 The appropriate and the most rich way to think about mobility is to think of the interactions between family-specific influences and social influences. What I want to put on the table is something I have named the Memberships Theory of Inequality. Now the Memberships Theory has three distinct components. First is the idea that when we think about an individual's beliefs, their preferences, their opportunities, they're all conditioned by group memberships. And what I mean by that is we think of individuals making choices, they're constrained by the opportunities that they face. When individuals make choices, they have preferences over different outcomes that can be associated with the choices. And similarly, people have to have beliefs about the effects of their choices, about the world in which these choices are going to manifest themselves. And so the proposition is really very simple. If you describe my preferences, describe my beliefs, you describe my opportunities, knowing the groups I'm a membership of matter. Now some of those memberships can be ethnic, some of them can be gender, some of them can be religious. And those in some sense we think of as predetermined with reference to socioeconomic outcomes. In contrast, other memberships are determined by the society and that would be things like schools and neighborhoods. But the key feature is this mapping, this relationship between the location and the social milieu and the way that an individual makes decisions. Now within economic theory, and I should say social science theory, it is typically assumed that there's complementarities. If my friends are going to do something, for example, I'm more likely to do it because I find it relatively more attractive. So the complementarity idea is important because it explains something about correlatedness in behaviors within similar types in a population. But the key idea for proposition one is that the way that individuals make decisions is socially conditioned. And so in that sense it moves beyond methodological individualism, a hallmark of many economic models, to try and in a good faith way to introduce sociological ideas. The second proposition is that the group memberships in some cases evolve in response to the fact that these interdependences matter for individuals and how they think about the world and how they make choices. Now obviously when saying that I'm focusing on certain types of memberships. And so the obvious ones would be things such as neighborhoods, schools, friendship networks, and of course the organizations in which people work at adults. The key implication of those models is that incentives exist for people to make choices that can lead to types of socioeconomic segregation. One type of socioeconomic segregation is the income segregation of neighborhoods. A fundamentally distinct one can be ethnic segregation. This stratification affirms by the skill levels of their employees. But in each case the key idea is that the social influences that was proposition one have consequences for thinking about the degree of segregation in a society. When you put propositions one and two together that leads to proposition three. The interdependences of individuals combined with the forces that lead towards segregation actually provide a theory of persistent inequality and intergenerational immobility. To put it very simply, if there are incentives for rich families to isolate themselves from less rich families and collectively more affluent individuals have incentives to isolate themselves from the poor. What is happening over generations is that the segregation between rich and poor creates very different environments in which the children of the rich and poor develop and hence relative affluence, relative deprivation are translated across generations. And so when you put the three propositions together you end up with a vision of intergenerational mobility and persistent inequality in which the fundamental determinant is segregation. And so you can have segregation and phenomena leading from marriage patterns in which the children of the rich marry each other so to speak. In which highly educated people tend to marry each other to neighborhoods in which you get socioeconomic and racial segregation, to occupations, to even within firms the stratification of firms by skill levels and the like. So what sort of evidence exists that this perspective speaks to persistent inequality, speaks to immobility across generations? Well there I think I would want to divide the evidence into different categories, one of them would be racial segregation. Now for racial segregation the evidence is very long standing that there are adverse consequences for African Americans so in some sense that's not a surprising thing to even raise. David Card and Jesse Rothstein showed for example strong negative correlations between the degree of segregation in school districts and the black-white test score gap. And so the issue in thinking about racial segregation is that it is generating persistent disparities in education and for reasons I've talked about in the previous lectures that has long term consequences for future education as well as for labor market outcomes. In thinking about racial segregation it's not just a matter of asking whether school quality is different or asking whether or not the tax bases are different. At a fundamental level one can find very deep differences in the natures of the neighborhoods that blacks and whites live in. And so Robert Manduka and Robert Samson have most effectively demonstrated that looking across Chicago neighborhoods. And these are histograms that basically tell you the empirical densities of the levels of the harshness for majority black and majority white neighborhoods. Levels of crime and levels of incarceration of the community. And what you see here is they scarcely overlap. And so when one hears a terrible story such as what happened in Flint, Michigan that's the extreme form of a more general relationship between racial segregation and exposure to environmental hazards. It's merely another piece or another side of this multi-dimensional set of adverse influences as segregation generates. Second thing I want to say is that in thinking about segregation part of what generates it is economic inequality because it changes the ability of less affluent families to live in more affluent neighborhoods, ones that have better schools and the like. There's a strong negative correlation between the levels of inequality parents experience and the likelihood of upward mobility of their children. So in other words the more inequality parents experience greater the adverse effect on their children. Another example of this calculation and this is due to Carney and Levine was that you take different U.S. states and you look at measures of inequality within the states and ask what's the association of the inequality in those states with high school graduation. And what you find is perhaps unsurprisingly the states with greater inequality are associated with lower rates of high school completion. Now for every one of these bivariate relationships I've shown you the bivariate empirical relationship is not demonstrate causality of course. But they're the sorts of stylized facts for which we have theories to explain them that suggests that the correlation in fact has a substantive interpretation. Another statistic that Carney and Levine calculated is very interesting that asks the question what's the rate of return of additional year of education conditional on a mother's level of education and conditional on the level of inequality that's one experiences in the state. And what they found is once again conditioning on the mother's education the value of that additional year is sensitive to the degree of inequality that exists in the state. And notice that's a fundamentally social mechanisms a statement about the entire income distribution providing explanatory power for individual level inequality. Let me finish this discussion about segregation is just to emphasize the fact that you have extreme segregation at very micro levels. In other words this is the city of Chicago and just as I showed that the majority black neighborhoods had substantially greater levels of social ills than the majority white neighborhoods you can see the concentrations of African Americans versus whites in the population. And so these levels of segregation are powerful and extreme and they are empirically important. What I've done in thinking about these social influences I wanted to identify got a facts about racial segregation and its association with adverse consequences facts about income inequality and adverse consequences. And now I want to talk a bit more about the the micro mechanisms. So one example would be school finance. This is California Texas in New York in which you can ask the question how much heterogeneity is there in per pupil expenditures. New York has extremely little. California and Texas have very high levels. And the point of a picture like this is to demonstrate the basic fact that local public finance of education has meaningful consequences for the resources available per student. That's a very basic idea in thinking about the consequences of socially economic segregation. And that is that economic segregation begets differences in school expenditures. Now what about interactions between these? One perspective in thinking about adult outcomes is to understand the family background the family incomes. That's what I talked about in the previous lecture in giving you this very high level overview of the membership theory of inequality. That's a suggestion that neighborhoods matter. And so measures of neighborhood quality in terms of unemployment in terms of income etc should have distinct influences there. What Watke Harding and Elworth did is they said that we want to think about both of them affecting individual outcomes. The example in this picture is understanding the probabilities of high school graduation. What their paper does is it establishes that for very poor families neighborhood quality and neighborhood disadvantage have particularly adverse consequences on their children. Furthermore in their paper they found evidence of some change in the strength of family influences versus neighborhood influences as children move from childhood to adolescence. That won't surprise any parent. Substantly and seriously for this matter it gives this a vision in which there's a complementarity between family income and neighborhood quality. So that when they interact in particular ways if you have both disadvantage of the family level and disadvantage of the neighborhood level that's a bottleneck. Those are the cases where much harder much less likely to graduate from high school so on and so forth. And furthermore what the paper establishes is we have to think dynamically about how different influences intergenerationally within the life course of childhood and adolescence affect children and teenagers. My own view is that this speaks to the idea that if you think about what family income does for children we have direct purchases of things direct investments as it were in early childhood. But income becomes the mechanism that identify school and neighborhood quality in adolescence and that's consistent with that intergeneration mobility curve that I showed you in an earlier lecture. So that's really what I wanted to say about the memberships theory just to repeat the key ideas. Our outcomes are socially determined because of their effects on our preferences the way we think about the world the opportunities we face. The fact that those influences exist actually creates incentives for socioeconomic segregation. And once you interact the importance of social influences with the incentives for socioeconomic segregation those provide a theory of intergenerational and intergenerational persistence.