 Inequality is an incredibly complicated phenomena. All of these different dimensions, location, family, age, what I'm going to focus on is one set of dimensions of inequality, which are intrinsically temporal. Think about a society in which half of the population is rich and half of the population is poor. The children of the poor are always poor and the children of the rich are always rich. Contrast that with a different society. But this one has a very unusual social structure. The children of the rich are always poor. The children of the poor are always rich. Notice in both of them, family determines destiny, but it does it in very different ways. But it's not clear that that's the best way to think about mobility. Suppose it's the case that it doesn't matter what your parents are like, it's 50-50 whether you'll be rich or poor. You might say that the society where there's no predictability of the children given the parents, that's an extremely mobile society, but I think most people would prefer society one. In other words, in thinking about mobility, there is a deep issue in asymmetries. Our ethical considerations have to do with the desire to make sure that people who are less advantaged are not unduly inhibiting the possibilities for their children, and it's a very distinct question to ask whether or not more successful families can lock in success. Measuring mobility is not an obvious thing. There's many different ways we may want to characterize mobility, and part of the complication is deciding which mobility statistics are the ones that are ethically or policy salient. Being concerned about the mobility in society, it often has to do with the concept of equality of opportunity, an argument that we want a society to be one in which there are large opportunities for individuals outside the accident, as it said, of birth. Sometimes it's, in fact, argued that there's a conflict between equality of opportunity and equality of outcomes, whether or not a society which has very broad opportunities for individuals is also one that's going to generate substantial degrees of inequality cross-sectionally. This is sometimes even referred to as an explanation of American exceptionalism. It's kind of an old chestnut problem in social science as why does the United States have a less generous social welfare state than Europe, and one of the classic arguments was that Americans prefer opportunity, mobility, to that of equality of outcomes. There are not good reasons to believe that there is a trade-off between mobility and inequality. In other words, in my discussions of mobility, I'm not going to be implicitly saying that one is not interested in inequality cross-sectionally, or am I going to say that mobile societies are logically ones that have higher degrees of inequality. Now, why do I emphasize this? It's because the idea that there are equity efficiency trade-offs, these types of phrases that often appear in public policy discussions, I don't think have strong justifications. They aren't true at a theoretical level. If you asked me as an economist, can I write down a mathematical model that generates a situation where the more mobile the society is, the more unequal it is? The answer is yes, and there could be explanations for that, but I could just as easily come up with different explanations that had them move together. But perhaps most important, I want to emphasize there's no evidence, at least at a very aggregated level, that you have to trade them off. Economist Miles Korak took a measure across different economies of the degree of mobility, or I should say in this case the degree of persistence between parents and children, and he took another measure of cross-sectional inequality and he simply asked, what do they look like if you correlate them? And what he found was that societies that have high levels of inequality also have low levels of mobility. In contrast, the societies that were very concentrated in terms of having relatively narrow income distributions, they also were more mobile. And so the reason the Greg Gatsby curve is of interest is it says something about the relationship between the subject for these lectures which is mobility and the other phenomena we care about which is cross-sectional inequality. I think the key thing that we want to focus on is the extent to which one sees persistence within the life course. In other words, whether certain successes, certain achievements at one age make other successes other achievements more likely later on. And so one example of that would be relative status. A question one could ask is that suppose that an individual is in the lower part of the income distribution at one age, how likely is it that they're going to move out? There's one reason why you might think it's very likely, and that is simply because the wage trajectory for individuals is increasing in age and so you would think that individuals, as they get older, they make more and they shift in the income distribution. If you say to think that relative status has to be persistent, nevertheless, the data actually is very suggestive of substantial persistence. If you look at individuals in the bottom 20% of the income distribution, how likely were they to be in that same location four years later? And what you see, of course, is that it was almost two out of three that they would be in the bottom quintile and conversely, extremely unlikely to be in the top 20%. Interestingly, these data do not suggest that there was much of a temporal change between 76 to 80 in 2010 to 2014, and that's another thing one wants to think about in measuring mobility. One question is to look at a society over the last 50 years, 60 years, and others to ask in shorter time units has there been evolution of the degrees of mobility? And that's a topic I will say a bit more about later. A second example of intergenerational persistence, I'm going to call education, and thinking about education as a stepping stone. One question to ask is, if I have parents that are relatively unafluent, what does that do to me in terms of the next stage of life? Does that lock me into different types of colleges? And so there the evidence is very strong, and that is there's a very strong substantial relationship between the type of college the individual attends, whether they attend college, and their family background. This is from Raj Chetty's and Nathan Hendred's team at Harvard, and so what these interesting calculations demonstrate is the extraordinary extent to which elite universities are dominated, at least in enrollments, by children from the top 1%. That tells you something about the first stage, parental income would be get college quality. The second observation to make is that if you look at the graduation rates for students, the more selective the college, the higher the graduation rate. So that's what I mean by lock-in. First thing that happens is the parental income affects where you attend school. Second thing that happens is where you attend school affects the probability of graduation. And of course I could go on. The next thing that's going to happen after college graduation is persistent effects in the labor market. And so I just want to put on the table this idea that you have substantial lock-in across trajectories for individuals. The second example that I think is worth thinking about is early childhood investment. There is substantial and I think empirically compelling evidence that early childhood investments, when the programs are of high quality and they are intensive, have very persistent consequences for children. And so James Heckman at the University of Chicago has been the pioneer in calculating rates of return on investments in different ages and has led to the generation of a literature on these types of questions. And there I would simply put on the table you find persistence within the life course because of the fact that the influences of these investments, of these intense programs in ages three to six can be seen decades later. Again, this idea of intergenerational persistence. Intergenerational and intergenerational persistence and by implication mobility, they're deeply interconnected. The influences that child fields or in adolescent fields, those translate into adulthood in fundamentally important ways and parental socioeconomic status influences childhood and adolescence. And those trajectories are influenced by the status of the parents. And so you have these deep connections between the process of intergenerational mobility and what's happening across generations. A baseline question is to ask in a society how predictive are features of an individual's childhood and adolescence for their outcomes as an adult? The most common statistical measures of mobility focus on that exact question. Economists typically ask the question how predictive is parental income for the adult incomes that their children will experience? Sociology has historically been much more focused on occupations, asking what's the relationship between the occupation of an adult and the occupations of their parents? But each of them is asking the same persistence question. The second thing I would want to put on the table and this is something which we have theories of this phenomena I'm calling bottlenecks but I don't think the empirical evidence has been well developed. And so if you think about a bottleneck metaphorically you have individuals flowing across time and is there some configuration where it's just hard to get through it? And so I put this on the table because in thinking about the statistics of mobility and how do you characterize a mobile society one thing is to look for broad empirical regularities. A different question is to ask are there particular family configurations that make it especially difficult for children to advance? And so I've alluded to something that sounds like a mathematical issue and actually it is and that is much of the work on mobility is things in a linear fashion. The income you give to a parent has the same effect on a child whether or not the parent is already a fluent or not. The nonlinear view is different. It says that there's something special about the certain configurations that can be very sticky in terms of the families but for other families you can have substantial mobility. And in thinking about those I said that's partly how we want to understand the facts of mobility but it also tells you why it is difficult to characterize mobility and do comparisons. If you think about the Great Gatsbyker one of the features it illustrated was that Scandinavia is substantially more mobile intergenerationally than the United States. In doing that comparison though that's looking at the entire population. It doesn't ask the question what society is more mobile for immigrants. If you think about very heterogeneous complicated countries such as the US there's many mobility questions that are built into it and the Scandinavian countries which of course have their own degrees of heterogeneity are nevertheless qualitatively different in that regard. And so I simply would urge everyone to think of asking about mobility for who.