The Importance of Being Gary Becker: Economics is Everywhere

Oh the founding father of at least two very important fields within economics labor economics and law on economics and in 1992 he was awarded the Nobel Prize in Economics for having extended the domain of microeconomics to include a wide range of human behavior and interaction including non market behavior so Becker was instrumental in the development and application of microeconomics and I thought I'd start off by telling you a little about what micro is so micro economics is really the heart of economics because it develops the economic theory of how people make decisions so what's this theory of decision-making it's actually super simple it's that people do the best they can with what they have I mean more technical terms perhaps mathematical terms would be in any given situation people figure out what their objective is and then they try to maximize that subject to the constraints that they face and slightly less technical but still technical terms it's that people decide what to do in a given situation by figuring out what all their possible courses of action are in that situation then figuring out their benefits and their costs through the different possible course of action and then picking the one that yields them the highest payoff or the payoff would be the difference between benefit and cost and this micro economics is we can use it that theory to shed light on any economic social or policy issue of interest the theory yields implications for how people's decisions will change as their costs or their benefits of their possible course of action or as the set of actions that are available to them in a given situation or as the information they have at the time when they're making a choice among the set of possible actions available in a situation change because of changes in policy or because of changes in social norms and so um this way micro economics in yield predictions about the effects of a wide range of policies on people's choices and ultimately on a great outcomes that are important to society and once we have these predictions we and test them with real-world data and Becker in particular was showed how widely and fruitfully applicable this micro economics can be micro can be applied to any kind of human behavior be it crime or sex or marriage or discrimination or education these are all types of human behavior and they can be explored using microeconomics and so but today I'll show you a little bit how Becker explored some of these topics to to great effect and we'll start with crime and then I'll just like in Oscar Wilde's The Importance of Being Earnest where everybody is pretending to be earnest but really isn't I'll pretend to be Becker for a minute and show you how a similar analysis would be applied to the topic that I'm short of great interest to you sex and that'll lead us into a a Becker's analysis of fertility and marriage and the economics of the family and then again pretend to be Becker for just a minute and cover some of my related work on money in marriage and then we'll turn to Becker's analysis of discrimination taste based discrimination and we'll end with another topic that hopefully it's of great interest to you as well education okay so I have a lot of fun and let's go ahead and start the fun so what would motivate an economic analysis of crime so this is going to be based on a very seminal paper that Becker published in the journal political economy in 1968 titled crime and punishment and economic approach and well it's easy to motivate an economic analysis of crime actually there's a pretty recent paper by Anderson and the journal law economics titled the aggregate burden of crime in which he estimates the aggregate burden of crime in the United States though a great cost to crime in the United States at one point seven trillion dollars and this includes expenses for the legal system victim losses crime prevention agencies the opportunity cost of time lost by victims by criminals by there's private deterrence and cost due to a fear of being victimized you could take out from that number transfers due to property crime by steal money from you then you lost that money but I gained it so if we only care about everything on the whole that wouldn't affect things and if you do take out those kinds of property transfers then the estimate of the total cost of crime annual in the United States is still 1.1 trillion dollars so no matter how you see it's pretty high and this would motivate trying to think about strategies to reduce the crime rate right now we wouldn't want to reduce the crime rate to zero as a first observation why wouldn't we want no crime whatsoever because it would be extremely costly to society to have a zero crime rate you would have to give up all of your civil liberties you would need a police state nobody would be okay with that so we're not thinking we want crime at zero but we might want it a little lower than it is now with these tremendous costs at over trillion dollars and so what Becker argued in the paper is that it might be illuminating to think about different strategies to reduce crime rates if we think about the decision that underlies all of crime and that decision is a decision by all of us that we make every day about whether or not to engage in criminal activity so on the way home from school if you're driving a car he'll make this decision you know scores of times every time you decide not to run a red light or a stop sign you'll be choosing to go down the path of no crime alright in this model so I'm gonna present to you kind of a Bukharian model of criminal activity and from it try to derive some policy implications that might be of interest um so this is a simple decision tree model you have the individual the representative individual this can be us and the individual can go down one of two paths could go down the crime path commit crime or go down the lower path the the no crime path on top or go down the crime path on the bottom and if the individual chooses to commit a crime then nature comes in and is going to choose whether or not to bring the individual that's committed crime down the conviction pass ie the result will be that that individual after committing crime is apprehended and convicted or nature will choose that the individual who has committed crime is not convicted and obviously there are benefits to committing crime and there are cost to committing crime and so will denote the benefits from community crime is B we'll talk about elements of the course of committing crime one important one obviously is the fine which will denote by F and this can be well this is the punishment and it can be a fine and monetary fine but could also be a sentence like a prison sentence if you're caught and convicted and there's a probability P that you are convicted and so now we can apply microeconomics to figure out what this individual would do the theory of decision making is that this individual is going to compare the payoff of no crime to the payoff of crime and pick crime if the payoff the crime is higher right the super simple model of decision-making so what is the payoff to no crime here we're going to normalize it to zero okay I want to think about payoffs relative to zero if you commit crime with probability P you're going to get the benefit of committed crime if you get to enjoy it before you're caught because P is the probability that you're caught - the punishment that you'll have to incur and suffer because you are caught so with P probability you're gonna get B minus F and then plus with 1 minus P probability you will not get caught and convicted in which case you'll enjoy the benefit B of having committed the crime which could involve you know revenge or increasing your wealth from having stolen something and that will be the payoff to committing crime here we do this algebraically ok so 0 for the expected payoff of committing crime and that for committing crime would be P the probability that you're caught and apprehended times the benefit minus the fine plus 1 minus P the probability you are not caught after you committed the crime times B the benefit here we're assuming that you're enjoying that benefit of committing crime regardless of whether you're not you're caught if you in fact did commit the crime and that's an assumption they might not hold if for all types of crimes I mean obviously if you steal some money go to the Bahamas and you enjoy yourself and then you get caught on the way back to the United States then you would have enjoyed that B regardless but if it was a different situation like you were a terrorist and you're thinking about committing a terrorist act and you're stopped before you carry it out so it's a pre-emptive a pre-emptive act that stops you from committing the crime then you will only enjoy those benefits B if you are not caught and convicted so the B wouldn't go in the first parentheses but would remain in the second the analysis that Becker went through would go through no matter what perspective you take what type of crime you're thinking about so if that's the situation then the expected costs of crime are very the it's partly the expected payoff from crime is very simple it's the benefit B from committing crime minus the expected cost of committing crimes the probably you're caught and convicted times the fine and so Becker would assume applying this very simple theory of decision making that the individual would commit the crime if and only if B is greater than the product of P and F okay so this condition this necessary and sufficient condition for an individual to commit crime already suggests three types of policies that you might employ to reduce the crime rate one thing you could do looking at this is you could reduce be alright if somehow you can reduce the benefit of committing crime to individuals you would tend to reduce the crime rate and so how would you do that how would you reduce B well you could make prospects more attractive in the legal sector for example if people are able to find jobs in the legal sector then they're less likely to move to the illegal sector for their livelihoods right and so that would be a strategy is a general economic improvement of opportunities for people in the legal sector would lower the relative benefit of committing crimes so B would be lower this condition be less likely satisfied and crime rates will go down on the right-hand side if we're able to raise the right-hand side we have two options to raise it we can raise P the probability of apprehension conviction or it can raise F the fine right so those are two other types of strategies we might employ to reduce the crime rate so how do we reduce P well the way to reduce P would be to increase funding for crime prevention agencies and expanding the legal system right and that would increase the probability that individuals will commit crime are apprehended and convicted but that could be very costly right to expand the size of the legal system expand the police force and so forth on the other hand you could increase F an increasing F might actually be revenue generating to the government if the punishment F is a fine right I would bring in money even if of course if it's a if it's a prison sentence then it would be costly right keeping people in prison is costly but it could be less costly overall than expanding the size of the legal sector the legal system and increasing funding for crime prevention agencies so already there's an interesting implication that Becker drew from his model is that you know perhaps we should increase F before increasing P since increasing F could be less costly than increasing P and another observation he made that's particularly interesting is that we could reduce P and Inc increase F so that the product of P and F remains the same and that way we could maintain a given level of criminal deterrence but do it at lower costs since you have a lower P and a higher F alright that's also an interesting implication this simple model well taking this to the extreme this argument to the extreme and it was taken to the extreme following his paper in in papers that built on his Pathan his seminal work why don't we increase F to its highest possible level that way if F is really high we can have a low P and we can deter crime and do so at a pretty low cost social cost all right we don't need a lot of lawyers we don't need a lot of policemen we don't need a lot of we don't need the size of legal system to be too large and even sending this argument at absurdum what do we have the death penalty even for minor crimes like stealing candy you know it would seem to be efficient because if you had the death penalty for stealing candy nobody would commit crime of any kind right they'd be terrified to commit a crime and then you wouldn't even have to administer the death penalty right so there would be no trouble at all about a very low crime rate and nobody really put to death because everyone is afraid of committing of the crime and very low cost of enforcement and perhaps that's the efficient policy so he proposed this as a conundrum it's known as the Becker paradox and it sounds absurd right probably sounds absurd to you and it is absurd to some degree and the resolution of that Becker paradox our paradox lies in the fact that we need deterrence at the margin I mean what would happen if we had the death penalty for somebody who steals candy and what would happen if despite this death penalty for stealing Pammy somebody went ahead and stole some candy then what would be their incentive to not kill all of the witnesses in the candy store they're already getting the death penalty right so might as well eliminate all the witnesses in the candy store and all the witnesses outside the candy store and go on a rampage right you need deterrence at the margin and so this is one reason why we have an eighth amendment to the United States Constitution against excessive fines and punishments and we there's a rationale provided in this simple model for the punishment to fit the crime and so you can see even you know from an extremely simple economic model we already get interesting legal insights another reason another resolution to the becker paradox going back to the slide that gives the necessary and sufficient condition for an individual to commit crime is that unlike Becker's assumption P and F the probability of conviction and the punishment they're not really independent in reality it's not as though you can move one without moving the other what would happen if you were a juror and anyone consumed the jurors are making their decisions just like individuals making decisions by weighing the benefits of the cost and choosing the one with the highest payoff what would happen if they knew if you were jury and you knew that upon conviction the individual the defendant would be put to death would you be more or less likely to convict than if you knew that upon commission the defendant would not be put to death would have a serious penalty but not the death penalty you'd probably be a whole lot less likely to convict all right because the jury is going to rationally weigh errors that they could be making there's two kinds of errors a juror could be making type 1 error and a type 2 error a type 1 error would be like letting a guilty defendant go free a type 2 error be like locking up an innocent person and the cost of locking up an innocent person that's tremendous if that person upon conviction is going to be put to death right that cost is really really large if you send an innocent person to jail and they're going to be put to death that cost is tremendous for you in terms of as a juror thinking about which way your verdict should go on the other hand the other mistake that you could be making a letting a guilty person go free that's costly to write as that person could then harm somebody else but that that's not more costly if upon conviction the defendant would be put to death right because this is a cost that happens when you're quitting rather than convicting all right so you see that one cost is rising tremendously with a very severe punishment upon conviction and the other one is not so when there's a very severe punishment upon conviction you expect the probability of conviction to go down and and so you could even get from an increase in severity of punishments actual increases in crime rates because juries will nullify that increase in the punishment by reducing the probability of convictions is known as jury nullification they see so many interesting legal insights from the very simplest possible model and so now I'm going to go a little crazy and show you how how this model could just be applied in two different in different contexts and we'll go straight to the context of of sex so what could possibly motivate and I'm going to do the exact same what could possibly motivate an economic analysis of of sex well about 19 million new sexually transmitted infections occur each year in the United States annually costing the health care system well over a 16 billion dollars and so you might want to think about ways to reduce the spread of STDs right and it's just like with crime there's a decision that underlies that the decision is whether or not to be sexually active and so I'm gonna write down the very same model but change the variable names and see if we can get any policy implications from this just like with the crime model so you got two options you're an individual I'll have this choice you can abstain from sexual activity the top option or you can be sexually active the bottom option now obviously being sexually active is contingent on somebody else being willing to be sexually active with you but in order to model that kind of strategic interaction we need more advanced techniques in microeconomics those are the purview of game theory I hope some of you will take game theory classes later on we offer many of them in the economics department so if you choose the bottom option of being sexually active that nature again is going to move and things could go well or things could go poorly thanks for your poorly in the sense that you could get an STD we're gonna say that the probability that you get an STD is pee and things can go well in the sense that you don't get an STD and of course there are benefits and there are costs to having sex just like there are to committing crime will denote the benefits of having sex by BS and will denote the cost of getting an STD by D and D stands for death I'm just kidding you not death I finally got you to laugh serious crowd and so just the same as with the crime model we can go ahead and figure out what this individual would do if they were picking the choice optimally the expected payoff of abstaining from sex we will normalize to zero even though it's zero you know this this can be a very fun option if you become sexually active you take the bottom path then your payoff is that with probability I'm gonna get be s minus D and with 1 minus P probability I'm gonna get BS you're gonna get the the benefits from sex bs regardless if you do if you do become sexually active but you could get that that cost that terrible cost of getting an STD and so we have these expected payoffs and once again individual is going to engage in this risky sexual activity if and only if the benefits bs to engaging in it exceed the expected costs which are the product of the probably you get an STD and the cost of getting that STD D okay so we have the very same framework and let's use this Becker inspired a model to evaluate the effectiveness of a policy that we might consider to reduce the spread of sexually transmitted diseases say among the teen population that be most relevant to us one possible policy we could propose is to make condoms freely available in schools all right in the bathrooms so what effect would that have I mean looking at the model here you know it would reduce the probability P that somebody would get an STD and that would seem to be a very effective policy right you're reducing the probability of transmission of an STD and so we should go with that perhaps it could be a very well-intentioned proposal right but looking at this condition here from the Becker inspired model what's going to happen if we increase if we reduce P if we reduce P we're gonna make that right-hand side lower right which makes this inequality more or less likely to be satisfied more likely to be satisfied right and so the condom availability program is going to encourage people to become sexually active okay and so that program will then have two kinds of effects it'll have both intensive and extensive margin effects so let me explain a little bit at the intensive margin you'll have the people who have a very high BS these are people that want to be sexually active and would be regardless the risks for those people the condom availability program is just going to make sex safer right and therefore reduce their probability of getting an STD that's wonderful but you also have an extensive margin effect of this program at the extensive margin you have people who have a very low BS so that this inequality is not satisfied they would not be sexually active given the risks that are involved right especially as young people and for them the condom availability program is going to put pressure on them and it could be literally peer pressure right there condoms everywhere this is peer pressure ring gate for becoming sexually active Oh put pressure on them to become sexually active and for them the probability of getting an STD can only go up because they were up staining before right and so they probably get an STD was zero and now they're sexually active because of the condom is providing some insurance but condoms are not foolproof I use them regularly they can tear and so forth right so for them the probably get an STD can only go up and so ultimately the effect of the condom availability program on the spread of STD it's going to depend on the distribution of these benefits from engaging in sexual activity bs in the teen population and you can go ahead and you know estimate the effects of these condom availability programs they perhaps vary by school by state by overtime and you can look at variation in that and look at how it correlates with teen STD rates maybe syphilis or gonorrhea or chlamydia rates in the teen population and if you were to find that the two are positively correlated that condom availability programs are positively correlated with STD rates in the teen population then you could probably infer that at some degree the distribution of the benefits of sex in the teen population are skewed towards zero in which case the program could could backfire in terms of its primary objective of reducing STDs and so you see how even a super simple Becker inspired model of of sex already yields some interesting Public Health sighs now the benefit of sex you know there's pleasure but there there's another benefit of becoming sexually active its procreation right having kids and so I'm gonna move now to Becker's theory of fertility and population growth and this is very interesting as well and in order to understand Becker's contribution on this issue we should go back to Thomas Malthus Thomas Malthus is the founder of the field of demography as in a political economist as well and a moral philosopher and he has a very famous essay the essay on the principle population which he wrote in 1798 and in that essay he argued famously that income and fertility are positively correlated and his theory was pretty basic and a pretty simple and I can summarize it in these two bullet points so first as incomes rise above subsistence level above the minimum that you need in order to survive men and women lose their morals this is the moral philosopher in him then they have more sex they marry younger they have more kids and then population grows but given that resources are scarce now he's putting on his economists hat as population grows incomes per capita fall men and women then regain their morals they have less sex they marry older they have fewer kids and populations shrink so one prediction of course of this theory is that population should grow as as incomes rise right okay this this prediction didn't turn out to be exactly accurate as countries develop fertility rates tended to decline income and fertility are actually negative generally negatively correlated always but inversely correlated both within and between countries there's a J aspect to it where if you have a great wealth and you can have more kids but now generally speaking inversely related population growth rates are negative and in many countries in the developed world Japan Germany all negative tricking populations population growth rates in the United States would be negative if you took out immigration that is birth rates among non immigrant women are below replacement level and immigrant women have more children generally then than non immigrant women so why was mouth is wrong okay and so this is a very interesting question that Becker wanted to apply microeconomics to understand and he did so in a paper I believe again in the journal political economy published in 1960 in economic analysis of fertility and this is this is a Greek paper and was was tremendously controversial at the time so Becker resolved the Malthus paradox by treating kids as just another good in the basic theory of consumption demand from economics so it was a it was a modest proposal you know not not in the Jonathan Swift's sense of consuming babies for breakfast to avoid famine but but but in the economic sense of you know people optimally allocating their resources across Goods and kids being one of those Goods we spend a lot of money on a great proportion of our resources on right and it's true that was verse he pointed out that that summarizing Malthus is theory he point out the Malthus was essentially arguing that children are what economists call normal Goods those are goods the demand for which increases when your income increases you could probably think of many goods that you will demand more of when you get wealthier right so children are one of those goods and they are normal Goods no question about that but in the theory of consumption demand it's not just income that affects your demand for a good it's also the it's cost or price and you know demand curve is a relationship between the price of the good and the amount demanded of it and demand curves tend to be downward sloping and so when the cost or the price of a good Rises people demand less of it and Becker noted in the paper that the cost or price of having kids has increased substantially as countries have become more developed for several reasons first the most obvious one is that you know in poorer countries children are cheaper literally though those words cheaper because they start working earlier in the family business or the family farm whereas in more developed countries children rarely not not not always i mean sometimes they do rarely help with production and so until they are much older and second with increased female labor force participation which is one of the key catalysts to economic development the opportunity costs to a woman of having a child is much much greater in the u.s. one interesting statistic is that well over 30 percent of wives out earn their husbands there's a recent statistic a couple years ago and i am absolutely sure that that statistic is underreported because there is stigma attached to a man being out earned by his wife still and so it's probably underreported both by the men and by the women and you know the the trend is likely to continue at a marine and higher end in general there are more women than men believe it's 55 or 56 percent I mean at Emory and so we can expect higher salaries for women going forward and so given this increase in the opportunity cost of having kids you have to give up part of your career and so forth that that that affect the price effect could trump the income effect and resolved the Malthus paradox that price the price effect is stronger than the income effect and the price effect would lead to a reduction in the demand for children and his terminology and slower population or population declining and of course the cleaning populations they have very important implications for the viability of our social programs and for continued growth and development you know a pay-as-you-go social security system for example where the young finance the old is more difficult to sustain when there are fewer young people and more old people older people and in addition you know innovation could be affected expenditures on innovation tend to be more profitable when the total demand for your new product is greater because population is greater since it's then easier to recoup the often very very sizable initial costs of developing new products for example in the pharmaceutical industry it cost the average for four major drugs is it can be a billion dollars of R&D for a company to a research develop it and go through the various stages of FDA approval there better be a whole lot of people who will be demanding that drug once it's developed otherwise the companies won't have incentive development in the first place and so this is crucial for innovation as well now I hope you'll be ok with this transition kids are sometimes not always had with in a committed relationship or marriage ok so that's my transition to the economics of divorce and marry marriage duration and divorce I'll cover this is based the material here we based on Becker's very controversial in seminal paper that he published in 1977 an economic analysis of marital instability this is a fascinating fiat theory were paper where he formulated an economic theory of marriage and divorce and he went on to formulate more generally an economic theory of the family and so he used both microeconomic theory and empirical analysis to analyze what would be the factors that would affect marriage duration and interestingly his theory predicted that both kids and income would be negatively as associated with divorce positively associated with marriage duration and more specifically his theory was or predicted that an increase in the number of children that somebody has particularly younger children from a first marriage would reduce the probability of dissolution of that marriage and the reason is that young children are in his terminology marriage specific capital first children are consumption goods and now their marriage specific capital and such capital because it's specific to a marriage is worth less in another marriage you know concretely it might be harder for you to remarry if you already have 12 kids than if you have none and so his his theory also predicted interestingly that greater earnings more income would reduce the probability of marital dissolution and conversely that lower earnings would increase it and he had two very interesting channels that he proposed for this effect one is an uncertainty Channel and one is an assorted assorted of matching channel so let's start with the uncertainty Channel which is particularly interesting you know since there are great costs to divorcing right they're emotional and there are financial cost to divorcing and they're pretty large people presumably would prefer to stay single than enter into a marriage that is likely to dissolve within a few years right it's a sensible assumption but from that assumption that seems sensible he deduced that well then the majority of divorces would have to result from uncertainty and unfavorable outcomes that were unanticipated at the time of marriage right that's a very interesting observation so for example a big decrease in income or in earnings such as might occur if one of the spouses was fired from their job you know could cause a deviation between expected or anticipated actual earnings and put a serious strain on a marriage and perhaps cause divorce and so lower earnings could then be associated with higher divorce probability the assorted of matching channel argues that well higher male earnings could also be associated with lower divorce probability individuals so he made this following observation which makes total economic sense individuals are more likely to stay married if they're expected gains from marriage are greater right sensible assumption again and an increase in the value of variables that are positively sorted in the optimal sorting of mates like earnings for men or fertility for women those increase the expected gains of marriage remember he's writing this in 1977 so if you have a higher earning man matched with a very fertile young woman then presumably both of them have a lot to gain from marriage right so they'd be less likely to divorce whereas if you have a higher earning man matched with a higher earning woman the woman doesn't have as much to gain from the marriage right and again this in 1977 he got a lot of flack for this but but if you have a so the theory predicts that higher male earnings would be associated with a lower probability of divorce but higher female earnings would be associated with a greater probability of divorce that was another implication of his model very interesting and and analyzing data from a survey of 30,000 households conducted by the US Bureau the census ten years earlier in 1967 he found evidence consistent with his predictions strong and evidence namely the kids in income are both particularly male income though at the time male male earnings was the bulk of household earnings were positively associated with marriage duration okay so inspired by Becker I too wrote a paper on the determinants of marriage but including the relationship eating money in marriage but I was interested in I won't spend too much time on this and the money spent on the engagement ring and wedding ceremony so the papers title the diamond is forever another fairy tales the relationship between wedding expenses of marriage duration and so what motivated is those ads on TV by the diamond companies particularly the ones by De Beers you might have heard of the slogan a diamond is forever biers also had another interesting slogan isn't two months salary a small price to pay for something that's going to last forever that later went on to be three months salary it might be more at this point the those ad campaigns were hugely successful these are some of the most successful ad campaigns of all time prior to World War two fewer than 10% of engagement rings contained a diamond in them and nowadays you're more towards eighty to ninety percent you asked your grandparents for their engagement ring you know they'll probably show you this ruby or opal ring or no ring at all more likely than not so very effective and what these ads are trying to do is suggest in our minds subtly or unsettling but there's a positive correlation between how much you spend on the ring and how happy your marriage will be but nobody had ever tested that correlation perhaps an expensive now beforehand I could think of reasons why the relationship could be positive or negative I mean it could be that an expensive wing or an expensive ring pardon me or wedding ceremony for that matter would be a signal of commitment the couple is really committed to one another spending that much on it which would suggest a positive correlation between spending and marriage duration but it could also be that if two people are a perfect match for one another then they don't need to spend as much on showy things like an engagement ring or expensive wedding ceremony also spending a lot on a wedding ring or wedding ceremony could put couple especially younger couples in debt which could put stress on their marriage according to a number of industry publications the average cost of a wedding in the United States now is over $30,000 I remember looking through editions of brides magazine this is a very popular magazine that you'll find most you know dentists offices and all over the place and if you look at older editions of it go back to the 1960s they published a list the tasks that couples who are about to be married should undertake before they get married and their initial list was somewhere around there's 20 things you need to do right and you should budget at least two months before your wedding ceremony to complete these tasks and if you look at today's edition of brides magazine the list is now a 44 tasks that you need to complete before getting married and that you should budget at least one year before getting married to complete all those tasks I guess playing the 30,000 figure and the fact that the diamond industry in the United States is almost 7 billion dollars worth so I thought it'd be interesting to see which way the correlation goes and so this is a paper that I co-authored my former colleague Andrew Francis tan Andy and I surveyed over 3,000 ever married people we asked them all kinds of questions you know about their marital status their marriage duration how long they had been married or if they were divorced you know how long that lasted the number of children they have the length of time they dated prior to getting married feelings in attitudes at the time of the wedding proposal so you know things like was income a major factor in your decision to get married or was I mean money or was it looks an important thing and your decision to get married and of course you know wedding attendance engagement ring expenses total wedding expenses whether they went on a honeymoon and all these demographic characteristics their age at marriage their gender education employment household income religious attendance differences between the partner in age and education between the respondent and their partner so is there a big age difference between them and controlling for all those demographic and relationship characteristics including income we found so we were looking at you know couples within a particular income bracket we found that a marriage duration is inversely associated with spending on both engagement ring and the wedding ceremony so our findings really didn't support the validity of the wedding industry's general message that was connecting expensive weddings and rings with positive marital now our analysis doesn't prove that hi wedding expenses cause divorce only that they are positively correlated with divorce holding constant a number of demographic relationship characteristics including income but it's possible that the relationship would be causal in some ways in the paper we do present some evidence that those who spent a lot on their wedding we're more likely to report the debt resulting from wedding expenses cause stress on their marriage and there's a sizable literature in economics coming from the work of Gary Becker linking economic stress with marital instability we also found a host of other interesting correlations that were intriguing interestingly income and number of children were both very strongly positively associated with marriage duration quickly number of children and that's consistent with Becker's original analysis using data from 1967 also interestingly our relatively high wedding attendance having many people at your wedding and going on a honeymoon we're also positively associated with marriage duration so you know that might suggest that having a wedding ceremony that's a backyard barbecue with unlimited hotdogs where you invite everybody in the neighborhood and then go on a honeymoon around the world be the way to go but again these results are correlational they're not causal you know the fact that having more people at your wedding is Possley correlated with marriage duration could be an indication that of a some kind of community effect where if you have lots of friends and family then you know they can they can help you in your marriage during tough times that would be a causal channel but it could also be that the type of people who have lots of friends and know everybody in the neighborhood are just types of people who are less likely to divorce alright so okay we can't disentangle those two things all right so enough of pretending to be Gary Becker and go back to more serious topics of his all right so two other serious topics that he applied microeconomics to our discrimination and education so start with discrimination and this is actually based on his PhD thesis his PhD thesis which he published in 1957 was on the economics of discrimination and he modeled discrimination in a very intuitive way he modeled discrimination as a prejudice or taste for hiring or a taste for against associating with a particular group so for example an employer might have a taste for hiring a white candidate which would be evident right you would know that this is a prejudice of that employer if the employer hired a white candidate over a more qualified black candidate all right then it would be evident that it was evidenced that it was prejudiced now an interesting implication that he drew right away from that is that [Music] discrimination then it might be unprofitable if an employer has a taste for hiring a white candidate over a more qualified black candidate then the employers forgoing the opportunity to have a more qualified workforce all right that's one observation I have better qualified candidates work for them and so a competing firm that doesn't have this prejudice in hiring would then have a more qualified workforce and could out-compete the firm that is prejudiced compete them perhaps out of the market and so an implication of Becker's taste based discrimination model is that increased competition in markets could drive out prejudice and or conversely conversely reduced market competition would be associated with more prejudice and that's a very interesting prediction that is potentially empirically testable empirically if a market is not competitive it's highly concentrated there are very few firms in the market very few employers in the market then we might expect to observe large se gaps maybe in wages between blacks and whites or gaps in wages between men and women the male-female gap and there's this interesting paper by black and brain art in 2004 we published in the journal law economics that examines data from the current population survey and they indeed find that female male wage gaps are lower in competitive industries than in historically concentrated industries so an example of a very competitive industry like real estate a concentrated industry would be like soft drinks you know Coke versus Pepsi and also they used data trade data from the National Bureau of Economic Research and they found that historically concentrated industries that were more opened up to trade which would make them more competitive also had lower wage gaps than concentrated industries that weren't open up the tray but Becker did note in his paper and this is a very important caveat that prejudice need not always be unprofitable it could actually be profitable for you to be prejudiced as an employer if your customers are also prejudiced right for example when a company is selling to white customers who themselves have a taste for interacting with white employees or for doing business with the firm whose workforce is predominantly white but then this suggests another testable implication in that case we might expect employers to discriminate less against blacks a if they are located in a neighborhood with a greater proportion of black residents and therefore are greater proportion of black customers and so there's this fascinating paper by Bertrand mullainathan in the quarterly Journal of economics in 2004 that sets out to test this implication of Becker's PhD thesis decades prior and they do so using one of the most daring datasets you ever see assembled in economics they actually crafted resumes with fake resumes with white sounding names like Emily and Greg and black sounding names like Lakeisha and Jamal and then they controlled everything about the resume its content the level of education all this fake the word that level the work experience certificates that the individual had earned and then they sent out these fake resumes to real businesses in different neighborhoods and sadly and then they compared the callback rates I want rate where you called back depending on on your resume and its content and sadly they found a 50% overall higher callback rate for the emily's and the Gregg's and they did for the lakeisha's in Jamal's and they found a significant favoritism for the white sounding names in every treatment that they had in the paper and what's even more depressing is that they found some evidence or positively associated with the callback rate for those with white sounding names than for those with black sounding names are actually improving yourself getting a certificate and having more work experience that will benefit you a lot more if you're white than if you're black and in addition they did find some evidence that those racial gaps in callback rates were significantly smaller in neighborhoods with a greater proportion of black residents and therefore black customers and all of this you know broadly consistent with Becker's predictions from his PhD thesis in 1957 of employer and customer prejudice now Beckert tastes based model discrimination it can be contrasted with another economics Nobel Prize winners can Aero statistical discrimination model so under this model in the absence of direct information about a job can its ability the employer is going to substitute or could substitute information about average ability in a group that the candidate belongs to so my favorite example of this is you know an employer could deny employment to candidates who don't have a college degree all right it's sensible on the basis that people who don't have a college degree are on average lower ability than those who have a college degree and what this type of statistical discrimination is going to do is it's going to be unfair to a typical individuals from the disadvantaged group so if you have a very smart but uh ninja gated person this kind of statistical discrimination can be highly unfair to them right now this type of discrimination is often use even tolerated unlike the taste based discrimination that Becker was talking about also statistical discrimination can clearly be profitable to employers right in the absence of a low-cost mean of assessing abilities you know using resumes and certificates and college degrees is if is efficient also you know statistical discrimination is not likely to be competed away like like a taste based discrimination yes and so the two models have very different implications testable implications all right but I want to get to my to the last topic you know since we raised the issue of college degrees and I'll end with this let's turn to Becker's analysis of Education and draw some implications that are particularly relevant to us and so this is based on is hugely influential paper human capital theoretical and empirical analysis as a book to on human capital publishing the 1960s and uh I'll ask a question you're probably expecting your Amory college degree to pay off right yeah you hope so yeah you're hoping so right right right and it will I promise it will all the data indicate that it will on average since 1970 those with a bachelor's degree earned around 65,000 per year and those with just the high school diploma earned around 40,000 per year so those are the bachelor's degree of tended to earn you know sixty percent more per year than high school graduates and the returns to college they last a lifetime if you look at lifetime earnings as well and despite entering the labor force at a later age workers with a bachelor's degree on average earned well over 1 million dollars more than high school graduates during their working lives a plug for the economics department the median earnings for economics majors if you're interested in economics is 3.5 3.3 million but but where does this return come from it's very important to think about where this return comes from well we have two basic theories of where this return comes from one of them the human capital theory of education pioneered by by Gary Becker and the other one is the signaling theory of education developed by another Nobel prize-winning economist Michael Spence so the human capital theory is very very easy to explain I mean concretely the human capital theory is that we learn stuff we get smarter we develop new abilities during the four years that were here in college and employers are willing to pay a premium for our higher abilities or improved human capital nice and simple the signalling theory is a little more complicated to understand and if you really want to get through it it's good to do it within the framework of game theory but the base the basic idea is relatable the signaling theory is that getting a college degree is hard you guys know how hard it is I mean you have so many exams to pass so many papers to write only people are already smart will be able to complete the degree with all of those hurdles there's a lot of hurdles in which case having a degree will be a signal to the job market that you were smart to begin with and be a quick and easy statistical way for employers to infer that you're smart right notice that under the sea there's still a return to your education here at Emory even if you learn absolutely nothing for the four years that you are here or forget everything that you learned here the second you graduate I mean it's simply that a college degree is harder to obtain for somebody that is not smart than for somebody who is smart so obtaining the degree is a signal that you're smart even if nobody is getting smarter for the four years that they are here so the question is I'll end with this do colleges help people become smart or do colleges exist to help people who are already smart prove that they are smart do the observe returns from education come from an Inc in human capital Becker suggested or purely from signaling through that piece of paper certificate Spence suggests and empirically the two theories can be distinguished and is there's actually a very interesting paper by Tyler at all in the American Economic Review published in 2000 they have a very smart way to distinguish the two theories they focus on the GED with is a general education development certificate and they exploit the fact that the GED test is the same but the GED passing standards are different in different states so for example you would pass the GED test with a score between 40 and 44 if you were in Texas but with that very same score between 40 and 44 you would not pass in New York so you could compare people in Texas and New York with the same score and see passed in Texas are doing a whole lot better than those who didn't pass in New York but with the same score alright so clever way assuming they're people with the same score on the test or of the same ability which is elite but nonetheless that's an assumption that's somewhat reasonable this would be evidence of a pure signaling return to education right no difference in ability only a difference in having the piece of paper the certificate and what Tyler and company find is very large signaling returns estimated at 20 percent earnings gained after the first five years for those who got the certificate versus those who did not so there's a substantial fraction of the returns of Education there due to signaling not human capital development now what are the implications of this if education is mostly about signaling then this has very serious implications for its social value right in particular if it's about signaling then education is a social waste resources are being spent on it all these resources and nobody's abilities are improving nobody's getting smarter from it so has the waste manifested its manifested and the fact that y'all could be doing something more socially productive perhaps you could go be fighting fires and I could be doing something more socially productive I could be delivering mail right rather than going through this enterprise where nobody's getting smarter and we're just trying to separate the smart from they're not so smart and it's even worse it gets even worse than that if colleges exist simply to help people who are already smart prove than it that they are smart and colleges are actually increasing social and economic inequality and not in a benign way under the signaling theory education sadly is all about leaving children behind passing the tests writing the papers staying in school long and doing so at a top school like Emory that's gonna prove that you're smart no doubt but it's also going to reveal that those who couldn't do all of that we're not as smart as you all right and so by doing all that you'll be increasing your earnings later on through the signaling effect but at the expense of their earnings that will go down so it's a cynical theory I'm in the Becker camp I believe we can learn a lot in school that we have opportunities here to greatly improve our human capital but the contrasting Becker human capital and spent signaling models offer a valuable bit of wisdom and I'll end with this where we're all you know smart in this room you were admitted to Emory at the top school my colleagues and I were hired here so this means we're pretty smart right but we have a choice to make with the incredible opportunity that we've been given and that we've earned we've earned this opportunity and we have a choice my colleagues in I we can choose to not just teach you the bare minimum that it takes to showcase short-term proficiency in our fields but try to impart you a deeper knowledge that will last a lifetime and to keep to invest in you beyond your grades all right to keep connections with you beyond our classes we can make that choice and you you all can choose to not just do the minimum to pass our tests not just study for your grades I really try to learn something for life and to invest in yourselves and take advantage of all the amazing opportunities here at Emory to connect with others and to the broader community and if we do that then our endeavors here will not have been a social waste we will have improved our human capital as Becker suggested and with this improved human capital we can change the world in in many ways that can benefit people far and wide [Applause]

Channel: Emory University

Views: 6,523

Keywords: Gary Becker, economics, crime, sex, population, discrimination, education, cost, benefit

Id: I47Gisb_iKI

Length: 61min 13sec (3673 seconds)

Published: Tue Nov 13 2018