the title of the festival is identity and global crisis we will focus on the first of these topics so that is identity and we will speak about identity and personality which are two very similar subjects identity is something that one chooses and personality is the result of many factors while there can be more identities if you want there can be different identities depending on the problem tackled uh they can be father or son in the same day and personality is unique of course there are cases of double personalities but normally there is just one so the notion of personality and identity has been another topic both of economist research works and also psychology research works and professor heckman is perhaps the person who has a most focused on psychology neurology biology economics um so it's been a very broad-based synergy and cooperation and it is the scientists themselves who look forward the contribution of professor eckmann and all this is due to the rigor rigorous work of the james hackman this is something which has always characterized his academic life james heckman is anti-ideological he was at the university of chicago where it is commonly known that there is an approach which is against public interventions but on more than one occasion and professor heckman actually recognized that markets need regulation and that public interventions are necessary for some cases then he is a very critical man also self-critical man i remember a famous declaration once he said many things are made by the economist explain things that people already know and the reason why he received the nobel prize is for his contribution in the area of micro data analysis trying to understand that some of these data actually hide information and so it's necessary to remedy these situations of errors of biases because there are samples which do not give us full information hackman's work has been a guidance for all of us he also focused on data saying that it is necessary to have also the theoretical work of economists and this work is only some decisions work and this is very important because perhaps in the past economic economists looked at data only as pure statistics without asking the right questions to data today we will uh focus on the distinction between cognitive and non-cognitive abilities and the processes which lead to the formation of abilities or capabilities now cognitive capabilities are those which are commonly tested for example when you have a pisa test or many school tests school exams you test the level of cognitive abilities or iq tests for example those are tests which look at cognitive abilities but then there are also non-cognitive capabilities for example perseverance consciousness or self-esteem well these non-cognitive abilities have been focused on by professor ekman and our major uh capabilities which affect the behavior the performance and the success of the person and they also affect the acquisition of cognitive capabilities well but apart from his theories it is his activity which demonstrates that non-cognitive abilities are important james hackman an unterrible person a person who continues to have multiple interests and he has an agenda of research which is really unlimited and when you invite him and i'm very happy that he accepted our invitation you feel a bit guilty because it is you say well we took uh time away from him which he would have used to go on with his research projects and my guilt is double because i'm taking away time from his presentation and so i give him the floor straight away uh thank you very much tito for that kind introduction i'm not sure uh i and you're gonna i'm losing that much uh productivity but we'll see i'm very happy to be here at this conference happy to be here with george akerloff and looking forward to his talk and the general theme of identity and identity in economics is of course something that george has done fundamental work on and i view personality as being part of the aspect of understanding identity and i want to give some aspects of identity and psychology and psychology of personality that is maybe not so familiar so let me give a little overview about the relationship between economics and psychology at one point economics and psychology were closely united this was like 150 years ago and i think anybody who knows the history of economic thought will know that a major development in economics in the first half of the last century was the weaning of economics from psychology this was viewed as high level of progress and it culminated in reveal preference analysis that replaced any concept of measurable utility uh any cardinal utility as it's called with ordinal concepts and minimalist assumptions but i think what's happened in the last 60 years is a gradual reattachment of the two fields and since the middle of the last century economists starting with von neumann and morgenstern have returned to certain versions of cardinal utility the measurable utility to understand choices that people make in situations where there's risk and we know and there's a strong group here in trento uh in behavioral economics that's extensively drawn on the psychology of preferences the work of kahneman and tversky notably and others has introduced a variety of new preference specifications into economics to explain behavior behavior that's found not only in the field out in natural settings but also in laboratory and this is a major development and an important development for our understanding of phenomena and what we've learned from this subject has been new concepts of what we mean by risk aversion how people perceive the future time discounting notions of loss aversion and ambiguity aversion and social preferences that actually explain a lot of behaviors that standard theory cannot explain in that sense it's a major scientific advance and it's an advance that continues there's also another line of work by economists and i mentioned a dutch economist of some 30 years ago bernard von prague who working with psychologists have started to measure happiness and of course there's a large body of work done by kahneman alan krueger and richard laird and others that quantify a version of bentham's social utility bentham is associated with a concept of the greatest good for the greatest number but he had a real sense i think that a utilitarian approach could be captured by some quantitative utility and that that approach is still used but what i want to talk about today is a more recent development and that's something that i think is new and exciting and that is that economists have drawn on and contributed to an aspect of psychology that is what i call personality psychology i will i will define that in in in right now but i think that personality which has an intuitive concept is one aspect of identity and it fits well into the theme of this conference so let me do a little background because i want to place how personality psychology entered economics is a little different than the way other components of psychology entered and i think there's an there's an interaction between the psychologists and the economists in personality psychology that's a little more unusual than the traditional relationship so let's go back and review the basic economic model of choice and and to see how i see personality psychology contributing so the standard model that many of you have probably seen or you've seen versions of or know or teach is a model that talks about preferences that people have over choices and then constraints that people face so we always live in a world of constrained environments and we also have preferences some people for more prefer more of this and less of that and are willing to trade off and so this is a central idea preferences but then also constraints modern development in economics has broadened the concept of constraints to include things like information and expectations so we've learned in a major development and of course george participated in this development in revolutionizing economic theory is systematically incorporating information and expectations into current information and these also constitute constraints on the individual individuals who can't really fully know what goods they're buying what world they're facing these are important constraints and i think that the motivation by economists however in personality psychology came more on the constraint side than on the on the preference side and that makes a little different from behavioral psychology so let me give you a brief history so here i'm giving you an econ 101 lesson and i hope i'm not boring you or insulting you when i do so but this is a standard diagram that all of us teach and it's basically a diagram that says that individuals are constrained by resources they're two goods good x1 good x2 and we can only live within the area beneath that constraint and we we can't live outside our means and that is a very fundamental notion in economics now we add to that notion of constraint preferences and so preferences determine where we settle on the budget set so the argument is well higher levels of this curve i call an indifference curve a curve that says how i trade off one good for another in preferences higher levels associated with better well-being and agents would seek to try to achieve a higher level of well-being and that's the choice selected so the choice selected is one that maximizes utility that's so standard it's just part of the central body of economics and the standard exercise is suppose we make good one much more expensive and so what does that mean it means that for the same resources you have you can buy less of good x1 so that's why the budget sets shrunk towards the origin it means you now have fewer possibilities for buying good one you can spend the same amount on good two but not the same amount on good one then we say okay preferences in a normal fashion we'll look at this and we'll see yes what do we select we select this point here in the new regime and the fundamental uh proposition is as the price of good one goes up we buy more of a price of good when we buy more we substitute towards good two that's the law of demand that's a fundamental principle in economics but i think we should go back and step back from this and this helps motivate some of what i'm talking about and i think motivate some of the discussion in the conference long ago a colleague of mine gary becker showed something that really shows the fundamental nature of constraints and he showed however irrational people are they operate under constraints and he also showed that many of the traditional implications of economic theory are consequences of constraints in particular the law of demand so here let me go back to the same picture this is the our budget set what becker said is supposed that people were completely irrational they're like molecules they're like clouds of gas the one thing that a cloud of gas if just the gas in this room or the gas in this budget set has to respect is that it can't consume more than it has it cannot has to live within its means and so out of that proposition he said well you may not be at the frontier but you will have to live within your means and so that the idea of irrational behavior is that no matter what we say about preferences we're still going to have people constrained and the budget set affects a lot of our are a lot of our choice so let's go back to our experiment where the price of good one has fallen we ask what happens in that situation well if the price has fallen we know that people can't attain the red dot anymore they have a new budget set and so rational or irrational on average if they distribute themselves they're going to tend to move towards good two and that's becker's theory of irrational behavior i don't want to dwell on that theme except to say that constraints are a very natural way to approach the study of many economic problems they're more easily measured in many cases and economists have borrowed from psychology in another powerful way in a way that's not so much the way behavioral economists have talked about but in the way that uh moby labor economists and people borrowing from the iq literature have thought about and that is that iq is a principal determinant of earnings in earnings the money that you have is basically a major determinant of this constraint how much money you have so iq is actually affecting choices in the sense of affecting your budget set recently in work that actually started by some marxist economist some 30 years ago sam bowles and herb gentis in particular studied personality as a determinant of earnings as well as information and preferences and the like and i want to talk today about this notion of personality as a set of constraints like iq that affect the options that people choose and in that sense i want i like to think of this as somewhat complementary to the activities that go on in other parts of psychology and economics but it focuses somewhat differently on this constraint side the problem so i i i i like the idea and i think many economists like the idea of loading as much of our explanation into constraints as we can and lessen their preferences because it avoids the temptation which most economists resist of saying well you know if i have a new situation something has changed the preference has changed that can be a very troubling and too easy strategy here if we simply isolate the way that changes have occurred because of budgets we can sometimes have a less tautological explanation so i want to talk about this body of work because it informs the way we think about under human development but i also want to consider the origin of preferences so i don't want to just focus only on constraints personality and cognition and how families and other social institutions affect these things this i think is a very powerful tool for understanding the origins of inequality and in helping to devise policies to reduce it and it helps us to take us away from some of the foci of focus of a lot of public policy discussions let me give you an example of the power of personality traits that are frequently neglected in public policy discussion in the united states we have a program it's called the gd program the name is not so important but it's a program given to people who are secondary school dropouts secondary school dropouts who don't complete the ordinary curriculum in a school can take a test and certify that they are the equals in the test of people who ordinarily complete secondary school so this is a way for a dropout to be made whole and in the united states this program now accounts for close to fourteen percent of all secondary school degrees that are presented and it's growing and so it's a huge program it's been growing until recently what this is is a test of cognitive abilities it's not much different from the pisa scores and the scores the iq tests are a little different but the achievement tests sometimes called iowa tests or pisa scores are very similar to what the ged says let me show you that this test is actually very successful in the way that typical public policy measures things here's a study from work that i did with jonah rubinstein and jing jing c some seven or eight years ago now where what we did is we looked at the density of age-adjusted armed forces qualifying think of these as pisa scores these are achievement test scores that are used daily to decide whether schools are succeeding or failing who is succeeding who is failing so we have a distribution on the left for white males and the right for white females the male female difference isn't the issue here where these are ordinary these are the scores of ordinary secondary school graduates where do the geds fit in this distribution these are the people who couldn't finish school but nonetheless took the test where do they fit almost perfectly a little bit below but almost perfectly they have the same test scores as people who are ordinary high school secondary school graduates and that looks very good and by the standards that we currently use to measure and decide policy what this is is that these people have been made whole we've actually certified a group of people who are just as good as ordinary secondary school graduates yet actually they earn at the rate of dropouts so even though they actually have the same test scores they're actually earning at the rate of dropouts something is missing in our current thinking and i argue that they're as smart as ordinary high school graduates but they lack non-cognitive skills and that's the theme that i want to develop and in fact you can look at these geds and in fact these are the wise guys who can't finish anything i'm writing a small book on this actually it's fascinating be simply because in so many areas of life they drop out of marriage they drop out of the army they drop out of almost anything they start they start jobs but they drop out much more rapidly so there's a treat missing that the test score doesn't capture so what i want to argue is that these is just one of many examples that i could give and i'll talk a little bit about these examples but what i want to argue is that incorporating personality enriches economics that a core low-dimensional set of capabilities explains a lot of socio-economic outcomes and cognitive and non-cognitive abilities can be measured they can both be measured sometimes non-cognitive abilities are called soft skills the argument is we can't measure them well but that's simply false we have developed measurement systems and we have a lot of indicators that give us as hard a piece of evidence on their causal effects as for non-cognitive skills and so i want to show you a little bit of that evidence today and the other part that's very important is not only did they determine many outcomes so when we think about what determines success and failure in life incentives matter for sure cognitive and non-cognitive capabilities personality as well as smarts both matter but we also know that their origins are not strictly genetic it's not a question of who you're born to it's not a question of your father's genes or your mother's genes it's really a question also of social environments and that's the exciting part of this work and what we've also learned is that there are sensitive periods for the development of these skills and at many of the gaps between the rich and the poor the haves and the have's nuts started very early ages long before people enter school but it's not purely genetic and we've also found that these sensitive periods for development are earlier in life for cognitive capabilities and later for non-cognitive capabilities and that's important too for the way we think about social policy exploiting this new information so what i would argue too is that this has substantial interest for the fact that we should treat people not as identical i think that's a bias that many economists have a traditional representative agent model and what we've learned from a large different collection of bodies of work in psychology and economics is the heterogeneity that's manifest and we need to think more broadly so let me just to go very briefly over some of these abilities and show you what the evidence is so we need to understand the multiplicity and power of capabilities so the one thing we've learned and this is something that we've come from a large body of empirical research with long large longitudinal data sets with very detailed information that ability matters and abilities are multiple in nature what do i mean by cognitive abilities well psychologists will frequently distinguish two different ideas in cognition one is the notion of crystallized intelligence which is kind of knowledge how well we can command a body of recognition a set of techniques that we've learned a set of reactions to sort of common problems and then fluid intelligence the ability to solve a problem de novo and we know that there are different age profiles for their development and in fact unfortunately for most the people in the room fluid intelligence peaks out around 19 20 and then it's downhill and crystallized intelligence allegedly keeps growing so there's wisdom with age at least i hope so so cognitive abilities we know and they've been studied and there's a large body of work we've had iq tests and achievement tests around for more than 100 years less well-developed and still somewhat of a hodgepodge is this concept of non-cognitive capabilities they involve traits that are sometimes thought of like perseverance that our avoidance of risk our desire for leisure our motivation to achieve certain things our self-esteem how hard we work on problems what is our what is our the relationship that we have for the future how much do we discount the future and what way do we discount the future and self-control aspects of self-control which behavioral economists have looked at quite quite extensively and the notion of forward-looking behavior the one thing that i would urge that we recognize and this is a very important body of work in personality psychology is there's a view out there that some people have argued some economists that these traits especially the psychological the personality traits are very situational specific that they people respond specifically to the situation that they're in and what i want to argue is they aren't solely situational specific situations affect the manifestation of these incentives but they do not shape entirely what these traits are these traits evolve over time they're affected by the environment but over time there's persistence in levels of these traits especially when we standardize measurement situations and the key thing is these traits are causally affected by family and social environments and so what we've learned is that both types of capabilities have direct causal effects on a wide range of behavior so there's nothing inimical to economic theory in fact economic theory has always said that a low dimensional set of traits preference traits may be enriched by the new personality and psychology traits that have been introduced recently can explain a wide range of behavior and i just point to these wages schooling crime let me just show you some examples this is an example taken from some work in psychology i don't know if people can see this figure i know you have a handout so maybe you can look at that but this shows uh for different traits leadership job performance longevity college grad grades and years of education the importance just in a simple sense of correlation between iq which is what we mostly focus on and success so for example if you look at leadership you'll see that iq has actually got a lower correlation with success and leadership than conscientiousness and other psychological trait and the bunch of these psychological traits is much more predictive we get that iq is somewhat better in job performance but not so much different than conscientiousness if we look at something like college grades well iq matters of course but so does conscientiousness and years of education and just to show some examples from some work by economists if we ask the question well what's the probability of somebody ever having landed in jail at age 30 by measurements of these traits taking at much earlier times if we go down the distribution there's been very commonly uh noticed by many psychologists and economists that less capable people people with lower cognitive skills are much more likely to commit crime you can see that when you go from the bottom of the distribution which is the left side of the figure to the right side of the distribution you'll see the probability of being in jail goes down cognitive skills are important non-cognitive skills are also important and these are causal effects i can't get into the details of that but this is and so what we find is both cognitive and non-cognitive skills are important i want to come back to this figure look at teenage pregnancy yet another problem which is usually associated with disadvantage for the children the next generation and for the mother cognitive skill huge difference 10 chance of being single with a child and a teenage child pregnancy very high if you have low cognitive ability very low with high cognitive ability if you look at non-cognitive ability you find exactly almost the same trade-off so they're equally predictive and i could produce hundreds of these graphs in fact i did have hundreds and tito edit them out so i'll stop but i will say that something that economists worry a lot about wages again if we look at the gradient the response of what happens as we move from the bottom to the top increasing wages we see equally powerful effects of cognitive and non-cognitive skill in the sense of moving from the bottom to the top so what do we know this is a very exciting findings we have a lot of predictive evidence and this shows that these traits matter we also know that there are two parallel system of preference parameters that have been developed in psychology and economics and the research in this area is very active putting these together so right now i think it's fair to say that economic preference parameters like time preference risk aversion inequality aversion altruism reciprocity these traits we know are important they do predict there's no question about it but they can also be related to these psychological traits and so what you're seeing is not only a theoretical exercise of putting two bodies of work together but you're uniting in one place a tremendous body of findings from two literatures which will reinforce each other and we can write but we still are in the process of synthesizing these literatures now a third finding from this literature and this is very important this has a huge implication is that when we measure these traits in stable measurement systems over the age of people and we follow people over time i'll just give you one example that gaps and differences between people in these traits open up very early in the lifetime so let me give you an example this is a program that i'm studying with a developmental psychologist and and several others uh including greg duncan economist at irvine and what we're finding is the following test scores play a very powerful role in explaining who goes to college and who doesn't we know that and in the us when we looked at this we find that once you control for the test score the family income or the tuition at a college plays almost no role in explaining who goes to college who doesn't the test score is extraordinarily important but gaps in these test scores open up long before schooling begins so let me show you this this shows you the test score gap starting at three which is about as early as time as we can measure it up to age 18. and this is now classified by the measure of advantage in this case the mother's education and what you find these people were randomized into these there's a lot of detail here but the important thing is the gaps between people at age 18 which are on the right side of this graph are about the same as the gaps that are there at three and certainly by age five so what this means then is that we can't in any fundamental way account and say that schooling or things that are going on in schools are explaining the gap what we found is that schooling after the first few years plays only a very minor role in creating or reducing the gaps and we get a similar story for non-cognitive skills so we i think it's very important to reorient these th these factors are important and gaps between the haves and the have-nots are opening up very early in the lives of children long before they enter school now this kind of information that gaps open up and when you control for these factors in a statistical sense they can eliminate them leaves open a very basic question it doesn't rule out a genetic explanation you could say okay it is the manifestation of genes and some people would say that some people would say no no it has to do with family environments or family investments more generally and what i want to show you is some brief evidence from intervention studies that suggest an important role for investments in family environments in determining these capabilities so it suggests yes genes play a role what we've also learned is there's a powerful gene environment interaction and we've also learned that we can supplement families and reduce some of these gaps in some cases substantially so let me just make an observation that with the data that we've uncovered we can study how families differ and families vary a great deal the amount of investments they make in children disadvantaged families spend substantially less time with their children give the child substantially less advantage in the last 20 years as knowledge of child development has expanded more educated women while they're working more have actually been spending more time in child development precisely because they recognize the value less educated women also working more although the growth is a little less but also not spending any time more so a gap is opening up and the resources available to young children and so what i would argue is that a divide is opening up between the advantaged and the disadvantages in many countries around the world certainly in the us and that those who are born into disadvantaged environments are receiving relatively less stimulation which we know to be important and it leads as a major source of inequality now what have we learned we've learned that critical there are critical and sensitive periods more sensitive periods in critical periods but a large body of work in epidemiology and neuroscience has actually documented that there are sensitive periods where certain investments and certain activities have a huge payoff for lifetime outcomes and if those activities are not made then there's substantial disadvantage that accrues some of the most dramatic would be biological insults if a child is born with substantial vitamin a deficiency in the environment that person may have blindness permanently without any known remedy a small drop of vitamin a can eliminate that less dramatically their language and other other traits so we know that this is uh there are critical and sensitive periods but we also know and this is the positive part of all this that enriching early environments can compensate for the risks from disadvantaged environments but what have we learned and this is another key lesson and this is why it's so important and i leave you with one lesson i hope i leave you with several but i want to leave you with one and that is the power of these non-cognitive social emotional traits which are sometimes called soft traits a main channel in which many interventions operate is through enhanced non-cognitive capabilities let me just give you one example the united states has been very active in promoting early childhood development there's a famous experiment run somewhere 40 years ago the perry preschool program these were kids who all of whom were disadvantaged all of them african-americans all of them were chosen to have sub-normal iqs and they were given initially separated into two groups some of them were randomly assigned in rich parenting environments others were inside just their ordinary environments enriched parenting simply meant the parents got a little stem a little help on how to parent the children were enrolled in programs given opportunities they normally didn't get these children were randomized into two groups and they've been followed ever since they're close to 50 years of age in fact i'm participating in a study to actually look at these people at age 50 they're turning 50 as we as we speak so the people are followed 40 years and the interesting thing that came out of this program and true of many of these programs is that there was no effect on iq none the way that we normally value social policy said this program was a failure there's another program in the united states called head start and that program also was viewed as a failure it was viewed as a failure why because it didn't roost the iq yet what happened is when we look at this program i recently recomputed the rate of return and what we find is a strong statistically significant rate of return about 10 percent per annum for both boys and girls a huge rate of return even if you take a normal equity market if you look at the uh stock market returns it's about 5.8 percent uh up until uh recently anyway so what we've learned is that in fact there is a way to improve but it's through social emotional skills far more than through iq and that's something that current policy doesn't really embody so the whole mentality of the pisa scores the whole mentality of the iq and the wealth of nations i think ignores that so let me talk about these traits since many people think these traits are stable traits let me show you a few pieces of evidence on this these traits are not set in stone they can be affected by investment peri is just one of many experimental studies with long-term follow-up and these traits also are not just matter of determined solely by the situation in which you find yourself this is the graph that i was telling you about that if you look in the graph that's sloping up crystallized intelligence is increasing fluid intelligence is decreasing over your lifetime and but iq which is a quantified some composite of the two is roughly stable you can also look at how conscientiousness changes over the lifetime and not surprisingly people become somewhat more conscientious as they get older many of the other traits do evolve over time and these are longitudinal studies on these traits so what i would argue then is that what we have in in current thinking today is that later that that what we've come to understand is the cognitive and non-cognitive traits matter they're very predictive of a large variety of behaviors but we under value the role of personality not only in producing these behaviors and in our capacity to influence it in shape to motivate people and to create ambition and to create a sense of larger social engagement what i would also argue that if we think about other kinds of public policy like for example active labor market programs which aren't so prominent here in italy but are in oecd europe if we think of classroom size reductions and a lot of the traditional policies that we think about in terms of education and boosting the workforce skills what we find is that these programs are much less effective in terms of rate of return i'll give you an example many people look at reduced classroom size there's a famous paper by alan krueger and david card and alan krueger that shows the effects of reducing classroom size on raising adult earnings and that certainly a major channel is through improving schooling but what they didn't do and what was done later was actually do a cost-benefit analysis of their study you say okay i'm going to reduce classroom size i am going to raise earnings but it's going to cost money to hire teachers and it may even bid up uh the cost of teacher salaries that's a larger effect when you do that you actually find that the rate of return to reducing classroom size is either zero or negative very small and when you go through a whole range of programs that are traditional in productivity enhancing programs like adult literacy programs public job training programs tuition reduction policy rates of return on these are really quite low so a general pattern has emerged that when we start trying to solve problems of social inequality especially in the adolescent years especially targeting cognitive skills the returns on these programs are very low and they tend to be higher for both cognitive and non-cognitive individuals with higher levels of cognitive and non-cognitive traits and the sad fact is that there are lower returns for the less able people which is something people don't like to admit however i would say as a caveat motivational programs that operate on social skills have some real effects in adolescent years so a major finding from this literature is that there are substantial equity efficiency tradeoffs for the less able when we start remediating later in life so even we think about going to the schools or into the secondary schools what we do is we are actually starting way too late but we also know that there is no equity efficiency trade-off at least in the sense of saying targeting the disadvantaged the less able disadvantaged those who are coming from disadvantaged situations early in life leads to high level of economic efficiency and helps reduce social inequality i don't know if i should get into this but i will i'll try so let me try to formalize this i'm an economist so i should put a few equations down right there's a there's a that's an occupational hazard uh but i i will try to avoid too many equations but let me just think about this vector theta the theta here is a standing for a bundle of capabilities and i really mean capabilities in the sense of a march of sin and martha nussbaum i think of these as things that enhance the freedom of individuals that enhance the potential of people and you can add health although that's another trait that i'm really not talking much about today cognitive non-cognitive trait and when we think of outcomes we have to understand that the outcomes which i call ykt are produced not only by these capabilities but by the effort that people exert in the situation to manifest the trait so what we have to understand and this is very important in sort of organizing the psychological evidence interpreting the evidence is that its situations have to be standardized we make comparisons if we're going to look at traits we're going to look at the manifestation of traits we literally have to understand that we have to standardize for what the rewards are to manifest the trait so a person who is punished heavily for being very aggressive may be very meek in one situation but the situation may call in other cases for very aggressive behavior so we know that so a lot of the discussion and personality psychology recently has been precisely to standardize environments so we understand that these traits actually are more stable but we have to adjust for the environments but the main thing that i want to take out of this is that and this is very important for social policy is there are many ways to achieve performance on a given task cognitive and personality traits both determine earnings i showed you that a minute ago and you can compensate for one shortfall by having greater strength than the other and we also know from basic rule of comparative advantage of concept borrowed from trade theory but applied to labor economics that different tasks require different capabilities and that people pursue their comparative advantage by sorting into different tasks and this equation essentially suggests that equation one is a model which is allowing for the fact that we can compensate in different dimensions that we can incentivize people to perform and that these different tasks these different capabilities are all producing excellence in performance in different dimensions and so we have to adjust for these and making comparisons across these tasks but the but the mythology that's been out there that these psychological traits are only situational specific is i think a complete myth and i think we have to really understand that there are some stabilities and traits once we standardize for incentives what do we know about the capability formation process i won't give you many more equations but i'll give you one more what we've learned is that there are stages of development in the life cycle of a child and in using a term which i've developed with flavio clunya a graduate student of mine who's now at the university of pennsylvania he and i have developed what we call a technology capability formation and that's equation two and what it says is that capabilities are produced partly by your capabilities last period investments in social and parental environments that are in this in this environment and so the structure of the uh so what we've learned then is that and we've estimated these technologies that show the power of investments the power of self-productivity and the power then of these traits and a crucial notion in this literature is the complementarity and that is that higher levels of stock of theta the more your level of the capability the higher the productivity of investment at any point in time and what we've also learned is that is you increase the stock of investment so as investment goes up you increase the stock of skills but when you have a higher base to start with you learn more so what this means then is that the getting getting it right early on in life plays a huge role because you can remediate in early on in life when people are fluid when there's a lot of malleability and when you're building the base for later investment but later in life so there's a synergy that's going on and this synergy explains a lot of the behavior it explains why early nurturing environments affect the ability of animals and humans to learn so we go back to this technology theta has multiple aspects and what we've learned is motivating the kids like the perry kids making them more interested in schooling actually motivates their achievement in schools they're more likely to finish grades even though they may not be any smarter they're more motivated and it turns out they have a much higher level of achievement and many other skills and this explains why investment in disadvantaged young children so productive because they enhance the productivity of later investments it also explains the dismal evidence about later active job market programs programs that target disadvantaged people but much later in the life cycle so let me just conclude i really really will bring this to conclusion but i wanted to summarize i have a large body of empirical work and i just want to show you one or two policy simulations with this what we found and this is work that flavio and i have and some work with suzanne chanak as well is though we find is they're relatively more productive for investment and cognitive skills early in life and relatively more uh productivity for social emotional skills in later adolescent years and that's partly associated with the development of personality as we're human beings we develop much richer traits and as these personality traits emerge the opportunity for working and assessing those traits and enhancing those traits as they emerge is enormous so let me just focus on a childhood that consists basically of an adolescent period and an early childhood period that's it just two periods for simplicity and let me suppose that i'm a social planner i'm just trying to maximize the amount of schooling this is the kind of result that you get out of our studies so you look at all the conditions of disadvantage in american society and you ask what should be the optimal policy i don't know if you can see this figure here this figure on the left is a figure that gives you a measure of child disadvantage so as you go back in the figure you get more and more disadvantage so one means high level of advantage minus one means low level of advantage okay and the same is true child initial cognitive capability so these are the endowments the child is born with the light figures suggest what a social planner somebody who is only interested in maximizing the total schooling would do in terms of investment in in children and you can see in the thrust of this diagram on the left is that the investments are actually greater for the most disadvantaged even though there may be some complementarity the huge productivity that comes from remediating downstream problems is advantage is investment in the in your early years so the the black bottom lines are showing less investment the top lines more investment if you go to the second period it's reversed that's where the equity efficiency trade-off occurs you're actually finding more investments for those who started off earlier on so if you wait too late then you really do have a trade-off where you have to be an economist has to have this kind of dismal science quality that you can't argue well i want to help poor people but doing so runs against economic efficiency arguments early on in life the investments will actually both reduce inequality and produce economic productivity in this case enhancing schooling so let me let me skip past that and make one other point and that is the following what we've also learned and this is something that's very important depending on the outcome that we study cognitive and non-cognitive traits are differentially affected that was my outcome equation remember i had cognitive traits and i had non-cognitive traits so cognition turns out to be relatively more important in explaining who graduates school the non-cognitive traits non-cognitive traits are important but non cognitive traits are most important i saw that figure that i put up earlier crime is actually weighted much more heavily on non-cognitive traits and so if you look at what a distribution of optimal ratios of early to late investment would be if our target is to reduce crime it might be to actually target more towards the later years oops if our if our goal is to reduce is to promote education it might be to reduce relatively more in the early years and the reason why there's a distribution here is that an optimal policy would recognize the different conditions of advantage and disadvantage that people have the different conditions the parental background the different initial endowments and i'm sweeping all that down on these figures but that's why we have a distribution here so the bias is relatively early if we're trying to maximize education relatively late investment more late not not zero one not all late all early but relatively more in the later years if our goal is to reduce crime and that's just a consequence of the fact that non-cognitive traits flourish more in the later years and optimal policy can actually target that in the adolescent years more than the early years so the timing depends on the condition we're trying to maximize what we're trying to say so let me just summarize first of all what i've tried to do maybe tried to do too much but i nonetheless i've done it so i'll i'll conclude that cognitive and non-cognitive capabilities produce a variety of behaviors and i really want to stress that these personality factors which are considered soft and fuzzy are not and this integration of personality psychology and economics really does enrich our thinking of policy there's an emerging field of psychological measurements of personality and cognition which is linking together traditional economic preference parameters with the psychological preference parameters and i believe just as we've enriched the normal set of behavioral parameters in in standard economics like the social preference parameters we're going to have a deeper understanding of self-control our deeper understanding of aspects of time preference by drawing on the psychological literature we also know and this is something that's extremely important is the comparative advantage is an empirically important feature of economic and social life people specialize and we can trade off we all know the people who are very smart but can't wake up in time those are the geds they just can't put it together and they're also people who are tremendously well organized who aren't that bright but nonetheless succeed so we have to understand that there's a lot of comparative advantage out there in the world and there are traits and activities where people can use these traits and that's an important and neglected dimension of social policy recent work on the technology capability formation provides an operational empirical framework capabilities are not something that's just genetically determined they are causally affected by parental investments and they're not something that is just situational specific although i think at one point in the behavioral economics literature the view was that personality was an ephemera that was totally situational specific that's just not true but we need to do and when you understand that however we have to standardize the measurements that we take and what this technology capability formation shows is it rationalizes a large body of evidence in economics psychology and neuroscience that these capabilities are self-productive health actually is a major determinant of education cognition and also uh the cognition and social emotional traits affect health these synergies explain in a very important way why it's productive to invest in the cognitive skills of disadvantaged young children but why later remediations for cognitive skills may not be very effective there's no equity efficiency trade-off for a distant for investment in young disadvantaged children but there is an older disadvantaged children so i would conclude that i mean i hope i've had some influence on your thinking about this i think it's an exciting area of work where personality psychology in addition to the ordinary cognitive psychology that's influenced so much of the thinking behavioral economics as well as the work and happiness and the work on risk aversion will come together and help us with a richer economics and a social policy that's guided by fact and by by a greater understanding of the uh the of what what is possible for human beings so thank you very much it is indeed true that while jim was preparing his light i i suggested that he should cut down a bit because i was worried about the time that it was not taking into account the speed by which he was able and the clarity but i think he established a new record of a festival i counted about 15 seconds per light but always very clear very clear and very well structured so now we have also time to uh to take uh two questions um i moved to italian um think of the effects of efficiency and equity think of cognitive capabilities think of the role of families and the bad news is that disadvantaged families transmit less of these capabilities but then there is also good news for example the perry preschool program and this is an encouraging message in order to reduce the gap which is created very early in life and then a second major thing an important application of what has been said is the following much is said about a meritocracy well here we have a fundamental dimension that is non-cognitive abilities which are often ignored by tests and by measurements this is just to give you a couple of uh points well we put the slides on the website of the festival if you want you can download them because they are very very interesting there are many implications which perhaps have been lost because of the speed we have a couple of questions only because we don't have much time i'd like to know what are the criteria for selecting children where to intervene because the criterion the selection is fundamental otherwise you risk choosing or selecting a child who is not ready for that specific investment so what are the criteria for selection the criteria of the child there are a lot of child inventories that have been developed for looking at children who are at risk but it's a combination of risk factor that it concerns the child's abilities the child's cognitive and social emotional skills but also the parenting environment so the parental environment that's those are the so-called risk factors so what we've come to understand and you can there are lots of lots of pieces of information but where the family structures are very weak what you see is that less of what is normally done to middle class children is done to those children so you'll find fewer amounts of encouragement less help with homework less taking children out going to the zoo reading to the child and so the measure of inventory is a measure of disadvantage both in terms of of parenting and it's more it's more parenting than it is money per se and also in terms of the structure of of the own of the child's own abilities and sort of shortfalls so uh the the question should really be more comprehensive that these are supplements to the family it's not a targeted program only for the child it's something that supplements the family and it supplements the resources of the family and sometimes in the case of perry and many of these programs is a supplement that's teaching parenting that's actually supplementing so for example a very effective program that the president obama has endorsed and i think there's some evidence supporting it is a program where nurses go to young girls all of whom are pregnant and mostly teenage and mostly unwed and explains to them the disadvantage they come from smoking and drinking when you carry a child and this program when embraced when evaluated and people are followed 20 years later leads to very large improvements in the well-being of the child well all that is is basically telling the mother some common sense facts that most middle class women would know and would adopt and act on and so in that sense it kind of equalizes the advantage so these programs have a very rich variety but they're supplementing parenting as well as working directly with the children so the measures of disadvantage now are pretty well documented you can get inventories of families and and so forth no i think i get it can you hear me can you hear me okay can you hear me okay so in the case of a new school which we could invent so what type of school could we invent in order to have the results that you mentioned i don't think it's a school so much as kind of an approach to way childs develop the child would develop i think it represents a a way you can think of it more as a family supplement i wouldn't view it as a school remember some of these gaps are showing up long before schools begin so these are supplements to families and these are really representing maybe you can think of it as partly a school you can think of it partly as a day care center but all you're doing and you have to work with the family i mean the one thing we've learned in in the last hundred years 150 years of failed experiments is that the family is the paramount factor in the life of the child that's that's an obvious point but when we remove the parents and there have been times in australia in canada in the united states where indigenous children are removed from their parents that have had catastrophic consequences so we have to supplement the life of the child we supplement the life of the child in the family and you work with the family but i wouldn't call it a school because it's not the formal division in fact i think the mistake that's been made at least in the united states and i think also increasingly in italy is the presumption that somehow schools are the source of skills in the society and they're not as i showed you before those gaps in in in test scores which we which we feature schools aren't adding that much a very famous sociologist named james coleman who was my colleague and we taught courses together in chicago for years he had a very important finding which was most of the gaps between the haves and the have's knots in test scores came not from the school quality so much as the parental quality the family homes so what it does is it basically suggests that we refocus policy towards the family sometimes people don't like to think about that but every family is under great stress in the united states we have close to 26 percent 27 percent of all kids grow up in single parent homes about 15 percent are in homes where the children have never seen the father father's never been a factor and we know that that is associated with lower levels of wealth lower levels of resources lower levels of parental time i'm not castigating the mother i think the the concerns are real but the resources given to those child those children is lower so recognizing the way the family is evolving we need to devise policies that recognize and supplement families i wouldn't think of it as a school i would think of it as a family supplement something that works with the family and does not intrude on the family either you don't want a situation of telling people what to do you want to say here we want to supplement the life of the child we want to think of an institution that goes beyond and and includes and builds a stronger family maybe you can think of it that way supplementing the resources of single-parent families and i know that in italy i was recently in reggio and uh we were looking at the famous program there and what was very interesting to me it was that the italian uh thinking about this was very similar to the american thinking 30 or 40 years ago namely the assumption was that all children came from healthy functioning families and yet many of the immigrant families in reggio simply did not have the traditional italian family structure they did not have two parent families they did not have the same kind of access to italian and to the basic language so that they were growing up with a tremendous level of disadvantage and when they entered the famed reggio schools they were at a huge disadvantage and so i think we just have to realize that all societies are seeing increasing number of children facing this this important aspect of disadvantage so it's it's a family supplement not just a school gracious i am an expert in crime at the university in trenton we are working with professor tremblay richard tremblay whom you certainly know who applies his theories to crime prevention and for some months we have implemented an early prevention scheme here in trentino it was not easy to start that project for various reasons often when these theories are communicated they are interpreted in a sort of deterministic way and then the second point is that very often it is difficult to communicate because there is an investment in young generations and it is difficult to communicate these investments public spending is not concentrated on young people so there is a political resistance to these projects to these interventions i'd like to know from you is it then easy to apply these theories in practice well uh let me take your second point first obviously there's political reality anywhere i was just in the uh with the uh domestic policy group in uh washington last week and then the white house group and there are real even though this new administration in washington has had a very profound interest in these topics there are political budgetary realities everywhere not just in italy but i would say that if we prioritize expenditure this is something that's kind of gone out of style in economics but it's unfortunate it would have helped us a lot last spring or last winter when we were starting to build a stimulus package that will you start doing rates of return and start understanding which programs have high benefits and which ones don't i would argue that everything we know about early childhood programs at least what i know has suggested extremely high rates of return not not phenomenal not not 100 percent rate of returns but on a pure competition for budget items it will win so given a fair trade when you ask for example peter orzag who's now head of omb the office of management and budget when he was head of the cbo and was doing studies did a rate of return analysis on fixing potholes holes on the streets and i can assure you the rate of return on fixing potholes is much below the rate of return on fixing kids and fixing disadvantaged children and i would say that a lot of the rate of return on infrastructure that's received so much attention has actually got a lower rate of return i mean there's some return i'm not going to argue that one focuses all of the budget on one item of course not that would be insane but i do think that this does compete if you actually give it a fair chance to understand sort of how effective this is relative to a lot of the other interventions even the standard interventions in education programs like reducing tuition or as i said that to me a staggering finding is that the minority majority gap in college going is reversed black children are more likely to go to college than white children in the us once you control for those ability factors at 16 and 17. tuition plays a very minor role in explaining the difference it's the ability to actually benefit from the college that's playing a big role and when we understand that those ability factors can be supplemented then a powerful lever is available for social policy so i think it does challenge our thinking because we have to go outside the standard channels and and and ask that we think more comprehensively about how skills are produced in italy and in the united states and in most countries we think of different aspects of the life cycle as different cabinets different functions so we don't realize that if you want to reduce crime and you want to reduce illness that you start back with a certain set of social emotional and cognitive skills relatively early in life that essentially avoid a lot of the burden that is created if these skills aren't fostered early on so you need to think much more creatively so that's that would be the pitch but i realize that requires a lot of a lot of further education of the public now speaking about tremblay of course rashard and i are together we collaborate on projects i have a very high regard i knew that he was working in italy i did not know until now that he was working here so that's great he's a first-rate person i would say that you're perfectly correct that there is no determinism here what you're doing is you're shaping people's abilities to respond dynamically so it's not like i can suddenly enroll kids and i get a homogeneous response or that there would be one uniform reaction but what we can do is we now know ways to encourage children to enlarge their lives and in that way we can reduce a whole series of problems that are previously thought to be unrelated so although i i think richard himself can be a little too deterministic at times in his work with daniel nagin but nonetheless i do think that this is one of many forces but a powerful one that has not been addressed namely supplementing the early lives of children and seeing what the consequences would be although i didn't mention it reduced crime is a huge activity i showed you one figure but in the cost-benefit studies of early childhood programs one of the major determinants has been one of the major benefits has been the lower rate of return on the lower lower crime rate among people who participate why because they're socially engaged they have stronger measures of self-control they're more motivated to participate in the larger society and so i think that we see a broader notion of social policy if we if we understand that that we are talking about creating people we are talking about creating people and giving them the empowerment to choose much richer lifestyle so they can see possibilities for themselves they didn't see before they can break out of their identities that social class or other restrictions may have placed on them and so they can actually enlarge their capability sets so in that sense i honestly believe that it is a different approach than we've followed and that once you follow through the logic in the united states it's now becoming very much accepted five years ago it wasn't to the extent it is now but as the evidence has piled up and there's always room for overstatement there's always a lot of hyperbole but what we've observed is that increasingly groups from across the political spectrum are supporting these ideas of early investments why because they are essentially creating the capabilities that people of all backgrounds i think value and i think this is a huge change and i think it just requires a little bit of education and i'm sure that plenty of people are here to to to provide that education and the message was very clear grazer foreign