and the the empirical skills and for that purpose I think that this uh needs to be properly taught and unless it's in the curriculum you cannot expect students to study it let's face it I mean you know that that your generation is already suffering from having to do 65 internships before you even before that you can even apply for a job yeah you haven't got time yeah so unless this becomes an important part of the curriculum I think that that that you may want to study economic history or history of economic thought but you won't be able to you so let's leave it that that thank [Applause] you [Applause] very much now we have the other point of view from not doing too well here um I'm definitely not going to be as funny as uh I ch um I uh you know I like some of the jokes and I like lots of the Snippets uh from from the proponents but lo and behold when I was listening to you guys I really didn't uh hear you talking about the field I'm working in and the academic discipline that's that's around me um hen was talking about the a hierarchy of fields within economics with the more mathematical like mathematical economics and econometrics on top and development economics towards the bottom so I brought this graph um which has the Clark medal winners by decade the Clark medal is a medal given by the American economic Association um used to be every two years now it's every year um to the what the profession in the US thinks is the best Economist under the age of 40 so it's a pretty selective pricee actually more selective than than the Nobel and uh the blue ones were theorists the yellow ones were people were working with data empirical people the green ones are sort of empirical methodology um I didn't know exactly where to put them we could of course argue how to classify particular individuals but you can see in The Last 5 Years in the 2010s no single mathematical modeler has won the Clark medal in fact one of the winner was a development Economist at Esther defo um and you'd have to go back quite a bit before you find somebody um actually you know extremely mathematical when I did this GR I thought the math and the theory in economics used to be much more important and I'm actually somewhat surprised that even among the early winners um a lot of them were very involved with looking at data and looking looking at evidence um this graph is another bit um of evidence going along the same lines that actually if anything Theory I think in economics peaked around the 1980s um these are Publications and what most of us consider the three top journals to publish in in economics and uh this is a classification of papers into Theory and uh papers that uh look at data the the red ones versus the the blue ones um so in the 1980s about 2/3 of the best economics done or best economics that got published was Theory and now that's down to one 1 thir and 2/3 is empirical work um so in general it seems to me that economics has has gotten quite a bit away from just doing modeling and turned a lot towards looking at evidence and thinking about it a lot of the empirical work that's being done these days is actually not particularly mathematical or even particularly based um on models but having some theory in the back of of our minds and I think everybody actually agrees with that is sort of necessary in order to interpret the data as well and think about them um so the one point is that uh I think there's more empirical work and less theory in economics um than you might think I think the quality of the empirical work has also um gone up and uh empirical economics has actually delivered some important insights I want to give you just a couple examples um one is a study by Marian berant and sandel Mulan what they did is they mailed fake resumes to actual job adverts and what they did is on the the resumés were identical for the the different jobs but on some of the résumés they put names that were very wh sounding and on others they put names that were very black sounding that's easy to do in the US because there's a big uh cultural divide between the the races um on the white resumés they got uh about call backs for 10% of the resumés they sent out for the blacks they got only 6% of callbacks so blacks need to send 50% more resumés before um they get the same number of interviews um this work has been extremely influential it's been widely replicated in other countries in somewhat different settings with different ethnics groups with men versus women and so forth um and in many instances has thrown up pretty convincing evidence that there's some discrimination in the in the job market now you might have heard that uh very recently the Prime Minister announced that uh um yucas is actually going to remove names from um University applications and uh this is based on on This research um another aspect about this study that's interesting is that the methodology was not something these economists thought up by themselves um this idea of sort of sending two identical people who might look different um you know to buy a car or rent an apartment or get a mortgage or apply for a job that was actually thought up by sociologists and economists have learned it um from them and then um you know done their own wrinkles on this the second example I want to give you is uh research on the minimum wage um there's been a lot of studies particularly in the 1990s on um the effect of minimum wages and whether minimum wages destroy jobs and lo and behold a lot of that literature showed that either there are no employ mment losses or the employment losses are relatively small now the there's a big debate between those who find zero and those who find small but lo and behold the effects are small um and for policy makers the the differences that were found are probably not that uh important again these Studies have been fairly influential in policy and probably the UK uh national minimum wage in 2000 um was instituted it partly because uh of these types of findings now you know no employment effect of the minimum wage goes against the most standard um model of a competitive labor market if you have a downward sloping labor demand curve you would think that um raising the minimum wage destroys some jobs um but certainly the theories we have are flexible enough to accommodate either a zero finding or even employment increases as well you just need some frictions in the labor market where workers actually search for jobs um so you know it's not that these findings are refuting existing models necessarily um having the models in mind is useful because they do caution us that despite these findings these findings are in circumstances where the minimum wages are fairly moderate like in the US and the UK um that doesn't mean that you can raise the minimum wage to 15 pounds or 20 pounds and necessarily expect the same benign effects so to me really um a lot of Economics is fairly applied fairly in tune what's going on in the world policy relevant um and that doesn't sort of seem to chime with the description that uh we hear from the critics um of the discipline so if economics is not doing so badly why is the image of Economics so bad um I think the reason for that is that the teaching is not nearly as good as the [Applause] research I think the teaching is concentrated too much on the math there's too little empirical context and too little um applications of what we're doing and too little effort to to teach you um to actually make use of the tools that uh that we have to join outside are you you know my conclusion is basically um I agree with you guys there should be there should be fewer models there should be less math now I would say on the margin not wholesale you know I don't think we should throw out all the math and all the models um I think in the way that uh Franchesco has described them they're actually quite useful um I would like there to be some more history in uh in economics and in the in the economics curriculum I'd be fairly open to this I think we should interact more with other social scientists I don't see any anything wrong with that um on the other hand I'm skeptical that all of this is sort of going to lead to some Grand transformation of economics and all of a sudden it's going to be this wonderful very different animal at the end of the day we're social science we're dealing with complex social interactions um and it's just very difficult questions there are no easy answers um and I doubt that uh you know just by by having more interactions I think it's fruitful but I don't think it's going to be revolutionary and all of a sudden economics is going to turn into something great and very different um I don't I don't quite see that happening um even now you know I think economics is a pretty broad Church um it's not just some uh free market free Traders uh who all want small government um there's the whole range from Joe stiglets to Danny caraman to Bob Lucas with very different views and very different ideas of what to do within the current framework um uh of Economics so that's it for [Applause] [Music] me [Music] [Applause] thank you Professor P I would like to uh follow up on the point what I regarded the critical point made in that presentation that there is a difference between the undergraduate curriculum and what economists do for research it seems to me that this distinction is is worth uh discussing um in more detail and I imagine there are people on the floor who'll have strong views about that but could I just ask um the other panel members who um about whether they think uh a distinction between applying this qu this proposition to the undergraduate curriculum uh has more validity than applying it to economic research or advancement of knowledge any offers uh yeah I I think uh there is a certainly that problem you know I mean what is to in the uh classrooms is uh quite behind what uh happening on the academic front but I think uh you know if you think that only marginal adjustment like you know introducing a little bit of this little bit of that would help I think that uh we have a problem because that I think that the problem with the current undergraduate uh teaching program is unfortunately it has been designed on the assumption that the students are mostly likely to go into Graduate Studies but let's face it I mean 90% of you are going to go out in the real world work in the private sector working International organizations working NOS you know working the think tanks and for that I think the curriculum is just too kind of abstract and uh not uh uh related enough to the real world and I think that that is a serious point that we need to confront I mean it's no good pointing out that you know D is ass did that if uh what you see that in the classroom is a rather abstract you know Duality theorem and you know that the lots of math and you know student need time to learn other stuff yeah could I just reinforce that I mean when I was at War I spent number of years um marking MSC thesis um and um they all they were all formulaic really um I mean you start with with with an intuition of some kind you then um you then introduce the model in um in a mathematical form and then you test it in some way or other usually end up inconclusive but You' you've shown you can do what's required um and those are sort of things that um certify you as um competent now whether you do anything after that that's any good it doesn't matter I mean but but um that's that's that's the sort of the next stage up having having done one's undergraduate and of course and gone into that and and I would also I mean say one thing that I I I I don't recognize I mean I mean I think the the opposition drew a distinction between mathematics as modeling and the use of Maths for empirical work um and um I don't you you know I don't see that I don't see um that that um is is is is quite right because what I think members of our side were arguing was that a lot of empirical work is non it doesn't consist of a a Time series or comparative uh comparative series um it it is using history um in a non-mathematical way um to to tell us something about about what happened before and I think the attempt to confine everything within um the envelope of mathematical technique of one sort or another whether it's modeling technique or statistical technique is exactly where the subject goes wrong because it's got It's too exclusively reliant on it that would be the point that I would make I mean on this last point I couldn't disagree more I mean the the the wealth the wealth of knowledge that um empirical work by which I mean Pro sectional panel data perhaps a slightly less time serious uh but but even time serious has has yielded is it's just extraordinary and and I think Steve gave two very uh clear examples of that where uh we have used uh eon IST uh I've used econometrics to uh tell us things that are of first order importance about the world in which we live and and and uh I think econom etics is the right way to go for for many of these questions and we are blessed to have this tool and and uh it would be catastrophic if we lost it yeah well I mean we agree with that I mean is it the only way to go I mean you you seem to equate empirical with statistical I I think I've expressed my respect for history have you absolutely I think so what what history have you found worthy of respect can we uh I think we want to have a little order here I I would like to uh see if we can elicit some opinions from or positions from the floor we've got a short time so someone here can we get the microphone down to this rather Central position hello yeah so uh one of the issues that I I find with uh you know Professor skel's argument is really the centrality of Economics towards policym and uh by policym I'm defining it as something which is technical right so consider the problem of the central Banker who wants to decide to raise interest rates right there you need U don't you think we need some kind of a framework to tell us for example how much the interest rate should be raised or reduced history can tell us whether or not we should do this at all you know whether the FED should raise interest rates right now is a question that history should analyze in detail but without rigorous econometric models of some sort or the other without some kind of theoretical framework to back us up how much does the FED raise the interest rate when when it has to make this decision and what kind of basis would we give to such a question are you asking me specifically or power the en side that things mat should be thrown out if you know well of course I mean of course you you you have to have a a theory of what the impact of changing interest rates um is but that has been disputed it is I mean you've got to know something about some people argue that interest rate policy is actually relatively ineffective at doing either in its effect on prices or in its effect on the real economy I mean that has been a dispute going on now it seemed to have been resolved in some way in the 2000s when the whole of macro policy was really interest rate policy and you you know things certain things happened and that was a signal to raise your interest rate to raise Bankrate others to leow them it turned out that was that was actually not not adequate it was incomplete and now since the recession the arguments have started again you've got a zero bound now what do you do now the point is you've got to know something about those disputes then you come to a judgment of course data are very important and that includes quantif quantifiable data as well but then pure economic history and knowledge of the history of economic thought is absolutely essential for a central Banker to come to a judgment and in fact sensible Central Bankers who are not slaves to dead economists actually um find that they do use their judgment I mean I think that's a a very interesting example of uh how the statistical work and historical work have come together um the response in the US to the great to the Great Recession in particular was led by a man who has uh written mathematical models estimated statistical relations and being a scholar hiso his of the history of the depression and we were we were very lucky to have that guy and he was wrong you're talking of Bernard beran and as an academics graduate I can only summarize my three years of undergraduate academics education as too much you useless math and too little if any history I mean I am not anti- maath myself as I'm sure none of us here is I enjoy math a lot but I am against the excessive Reliance on one particular tool to explain the whole economy I mean when can we students stop learning only about constraint optimization which which which by the way is 19th century mathematics and start learning some other useful tools to understand the open complex adaptive system we call the economy I mean where is psychology where politics history sociology and of course also some real 21st century mathematics why we [Applause] [Music] added I think that's a statement rather than a question uh thank you I wanted to react to a comment made by Professor Celli and trying to uh differentiate between two debates the one on the assumptions in economics and the one on um the use of mathematics but I think they're very much Linked In in a way that history of economic thought and teach us that uh the the constraint by thinking quantitatively and wanting to do mathematical modeling in economics constrained the the assumptions that can be uh used um if economists are so fond of um rational expectations it's because they're very convenient in modeling the same way as if we assume selfish uh interest it's because it allows us to have closed models um and so um the same would go with some um Assumption of having rational Behavior with some be um bias it's still a simplistic view of human behavior that is very useful when you uh want to think in terms of uh equations and and Clos models can I just say you know before rational expectations even in mathematical model we had adaptive expectations which are much much easier to deal with so you know that that was a movement towards complication and and by the way I don't necessarily think that was an improvement in terms of the of the um posability of the Assumption so and and I think you know there are there are different ways you know there are okay so for example you know critics of Economics like a lot now something called Agent based modeling which is basically a mathematical model with no optimization okay that's that's Cuts away completely all the optimization part so that's much easier actually that's much much easier than having optimizing agents so I I think the statement that um the Assumption driven by mathematical convenience is just not not not right in fact you know abandoning adaptive expectation was a movement towards more difficult math uh more difficult models uh um using model with optimization instead of something like agent based models is more difficult thank you could I have another contribution from the floor please person right at the back yeah hi uh my question has to do with uh whether if say we do introduce greater theoretical pluralism in into the study of uh economics uh and if we do study more history uh during uh during economics that it will probably change uh what The Graduate economists know and when they try to enter into the labor market uh what their employers expect them to know now given that uh I will guess that employers really like the mathematical approach because it's easier to be implemented into a model and U in turn it's easier to sell uh to customers in financial Market you can create a model you can like predict things uh in theory at least uh so I'm what do you think that uh how how is economics going to compensate in terms of uh uh getting money into uh the study if we introduce more history and thus we kind of uh uh decrease the mathematical uh sophistication of Economics graduates yeah can I respond to that well first of all you know if uh well that one problem with this country is that the most employers are in the financial sector so they are working with something more prone to quantification if you're working in the car manufacturer or you know other types of companies that uh that kind of mathematics might be less useful I mean you will need math for engineering but not for uh the the financial management but uh more importantly you know if they really want good mathematician why would they hire Economist no because there are real mathematicians they are physicists they are engineers and then we economists are like number four you know so we are four best mathematicians yeah they don't necessarily want math from us I because if they they really want math that that they can hire real mathematicians you so Economist need to learn other things that that in order to compete with those people and I think that the real world employers actually find people with broader education you know I'm in Cambridge at that also involved with the center of development studies and I mean we produce phds in development studies which is uh interdisciplinary study and those people have no problem in getting jobs in the private sector in the the the World Bank in the United Nations because actually that they know a bit more about the real world than your standard Economist and I think that can be a strength so this curriculum change yeah it might need a kind of adjustment period because the employers are going to say wow this guys are different are they necessarily better you know more useful but then after a while I think they'll be persuaded thank you the lady Judi Shapiro LSC uh I'd like to pick up on something Robert skeli said uh that is development economics needs little maths because the problem is governance and that is exactly why I could not possibly support that side I'd like your reactions to that despite the fact that we've had agreement across the panel that we don't want 20th century economics we want 21st century economics and we want to bring it into the uh classroom the seminar room and the lecture room and and it's not just because they're colleagues of mine and I have to see them tomorrow and they're the best teachers we have at LSC it's because I believe that hajun Chang has been tirelessly arguing and very well and not necess and not using econometrics for the causality flowing the other way that's how I understand his work and it it was relentless and it made an impression on me uh I'm not necessarily going to apply it to Africa I'd like to see an argument about that but he said the evidence is the causality goes the other way today's rich countries acquired most of the institutions um that today's dominant view that's Robert skel's considers to be prerequisites I also don't know what there is about governance we need to unwrap that which is what I think some of the random control trials try to do is it political democracy because Singapore doesn't have political competition but it doesn't have corruption on the other hand is corruption important and that's where I think the kind of econometrics that Steve pichy's books talk about and that is practiced is one of our ways forward and I also then finally want to get your reaction on this question of not going from the particular to the general because we haven't got very far with the general so it would be very nice to see some of the particular where we drill down and find out for example one of the things that Franchesco casselli works on the natural resource curse what happens there not just is it a curse but why and I don't think we can do that without econometrics but of course you need smart econometrics you need arguments the internet which hun Chang thinks is less important than the washing machine I stand by this statement yes okay let's have a debate about it you and I sometime okay I shouldn't have got into that let me finish because I want other people to do it with that idea that I can't trust you with my curriculum because I think you want both of you economics by assertion and by the way I don't think your assertions would then win and I think evidence is very important and not just models but we need to test them we'll never get a right answer but at least we'll have a debate that's based on evidence not on who has the toughest convictions which by the way is one of the reasons we have five males up on the platform [Applause] thanks I mean emphatic statements based on false either or propositions will always win support in a political argument that was a political speech and it it didn't really I mean it was great I mean I enjoyed it very much but I don't think it really addressed the issues of no one's against evidence and no one is against using um using uh whatever tools you have available um to get to get evidence to support your propositions I mean no one is against that so I mean to say that we're against it what what what what I think I was saying when I when I made that statement about development economics not needing as much maths as some other branches of EC ecomics was that the problem um which many development economists have identified is the problem of governance and that is not something you can necessarily um deal with or or resolve just by lots of data um you just can't do it I mean these are problems of political Theory these are problems of Institutions institutions how they work and and the context of those institutions they examples their their you know their their they examples of the incentives facing particular rulers um and so on so I mean I'm saying you just need some some a broad a broad array of evidence in order to come to a conclusion is that so controversial is that a reason for saying ah we can't possibly support you because you know you have said you need more than one tool that's the question I put to you and that's not I didn't intend that to be a political speech yeah may I just uh quickly add one point uh to that uh well I think what counts as evidence depends on your theory so it's not totally independent from Theory so that you know I was once listening to a debate on uh radio on the alternative medicine and you know this that the standard kind of medical person was insisting that we need RCT to show that that something works and this uh the acupuncturist he was British guy pointed out that well actually in acupuncture there's no point in body which where you can stick a random you know the pin and have no effect you every Point has effect yeah so there's no kind of uh acupuncture equivalent to PLO yeah so that if you have that kind of philosophy then the RCT doesn't necessarily establish evidence so well you know what I'm saying is that that basically you need different layers of evidence yes you need cross-section econometrics you might need that time series you might need panel data but then you need the historiography you need that anthropological approach you need the surveys you know that you need the structured interviews you know all these things uh need to work together to create that the robust evidence unfortunately we are these days are saying you know unless it's econometrics so RCT is not evidence I I find it very problematic yeah are just can you explain to me how do the other methods resolve the acupuncture problem no no no that that is just an illustration but you know I already gave an example was a bad illustration no no no how do your message solve it no that that was an illustration of how your theoretical position actually affects what you count as a valid evidence yeah so that was a anywhere and it does something then it's very hard to figure out what acupuncture does sorry what what can't you have some guys you don't stick a in and then the guys in no but their bodies are different you know that's the problem yeah but more seriously you know that these other evidence are necessary I already give you an example of Korea having this 40 exchange rationing by the government which makes that tariff rate meaningless you so unless you know about how these different countries work you don't understand you know I mean there were people in the 1980s running regressions trying to show that the share of that sorry spending on Industrial subsidies as a proportion of GDP had negative or positive effect on the the industrial performance yeah unfortunately they that what they didn't understand was that in countries like Japan industrial policy was not mainly conducted through subsidies as it was in Europe yeah so if you don't know that evidence you will treat Japan exactly as same as European countries which creates problem but so you need those other knowledge to even to conduct intelligent econometrics investigation but that's just saying better done uh academic work is better than poorly done work that we can yeah yeah yeah no but what is better I mean there's a a question on the end here yeah um well um probably uh because of this reference to the eist uh but I got in sort of interested on this distinction between the particular and the general and I cannot help but you know remembering uh either the work of Carl poer or Berton Russell and that's the story of the inductivist turkey ke so I want to ask Professor geli uh because I think that after all this is a kind of false uh dilemma history and economics uh I think that there is something prior to that and is the empirics of it and and you mention you you you you started your presentation talking about history like well staff or the empirics the staff happens uh is the data given is to consider that the way to economist or social scientist would take the empirics or the data as given is that what you suggested yeah I mean the the way I think about it is a is a journey that starts with data and ends with data um you uh you start out with the observations which are often no I understood that my question is if data is given or if it's a historical construct or if it's it is you know part and partial of the epistemic of it and that's the story of inductivist turkey or the fail aunist uh uh what do we start from the observation is that observation exogenous completely external to us right uh or is historically determined you know and that is a problem with economic history because we cannot witness the data we have to work with data that is given to us yeah yeah I'm I can I the question was to me so I'll try to um I mean look I um there are um we could spend hours and hours uh discussing all the things that are wrong with the data that we have um and and and then we can also spend many hour thinking about whether the things that are wrong about the data are wrong by Design or they're wrong by accident um I I I mean just um I don't have the patience to sort of um my my hunch is that uh you know I'm a microeconomist so a lot of the data I work with this national accounts data my hunch is that the national accountants do the best they can uh with with enormous amount of problems and uh they don't get it perfect but I think there is information in the that they give us if I may uh I'm not going to go to the 1990 uh Gary K dollars but I will ask you then why is that the case that all economics papers uh end up predicting I mean the results are as predicted and there is usually no attempt to bias or falsify the the model or the the hypothesis that's just not that's that's just not not not not the case I mean that's that's not it I mean the you mean empirical paper you mean empirical papers yeah not the I mean there are lots of negative results people say the deficient is zero get in I mean I think uh illustrates a sort of point of important difference you say economics starts with data and ends with data I say starts with Theory and ends with Theory um and none of the uh e economic theories we are are um familiar with came out as a result of direct observation at all um uh they they just didn't it was and and a very good example is adaptive expectations ends with rational expectations I it's one Theory um contesting another theory and the whole debate takes place within the world of theory and observation had very very little to do with that particular transition it is my S duty to wind up the debate uh I want to call upon uh both sides to offer a very succinct summary of of key points before we take a vote on it so I I I'd like my take my short time to to rebut your last [Applause] point I just completely disagree um probably the most fundamental change in economics in the past 30 years is behavioral economics the way economists have started thinking about how people make decisions um what type of information they have how they process it um what goes what type of preferences they make may have and that came about largely because um when economists in the 70s and 80s turned towards Game Theory other people land ran experiments in Labs again something we didn't invent but picked up from the psychologists and L beh hold a lot of game theoretical simple results get rejected if you have people play it in the lab so after a while some IST get fed up with that and they started thinking about different models and that's been the biggest change in economics since I've been in the profession and a large fraction of the work today does behavioral economics tries to incorporate those insights into more mainstream um modeling and so forth so I think this view that economics is just Theory not responsive to evidence it's just so false [Applause] [Applause] the last word is that what you're saying well well somebody's got to have the last word okay so so one thing I agree with is that probably true that the washing machine was more important than internet that's that I have sympathy with he look I think the the fundamental problem that I here it's the same as Steve said which is and I have it this very often when I um listen to events like today which is that the description of the economics as I leave it dayto day uh is is very different from the description that I hear from the critics um it's not this Theory obsessed math obsessed discipline I mean Steve showed you the data it's it's actually right now is a empirically obessed discipline if anything and uh and it's not uh it's not driven by math it's it's driven by observation uh and it's driven by a sincere desire to explain facts that happen out there and the mass is just incidental to that process thank you well over to those in favor of the proposition now for their final succinct summary yeah um two two observations I mean I think that you know some of the examples given about for example the effect of rising wages on on work effort I mean um you know you think one thing you think that people will work harder hours will work longer as their wages go up but then it was found out that um um there is also a Leisure option in all this now you don't need any modern economics or contemporary economics to understand that people actually do um uh have um uh make these choices John Stewart Mill realized that in 1844 um without any maths whatsoever and therefore I see a lot of this a lot of modern work as Reinventing the wheel an endless reinvention of the wheel that's the nature of the subject full stop and every generation people haven't been doing economics for long but just studied in the 80s I mean they will say no look we've made terrific advances I mean all that stuff that was in invogue 10 or 20 years ago that's all irrelevant now because we've got behavioral we're now doing different different types of better types of econom econometric work and therefore we advance I think we will um uh this is cyclical and we will uh so I don't I don't accept that SEC second point is I noticed Professor pisher he used the word frictions at one time in talking about the labor market where does the word frictions come from of course it comes it comes from physics and engineering and that is the model that um the economists will always have if things doesn't go if things don't go away the way that maybe the theory will tell them they should they will always introduce frictions and that is exactly the mindset that I am disputing thank you right well let me yeah let me make two points you know the proposition is that there's too much math too little history yeah you know we are not sayfe that we should abolish math you know we should turn economics into history no I'm not saying that you no I'm that pluralist both in terms of theory and methodologies you know I I mean I have you know never use any math in my work but I don't say that people shouldn't do math yeah I mean people should do math yeah we just need division of labor yeah that's our way economics that's where economic started wasn't it yeah what uh the first thing that Adam Sim wrote about is division of labor so that's the that origin of Economics yeah so we should practice that yeah yeah so that you some people should do math some people should do econometrics others to survey interviews historiography everything yeah and we need all these different theories to fully understand the world yeah so actually there's nothing to dispute in the proposition yeah because that that's that that only saying that given that we are actually doing too much math yeah yeah and I think that that very few people will dispute that you know because you know yes if we had nine lifetimes why not everyone become you Top Class mathem mathematician as well as historian as well as uh you know the interviewer and so on but we don't have enough enough time for that yeah so that somehow we don't have enough time for that yeah thank you yeah one last comment I mean we need to think about the cost benefit ratio that we get from this math yeah I mean are we getting enough benefit given the cost that that the student have to expend to learn all this stuff let's leave you that thank [Applause] [Music] you on my schedule I'm I'm required to deal with some voting I'm not quite sure how but would those who are in favor of the proposition too much math and too little history raise their hands and I'll count speak up yeah if you're in favor of the proposition there's too much math too little history but after a quick count I think we should uh consider those uh who are against the proposition feel good at estimating what's your intuition well I think it's a nice clean fight and I'm not sure the best side won but uh we um I'll leave that to the students to make the announcement I think thank you very much for coming thank you [Music] speakers so thank you again to our distinguished panel who gave up their time today to uh battle it out uh leaving no blood on the floor I'm very pleased to see if you wouldn't mind just moving out to the auditorium now taking all your stuff with you