Economic Inequality and Political Power (Part 3 of 3)

In response to the disproportionate influence of the well-off that I document in Affluence & Influence some have argued that policy reflects (and perhaps should reflect) the preferences of the affluent because affluent Americans are better informed, or because they care more, or because they are more likely to engage politically by voting, contacting their representatives, or working on political campaigns. None of these explanations can account for much of the representational inequality I found.

Faith in the wisdom of the affluent to guide public policy has been sorely tested by the enormous costs in money and human suffering resulting from the Great Recession. (Affluent Americans have been more enthusiastic about deregulation in general, and of the financial industry in particular, than the less well-off.) My data cast further doubt on the notion that representational inequality arises from the greater knowledge or better judgment of those with higher incomes. If government responds not simply to the most affluent but to the most knowledgeable citizens, we would expect education to be a stronger moderator of the preference/policy link than income. In fact, I find the opposite: when both are taken into account, income is a far stronger determinant of influence over policy outcomes than is education (see below).

As far as caring about policy outcomes is concerned, there is no evidence that I’m aware of that suggests that middle-class or poor Americans care less about policy outcomes than the well-off. When the 2004 ANES asked “How important is this issue to you personally?” with regard to half a dozen prominent policy issues, there were no differences across income levels. Similarly, my own data show no difference in the proportion of respondents at different income levels expressing “strong” versus “somewhat” favorable or unfavorable views toward proposed policy changes.

Poor people do vote at lower rates than middle-income or affluent Americans, and other forms of individual political engagement, like working on campaigns and contacting elected officials, also vary by income level. But Larry Bartels’ analyses of Senators’ voting patterns and the preferences of their low, middle, and high income constituents shows that little of the representational inequality he documents can be accounted for by class differences in turnout, political knowledge, or contacting elected officials (see Larry’s masterful book Unequal Democracy).

Moreover, the relationship between income and political engagement does not parallel the relationship between income and influence over policymaking. As the charts below indicate, it is the poor who are distinctively low in voting and campaign work, but as my first post showed, it is high income Americans who are distinctive in their ability to shape government policy.

What most distinguishes the affluent is, of course, their money. And unlike other forms of political participation, donations to campaigns, parties, PACs, and lobbying organizations are highly skewed toward the most affluent Americans:

Electoral campaigns are extraordinarily expensive, and they have become considerably more so over time. The “professionalization” of political campaigns, with their increased focus on paid advertising and sophisticated targeting operations, has increased the value of money relative to other resources (like the ability to mobilize large numbers of volunteers). The growing need for campaign cash has combined with a growing economic inequality that has concentrated more of the country’s wealth in the hands of fewer of its citizens. Together these developments threaten to produce a vicious cycle in which affluent Americans shape policy in ways that enhance their existing advantages and further their ability to determine the course of government policymaking.

Can anything be done to make policymakers more equally responsive to the preferences of all Americans? Campaign finance reforms that reduce the role of large donors are one avenue to pursue. The current climate does not seem auspicious, but Citizens United was a five-to-four decision and perhaps a future Court will be friendlier to campaign finance reform efforts. In addition, competition-enhancing reforms like non-partisan districting might produce more competitive elections and induce policymakers to attend more closely to the public’s preferences. Finally, advocates can focus on those policies that are supported by the affluent and poor alike. Majorities of affluent Americans support increases in the minimum wage, spending for education, job training programs, Social Security, and Medicare (albeit with somewhat less enthusiasm than the less well-off).

Reducing representational inequality will not be easy. But the stakes are so high in this era of enormous social and economic challenges that we cannot be content with a government by the few and for the few, especially when those few are distinguished solely by the size of their bank accounts.

A non-technical description of this project (with responses from 10 commentators) is in The Boston Review, some early findings and technical details are in Public Opinion Quarterly, and much more (including discussions of the quality of mass opinion, the role of interest groups, the difference between Democratic and Republican party control, and elite manipulation of public preferences) can be found in Affluence & Influence: Economic Inequality and Political Power in America.

10 Responses to Economic Inequality and Political Power (Part 3 of 3)

  1. LFC August 15, 2012 at 10:42 am #

    Interesting and persuasive.
    I’m a little puzzled, though, by:
    the relationship between income and political engagement does not parallel the relationship between income and influence over policymaking. As the charts below indicate, it is the poor who are distinctively low in voting and campaign work, but as my first post showed, it is high income Americans who are distinctive in their ability to shape government policy.
    Doesn’t the second sentence here contradict the first? If the poor are least engaged politically and have the least influence over policy, then doesn’t “the relationship between income and political engagement” indeed parallel, at least to some extent, “the relationship between income and influence over policymaking”? I understand of course that “parallel” is not “cause”; also, I suppose that the ‘middle’ group may have less influence over policy than its political engagement rates might suggest it should.

    • Marty Gilens August 15, 2012 at 2:04 pm #

      Your last observation (“…that the ‘middle’ group may have less influence over policy than its political engagement rates might suggest it should”) was my key point. Although there is a positive association between income and all three kinds of political activity shown in those charts, only donations show the same pattern as responsiveness (low for both the poor and middle-class, high for the affluent). If voting or volunteering were important in explaining representational inequality, we would expect the income/responsiveness association to look more like the first two charts, rather than the last one.

      • LFC August 15, 2012 at 4:16 pm #

        That clarifies it for me – thanks.

  2. Justin Buchler August 15, 2012 at 10:48 am #

    Thank you for these posts, which have convinced me that I need to read your book. A couple of quick comments, though. Your post above suggests that “non-partisan districting” might improve policy responsiveness. In a 2009 article in Electoral Studies, Tom Brunell and I examined the empirical association between constituents’ opinions and their Representatives’ opinions and found that competitive districts do not reduce the distance between constituents and their Representatives. Instead, agreement between constituents and their Representatives is maximized when districts are drawn in such a way as to minimize the number of voters who are represented by the candidates whom they opposed. That happens when districts are as homogeneous as possible. The finding was consistent with the arguments that Tom and I had published separately. In my recent book, I argue that the solution is something that you hinted at in your previous post.

    http://themonkeycage.org/blog/2012/08/14/economic-inequality-and-political-power-part-2-of-3/

    In that post, you wrote, “But when political conditions threaten too high a price for pursuing unpopular polices, the parties cleave more closely to the preferences of the public and more equally to the preferences of lower as well as higher income Americans.” The key word there is “threaten.” My argument has been that the key to forcing responsiveness is not the occurrence of competitive elections, but to pose a credible threat to “fire” incumbents who behave in an unrepresentative way, just like with any other bad employee. However, if this threat actually works, then the result is that incumbents behave in representative ways, making it unnecessary, and in fact, unwise to fire them. The threat of losing, conditional on behavior, is the key to responsiveness, not competitive elections. The book is called, “Hiring and Firing Public Officials: Rethinking the Purpose of Elections,” (Oxford 2011).

    • Marty Gilens August 15, 2012 at 2:57 pm #

      That’s an interesting point (which would seem to hinge on whether incumbents in safe, homogenous, districts with non-competitive elections face the same degree of credible threat of being unseated as those in competitive districts). And I agree that electoral uncertainty is key in constraining policymakers to respond to public preferences.

      My focus in this project, however, is not on the dyadic link between constituents and representatives, but on the aggregate association between the electorate’s preferences (or a subgroup of the electorate at one or another income level) and policy outcomes. As I discuss in chapter 1 of Affluence & Influence, a strong association between representatives’ votes and their constituents’ preferences might be conducive to overall policy responsiveness, but by no means guarantees that policy outcomes will reflect the electorate’s preferences. If nonpartisan districting can produce legislatures that more closely match the partisan distribution of the electorate, policy outcomes may more closely reflect voters’ preferences–even if individual voters’ preferences diverge more frequently from that of their own representatives’.

      • Justin Buchler August 15, 2012 at 3:36 pm #

        As linguistic point, you may be using the term, “uncertainty,” differently than I use it. In my book, I use the term to refer to elections whose winners cannot be determined in advance, which is different from the credibility of the threat to remove an incumbent. If an incumbent faces a credible threat of losing for voting the wrong way on legislation, and then votes the right way and easily wins reelection, the result is not an uncertain election by my terminology, despite the fact that it resulted from a credible threat. My argument in the book is that “uncertain” elections are actually quite bad for democracy, but a credible threat to fire incumbents who misbehave is good for democracy because it prevents uncertain elections.

        On the other points, though, consider the following. Paul Ryan is getting a significant amount of attention for his conservative voting record. In the 111th Congress, his DW-NOMINATE score was .805, which is pretty far to the right. Going by Jacobson’s data set, Obama carried that district with 52% of the two-party vote in 2008. Clearly, that relatively close district hasn’t posed a credible threat to remove him from office for being too conservative. Now, let’s turn that around. Cynthia Lummis had a DW-NOMINATE score in the 111th of .884 (yes, I know that Wyoming is At Large– she just happens to be near Ryan in the spreadsheet, and she demonstrates the substantive point). What do you suppose would have happened to her if she had voted like Snowe or Collins? One lump, or two?

        As for policy outcomes, that was what I found so interesting about your argument. I have not looked directly at policy outcomes, but in a 2005 article in Journal of Theoretical Politics, I argued that an effective bipartisan gerrymander actually produces a legislative delegation with an ideological distribution closer to the state than a competitive redistricting plan. If policy outcomes are determined by the median legislator, then my argument still applies (insert parties-in-Congress debate here). When a district that goes for Obama by 52-48 can reelect Paul Ryan with a NOMINATE score of .805, then we can’t count of a competitive redistricting plan to force ideological convergence.

  3. RobW August 16, 2012 at 3:26 pm #

    On the whole, this has been a very interesting series. I wish to make one observation however. It is important to recongnize that the following:

    Faith in the wisdom of the affluent to guide public policy has been sorely tested by the enormous costs in money and human suffering resulting from the Great Recession. (Affluent Americans have been more enthusiastic about deregulation in general, and of the financial industry in particular, than the less well-off.)

    Is offered itself as a matter of faith. It assumes prima facie that since affluent voters tend to favor less regulation and we can point to some examples of deregulation in the years prior to the financial crisis, deregulation must have caused the crisis and thus affluent voters helped enact policies contrary to the majority of Americans. It bares repeating that this is by no means clear. The financial and banking industries have been and continue to be highly regulated. To assert that deregulation was the cause of the financial crisis without qualification is, at base, a normative conclusion and thereby somewhat sloppy in an otherwise solid piece of work.

  4. Tracy Lightcap August 17, 2012 at 11:30 pm #

    Martin:

    I assigned the Boston Review symposium to my Congress and the President class, so don’t take this the wrong way.

    Why are you trying to present results with a 3-D bar graph? The only thing worse is, of course, the dreaded 3-D pie graph, but this is pretty donegone bad. I tried to get an idea of the actual, you know, readings and, as usual, I couldn’t.

    I know that 3-d graphs are now, God help us all, part of the data analysis landscape, but look at this one. Wouldn’t a simple table be a lot easier to read and understand (hint: yes, a thousand times yes)? A small multiple would have been even better.

    Don’t do this to us. The rest of the presentation is just about perfect. Follow your own examples and keep us happy and informed. Thanks.

  5. Xavier August 24, 2012 at 8:06 am #

    I discuss a sort of impossible solution to this problem (a voting system where formal power is inversely proportional to income) here: http://abandonedfootnotes.blogspot.com/2012/08/impossible-political-systems-further.html. The idea came from reflecting on Rawls, not this; but I’m wondering whether anybody has proposed similar ideas?

  6. Alex August 26, 2012 at 3:22 pm #

    Isn’t what does the work here just agenda-setting? Economic majoritarianism just isn’t offered seriously.