 In economics, most of the explanations for mobility has focused on family factors. I'm going to look at explanations that focus on the influences that parents directly have on their children. Families make investments in their children. Now, investment doesn't mean that they're doing it with an explicit cost-benefit analysis, the way that you would invest in a stock or a government bond, but rather they're allocating resources to their children based on the desire to augment their future earnings capability. The key idea, therefore, is that the parent's income is going to have some influence on the level of education of a child. We can think about that in two ways. One of them is that I directly purchase education in the sense that I pay for college, or I could think about things such as family tutoring that I provide, or the number of books in a house. In other words, if I'm thinking about both levels of education and the overall human capital, the cognitive skills that a child has, all of those are being influenced by these investments. So parental income determines human capital and education. Education and human capital determine future income. And that becomes not just a correlation, but an explanation. In other words, a description of mechanisms. So what do we know about human capital and income and the likes? Well, here are the basic facts. If you look at the relationship between the returns to education and the persistence of intergenerational status, they tend to move together. In other words, in periods of time, when the return to education is particularly high, you also have more persistence in income across generations. So what that tells us is that it's meaningful to think that there's something about the intergenerational elasticity of income and the ability to invest in your children. This doesn't prove a causal relationship per se, but it tells us that there's something important to think about. There's actually a lot of evidence about investments in children across the life course, and there's a famous curve due to the Nobel laureate at Chicago James Heckman, which is called the Heckman Curve, which basically calculated the rates of return on investments at different stages of childhood and adolescence. If you looked at the rates of return and what's the rate of return on an investment, it's asking how does it affect the future earnings trajectory of children, as well as whatever other factors you want to put in in terms of social benefits. What he concluded was the rates of return are especially high for preschool and a comparatively low for job training. That doesn't mean job training is not helpful. It doesn't mean that schooling is not so important. Rather, it's said that given the observed levels that we have in society, the marginal value of a dollar seems to be especially high in the United States with respect to human capital for early childhood investments. And so if you put these types of facts together, in other words, the role of the college wage premium being positively associated with the levels of intergenerational immobility or persistence, as well as the Heckman Curve, you can see why this is a natural explanation for intergenerational mobility. Income constraints that parents face in the creation of human capital for their children has dynamic consequences. Now, I think it's fair to say that the modern literature on thinking about family investments has taken a somewhat broader notion than the traditional one of cognitive skills and isolation. And again, this draws on work by Heckman Flavio Cunha and others. And this modern work has opened up a broader vision of parental influences. In other words, that the inputs that parents provide for kids are not just money, but rather the characteristics of the parents in terms of what they know about navigating public school systems, so on and so forth. And so the upshot is to say that there's a broader set of characteristics of the parents that are affecting a broader set of characteristics of the kids, which are going to summarize with the terms cognitive skills plus personality traits. Now, one way to see the importance of personality traits is to recognize that in a modern economy, if you look at the skill requirements as they're evolving, they are increasingly emphasizing social and service abilities, the ability to engage in routine tasks. How important is that for current occupations versus mathematical skills versus social skills versus the ability to provide services. And of course, the upshot of the say that services and skills are increasingly important in what is known as the service economy. And consequently, there's an implication there. And that is the personality traits are going to have a role that's distinct from the ability to multiply numbers in one's head. And so I put this on the table simply to emphasize that sort of same personality matters, that's not a frivolous idea, it's a fundamental idea. And that is what people bring to the process of work is a whole range of mental abilities. And some of them we traditionally call personality as opposed to education or cognitive ability as traditionally understood. One can identify similar relationships between earnings and both cognitive skills and non-cognitive skills. If anything, it would appear that the non-cognitive skills are pretty tightly correlated. First of all, both cognitive skills and non-cognitive skills are associated with higher incomes, higher earnings, and second, there's actually a pretty high correlation between cognitive skills and non-cognitive skills. The latter is important because it says, and this is again the skills idea, that levels of skills and personality traits at one period of childhood or adolescence, they feed back into the changes in both across time. An intuitive example of that is conscientiousness. In school, that's a personality trait. It has the implication, however, that the student may learn more, a child may learn more, and that becomes a cognitive skill. And so in fact, in modern societies, both cognitive and non-cognitive skills are co-evolving. So what I went on the table in other words was to say, one, parental income has direct inputs. Number two was to say that there's probably a more broad conception of how families are influencing skills. Number three, I wanted to say something about family wealth. You know the statistics probably about family wealth as I do, and that is that it is extremely skewed in the United States and the skewness has dramatically increased in the last 50 years. And so wealth is going to be an additional contributory factor to intergenerational income persistence for the obvious reason that not all income is generated by labor market earnings. If you're actually looking at the incomes of individuals, you know the dominant form of course for most people is going to be labor market earnings, but for some it's also going to be that. The final thing I want to do because I think it's incumbent on me is to say something about genetics. I don't think there's any serious disagreement that in an individual level there's something about the genotype of an individual that matters for socioeconomic outcomes. If you engage in more fine grain types of activities in society, it becomes, you know, it would be silly to argue otherwise. Whether it's Beethoven or Duke Ellington, there clearly was a genomic component to their musical talent. And so to deny the existence of genetic influences on outcomes it's an unserious argument. At the same time, I'm going to also say that any claim that group differences are explained by genetic differences with respect to income or wealth I think is also an unserious argument. It's both intellectually unserious and I'll be blunt, it's morally unserious. If you go through every argument that's been made of that type, they're easily rebutted. It's morally unserious because it is to be blunt, shocking to me that somebody would ask the question if I have two groups of people and one of them has been viciously mistreated for 300 years that after 60 years of a diminution of the mistreatment I would attribute residual differences in their income or wealth to a genomic explanation as opposed to the persistence of the horrible treatment as well as contemporary mistreatments. So I want to be clear on this that I think that if somebody says genes don't matter of renaquality they're not being intellectually honest. It's not a sensible argument for the reasons I've alluded to. At the same time, any claims about groups in my judgment are not starter. The third thing I want to say is that even if we agree at an intuitive level that there's something about the genotype of a person that matters for socioeconomic outcomes that doesn't answer the question as to how do you measure the significance of it. After all, to say that there are different factors doesn't answer the question how are they contributing in magnitudes to the overall degrees of persistence we see across generations. The reason that the measurement of these effects is problematic is what is called an identification problem in econometrics. So let me just give you a quick example to give you a sense of why this is very hard to do. It's often argued that monozygotic twins are more similar than dizygotic twins and so that's suggestive that it must be genes. The answer to that is that it's not necessarily so because after all you have to it may be the case families raise monozygotic identical twins more similarly than other twins. And so is it environment? Is it genes? It can't be disentangled from that type of argument. And so I want to emphasize I think that that's an area of fruitful research but at this point there's not much that can be said on that. If somebody wants to make an argument as to the role of genes it's going to require pretty heroic assumptions. Now that said there have been efforts to decompose the different factors associated with the intergenerational ability into these mechanisms that I've referred to. And so here I want to tout a paper by Samuel Bowles and Herbert Gintis really two of the great social scientists of the post-World War II period. They wanted to basically ask the following question could they decompose the relationship between parental and offspring earnings and parental offspring income to the following factors and then they asked how much can they explain? The way that they tried to measure the genetic input implicitly was to say assume that you attribute all correlation in IQ to genes which they don't believe by the way that's just giving you an upper bound but if you go through this exercise what you found was a following of the .32 coefficient for example in income an eighth would be explained by their upper bound on IQ in contrast nearly twice as much as simply the racial marker and so the upshot is that I think that these are actually pretty reasonable numbers so what I take from these decompositions is number one that parental income via the education channel is an important factor much of it has to do with the schooling that's produced number two wealth can also matter wealth itself if you transmit that to kids that's going to pay off number three even under the most heroic assumptions in favor of a genetic component it doesn't seem to appear in the data nearly as salient as other factors and finally race is a determinant of intergenerational status which has to be accounted for if you understand the levels of persistence in the United States in 2023