“Equitable Growth in Conversation” is a recurring series where we talk with economists and other academics to help us better understand whether and how economic inequality affects economic growth and stability.
In this installment, Equitable Growth Senior Policy Advisor Liz Hipple talks with Fabian Pfeffer, an associate professor of sociology at the University of Michigan in Ann Arbor. His research focuses on social inequality and its maintenance across time and generations. His current work investigates wealth inequality and its consequences for the next generation, the transmission of inequality across multiple generations, and the maintenance of inequality through education.
In an in-depth conversation about his research and its implications for public policymaking, Hipple, who leads Equitable Growth’s work to explore the relationship and connections between inequality and economic mobility, and Pfeffer discuss:
- Why wealth is different from income and other measures of inequality
- The importance of multigenerational transmission of wealth
- The racial disparities in the transmission of wealth
- The intergenerational transmission of wealth earlier in life
- The psychological processes underlying the intergenerational transmission of wealth
- Policy recommendations for more equitable intergenerational transmissions of wealth
- Ways forward in wealth research
Liz Hipple: Thank you, Fabian, for taking the time to talk with me today. I’m really looking forward to it. In particular, I was interested in talking with you as part of our work at Equitable Growth on intergenerational mobility because I know a lot of your research focuses on wealth, as opposed to income, which is usually the metric used by a lot of economic research on intergenerational mobility. So, I was just curious, why do you study wealth? What does it tell us that income can’t?
Fabian Pfeffer: That’s a great question. One of the main reasons why a lot of research on socioeconomic mobility has been focused on either income or some occupation-derived measure is that the data just hadn’t been available for the study of wealth. Fortunately, several studies, including the Panel Set of Income Dynamics [the longitudinal panel survey of U.S. families, conducted by the Survey Research Center at the University of Michigan] started to collect data on wealth beginning in the 1980s. And today, we’re beginning to be ready to analyze those longer-term effects of wealth across generations. I think it is important to study intergenerational mobility and wealth because wealth is something empirically and conceptually different from income.
When we talk about wealth, we normally talk about all the financial and real assets that individuals hold, such as savings accounts, stocks, a house, a vehicle, minus all the financial obligations, such as student debt, credit card debt, medical debt, and so on. All of that together we call net worth. Now, it has been acknowledged for a while now that income and wealth do not correlate as highly as some may expect. They’re empirically not exactly the same.
But what is more important, I think, for the kind of research I’ve been doing is that the effects of wealth or the associations of wealth with the outcomes of the next generation also are partly distinct. So, for instance, when we study socioeconomic inequalities in educational attainment, if we only look at parents’ income, or only at parents’ education, or only at parents’ occupation, we’re not getting the full story. Wealth contributes independently from those characteristics to the advantages and disadvantages of the next generation. That’s what my own research and that of other contributors has shown over the past couple of years. To me, this research strengthens the case that wealth is something conceptually different. I have a few explanatory approaches of why wealth may be different. So, I’m happy to talk about those two.
Hipple: Please do. That would be great.
Why wealth is different from income and other measures of inequality
Pfeffer: So often, in my work, I distinguish between two broad mechanisms through which parental wealth may translate into advantages in the next generation—say, educational advantages or advantages in the labor market. One is what I would call the purchasing function, which arguably is the most intuitive, at least in the American context: We know that it often takes money to invest in human capital, for example, to cover the high cost of college tuition or get access to high-quality primary and secondary education by buying into certain neighborhoods. I think it’s very obvious how wealth can often directly purchase access to these valuable goods.
In addition, however, I believe that family wealth also has what I call an insurance function. Wealth functions as a safety net for these types of investments, as it can fundamentally change decision-making processes wherever some form of risk is involved. Take the example of education again. There’s always a risk in these human capital investments. If you go to college, for example, the risk is that you may need to drop out and be stuck with student loans that you have to pay. If you don’t have the college credential to do so, that’s a big problem. This is one of the big risks you’re facing when you’re deciding whether to go to college. But, in reality, the risk is really dependent on whether you have a safety net that is partly provided by your parents’ wealth. Basically, is there someone to bail you out or not? If there is, then that puts you into a very different risk situation and, by extension, may alter your decision-making process.
Some people critical of the value of studying wealth have a concern about wealth just being a reflection of families’ levels of risk aversion. But I am not making an argument about different levels of risk aversion here: Children who decide whether they should go to college may well have the same level of risk aversion; the point I’m making is that the actual risks that they’re facing are really different depending on whether their family is wealthy. This is just one example of an insurance function of parental wealth that I think is really fundamental and that may account for quite a bit of the association between wealth and the opportunities of future generations.
The importance of multigenerational transmissions of wealth
Hipple: I have always found your use of the insurance function and the purchasing function of wealth to be a really interesting way to think about it. And so one thing I wanted to ask you about is not just the intergenerational transmission of wealth between a parent and a child but also the even bigger picture between grandparents and grandchildren. How does multigenerational family wealth shape people’s outcomes in life through the purchasing function and insurance function that you’ve talked about?
Pfeffer: There is a growing body of literature on what we’d call multigenerational social mobility. Here, we leave the standard focus on just parents and their children behind and expand to not only multiple generations but also to extended families that may include uncles and aunts and others in extended family networks. While the research on multigenerational mobility is at a somewhat early stage, there’s a lot of debate about whether it actually adds a whole lot to our understanding and the conclusions that we’ve drawn from a two-generational perspective.
Expanding the view to three or more generations, however, does provide us some additional information in specific contexts and probably for specific types of resources. I think wealth may be the most obvious one where we can trace the longer-term persistence of advantage through family lineages. In collaborative work with Martin Hällsten at Stockholm University, based on high-quality Swedish administrative data or registered data, we show that even in Sweden grandparental wealth is a quiet strong predictor of the educational achievement of their grandchildren, even when taking into account a lot of characteristics of the parents.
Stated differently, children with parents who are similar on many, many observable characteristics are still distinct in their educational achievements depending on their grandparents’ wealth level. This does not only hold for Sweden. In my recent work with Sasha Killewald at Harvard University, we’ve also described the persistence of wealth across three generations in the United States.
The same pattern holds when examining racial inequalities in wealth and its transmission. Racial disparities in wealth transmission are also stark once you expand beyond just two generations. In this part of our work, we measure wealth as home ownership and home values among white and black families in the late 1960s. Black homeowners of that generation, if you think about it, were a quite advantaged segment of the African American population. But it turns out that their grandchildren have lower homeownership rates than the grandchildren of white nonhomeowners. So, across three generations, racial gaps in wealth are quite persistent and really striking.
The racial disparities in the transmission of multigenerational wealth
Hipple: I’m glad you brought up your work on the wealth gap and its intergenerational transmission between black and white Americans because I think that wealth gap has started to get increasing attention on the national policy stage, even coming up in the recent Democratic presidential primary debates. Looking at how racial inequality is perpetuated across generations, is there anything from your research that suggests policy solutions that could close that black-white wealth gap?
Pfeffer: As a general observation on the debate about the racial wealth gap, I think it’s important to establish or acknowledge that, first of all, the reasons for why these wealth gaps are so large, especially between white and black families, are twofold: They arise, obviously, from long-standing historical patterns of active discrimination and often brutal exclusion of African Americans from asset-accumulation strategies such as homeownership. In addition, there are the continuing processes tied to institutional discrimination that make asset accumulation harder for people of color.
Since both of these things are true, I believe policies will have to be respond to both a historical obligation to correct past discriminatory policies and to ongoing discriminatory processes that still disadvantage African American families and other families of color. Reparations, of course, are one way to respond to the historical obligation, but even with those in place, other policy solutions responding to ongoing discrimination would be needed.
The intergenerational transmission of wealth earlier in life and economic mobility
Hipple: In your research you talk about the role of wealth in shaping outcomes and how important it is early on in children’s lives. You talk about, for example, the importance of the purchasing function because it provides families with the ability to buy a home in a good school district. When researchers think about the role of wealth and wealth being passed from one generation to the next, we tend to think about it at the end of life and how it could be transmitted in a will, as a gift upon the death of someone.
But actually, it seems like the transmission of wealth starts to play a role very early on. I’m curious if you could say more about that. And, in particular, from a measurement perspective, how do you actually capture that transmission? On the one hand, it seems like it’d be pretty straightforward to be able to measure the transfer of wealth upon death from a parent to a child, but a little trickier to figure out how grandparents were able to give the parents some money for a down payment on a house in a good school district. How do you untangle that?
Pfeffer: That’s a great point. A lot of research has been focusing on the role of inheritance and bequests—that is, the passing of sometimes large chunks of money at the end of life. That’s understandable because passing on wealth is a lot easier than passing on other aspects of socioeconomic standings, such as occupation or education. In that same recent paper with Sasha Killewald, we argued that these large transfers of wealth at the end of life are not unimportant, but they are, in some ways, the cherry on top for those who have already derived a lot of advantage from parental and grandparental wealth during their earlier life.
The way that we captured that indirectly was to look at what you may call mediating processes, or things that parental wealth can provide to you that also allow you to accumulate assets. Education is one example. So is homeownership, as you already mentioned. As parents use their wealth to provide assistance with down payments to acquire homes, and as home ownership is still one of the main asset-accumulation strategies among U.S. households, then access to home ownership has such a mediating role in the intergenerational transmission of wealth. And, like housing, we’ve found education to play this role, too. So, the ability to go to college and graduate from college helps accumulate assets, as does the forming of businesses, marriages, and so on. And these are all earlier life processes. By the time the inheritance comes around, which today in the United States is expected to be at about age 50, the inheritance really is the cherry on top.
Now, you also had a question about measurement, which I think is a great question because in principle, at least, you would think that you can empirically capture these transfers. Researchers should be able to, and many studies try to, ask parents whether they provide financial assistance to a child. But it is not quite clear to me whether we will really ever be able to capture all types of transfers that go on. If the parent or the grandparent basically writes a check each month to take care of the electricity bill, phone bill, part of the rent or tuition, for example, then the question is whether parents and grandparents reliably report those types of transfers. I think many transfers are much less transparent and obvious and likely less accurately reported. So, there is certainly a measurement challenge.
In addition, however, and coming back to the insurance function, I will also point out that it’s actually even more complicated. If you believe in the insurance function, then the effects of wealth arise without any transfer because it may be the mere knowledge about the transfer being a possibility that has an influence on you. So, it doesn’t have to happen that you drop out of college and start drawing on your parents’ wealth. It may be enough to know that you could do that for you to decide to go to college in the first place and it affords you different choices once in college. That is, behavioral effects of wealth’s insurance function can arise even in the absence of transfers. And that, of course, makes it quite difficult to capture the full scope of the insurance function empirically.
The psychological processes underlying the intergenerational transmission of wealth
Hipple: That’s really interesting that you bring up the almost psychological role that wealth can play in shaping people’s decisions and the way they just approach their lives.
Pfeffer: Yes, this safety net and its psychological function have also been acknowledged and found in powerful ethnographic work. There’s a book by Tom Shapiro, The Hidden Cost of Being African American, in which he reveals how families think about what he calls these “real and psychological safety nets.” So, qualitative research can really contribute to our understanding here, too.
The other example I wanted to give is how I and my colleagues are identifying some of these insurance effects in an ongoing project, again based on Swedish registered data and in collaboration with Martin Hällsten. We’re showing that if bad outcomes do occur, such as a spell of unemployment, especially youth unemployment, then parental wealth shields you from large post-unemployment earnings losses, even in Sweden. There is a large literature on what economists call “scarring effects of unemployment” that finds, basically, after experiencing a spell of unemployment, workers get back into the labor market at a lower wage level. That is, unemployment causes a “wage scar.”
It turns out, in our ongoing research in Sweden, that this wage scar is reduced for children who come from parents with wealth, suggesting that parents may have activated their wealth to help their offspring secure a new job or keep looking for a new job. In fact, if we look at those parents’ wealth later on, we also see that their own wealth has decreased. That is, they have traded off their children’s wage scar for their own wealth scar. So, there are ways to empirically observe the insurance function in action.
Policy recommendations to make the intergenerational transmission of wealth more equitable
Hipple: What are some policy recommendations you would have for policymakers based on your research into wealth and the way it influences outcomes to better ensure more equality of opportunity for kids?
Pfeffer: One way is to implement policies that spread asset ownership more widely. And there’s certainly a lot of excitement in the policy community, and more specifically in what is called the asset-building community, to make sure there is financial inclusion and that all families have access to fair banking products to support wealth accumulation.
And there are indications that some asset-building policies work even among very low-income people, enabling them to accumulate some assets—say, $500. In many circumstances, that can be quite important: If you blow a tire on the way to work and you have the money to fix it so that you can get to work the next day, that can make a big difference compared to losing a job due to a blown tire.
So, I’m not trying to suggest that these asset-building strategies focused at the very bottom of the distribution are not meaningful. But I think if you look at my research and other researchers’ contributions, you realize that wealth inequality is, of course, a much greater problem than just the lack of assets at the bottom of the distribution.
I think another perspective on which policies to engage with when it comes to wealth inequality is to ask how should we deal with this inequality across the distribution? One perspective looks at the institutional context that allows wealth to take a such significant role in transmitting advantage. Once can design systems that instead rely on forms of public wealth, such as tuition-free higher education. Although it may seem utopian in today’s United States, such systems may reduce credit constraints and thus the need for the purchasing function of wealth for higher education attainment. Of course, such systems do exist in other countries.
So, a focus on wealth as an independent dimension of inequality and advantage does not necessarily imply that all policymakers have to focus on is building wealth across the distribution. It may also suggest that they need to think about how to create systems that require less private wealth, less purchasing through wealth, and less private insurance. Of course, the next question will be how to pay for that. And I think there are two answers, both of which have received more attention in recent debates—namely, wealth taxation and inheritance taxation.
What’s interesting about these tax instruments is not that they’re going to immediately trigger a broad reduction of wealth inequality, largely because wealth is so highly concentrated at the top that you would have to very massively redistribute wealth at a level that I don’t think any of the existing wealth tax proposals do. So, it’s not that the main effect of wealth taxation or inheritance taxation would be to directly reduce wealth inequality. But these tax instruments would have a tremendous effect on revenues, bringing in a lot of revenue that could then be used for the kinds of public wealth investments pointed out earlier.
One more note on policy. Thankfully, there is a lot more attention being paid to the topic of wealth inequality these days. And a lot of this has been inspired by work that is focused on the very, very top of the wealth distribution, which I think is a very defensible and important perspective, especially if you care about political outcomes and political influence. It may really be important to understand just how much wealth is concentrated at the very top and what it could do to our political systems.
At the same time, I always point out that we shouldn’t forget that despite that great concentration at the top, there’s still wealth to go around among the remaining 99 percent. And that wealth is still tremendously unequally distributed, and much more unequally than income. That is important to keep in mind, too. Considering wealth inequality as a populationwide phenomenon should also inform how we think about the need for policy intervention.
Connected to the last point, I would also add a reminder that intergenerational wealth associations are surprisingly linear. That is, with each step on the wealth ladder among parents, there is some advantage to the children or the grandchildren. It is not the case that the advantages of wealth arise exclusively at the very, very top. They arise throughout the distribution and actually to a degree that is surprisingly linear, as I have shown in several recent contributions. This is another call for understanding wealth as a populationwide phenomenon because its intergenerational effects also extend across the population.
Ways forward in wealth research
Hipple: Any thoughts on the ways forward for research on wealth?
Pfeffer: Yes. A point that I have also made in my most recent publications that I think is important both from a measurement perspective and from a conceptual perspective: Most of the work I just talked about, which draws on an indicator of net worth as an encompassing, aggregate measure of wealth, can be replicated or approximated very well with an indicator of just home values. Not even home equity, which would be home value minus mortgage, but home values.
Now, that may not be entirely surprising because we know that housing is one of the main instruments of wealth accumulation in the United States. But I do think it brings up a really interesting conceptual question. To start, from a measurement perspective, this may be good news: It is much easier and cheaper to collect indicators of home values than full-fledged asset measures. But, I think, this finding also brings up really intriguing conceptual questions. Much of what I’ve been talking about, at least implicitly, suggests a genuine asset effect—an effect deriving from ownership of an asset.
But, given the centrality of home values, part of it may, in fact, be a housing effect, an effect of the characteristics of the house you occupy. And another part of it may be an effect of the houses surrounding yours—namely, neighborhood effects. I therefore believe that research on wealth going forward needs to more actively engage with these really broad literatures on housing and on neighborhoods to tease apart the degree of which these intergenerational associations are really driven by asset ownership or flow from other social processes.
In other, ongoing collaborative work with Nora Waitkus at the London School of Economics, I also compare wealth inequality across multiple nations and show that the United States is an outlier in terms of its high level of wealth inequality and concentration compared to other industrialized Western countries. But we also show that it is, again, housing and the distribution of housing in each nation, that explains international differences in wealth inequality. So, in this international comparison, too, we are again led toward housing as the crucial component. Processes tied to housing, such as the regulation and financialization of the housing market, may be crucial to understanding national levels of wealth inequality.
Hipple: Thank you so much for taking the time to talk with me today, Fabian.
Pfeffer: You are welcome.
In Conversation with Bhash Mazumder
Explore the Equitable Growth network of experts around the country and get answers to today's most pressing questions!