good morning everyone welcome i was talking to interchipoleta at 10 o'clock on a saturday morning seeing a big crowd here and a lot of people downstairs who may welcome thank you very much for being with us it's really moving today and uh during this festival we have been speaking about technology and jobs indeed we are really followed by uh the media and a lot of people we had meetings with freeman at the beginning then many meetings yesterday today we are going to have a wide approach to a very wide subject for a lecture by our speaker he is a friend of the festival already he visited us a few years ago and he is a very competent person ada turner is not only a university professor uh he was a lecturer and a professor the lses and others he also had very important positions in government bodies as well as in private companies and banks he was in charge of the um the federation of british industrialists and enzo chipoli that did the same for the italian uh confederation and indeed they know each other's very well uh maybe enzo will ask a first question after the uh lecture lord turner and he also was in a position in great britain as a person in charge the central bank he worked at maryland at mckenzie he indeed was a manager and an executive so he will expand about the changes in the present system from many viewpoints he wrote essays he wrote recently text between debt and devil by princeton he also wrote economics after the crisis which was also translated into chinese uh i had asked him to start the lecture which will then be followed by your questions by talking about the activities of the institute which you are a chair or the president i.e the institute for new economic thinking and this is my question one of the key issues we've been dealing in the uh festival of the economy for 13 years is taking snapshots of the changes in the economic thinking the economic crisis has questioned many of the economic theories we hope we are at the end of the crisis hopefully so i wonder is there a new orthodoxy or are economists in agreement with some theory or whether conversely there are so many different viewpoints with no convergence and maybe that convergence had never been there we used to have a washington consensus others called it single thinking so i'd like to know what is the new economic thinking if any giuseppe thank you very much it's a great pleasure to be here um to answer your question the the institute for new economic thinking was set up in about 2010 in response to the financial crisis of 2008 and it was set up in the belief that economics had done a very bad job at forewarning us about the possible financial crisis and that it had developed into a set of propositions which while intellectually elegant were not related to the real world i think it is fair to say that up until 2008 there was a gradually developing dominant economic orthodoxy which in theoretical terms we can call neoclassical economics in policy terms we can call it the washington consensus and the building blocks of that orthodoxy were the rational expectations hypothesis that i believe that people developed points of view about the future in a rational fashion uh the belief that from the rational expectations hypothesis you could derive the efficient market hypothesis that markets operated in an efficient fashion and from that derived a broad policy consensus that all we had to do to make sure that the world was stable and efficient was to have as free markets as possible and that that would bring us closer to a competitive equilibrium the theoretical pareto optimality which mathematics can prove uh is possible and it was that orthodoxy which i think led to first of all a set of political mistakes which has allowed the development of inequality and some of the stresses which rise to populism that give rise to populism but also that orthodoxy very specifically led to a laissez-faire approach to the regulation of financial markets which led to the crisis of 2008 and the institute for new economic thinking was set up after the crisis to encourage a more wide-ranging and more real-world economics now there had always been people who diverged from the neoclassical uh consensus i see from the poster outside that joe stiglitz uh is here uh joe stiglitz is very closely involved in the institute for new economic thinking and of course joe stiglitz's own nobel prize uh was for a set of propositions which challenged many of the propositions of competitive equilibrium rational expectations and deficient markets so there always had been alternative economic theories but i think they were marginalized in policy terms that the overall consensus was dominant now to get you so inet was established to provide funding for research in monetary theory in distributional theory in market theory which provided a richer understanding of the way the world works than the neoclassical consensus had um but to get to the answer to your question are we going to replace the previous neoclassical orthodoxy with a new dominant orthodoxy i think the answer is no because i think and but this is a challenge because i think when you truly understand the nature of the economy when you truly understand the uh the way that human beings work and that they are not entirely rational when you truly understand the way that markets are not entirely efficient you cannot develop one beautiful mathematically elegant model to describe what is going on you have to accept the complexity of uncertainty that there are some things that we don't know that indeed was the great insight in john maynard keynes's general theory at the absolute core of the general theory is the idea that the future is uncertain and once you accept an uncertain future i don't think there is a single mathematically elegant mathematically complete a answer and that i think is one of the challenges because i think there is a strong intellectual tendency and a political tendency for people to gravitate towards systems of thought which are complete those used to involve systems of thought which were complete in a marxist sense and appeared to provide the answer to what had happened in history and what was going to happen in history they we we then had systems of thought of the washington consensus which were also equally intellectually complete i don't think we should aim to have a complete unifying theory indeed my own intellectual load star i guess are two whether two intellectual guides who have guided through me through life uh one is john maynard keynes and the other is karl popper and i think karl popper's theory of the nature of our understanding of the world is that we cannot we should not and it is dangerous to try to construct uh complete a uh intellectual systems and that the core of living in as karl popper put it an open society is to accept the uncertainty the contention and the only partial truths that we can tell about social conditions and economic theory well thank you let's go to the lecture i'm happy that removal day is with us tomorrow this morning he will appreciate mentioning karl popper shortly and tito boeri was here he's no longer there were no chairs for him anyway i'm glad to remind a person who introduced me to him arendorf the first year of the festival and he is part of the liberal thinking you have the floor great pleasure to be here great pleasure to be here with enzo chipolata uh who we we worked together when he was running uh confined austria and i was running uh the uh the confederation of british industry um i'm going to talk this morning uh hopefully my slides will come up uh oops about capitalism in an age of robots and what i want to consider um is what is the implication of radical automation what are the implications for society and for economic theory of an environment in which i will assert we will eventually be able to automate all work activities i have called the lecture capitalism in an age of robots but just to clarify by robots i do not necessarily mean a thing that looks like a human being with legs and arms and a sort of smiley face and eyes by robots i mean any machine any combination of hardware and software capability that we can use to automate work activity and my proposition is that we are heading to a world in which we will eventually not have to work that we do have the capabilities to automate all activity and that that has profound implications and paradoxical implications for the future economy and just so that i get you thinking up front about what those paradoxical implications are and so that you can see where we are heading in this lecture uh let me highlight six things which are perhaps different from the conventional uh wisdom first i am going to assert that the faster the pace of underlying technological advance the lower the measured productivity growth rate is almost bound to be secondly i will assert that automation threatens income from activities which are essential to human welfare but does not threaten income which is gained from zero some competition and i will explain what i mean by that third i will assert that the more rapidly information and communication technology progresses the more that wealth and income will derive from inherently physical and subjective assets such as land in particular places or brands or physical beauty i call that the high-tech high-touch paradox i will argue therefore that increasing product measured productivity growth is no longer the key priority in developed economies i will argue that better skills cannot solve the problem of rising inequality and i will argue that aging and low fertility are no problem and that youthful and growing populations are no blessing so i hope you can see that i am going to use my theoretical constructs to try to provoke many aspects and to challenge many aspects of the conventional wisdom my lecture is set out in six sections and i think it may be that the last two sections i will have to cut for the purposes of this morning uh just in terms of the time that we have available but the text is available and i have to give you an incentive to read the text and i can do that by cutting off after the first four sections i'm going to argue that we are heading to a world in which we can automate almost everything that the question is when not if second i'm going to address what is called the solo paradox which is well okay if we keep talking about this automation capability why is the rate of productivity growth so slow in developed economies i'm then going to talk about what rapid productivity growth and automation means for the nature of our economy that and argue that things like gdp are becoming meaningless then i will talk about challenges for advanced economies and above all i will say that the biggest challenge is rising inequality if we have time i'll say something about emerging economies and about implications for economic theory let me begin my assertion is that we will eventually be able to automate away almost everything and that information and communications technology is having a truly profound impact on the shape of our economy now whenever one says that somebody says oh yes but we've had waves of technological change before we had the agricultural revolution we had industrial revolution we had the introduction of electricity what is different in information and communications technology but i think there is a reasonable case and i think it has been very well set out in a book by eric brinulson and andrew mcafee called the second machine age i think there is a reasonable case that there's something about information and communications technology which is different it's different in about four or five respects first the pace of progress in the fundamental hardware is faster than we have ever seen before moore's law tells us that the amount of processing power which we can have available from a computer chips or memory or communications bandwidth doubles every 18 months and if something doubles over 18 months after about 50 years it goes up by about 3 billion times because that is the extraordinary power of compounding so what we have in our mobile phones is thousands of times more computing power the nasa used to put a man on the moon in 1969 and in 50 years time we will have a million times more capacity again in our mobile phones this is an extraordinary pace and it is far more rapid than occurred in the progress of the electromechanical technologies secondly information and communications technology has the extraordinary feature of software which is that when you've made one copy of a piece of software the next thousand copies or next million copies or next hundred million copies really cost you nothing there is a zero cost of replication and that is different and i'll show you why it's different from the electromechanical era thirdly i think we are on a path in which the development of artificial intelligence is going to equal human capability partly because of the first two effects and as writers have suggested once we achieve general artificial intelligence we will then create machines which have super intelligence intelligence far beyond a human capacity and fourthly we have the extraordinary aspect of machine learning you do not need when you are trying to work out how to automate a textile function to observe how people's fingers work and then write the code for that what you do is you move a glove on the end of someone's hand and the machine writes the code as the glove moves that is machine learning and it means that we have an extraordinary ability to automate away almost all activities now that establishes the proposition i believe that we will eventually be able to automate everything everything or just about which we currently call work now of course it will occur at a different pace for different types of activities and there is a very good piece of work which has been done by mckinsey global institute and it was produced last year where they tried to think through which types of work would be automated earlier and which would be automated later to do that process they began by thinking about fundamental human capabilities like sensory perception or fine motor skills or gross motor skills and they said can a machine do those as well as a human being now if you look down at the bottom it is green shaded against gross motor skills gross motor shields means the ability of a for instance an automated pickup truck to pick up a pallet and put it exactly where you want that palette machines can do that better than human beings now fine motor skills are things like sewing at the moment we are a bit short of the ability to automate that but we will get there once you have understood where we are and where we are likely to go in terms of fundamental capabilities you can translate that into what is the automatability how much can we automate different categories of activity and mckinsey argued that 81 of all activities which are predictable repetitive physical activities could even with today's technology be automated activities which in total account for 18 of all hours worked in the usa at the other end there are certain set of management functions where only a small percentage can presently be automated but you can see the different degree of automatability that then turns into how easy it is to automate functions uh occupations sewing machine operators and graders of agricultural products or even pickers of agricultural products those will be automated psychiatrists and legislators you may rather regret the latter but psychiatrists and legislators will be a little bit more difficult to automate uh perhaps you do not think that your politicians perform absolutely vital functions or perform them well but whether or not they are vital functions performed well they are functions which are quite difficult to automate and that then takes us through to sectors where things like accommodation and food services are more automatable than for instance the professions or education services but lying behind this is the proposition that it's all a matter of time in which we move on from automating the easy functions to automate to a wider set of functions and mckinsey have set out two different scenarios of time scale the one in blue with an upper and a lower bound is how long will it take us to be able technically to automate away existing work activities and they say at the moment we could work out we could technically automate away 50 percent of work activities and that will rise to a hundred percent in an early scenario much of that will occur by 2030 in a late scenario by 2050 but it will occur the second green area is when will it occur given that the process is determined not just by technological capability but by the slow process of diffusion in the economy and because of the economic trade-offs that businesses make between capital and labor if labor is very cheap you sometimes don't automate something even if it is technologically possible to automate and again they have a range there but their range essentially says that by the end of this century it is highly likely that we will automate away almost everything and what i'm going to do for the rest of the lecture is not focus on if it will occur or even when it will occur but to assert that the question is when not if that we are heading towards the ability to automate everything and to ask the question when even if it's in 30 years 50 years 100 when we get to an economy in which we can automate away all work what are the consequences for society because what i'm going to suggest is that even if we are slowly only slowly heading towards that end point the fact that we are heading towards that end point already has pervasive and observable effects on our economy however before asking what are the effects i first of all have to address the solo paradox because economists will say at this point adair you are telling a story of amazing technological advance with all these machines that can automate away everything but i observe that productivity growth in the us in the uk in italy has actually slowed down so this chart is taken from robert gordon's book on the rise and fall of american growth in which he points out that the most rapid period of growth in output per person or an output per hour was between 1920 and 1970 which is before we all began to talk about the ability to automate everything through computers and it was in 1987 the the robert solo nobel prize winner in economics made the famous phrase you can see computers everywhere except in the productivity statistics he said everybody's talking about computers but why don't i see them in the productivity statistics but what i'm going to argue is that the solo paradox is no paradox that once you allow for three effects the solo paradox is exactly what we should expect to see those three effects are what i call first the bow ball effect in honor of economist william beaumont and an article from 1967 on the economics of unbalanced growth and the high-tech high-tech paradox second what i call the zero sum paradox and third the existing of nil the existence of nil or low-cost benefits which are not in gdp first then the beaumont paradox the high-tech high-touch paramounts i think when most of us think about productivity growth we all probably start with a standard paradigm in our brains which is about the transition from agriculture to manufacturing you start 300 years ago before the industrial revolution with a hundred self-sufficient farmers producing 100 units of food and then we have an agricultural revolution and we work out how to produce those 100 units of food with just 50 farmers and the other 50 workers go off to factories in the industrial revolution and they start producing cars and washing machines and televisions and measured productivity doubles and in this standard model which we have in our mind this process is endlessly repeatable because in the next period the 50 factory workers can also produce 100 manufactured goods the 25 farmers can produce 100 units of food 50 factory workers can produce 200 cars washing machines and televisions 15 of those factory workers go off to produce 60 units of computers mobile phones and software and 10 produce health care and you uh and other services and every time we take labor out of one sector we move it into another sector which can also be automated so that in every time period in this simple model productivity doubles it's an endlessly repeatable process and it's a process which we can apply as much to services as to manufacturing however it is important to realize that a different effect is possible it is possible that when we move from 100 farmers producing 100 units of food to 50 farmers producing 100 units of food that instead of those people moving to manufacturing they all become domestic servants paid half as much and producing just 50 units of value in which case agricultural productivity has doubled but the total economy productivity has only increased by 50 percent and if that occurs if what people move into is in itself not automatable what we have over time is an asymptotic process where in each period agricultural productivity improves people go off to be domestic servants more and more people go off to be domestic servants and eventually there is a mathematically asymptotic limit where productivity has gone up by 100 but that is as much as you get and whereas i have shown it there with them moving into low paid jobs it can also be moving into high paid jobs we can have the 50 farmers producing 100 units of food 45 domestic servants paid half as much to produce 45 units of value and five artists singers entertainers and fashion designers paid twice as much again the productivity growth will eventually come to zero now this is theoretically possible but i also think it's important to realize that i think it has happened to before i think this is essentially what happened about eight thousand years ago in the first agricultural revolution which began in the fertile crescent of mesopotamia northern iraq and syria as it is today about eight millennia ago we worked out how to get much more productive in terms of agriculture but there was no permanent revolution in productivity across the economy because all of the surplus went into huge numbers of domestic servants for the priests and the kings the building of pyramids uh the creation of artistic activities but there was no perpetual revolution and indeed many historians of that period believe that the standard of living of ordinary people actually declined despite the agricultural productivity revolution and was lower than for hunter-gatherers so it is possible and it has occurred now we tend to believe it will not occur in future because we tend to think and perhaps we are right to think that we live in a world in which we will always create new productivity capabilities we will always work out how to automate new activities whereas in the sixth and seventh century they'd worked out how to increase agricultural productivity but they then didn't have a manufacturing revolution whereas today we have organized processes of scientific discovery and innovation and entrepreneurship where we continually create new productivity but what hap but the determinant of whether we have this bowmore model or the standard model does not just depend upon whether we physically know how to automate new activities it also depends upon two other things first on the impact of the initial productivity spurt on the distribution of income and secondly on what the people who receive income choose to spend money on if the initial productivity revolution congregates income into the hands of a small number of people and if those people choose not to buy endless more cars and washing machines but endless more domestic servants then you will get a slowdown in productivity growth even if we physically know how to automate new activities now this is theoretically possible but is it occurring well in the uk we have very rapid new labor a job creation we have low unemployment and new job creation but we have very low real wage growth why is that that is because we are proliferating in london what i call 21st century jobs this is a man on that piece of 21st century technology known as the bicycle with a pizza on his back or anything else you want delivering it direct to your house and we are getting a proliferation of those types of jobs and if you look at the u.s bureau of labor statistics figures you will see that the jobs which are being created in large numbers are not software developers or software applications they don't even meet the top 10 forecast job categories there are a whole load of jobs in things like personal care aids home health aids nursing assistants and they are jobs which on average are paid much less than the average income and we also see this effect in india i noticed the other day in the economist magazine an article about the automation of tea packing a manager has explaining was explaining to the economist journalist that he'd got a new machine to do the packing of tea to replace a physically repetitive process and he said look round the factory that job's going that job's going that job's going but the manager insists that as in the past he will somehow find jobs for everyone as drivers or as watchmen or as people to bring the tea to management etc in advanced economies with rapid population with rapid productivity growth in some sectors of the economy you can have that offset by a proliferation of low productivity low wage jobs elsewhere the second effect at work is what i call the zero sum paradox and let's return to the standard model to understand that because back in the standard model we have another variant suppose we begin with 100 farmers producing 100 units of food they become more productive and now we have 50 farmers producing 100 units of food but the other 50 don't move off to be manufacturers and they don't move off to be domestic servants 25 go off and be criminals and 25 are employed as police to defend the farmers against the criminals now because somewhat arbitrarily within gdp accounting we count police as part of gdp and criminals as not part of gdp productivity will go up in this case uh by 25 because gdp will increase but i think we can all agree that there has been no increase in human welfare and we can get in economies a proliferation of zero-sum activities there are two questions to ask about those zero-sum activities one does total measured productivity rise in this case yes but not as by much as if everybody had gone off to be manufacturing workers and does human welfare rise well if the activities are zero sum no now it's theoretically possible but is this happening in modern economies i think it is i think if you look around the modern economy and look at what jobs people do you will see an extraordinary proliferation of jobs which essentially cancel each other out they are zero-sum they're necessary jobs once one person does one side there has to be somebody on the other side they're perfectly good jobs i've done those sort of jobs i suspect many of you do those sort of jobs but they cannot possibly increase human welfare anybody who works for a major company now knows that we are very very worried about cyber criminals lots of very very clever people are making income as cyber criminals and we are now within companies employing more and more lots of very very clever people to defend us against those cyber criminals but that is a zero-sum activity suppose we get cleverer and cleverer people going to become divorce lawyers will human welfare increase well no the income of divorce lawyers might increase but there will be a very clever divorce lawyer on this side and a very clever divorce lawyer on that side and it's zero sum it cancels it out that is true for a lot of corporate lawyers that is true for the political process which employs more and more people and in america now accounts for absolutely enormous uses of money and technology devoted to a political fight which must be zero-sum that is true of a lot of advertising trying to persuade you to buy brand a not brand b it is even true of some aspects of education where education is partly an increase in human capital but partly it's something you have to do to signal to the job market that you are clever enough to get into that university and in america this is unleashing an arms race of expenditure on education which is essentially a form of zero-sum competition in the lecture text i go through in greater detail what all of these zero-sum activities are but my argument is that we can expect them to grow in a modern economy their impact on measured gdp is arbitrary divorce lawyers because they are paid for individually are in gdp corporate lawyers count as an intermediate product which are not in gdp but that is sometimes rather arbitrary they are often highly paid and they are somewhat less susceptible to automation than the jobs which are really essential um uh legislators are less automatable than textile workers actually over time we will be able to create to use more technology to prove improve the effectiveness of lawyers but it will almost certainly not mean that we employ less senior lawyers it simply means that in a divorce case or in a corporate case one will use automation to search for every single ever relevant precedent one will simply be in an arms race of technological capability on either side one will improve the quality of the product paid for by client a and client b but one will not automate it away so we end up i think with what i call the zero sum paradox that in the end automation of textile workers jobs which is absolutely essential for the clothes that you have will mean there are no textile worker jobs at all whereas the application of i.t to legal services will simply mean an arms race of greater capability and skill and a greater computer power employed on either side ask yourself this it is almost certain that by 2050 it will be possible for a team of robots to beat manchester united at football but i don't think that will reduce in any way the pay of manchester united football players there is something about uh technology which threatens the jobs and the income of the people who are truly necessary for everything that we need but does not threaten the jobs of people who are involved in zero sum competition or games rapid technological progress could eventually automate away almost all activities which are truly essential for human welfare while supporting increasedly intensive zero-sum competition for relative income and status so that zero-sum activities account for an increasing percentage of employment and measured output over time so that sounds all a little bit pessimistic we have this amazing ability to automate things and what we do is we proliferate a whole load of zero sum activities zero sum competition or a set of low paid low productivity jobs the third factor which i'm going to consider might seem a little bit more optimistic and it is the argument that there are a whole load of benefits uh going on in our economy which we simply don't capture in gdp imagine this imagine that sometime this century a very small number of people create are able to be clever enough using artificial intelligence to create a wonder drug a wonder drug that enables everyone to live 200 with no health problems whatsoever no dementia no arthritis no early tragic deaths from heart attacks or childhood cancer we all have a hundred years of absolutely certain healthy life i think we can all agree that that will be an immense improvement in human welfare but if you run through the gdp accounting and work out by 2100 by which time that drug will be off patent and will be manufactured and will be manufactured in an automated factory for almost nothing what that wonder drug will contribute to gdp the answer is almost nothing one of the most important breakthroughs will hardly be in our gdp figures at all and again i explored it in more detail in the lecture how the precise gdp accounting works but there is a very good argument that what we have is a whole set of things like new drugs mobile phones and tablets streamed films and music computer games and social networks which are provided at a declining cost or in many cases just given to you for free search engines at zero cost where we have rapid productivity falling prices and increasing quality but they are inadequately captured in measures of real gdp and thus productivity growth marty feldstein has argued the text that you can see there the result is he says that the increase in real incomes is underestimated and that the common concern about what appears to be the low growth of average household incomes is replaced these low growth estimates he says failed to reflect the innovations in everything from healthcare to internet services to video entertainment which have made life better during these years now i'll come back later to a caveat of why i don't entirely agree with marty that all these innovations have made life better but what i think is clear is that there's quite a lot going on in automation which is delivering remarkable new services which we cannot be confident is being captured in gdp and in the text of the lecture i set out the three effects combined you don't need to look in detail through it you can look at the lecture but if you ask yourself reasonable parameters of how many of those beaumont type low productivity jobs are being created how many and is that growing how many jobs in the economy essentially zero some and is that growing and how much are we underestimating the value of our mobile phones our new drugs with reasonable parameters you can end up with the results over on the right hand side that you can have sectors of the economy where we are achieving rapid and indeed accelerating uh growth of automation but we can have gdp much lower and declining so my argument is the solo paradox is actually what we should expect to see what are then the implications of this for the shape of the economy and the challenges for society the challenge the implications for the shape of the economy are that we increasingly have meaningless standard measures of economic success and that value is going to migrate to new areas we live with a standard assumption that we know what we're measuring that we have technological advance can drive productivity improvement across the economy which shows up in gdp measures of output per hour and per capita and that that provides a good measure of the increase in human welfare and those i think were reasonable assumptions at an earlier stage of economic development but they are assumptions which are now collapsing in the long term i think they collapse entirely imagine an end point in 2100 in which we have solar-powered robots guided by artificial intelligence systems do almost all the work needed to deliver welfare enhancing goods and services but this work accounts for a very small proportion of measured gdp while almost all human work activity is devoted either to zero some competition lots of lawyers and a uh financial traders and politicians uh arguing or trying to get an advantage uh vis-a-vis one another or to the explosion and this might be rather attractive of the things we can't automate artistic craft design activities but non-automatable activities or low-wage activities of face-to-face service in those environments the growth or not of gdp per capita measured on current conventions will i suggest tell us almost nothing about trends in human welfare indeed we can raise the question of whether in that environment the economy even exists economics is meant to be about the allocation of scarce resources in production and in consumption but if robots do all the work is there any scarcity well the answer is i think we still have a world of measured income and wealth but it will have moved to some very different areas i think most wealth will derive and much income will derive from the ownership of locationally specific property from intellectual property rents and indeed a crucial determinant of the distribution of income will be what attitude we have towards intellectual property we will live in a world in which very small numbers of very clever people can create new technological capabilities new drugs new artificial intelligence and the impact of that on income will be entirely driven by what our intellectual property rules are whether we and how long our patent production periods are how long our copyrights are we will have enormous subjective brand values and rents fashion designers those individuals who are able to create the fashion which captures somebody attention will be paid enormous amounts of money and the ver there will be very high incomes a very small number of people skilled or lucky in i.t subjective value creation or zero-sum competition and my argument is that we can see some of this happening already this chart is taken from thomas piquette's capitalism in the 21st century now i'm sure that since you've all turned up on a saturday morning to listen to a lecture in economics many of you perhaps most of you have heard of thomas piquette i'm sure that some of you have bought thomas piketty's book and i'm pretty confident that a rather smaller minority have actually read all 700 pages of thomas piketty's book but it is a very good book and it sets out this remarkable fact of the amazing variation in the wealth to income ratio in modern economies and here are the figures for france where the ratio of wealth to income which was down in about 300 percent three times in 1920 is now about six times but what piquetty does not himself stress is the blue shaded bit on this graph which is that all of that increase in the wealth to income ratio is explained by an increase in the value of property and not by the construction of that property but by the value of the land on which it sits increasingly wealth in our high-tech economies paradoxically is migrating to the most physical thing of all land and to an asset which is not produced at all you can see this in uk national income accounts over the last 20 years the produced assets which economists tend to think about where the capital in year one is the capital in year zero plus the investment in the intermediate year those produced assets have gone from 2.4 trillion to 4.8 trillion but the unproduced assets which is essentially just land have gone 1.6 trillion to 5 trillion we end up i believe with what i call the high tech high touch paradox that the more rapid the progress of information and communications technology the more value will be placed on inherently physical assets and attributes desirable land sporting capability physical beauty of a supermodel or a a a film star and created subjective values fashion design brand celebrity what then is the challenge for advanced economies and i will end with a few comments on this so we have a few moments for questions some people think that the challenge is that there isn't going to be enough work to go around i believe that is wrong i believe that if you have a flexible labor market if people have to work to get income we will somehow create jobs we will proliferate new activities we will find things to do those may be zero-sum competitive activities or they may be this proliferation of what i call beaumont type jobs low pay low productivity uh service jobs but somehow we'll find things to do but it is highly likely i think that this will be an environment of rising inequality we have had very significant rising inequality in developed markets for the last 30 years that has been most dramatic in the us and whereas i think some of that derives from the globalization of trade and some perhaps from immigration i think probably the biggest single factor at work is technology what technology enables is very small numbers of people to create enormous equity value and income consider these figures these are the figures for the equity market value and the number of employees of those amazing internet and i t and software companies microsoft alphabet google facebook alibaba tencent facebook 502 billion dollars of equity market value 000 employees an absolute drop in the ocean of the global labor market if you look at that as a relationship between equity market value and the number of employees it is completely different from the world when general motors was the most highly valued company in the world when general motors back in the 1950s had the highest equity market value in the world it employed a million people these companies will never employ anything like a million people because they simply don't need to however we will create jobs as i showed earlier the us is creating lots of jobs but it is creating them in low paid low productivity areas so what do we do about this well marty feldstein's argument earlier was don't worry because actually although these people have low wages they've got completely free computer games and they've got you know free drugs that they never had before but the caveat i have with that is that i don't think you can assume that all those developments have necessarily improved human welfare i think the wonder drug that i imagined earlier would be a clear increment to human welfare but i think it is debatable whether the proliferation of ever more sophisticated computer games which lock our teenagers in front of screens all the day long is necessarily an absolute increase in human welfare and i certainly think that if you imagine a low paid person facing high property costs high rents and high costs of commuting to work and feeling that they are left behind versus other people are getting much higher incomes telling them it's okay because you've got three computer games is not going to be seen as an adequate answer nor do i think that skills is an adequate answer now let me be clear no one is more passionate about skills than me about educating people about enabling people to understand this world in which they live but however many people know how to code a computer game all the revenues from computer game writing will accrue to the minute percentage of all competent computer game coders who in that year create the computer games which capture people's imagination we have in these ict intensive industries a huge winner takes all tendency and the fact that we give everybody the skills to be an adequate uh competitor in that area will not mean that we deal with inequality so what do we do an american writer called tyler cohen in a book called average is over has set out a sort of dystopia of how the future will work he says what we'll do in america is that we've got lots of flat warm land on which we can put very cheap houses essentially trailer parks poorer people will live in tiny houses in warm climate where there's low land cost will provide them with adequate health care well they're not quite doing that in america at the moment but suppose they do and there will be close to zero cost entertainment meanwhile there'll be some very competitive people who want to go to london or paris or milan and compete like another and earn huge amounts of money uh and and they'll be involved in all these zero-sum activities or they'll be the guys creating the wonder drugs in silicon valley and then there'll be a third category of people which is talented bohemians who will opt out of the major city rat race and they'll go to live in cities which have you know a nice living style and uh relatively low cost because they've been left over by the past berlin is one of the favorites now in europe at detroit and they'll be artists or organic urban farmers or beekeepers and they'll go around on bicycles and they'll be very happy i mean i think it's a very attractive lifestyle but that's what he says and he says people warn that this era of inequality is going to be such rising inequality that the poor will come up and they'll you know they'll have a social revolution he says no no we really shouldn't expect rising income and wealth inequality to lead to revolutionary revolt the long-term picture will be fairly calm but unbelievably unequal now of course responding to that you've got to decide two things whether you believe that that's where we're heading and whether it's acceptable i don't think it's acceptable but i do think tyler cohen has captured something about how these trends will play out i think we have to try to make sure it is not as unequal as he is suggesting what are the policies that might make a difference i think we will have to consider redistributive taxation on real property wealth capital gains and inheritance because i think we live in a world in which real property is just going to rise in value i think we should consider less favorable treatment of intellectual property royalties and rights i do not understand at all why the us has found it necessary to extend the property right of the walt disney a copyright on mickey mouse we have intellectual property rights in order to encourage creativity and intellectual property creation well we don't need to give walt disney any new incentives to create mickey mouse because mickey mouse already exists and walt disney is dead so the extension of the copyright for walt disney is simply providing a distributional flow of income it has no beneficial effect economists don't often talk about economic geography but thinking about urban design and development is important because we need to find a way to make as many cities as possible and as many areas of cities as possible pleasant attractive desirable places to live so that we do not intensify this endless competition for the property in new york in shenzhen in hong kong in london in paris in milan which drives up property prices we may need to consider universal basic income to make sure that people have an adequate lifestyle even if they do not earn high monetary incomes and i think we have got to simply decide as a society that we provide adequate pay for basic services such as social care for the elderly which we need to expand which ideally we should want to do in a face-to-face version rather than a robotized version but which at least in the uk we pay people very very little to do and there are some favorable developments it would be a good idea if people took some of the benefits of increased productivity and leisure and we need an increased focus on artistic design and craft skills and by the way and i'm not going to go into the details on this i think in this environment it's actually quite good to have low fertility rates and potentially declining a uh workforces because i think it reduces the process of endless a downward pressure on labor so that's my thought every now and then i think i've had an original thought and then the trouble is i find that one man got there 80 years ago and that man is always the same man and it is john maynard keynes because what i'm saying is not fundamentally different from what john maynard keynes said in an article called economic possibilities for our grandchildren written in 1930 that once we have solved the problem of creation for the sort of production once we have solved the problem of production of how to produce things with very little work then for the first time since his creation man he ought to have said man and women but you know i mean forgive his uh historical tendency there for the first time since his creation since the first time since her creation woman will be faced with her real her permanent problem how to use her freedom from pressing economic cares how to occupy the leisure which science and compound interest will have won her to live wisely agreeably and well that is what i think information technology is delivering us and i actually think that solving that challenge is an even more difficult challenge than solving the problem of production which i think we are well on path to do so you will have to read the text to see what i say about the challenges for emerging economies and the implications for economic theory thank you very much benny um very well while lord turner was talking i was wondering whether amongst future scenarios there was also the one that mr dose evoked yesterday together with mr chipoleta claiming that a good representation of this world that may come is given by the first film blade runner where a super technological minority governs a vast majority of sub-proletarians i believe we've got about 10 minutes so we can accept a few questions but first i would like to ask mr chipoleta if there's any specific question for mr turner well first of all it was a pleasure to meet the unconventionality over there once again he's one of those economists that franco perot said managed to reason diagonally neither pro nor against the existing assumptions and i remember that when we were both the leaders of our two sister associations there published an article in the solivente cuatro stating that wages had to be increased i mean for a director of an association of entrepreneurs the idea of increasing wages is pretty clashing however this is proof of the fact of what a free spirit there is so i totally agree with the dare regarding the solo paradox a couple of months ago i gave a lecture in ancona and i said more or less the same thing with the exception of the zero sum activities theory what is actually missing in the growth of productivity team and i would like to know adele's opinion in this respect is actually growth i mean if we had more growth we could have also jobs with higher productivity but if there is no growth as in the past 15 years machines replace men and workers have to find any job possible and they can take on only low productivity jobs and productivity decreases in general if we had a stronger economic growth and a stronger demand there would also be a growth in wages and productivity would also go up because we would have higher content jobs because of the demand for consumption good and investments that would be stronger than what we have today so this is my opinion so i have a specific question for adair in relation to his position regarding the new thought in it i-n-e-t i mean the rational approach it's not satisfactory i mean people act following different impulses some are not rational for instance i mean if we move towards an automated world where algorithms decide about most of the choices and decisions to be made i mean if you buy a book on kindle you receive so many offers that have all been structured on your previous choice and based on people like him or her who have made these choices i just erase all that because i can't accept that an algorithm decides for me what i'm going to buy however in the world of finances algorithms are used to know how to invest your money so the economic theory at this point well should the economic theory investigate the behavior of people or the algorithms won't this become a vicious circle i mean if 90 of the economy is based on algorithms why should i study the behavior of people i have to study algorithms and the people have made algorithms and if i want to change the economy i just have to change algorithms rather than adopting economic measures is this meaningful and how will the economic theory be modified considering that a large part of our activities is no longer driven by our decisions but rather by a series of algorithms who have been decided by people anyway and even if they learn the algorithms learn on the base of algorithms that have been constructed so in this case human beings are like god in this type of creation well can you please wait a second before giving the answer shall we see if we've got other questions maybe you can take note of these two first questions so will economy in the future study machines rather than human beings that's the question second question and the first question refers to growth i'm a psychiatrist so you won't be replaced by machines that can deliver cognitive behavioral therapy for say depression so even psychiatrists can be substituted by robots but i think the robots are very slow in learning how to pay taxes because i pay taxes but the program automatic program to deliver cbt is not paying any taxes and i think as you suggested if this kind of robot can pay taxes this money can be invested in human development instead of just giving money to very few people this is what i think i'm gabriele veteran russell used to say that the ethics of work is the ethics for slaves and we do not want any slaves this is strongly related to the last part of the presentation well at this point i believe that the challenge is a cultural one i mean there is still a mindset which is anachronistic trying to pursue these ethics of labor of work viewing work as something positive as a principle however i believe that this mindset is still very widespread both on the left side and on the right side we have to overcome this kind of mindset how can we do that since this has become anachronistic well if there are no other questions we can leave the floor to there but actually i have a curiosity how many of you in this room come from trento seven or eight people ten well we have seen throughout the years and this is pretty impressing that the festival has grown significantly people attending the festival come from all over italy which is something beautiful also if you consider the fact that most of these lectures are also streamed they are also published on the website or the festival so before leaving the floater there just let me tell you that tomorrow if you're still here in trento since i can imagine that many of you will have to go back home prior to the conclusion of the festival at 5 30 in diato sociale we will have a meeting at three with mr soros the hungarian finance tycoon who's actually the founder of inet the institute of new economic thinking so he will come to trento he will speak with chipoleta and other economists about the europe we need in the future so if you're interested tomorrow three o'clock tetrali there you have received many many questions can you answer them in about 10 minutes thank you but please speak slowly thank you well just to say you were commenting on the remarkable achievement of the trento festival of economics when i was first invited to this festival of economics eight years ago i think it was i realized that the words festival and economics were not often combined um economics was seen as the the dismal science uh but you've achieved something quite remarkable here so three questions i i i i within enzo's question had lots of complexity to it but let me focus on on one thing um algorithms in financial trading are really interesting uh example of what you're suggesting um people are trying to develop algorithms which look at the patterns of financial trading and then automatically spot opportunities to make money other people are saying i have no idea where the market is going to go so i'm just going to hold a passive index of the whole market if every single investor in the market was either a non-human algorithm or a passive index investor the level of the market would become a sort of random number generator because you wouldn't have any sort of real judgment going on as to uh what the market uh ought to be so there are some real real challenges in the area of of algorithms about the circularity um and that applies in many different areas it applies in politics i think uh what we have realized uh in the us and the uk with the role of cambridge analytica is the way that the use of algorithms is changing the nature of political uh debate people are not being presented by a an argument which attempts to convince them the algorithms are trying to work out in advance which argument they are susceptible to and then making that argument to that specific a group of people and that is really changing the nature of politics in in a quite uh worrying fashion um so i don't know what the answer to this is but i think there are certainly some uh incredibly complex uh issues uh raised by uh this dominance of algorithms on the growth question and the going back to my comment in 1996 uh which was slightly taken out of context about paying people more it was in an environment in the uk where we had had several years of very low pay increases but i do think there is an issue for capitalism about the balance of returns for capital and labor and there's a very very good uh book by raghuran rajan who was at the imf and then suddenly subsequently the reserve bank governor in india and he wrote a book called fault lines about the origins of the financial crisis and he pointed out that one of the problems was rising inequality because rising inequality led to rising debt levels that when you had an environment where more and more income was concentrated in the hands of richer people richer people have a high marginal propensity to save because they've already got all the cars and clothes and washing machines and televisions they need and that can mean that unless something happens that accumulation of income in the hands of richer people will mean what larry summers has called secular stagnation there will not be enough demand in the economy unless the income of those richer people is picked up by the financial system and lent to poorer people who are trying to make up with the use of credit for the fact that their real wages are not increasing so if you have an environment of rising inequality it can be the case that the system only balances and only produces enough demand to drive growth it only produces that with a rise of debt which is sustainable for a period of time but then produces a crisis and uragu's argument was that that was one of the key origins of the crisis of 2008 and it is true that way back in the early 1920s henry ford was trying to work out who could possibly afford to buy his model t ford at the price he was producing it and he realized that the price he was producing it his workers could not afford to buy it so he wasn't going to sell many so he did begin to increase the wages of his workers so that they could afford to buy his model t ford and everybody else would hopefully increase the wages of workers so other people could produce other cars and afford to pay it there is a a problem of the balance of um a income and consumption and savings which can be upset by rising inequality and there are uh it's interesting when one talks about new economic thinking one sometimes has to return to old economic thinkers there is a an economist called kaleski who wrote many many years ago and neo-marxist about the balance of returns to labor and capital and they are not unimportant on the uh issue whether cognitive behavioral therapy can be automated whether it be effective uh or not whether robots pay taxes um as you probably know many people have said we should tax robots um the crucial thing i think here is to not have that in one's mind saying i'm going to look at something which has arms and legs and a smiley face and say i've got to significantly you know tax that particular robot um but as i said earlier when i use the word robots i mean it to represent any uh category of machine capability hardware or software and if you would put it not as we should tax robots but are we in a world in which we are getting higher returns to capital higher returns to intellectual property rights higher returns to real uh physical property and for many people lower real wages in that environment we do need to think carefully about our tax policies and broadly speaking it would be good if we could impose some more taxes uh on real property or intellectual property rights or copyright uh or or things like or capital returns and less on labor it is true to say that there are tendencies tending to push it in precisely the opposite and perverse direction and those derived from the fact that ordinary workers find it more difficult to move to avoid taxes than do capital intellectual property rights brand names etc once you've got a highly profitable brand name you can locate the ownership of that brand name in luxembourg and then your italian subsidiary can pay a brand use payment to luxembourg or ireland or the virgin isles or wherever is a low tax environment so we do have a problem that logically we should be thinking about shifting our taxation at least at the margin away from labor onto other sources of income but in a globalized world those other sources of income find it easier to avoid it than does uh labor uh the ethics of work the ethics of slavery um well i i think this is a very this is a hugely important issue which we we really need to to to to get to grips with at one level i don't think that we've just created a sort of focus on work i mean max weber argued that this was a peculiarly protestant thing uh to do the protestant work ethic coming out of calvinism uh that you had to you had to work to prove that you were one of the elect that you were going to heaven you know and that is part of the story you know that there are parts of the story in which certain forms of religious belief ethical belief and the needs of a capitalist society have created a sort of ethos of work but i think we also need to put it around the other way you know people will not be happy without a sense of purpose and a private sense of sociability one of the reasons why simply giving the unemployed money is not a reasonable answer to unemployment is that work is a social environment work is where you meet other people work is where you often meet friends it's often where you meet your partner become married it's and and we can't entirely replace that and work correctly done gives people you know for many people work is a sense of creativity certainly if we're talking about craft work artistic work a sense of pride in work so we certainly i don't think we can simply say okay we've got all these machines fantastic uh we'll all just have we'll all just work for five hours uh a week and we'll have 95 leisure because i suspect that human beings will that something is coded into us biologically i mean we are we are the product of an evolutionary process where for all of that evolutionary process we did need to work to live and so it's probably coded into us that we want to do something which we feel gives us you know meaning and satisfaction but certainly and i don't have the answer here i think we're increasingly going to have to think about this how do we create an acceptance i think would be better to have somewhat higher leisure but how do we also create a society where people do have a balance of of work some income from work some income from maybe universal basic income but crucially from that work get a sense of you know social purpose and value rather than it simply being you know work of slavery as as burton russell put it yes thank you very much thank you adeya turner enjoy the festival you