Social Position and Attitudes about Income Inequality

Over the last four decades the gap between rich and poor in America has widened. According to economists Emmanuel Saez and Thomas Piketty, the share of pre-tax income held by the top 1% of US earners has increased dramatically since the 1970s; in 2012, the top 1% held nearly a quarter (23%) of total pre-tax income.  To be sure, liberals and conservatives disagree about the causes of inequality and government’s responsibility to address growing income disparities. My contention here is not to weigh into these debates, but rather to examine attitudes about income inequality more broadly. What do Americans think about this growing inequality and what, if anything, should be done about it?

Public opinion data can provide some insight. For example, the General Social Survey (GSS) polls Americans on whether respondents think “income differences in the United States are too large.” The data suggest that over the past 25 years, a majority of Americans have consistently rated the income gap as too large. There has also been a modest uptick in public concern over that period, as actual income differences have increased. The share of respondents who “strongly agreed” that income differences are too large nearly doubled between 1987 (15.7%) and 2010 (27.1%), though the change in the overall share of respondents who either “agreed” or “strongly agreed” was not nearly as dramatic (58.5% in 1987 versus 64.6% in 2010).  Opinions on the matter have also become slightly more polarized.

Ascribing meaning to these public opinion data can be tricky though because Americans have varying understandings of the real distribution of income in the US.  One study asked respondents to first estimate the actual income distribution and then to construct their ideal distribution.  Both the estimated and ideal distributions of wealth were vastly more egalitarian than the status quo. And interestingly, the authors found remarkable consensus across demographic and partisan groups. But does this mean Americans want government to intervene to decrease income inequality? Although the ideal income distributions constructed by respondents looked more like Sweden than the United States, would more Americans support a Swedish-style welfare state if only they knew the true extent of income disparities? The evidence here is mixed.

My colleague Meghan Condon (Loyola University Chicago) and I are investigating this question. One idea we are testing is that frame of reference—how you view your own social position—will shape your attitude about inequality. If you see yourself as lower on the economic ladder, you will be more likely to view income inequality as a problem and believe government should do something about it. To test this theory, we included a survey experiment on the latest Marquette Law School Poll. Respondents were presented with the following:

Think of the people in the United States who are the (worst/best) off—those who have the (least/most) money, (least/most) education, and the (least/most) respected jobs. In particular, we’d like you to think about how you are different from these people in terms of your own income, educational history, and job status.

Respondents were randomly assigned to either think about the least or most advantaged in society (reflected in the parenthetical text above). Following this prompt, we asked respondents to place themselves on an income scale.  We then asked a series of questions about income inequality.

Our results provide evidence that frame of reference does matter. Respondents who were told to think about the most advantaged in society placed themselves lower on the income scale than those who were told to think about the disadvantaged. Further, those primed to think about the wealthiest Americans were more likely to say that income differences in the United States are too large. Our findings are consistent with other research showing that people are more likely to endorse generous donations to charity when they are induced to think of themselves as relatively lower than others in socioeconomic standing.

Most respondents believed that society should ensure equal opportunity for all (87 percent), but respondents were more divided on what explains income inequality and what government should do about it. Overall, 34 percent of respondents agreed that income inequality exists because it benefits the rich, and 40 percent agreed that the government should try to reduce income differences. Priming respondents to think about the worst or best off in society did little to move opinions on these issues.

Thus, thinking about social position relative to the rich (as opposed to the poor) caused people to express greater concern about income disparities, but did not change how people view the causes of or best solutions to inequality. Seeing yourself as lower on the economic ladder might make you more likely to view inequality as a problem, but would not necessarily make you more likely to want government to do something about it.

In the real world, perceptions of social position like this are influenced by media coverage, neighborhood contexts, and workplace interactions. How these experiences matter for public opinion about inequality remains largely unexplored. For example, the share of families living in neighborhoods that are mostly poor or mostly affluent has doubled since 1970. Rich and poor Americans are increasingly segregated from one another and fewer live in what we might call “middle-class” neighborhoods.  Researchers havespeculated that this spatial segregation of affluence will reduce wealthy Americans’ support of policies that benefit the broader public, such as schools and public transportation.

We find on the contrary, however, that individuals express the most concern about income inequality when they are primed to think ofthemselves as living lower down the income ladder, not when they are spurred to think about others living on the bottom rungs.

Amber Wichowsky is an Assistant Professor of Political Science at Marquette University.

Wisconsinites Like Truth-in-Sentencing . . . Sort Of

By Michael M. O’Hear

The latest edition of the Marquette University Law School Poll includes some interesting data on sentencing policy. I’m grateful to Professor Charles Franklin for collaborating with me in putting the questions together. The results are here (note that the sentencing questions start at Q25a).

The primary purpose of the questions was to determine the attitudes of Wisconsinites toward truth-in-sentencing, which was adopted by the state legislature in 1998. The questions are timely in light of recent political debates over new early release opportunities for prison inmates, which were embraced by the legislature in 2009, but then repealed two years later. Early release undercuts truth-in-sentencing by introducing uncertainty into the actual date that inmates will be released. Indeed, critics of the 2009 reforms complained — in what was probably a bit of an overstatement — that the new early release mechanisms “gutted” truth-in-sentencing.

At first blush, the new poll seems to provide strong support for the 2011 repeal and the return to a purer form of truth-in-sentencing: a decisive 63% majority agreed that “truth in sentencing should continue to be the law in Wisconsin.” (25c) Moreover, only 27% agreed that “many of the people who are locked up in prison do not deserve to be there,” and only 37% agreed that “many of the people who are locked up in prison could be safely released without endangering the community.” (27d, e)

But the story is a little more complicated than might first appear.

Take a look at the following:

26a: 85% agree that “Criminals who have genuinely turned their lives around deserve a second chance.”

26c: 67% agree that “Wisconsin should recognize prisoners’ rehabilitative accomplishments by awarding credits toward early release.”

26g: 55% agree that “once a prisoner has served at least half of his term, he should be released from prison and given a less costly form of punishment if he can demonstrate that he is no longer a threat to society.”

These answers, especially the latter two, are inconsistent with pure truth-in-sentencing, and are arguably more in the spirit of the 2009 reforms.

What gives? How could one person (as apparently many did) agree with both truth-in-sentencing and early release based on rehabilitative success?

I think there are at least two considerations that help to explain the apparent contradiction.

First, take a look at how we defined truth-in-sentencing laws: “Laws that ban parole and require prisoners to serve the full term of their sentences, regardless of what they do in prison.” (The full text of what was read to survey respondents is here. An explanation of the poll methodology is here.) By foregrounding “parole” in this way, we implictly invited respondents to equate a vote for truth-in-sentencing with a vote against parole. But parole has a lot of particular negative connotations in this and many other states.

If you go back and study the history of the debates over truth-in-sentencing in 1997-1998, as I have done, you will see that parole was blasted by critics from across the political spectrum as unprincipled and insensitive to public safety concerns. The parole spigot was turned up to full blast in the mid-1990’s to deal with prison overcrowding, then turned down and up and down again, apparently in response to various administrative and political considerations. The Wisconsin experience must surely have reinforced the pervasive, negative image of parole in popular culture as “revolving-door justice.” In any event, it seems safe to assume that few Wisconsinites regarded parole as a reliable gauge of rehabilitative progress.

The continued negative connotations of parole are evident in the 2009 reforms themselves, which created new early release opportunities that might fit certain textboook definitions of parole, but that carefully avoided the term “parole.”

Thus, when 63% of Wisconsinites say that they favor truth-in-sentencing, they may not be rejecting the theory of parole — early release based on rehabilitative progress — so much as the actual historical practice, which fell far short of theoretical ideals.

That’s one way of reconciling support for both truth-in-sentencing and early release. In essence, people are saying, “We like the idea of getting rehabilitated offenders out of prison, but we don’t like that awful system we had in the ’90’s that didn’t really pay any attention to rehabilitation.”

My second hypothesis requires some background to explain. Researchers have identified two distinct types of reasons why the public supports tough-on-crime legislation. (See my post here.) One type of reason is instrumental: the public simply wants to reduce the risks of criminal victimization. This theme was commonly sounded by some supporters of truth-in-sentencing in the ’90’s. Their basic idea was that TIS would reduce crime rates by keeping criminals locked up longer — no more violent offenders conning the parole board into letting them go early to prey on unsuspecting victims.

The other type of reason is symbolic. We like to be tough on crime not so much because we think new laws will really do anything to make us safer, but because passing tough laws is a way of communicating things about ourselves that we like to communicate: that we support crime victims, that we dislike criminals, that we support individual responsibility and accountability, that we subscribe to traditional moral values, and so forth.

Symbolic themes along these lines also figured prominently in the TIS debates in the ’90’s — my impression is even more prominently than the instrumental themes. Just think about the label used for the legislation itself: this was a law that purported to be about “truth,” not “safety” or some other term with instrumental connotations. We vote for truth simply because it is morally attractive to vote for truth — no other justification is necessary.

The new polling data suggest that instrumental and symbolic reasons both continue to play a role in support for truth-in-sentencing. In fact, more people support truth-in-sentencing (63%) than believe that it “helps to reduce crime and make Wisconsin safer” (55%, 25b). A whopping 70% recognize the symbolic character of truth-in-sentencing, agreeing that it “sends a message that society will not tolerate crime.” (25a) And a majority (54%) agreed that “even if [TIS] does not reduce crime, it would still be the right thing to do.” (25d)

It seems clear that some meaningful share, perhaps even a dominant share, of the public support for TIS stems from symbolic considerations — from the sense that by supporting TIS, one communicates one’s support for a certain set of attractive moral values: “truth,” individual responsibility, accountability, etc.

But these sorts of moral values do not necessarily have to lead to support for truth-in-sentencing; this association just happened to emerge as a result of the way that our crime-and-punishment politics have played out over the past 50 years.

In 2009, early release was presented to us in instrumental terms, specifically, as a cost-saving measure. But it could have been presented in moral, symbolic terms. Indeed, in a pair of recent articles, I’ve been trying to develop a moral theory of early release (here and here).

To be successful, such a theory (and practical reforms based on such a theory) must address common (not entirely unfounded) perceptions of prison and parole — that prison is merely a place of passive waiting for release, that prisoners have few opportunities and little support for doing the sorts of things that might help them to become productive citizens upon their release, and that parole release is normally handed out as a matter of course, without regard to whether an inmate has done anything but play basketball and watch TV while behind bars.

However we treat the passive, unproductive inmates, there is an argument that the prisoners who really do work hard for many years and accomplish everything they can to lay a foundation for successful reintegration into society should be released, not because they are “safe” or because it is “cost-effective” to do so, but because they deserve to be released. That, at any rate, is a way of thinking about early release that is not contrary to, but reinforcing of, individual responsibility and accountability.

Our poll questions were worded so as to associate early release with positive actions by inmates (not just passive waiting) and with moral/symbolic values:

26a: “Criminals who have genuinely turned their lives around deserve a second chance.”

26c: “Wisconsin should recognize prisoners’ rehabilitative accomplishments by awarding credits toward early release.”

26d: “Even if such an earned-release progam does not reduce crime, it would still be the right thing to do.”

26g: “Once a prisoner has served at least half of his term, he should be released from prison and given a less costly form of punishment if he can demonstrate [note that the inmate carries the burden] that he is no longer a threat to society.”

So, let me try to pull all of this together for a more succinct statement my second hypothesis. It is not really a contradiction to support both truth-in-sentencing and early release for inmates who work hard to rehabilitate themselves in prison, because both “votes” are really for the same thing — for an ethic of individual responsibility and a criminal-justice system that gives people what they truly deserve.

I’ve offered two preliminary attempts to make sense of the data. Both hypotheses resonate with me, but I certainly can’t rule out other possibilities. I hope to do some more intense number-crunching with the data later this year, which may clarify things a bit.

Michael O’Hear is Associate Dean for Research and Professor of Law at Marquette University Law School.

Recent polls offer contrasting views of support for mining legislation

By Amber Wichowsky

One of the big issues facing the State Legislature this session is how to proceed on controversial mining legislation designed to make it easier for Gogebic Taconite to open an iron-ore mine in northwestern Wisconsin.

The Marquette Law Poll recently surveyed Wisconsinites on their opinions about the mine. But so did another polling outfit during the same timeframe and they reached very different conclusions. How could that be?

Full disclosure: I am not a fan of the standard “environment versus jobs” questions that are often asked on public opinion surveys.  Such questions are typically a variant of the following:

Some people think it is important to protect the environment even if it costs some jobs or otherwise reduces our standard of living. Other people think that protecting the environment is not as important as maintaining jobs and our standard of living. Which is closer to the way you feel, or haven’t you thought much about this?

There are certainly cases where this question offers an illusory dichotomization of this issue. Earlier this year President Obama told the Environmental Protection Agency: “I do not buy the notion that we have to make a choice between having clean air and clean water and growing this economy in a robust way.  I think that is a false debate.” The President went on to note the domestic jobs created in the clean energy sector and the tourism jobs supported by having clean waterways.

But in other instances such trade-offs become more pronounced, particularly in political discourse. Consider the proposed iron-ore mine in northwestern Wisconsin. Last month the Assembly passed legislation that would streamline environmental regulations and speed up the mining permitting process in order to avoid protracted regulatory delays. Democrats and environmental groups criticized the bill for weakening environmental regulations, endangering wetlands and risking the contamination of drinking water. The issue is now being debated in the Senate, where some Democrats and Republicans are trying to find a compromise.

In the latest Marquette Law School Poll, Wisconsinites were asked about whether they support developing the mine. The poll interviewed 716 registered Wisconsin voters between February 16 and 19. Interestingly, Public Policy Polling (in consultation with the Wisconsin League of Conservation Voters) also conducted a poll on the same issue at the same time (Feb. 17 and 18).

The surveys revealed a stark contrast of opinions. Whereas the Marquette Law Poll found that 52% of respondents were in support of developing the mine, the PPP/League of Conservation Voters poll found just the opposite, with nearly half of respondents opposed to the mining legislation.

As Prof. Charles Franklin, director of the Marquette poll, noted in his most recent On the Issues appearance (Feb. 22), these two polls offered a “teaching moment about how the framing of issues is so terribly important.”

I decided to take Charles up on his suggestion to make this a “teachable moment” in my undergraduate class on Elections and Public Opinion. I began class by showing students the exact question wording from the two polls. We then compared the survey items and discussed how question wording affects survey responses. Finally, we tried to come up with a new item that would address the criticisms students had of each question.

First, the questions. (I have added the actual survey responses in parentheses.) The Marquette Law Poll asked:

There is a proposal to develop an iron-ore mine in northwestern Wisconsin. Supporters argue that the mine will create 700 jobs and long-term economic benefits. Opponents argue that not enough environmental protections are in place to preserve water and air quality. Do you support (52%) or oppose (33%) developing the mine? (15% of respondents said either “not heard enough” or “don’t know”)

The PPP/Wisconsin League of Conservation Voters poll asked:

As you may know, the Wisconsin State Senate is considering an open-pit mining bill. Supporters of the bill say that Wisconsin should streamline its environmental regulations in order to create more open-pit mining jobs in Northern Wisconsin. Opponents say that the existing water protections should not be weakened to allow out-of-state mining companies to expose Wisconsin families to chemicals such as mercury, lead, and arsenic. Which comes closer to your point of view? Environmental protections should be streamlined (34%) or environmental regulations should not be weakened (49%)? (16% of respondents said “not sure”)

Students felt that the Marquette Law Poll did not state the environmental risks strongly enough. They argued that putting in an exact number of jobs (700) focused respondents’ attention on the potential economic benefits of the mine, while the opposition statement was vague and open to significant interpretation. (What exactly does “quality” mean anyway?) The students noted that they didn’t feel informed enough to weigh the risks against the rewards.

Students were equally critical of the question wording in the PPP/Wisconsin League of Conservation Voters poll. Whereas a discrete number focused attention on “jobs” in the previous item, many of the students said they focused their attention on the particular contaminants highlighted in the PPP question (mercury, lead and arsenic). Other students said that “out-of-state” signaled something nefarious.

The students then were not surprised to see the differences in opinion between the two surveys.

But could we do any better? It was clear that what the students wanted was more information: What specific jobs might be created? What collateral economic benefits might the mine yield? What are the environmental risks associated with iron-ore mining? What are the health risks associated with ground water pollution? How does the legislation affect existing protections for the state’s wetlands? Why are wetlands important in the first place? By the end of our conversation, it was evident that what started out as a “simple question” was becoming a more “complex discussion” of a complicated issue. Surely, the time constraints facing pollsters would preclude any such item from ever making its way onto a survey.

But what was also clear was that question wording makes a difference. The class didn’t necessarily feel that one poll was better than the other (both had their limitations). Rather, students were left with a number of questions about public opinion surveys. Should pollsters try to maintain neutrality when discussing hot-button policy debates (as the Marquette Law Poll tried to do)? Or should pollsters seek to mimic the way that political actors frame issues? As Professor  Franklin rightfully noted: “Political debate is as much about deciding the terms of engagement as it is about the details of policy and the [associated] benefits and costs.”  Yet, even if pollsters craft questions that reflect how politicians and the media frame issues, how do those who write the questions decide which frames to highlight?  Indeed, given today’s segmented media market, voters are often presented with multiple arguments about public policy debates.

Of course, one could also have asked both questions on the same survey.  Looking at the items again, the results are not necessarily inconsistent between the two. The PPP/Wisconsin League of Conservation Voters poll asked survey respondents about whether they support regulatory changes to allow more mining jobs in northwestern Wisconsin. The Marquette Law Poll asked whether survey respondents support developing the mine.  One could imagine supporting the mine on the grounds that it would create jobs, but also favoring the environmental protections provided in the existing regulatory regime. Indeed, this position is quite similar to the compromise proposal recently introduced by Sen. Dale Schultz (R-Richland Center) and Sen. Bob Jauch (D-Poplar).

So what did we learn from the two questions on mining? Clearly these questions allowed us to more deeply examine the effects of word choice on survey responses. However, comparing these questions also opened up a broader discussion about what one can take away from poll results. Yes, framing matters. But framing also includes how the media report on particular survey questions and the context in which poll results are discussed. Polls serve a number of purposes. They can be great for taking the temperature of the electorate (is the country heading in the right direction?) and for getting a sense of public support for candidates and their positions. But if there was one conclusion my students unanimously drew last week it was that we should not infer the complicated preferences of the public on the basis of just one question.

But what about the nature of public opinion more broadly? Is there such a thing as “popular will”? That discussion will come later on this semester.

Amber Wichowsky is an Assistant Professor of Political Science at Marquette University.