This book is already being hailed as a landmark study of American political representation. It is a pleasure to report that, for once, the professional chorus sings mostly on key. Martin Gilens' main empirical findings add up to a counterpart in political science of the famous Michelson-Morley experiment in physics. They knock into a cocked hat a whole generation of claims about the decisive role of public opinion and the “median voter” in the formation of public policy. If Gilens' results are taken as seriously as they deserve, they should kill off “normal science” as many of us have practiced it.
The author builds his case by a careful and imaginative analysis of almost two thousand survey questions asked by different national polls between 1981 and 2002, along with others from 1964 to 1969 and 2005 to 2006. His guiding idea is simple, though its execution bristles with methodological and statistical difficulties: begin by focusing on questions that ask respondents whether they support, oppose, or “don't know” about “some proposed change” in specific public policies. Then, see how many of these proposals actually became policy within four years. The resulting array of opinions and policy outcomes is then analyzed by several methods, virtually all involving some form of logistic regression. The final outcome is a series of tables and graphs estimating how the probability of actual policy changes varies as support for them changes across the public as a whole and among subgroups defined in various ways, but especially by levels of income.
Gilens waves an abundance of yellow flags about methods and measurement problems. But in the end, he does not mince words. He shows that “higher-income respondents' views are more strongly related to government policy.” In fact, “the strength of the relationship between preferences and policy outcomes not only increases with each step up the income ladder but does so at an increasing rate” (pp. 76–77).
The Michelson-Morley result emerges from the author's scrutiny of cases in which citizens at different levels of income disagree about preferred policies: “When preferences between the well-off and the poor diverge, government policy bears absolutely no relationship to the degree of support or opposition among the poor.” There is, he concludes, “complete lack of government responsiveness to the preferences of the poor” (p. 81).
The much-vaunted power of the middle class turns out to be just as chimerical: “[M]edian-income Americans fare no better than the poor when their preferences diverge from those of the well-off … On the policy questions on which low- and middle-income respondents share the same preferences as those with high incomes, they are, of course, just as likely as high income Americans to get what they want. But when their views differ from those of more affluent Americans, government policy appears to be fairly responsive to the well-off and virtually unrelated to the desires of low- and middle-income citizens” (p. 81).
Gilens does suggest that not all domains of public policy resist popular will to quite the same degree. “Moral/religious issues,” he writes, “show the strongest levels of responsiveness to public preferences overall, while representational inequality is lowest for social welfare policies” like Medicare and Social Security (p. 97). Later, he observes that in the run-up to presidential but not congressional elections, poor and middle-class Americans are more likely to find some of their views enacted into policy, though he also notes that these measures are more likely to be watered down or withdrawn later.
Using the tools of conventional public opinion analysis to turn the customary conclusions of that vast field inside out is a great achievement. It takes nothing away from the significance of Gilens' work to suggest that not everything in its later parts is as persuasive as his fundamental findings.
A short review can only touch upon a few key issues. Perhaps most important is the author's treatment of interest groups. Gilens ascribes the relative success of protecting Medicare and Social Security to the fact that a powerful interest group—the AARP—sides with the public. And he admits a few similar cases. But citing other studies and a statistical analysis of his own that relies on a portmanteau index constructed from “Fortune Power 25” surveys and data for a handful of industry-wide lobbies, he argues that, in general, interest group politics are mostly unrelated to public opinion: “There is no evidence that the degree of interest group engagement on an issue either enhances or detracts from the public's ability to influence policy outcomes” (p. 138).
His measures of interest-group activity, however, share many of the difficulties I outlined in my earlier review of one of the studies he cites (Thomas Ferguson, “Review of Mark A. Smith, American Business and Political Power: Public Opinion, Elections and Democracy,” Journal of Politics 64 [August 2002]: 930–32). The Fortune index simply leaves out the major cross-sector big-business lobbies that take positions on many issues—the Business Roundtable, the Business Council, and the Committee for Economic Development. (Fortune, of course, regularly publishes pieces by or about the business leaders associated with these groups; only the Roundtable is formally registered to lobby.)
Gilens also places too much confidence in major media assessments of lobby activity, which (like many scholarly accounts) frequently rely too heavily on oral interviews vouchsafed by participants. No matter what Fortune may claim, the AARP is first and foremost the vehicle of an insurance seller. Especially recently, its record as a defender of Medicare and Social Security is nothing to celebrate (Yves Smith, “AARP Back in Bed with Effort to Cut Social Security and Medicare,” Naked Capitalism, March 18, 2012, http://www.nakedcapitalism.com/2012/03/aarp-back-in-bed-with-effort-to-cut-social-security-and-medicare.html, [accessed October 14, 2012]). For many years, the most reliable friends of these programs have come from organized labor. Gilens, who recognizes that labor unions play a vital role in securing fragments of popular power, should say so.
More strikingly, and far more fundamentally, individual business corporations barely figure in the book at all. Yet a giant share of all lobbying, as Gilens himself fleetingly acknowledges, is done precisely by individual firms. They are also responsible, directly and indirectly, for a plenary share of campaign contributions—a vital point that virtually disappears amid his final chapter's emphasis on individual contributions. He realizes that the apparent power of the top income groups he analyzes may really reflect the power of a much smaller group of superrich Americans. But his notion that surveys are the royal road for investigating the 1% illustrates how difficult it is to shake off the legacy of the past. If you want to find out about banking policy, you would be better advised to study what J. P. Morgan Chase, BlackRock, or Goldman Sachs are doing; and if you want to comprehend what Senators Harry Reid or Mitch McConnell are up to, better cast your gaze upon hedge-fund managers in New York and elsewhere than on any set of citizens of their own states.
This is simply to warn that the real “spectrum of political money” is considerably wider than Gilens acknowledges. In the period covered by his study, the number of think tanks funded by corporations and the superrich exploded, as did their budgets (Ferguson, “Legislators Never Bowl Alone: Big Money, Mass Media, and the Polarization of Congress,” presented at the Bretton Woods Conference of the Institute for New Economic Thinking, April 2011, http://ineteconomics.org/bretton-woods/paper-presented-bretton-woods-conference-thomas-ferguson).
While many of these institutions often target policymakers inside the beltway, the biggest spend considerable time and money trying to influence broader audiences directly and indirectly. And, as the British investigations of the Rupert Murdoch empire vividly illustrate, major media companies are in fact crucial political actors in their own right, not passive transmitters of neutral “news.” A substantial component of public opinion thus has to be considered endogenous, as Gilens himself recognized in a fine previous study.
Finally, I think that Gilens is ill-advised to use his data, however guardedly, as a Paris Meter Bar to gauge responsiveness across different presidencies. National polls are not cheap; they are probably best understood as integral parts of broader efforts to exert influence by various investor blocs, including media companies. They primarily ask questions that serve their ends and studiously avoid formulations that threaten investor blocs in general. Virtually no one, for example, publishes polls on whether the Federal Reserve should worry more about unemployment than inflation. Polls about public attitudes toward banks also mysteriously disappear from major media for long stretches. Nor do many surveys ask about public employment (typically quite popular), while only a few, at special times, allow citizens to indicate a preference for lowering their own taxes and increasing rates on the rich. Gilens, as usual, is aware of these pitfalls, but I think his splendid book would be even better if he foreswore the practice.
