Eds. Note: The Office Hours podcasts are among the most popular content on The Society Pages, and so, in a new special feature, we’ll be bringing our readers excerpts from some of our most scintillating interviews. In this first installment, Sarah Shannon talks with Joe Soss, the Cowles Chair for the Study of Public Service at the University of Minnesota’s Humphrey School of Public Affairs and author, with Richard C. Fording and Sanford F. Schram, of the new book Disciplining the Poor: Neoliberal Paternalism and the Persistent Power of Race. Click here to listen to the full audio interview.

In this conversation, as in his new book, Joe Soss traces the tremendous continuities, as well as the major changes that have occurred in welfare provision and poverty governance in the United States over the past 40 years. In particular, we take on how racial, political, and economic factors have shaped welfare policies in the U.S. and how, to address poverty in the long run, we may need more democracy.

Sarah Shannon: Joe, you have a book coming out soon: Disciplining the Poor. How did you get started on this project?

Joe Soss: The initial idea, actually, for the project was that we were just going to try and look at some issues around how this new approach to dealing with poverty in the United States was really operating on the ground. And so the project really started in a bottom-up way. We wanted to get some administrative data and also go do some field visits to look in the state of Florida, which was really pursuing a strong form of the new approach to poverty governance, to really look at what was going on down there and how things were operating. Because we all shared an interest in race and its role in poverty politics, we were particularly interested in how race operated within that new system. And once we began to analyze the administrative data, we began to see a lot of things that really resonated with some of the other work that we were doing and had done that we weren’t thinking of as part of the same thing, really, but it was really on why states were choosing the policies they were and what was going on nationally with public opinion. And we began to see a much larger and more coherent story coming together: there seemed to be the same sorts of things operating in similar ways at different levels of the system. Eventually, after working on several pieces on this and different articles, we decided to step back and turn it into a book project.

Shannon: “Neoliberal paternalism” is really important for understanding your book’s argument. Can you explain what it is?

Soss: Yeah, so we use this term “neoliberal paternalism,” which we’re not the first to have used (there are other scholars who have as well), but we define it in a particular way that I think is important. Essentially, our argument is that if you want to understand the direction of historical change in poverty governance in the United States, you have to see it as occurring at the intersection of two movements that have changed policy and politics. One is neoliberalism, which is a form of what we call “market fundamentalism.” In many ways, it places the state and lots of other social relations in a subordinate position to the market. It’s really about changing the way in which the state carries out its business, reorganizing the state itself around market principles, and redirecting the state to actually serve market needs in various ways. So, today we see that the state is used to open up markets, it’s used to construct markets, it’s used to absorb market losses (like through the bailouts and a variety of other issues that are currently very controversial in terms of the Occupy Wall Street movement)— the state is far from pulling back from its obligations and, in terms of being active in the economy, has expanded its activities. So, the state is actively involved, for the poor, in attempting to move people into the market, in attempting to immerse people in market experiences so that they will “learn” to conduct themselves in ways that are needed to be employees. It sort of builds these classes for people and creates incentives for people to go into the market.

Paternalism arose at the same time, in a sense, as the reform movement in poverty governance. As the name suggests, paternalism’s really modeled on the traditional father-child authority relationship. The idea that the father has an obligation to guide and help the child, because the child, in a sense, is not able to recognize what’s in their interests—or, if they recognize it, do not have the self-discipline to bring their behavior in line with their good intentions. And so the father, at various times, needs to set the agenda for what the child needs to do. In this view, the state needs to punish behavior that deviates from expectations, needs to reward and incentivize better behavior, needs to monitor behavior at all times…

Its domestic side is equally important, and it really has been an important part of the movement to shift poverty governance toward a model where the state is seen as having a fundamental responsibility to cultivate social order, to ensure social order in poor communities, and to also instill discipline in individuals from those communities. We see this rise of paternalism in the criminal justice system and the rise of mass incarceration in the United States, in the variety of changes in policing and criminal justice, and also on the welfare side, wherein we see a new model of providing aid on the basis of expectations, monitoring, and then meting out penalties for non-compliance and incentives for compliance.

Shannon: You use a particular model for looking at race—the “racial classification model of policy choice”— talk a little bit about that model and how you use it, maybe both as you’re looking at things at the national, historical level, but then also how you saw race operating on the ground.

Soss: There’s a real problem here of trying to understand how it is that we can observe such tremendous racial inequalities and racial disparities in treatment under the cover of official public policy in an era in which it is illegal to make racial distinctions and design different types of treatment for different groups in the letter of the law, in an era in which the vast majority of the people in the United States tend to endorse, at least on the surface, norms of racial equality. Our approach to understanding this is to begin from the premise that, really, it’s not explained by that old image that people like to fall back on (which still has some truth to it, and there are many instances where there’s a problem), but that sort of white people who have conscious prejudice and, in a sense, intentionally discriminate against people who are not part of their own group—a kind of conscious “us-them” dynamic, favoring those who are part of your own racial group or ethnic group and acting negatively toward those who threaten you as the out-group. That just doesn’t explain things very well in the context we’re looking at.

So what’s going on? Well, we take what we understand to be a basically structural view of race. We begin from the idea that race operates in a way that it structures social relations. It’s an important part of our politics, for example: people’s understandings of the parties and the kind of competition that we see between the political parties in the United States for example is aligned with the racial dimension of politics. The Democratic Party and the Republican Party are understood as having different positions on race, racial policy issues, and also on many of the social problems that people associate with, for example, the “urban underclass.” And so race, not only in terms of who votes for whom, but also in terms of the very structure of competition in our politics, just as one example, is structured around race. The economy also. Economic relations are structured in important ways around race, because race matters so greatly for where people live, the extent to which they’re segregated in where they live, the kind of education and human capital they acquire, and the kind of opportunities they have available to them in the market—we have, in many ways, racially structured labor markets.

So, because race matters so much for political relations and economic relations, it is, in a sense, inevitable that race will matter for the set of institutions that govern policy. That in a sense, race is just a creature of our political and economic life together. It’s constructed and will always, to some degree, reflect those relations. To the extent that race structures those relations, it’s going to matter for poverty governance as well.

But at the same time that we think you have to deal with race as a social structure, you also have to ask, “How is it that race actually affects the way people think about these issues, the decisions people make, and the actions they take?” Because, ultimately, the outcomes we get are not simply pure products of social structure. Human agency matters a lot; people act within those social structures, but they’re not determined in how they act.

Shannon: They’re not robots!

Soss: Yeah, they’re not robots. And so we want to understand how race ultimately matters. And if we’re saying that we can’t really understand that well through a model of intentional discrimination and prejudice, what’s going on?

The argument we make is that, when people attempt to design or implement policy, they tend to do so in a way where they think about the actors that they’re applying the policy to and try to figure out what might work. And that their impressions about what might work—“Is this a group that isn’t doing what we want because they don’t have the opportunity?” or “Is this a group that isn’t doing what we want because they don’t have the motivation and they need incentives to do it?” “Is this a group that would only respond to threats and punishments or…?”—that as people look at the groups they have limited information about those groups. One of the things they often have access to is racial information. They can identify that we’re talking about a primarily Latina group here, right, or an African American group, or they know the race or ethnicity of the person that they’re dealing with in an administrative setting. And that they use a kind of what psychologists call “heuristic reasoning,” in that they take the small amounts of information they have and extrapolate from that. They “fill in the details” in some sense.

And our model basically draws on a set of research that’s come from psychology and sociology that is often referred to as “implicit racism theory” or “implicit cognitive theory,” and it suggests that a lot of this goes on behind the curtainPeople are not aware that they’re doing this in many cases, but they look at who they’re designing policy for, implementing it for, and they make judgments about what is needed, they make judgments about (in implementation settings) whether a certain behavior indicates a real problem or is a one-time thing, they have to make these judgments, and there’s nothing wrong with making those judgments, but race is a potential basis for those judgments. And we argue that, in some cases more than others we see racial biases emerge in judgments.

You asked about the implementation setting. So, you know, a good example here is that some people may be aware of some work that Devah Pager, a sociologist, did, where she looked at the interplay of race and a felony drug conviction on the likelihood that someone would be hired for a job. What she found was, if you had people going out and applying for a job who were identical except some were white and some black, and then among those, one white person (male) had a criminal conviction and one did not and one black male had a conviction and one did not, the biggest effect in terms of not getting hired was when you attached the criminal conviction (that discrediting marker) to the African American male.

And we drew from that the idea that it’s important to think about race as a source of vulnerability. It’s not always the case that white clients and black clients in a welfare program are going to be treated differently, but under certain conditions that will emerge strongly. And our classification model is an attempt to specify those conditions. So, for example, we found in an experiment with case managers and corroborated this with administrative data, that when you attach to someone’s record (first in a hypothetical vignette, then in the actual administrative vignette) a note saying that this person has a prior sanction, that it has no discernible, statistical effect on the likelihood that a white client will be sanctioned. White clients would not be sanctioned at a greater rate just because they had that prior sanction on their record. But for black clients, we find a tremendous increase in the likelihood that they’ll be sanctioned.

Or, to take another example, people are sanctioned (that is, penalties are applied to them in welfare programs) more often in politically conservative regions than in politically liberal regions. But when you look at data from public opinion surveys in the United States, what you find is that negative stereotypes of African Americans (and, to some extent, Latinos) relative to white Americans run much higher among conservatives than among liberals. So the implication we draw from that is that the effect of shifting from being in a liberal region to a conservative region should be much greater for minority clients, and, in fact, that’s what we find. For white clients, it doesn’t really matter that much whether they pursue welfare benefits in a very conservative region or a very liberal region, but for black clients in particular, we see a very large increase in the likelihood of being sanctioned when you move to a conservative region.

One last example, one of the great racialized stereotypes of welfare in the United States is the idea that poor African Americans are more likely to prefer to stay on welfare for a long period of time and “avoid work,” in some sense, relative to white Americans. Taking that aspect of racial difference, we hypothesize that the longer people’s spells become—the longer they stay in the welfare program—the more that gap should emerge between the rates at which white clients are sanctioned and black clients are sanctioned. And, in fact, that’s just what we find. The differences are quite small (and even statistically insignificant) in the first months, but by the time you get out toward a year of participation in a welfare program, this gap has emerged, with black clients much more likely to get sanctioned than white clients. We think, again, consistent with the other findings, this is because, in a sense, their record gives a stereotype-consistent cue that seems to confirm the expectation that this person will be a “problem” in some sense. They now look like the long-term welfare recipient of stereotyping more, much like the person who has a prior sanction on their record now looks like more of a problem client if they’re a person of color.

Shannon: To finish, could you please talk a bit about your policy suggestions and today’s interesting policial climate?

Soss: Sure! I’ll try to be brief. There’s no magic policy design, there’s no way to design a policy that is going to serve low-income people well so long as we don’t deal with the underlying issues about the sort of weakening of democracy in the United States, the racial dimensions of our politics, the gender dimensions—all those sorts of things. And there’s no way to design good poverty policies, really, that are going to overcome problems in the market unless we deal with the fact that jobs are not designed to allow for caring obligations: “Bad poverty policies can be bad under any circumstances, but good poverty policies can only be as good as the underlying social relations that they serve.” So we really can’t divorce the effort to change poverty policy and poverty governance from the kind of broad efforts at reform that we see in the Occupy Wall Street Movements today, various things like that. We have to understand that these are intimately related, and you really have to deal with the whole.

But within that, I think we kind of owe it to the readers, after taking them through this long critical analysis, to offer some ideas about what better policies would look like. Much of what we argue is that there are certain aspects of what’s wrong with the current system that really can be improved through central, national policies, through national control. For example, a lot of the racial inequalities that we observe would not be possible without a devolved system of federalism, in which choices are being made in different places according to the racial characteristics of their populations. We also think there are a lot of dynamics that suggest that benefits levels ought to be set at the federal level for the poor. There ought to be, you know, federal child care policies and transportation policies to assist people in various ways, but that ultimately our argument is that federalism really can play an important role and localism can play an important role, even though we observe mainly negative effects from those things in the book. We think it’s the way it’s been designed, not necessarily inherent to the system. So what we argue at the end of the book is that we need to have meaningful federal floors, whether it be minimum wage laws or federal benefit levels for the poor for assistance or various service systems and supports, but that states should be able to rise above those floors and attempt to provide greater services and supports in various ways (but not fall below it). Once you establish those sort of foundations at the federal level, there’s also a set of issues that really can be handled at the local level and should be, because people want to be able to have the opportunity to govern themselves in different ways in different localities, different states. But the argument that we make is that ultimately, like many of the problems having to do with the financial crisis and many of the issues that are at the heart of the protest movements right now are actually much like the problems in poverty governance problems of democracy. They’re problems of insufficient democratization.

And we argue, in many respects, rather than the sort of top-down paternalist imposition of expectations and incentives and punishments, we really need to think about poverty programs as being a site where people are really addressing the most fundamental issues in their lives. The clients have a lot to say about what their problems are, they have a lot to add to the discussion of what kinds of benefits they need individually and what their communities need in order to flourish. We really need to design programs around a much more democratic set of values that draws people into the process. I think the evidence shows, actually, that when you have program designs like that, people actually benefit from them.

One of the things I’ve studied in the past is the contrast between cash welfare programs, which have these very paternalist designs, and Head Start programs, which have these participatory parents’ councils in which people have obligations to join together to talk about how the center is going to be run for their children. What I find of course (even though Head Start is in many ways designed for the children) is that people, parents who become involved in these participatory councils in the Head Start programs have higher involvement levels of civic engagement and political engagement as a result. And also, they actually do better economically—there tend to be positive effects on their outcomes in the labor market as well. So, in some ways, in giving people obligations that empower them is very different from and more positive than the kind of brand of “responsibility” that paternalists have argued for—that’s simply making people responsible for failing to meet the expectations that are being set from above. It’s a different notion of responsibility.

We argue for a system in which we start with broad reforms in terms of political and economic relations, then attempt to establish various federal floors that guarantee people certain necessary support in order to improve their lives, and then allow for certain types of local control that would let people to participate in the democratic process of the authority being exercised over their lives. But ultimately, there is no way to deal with poverty in the United States until we reform the kind of jobs that are available to people with low skill levels. Until we change economic relations so people can have access to jobs that offer wages they can live on, that offer the benefits that they need and their children and their parents need, that offer the kinds of flexibility that allow for care work, we’re not going to have a good fit between the realities of people’s lives and the labor market.