Preferences Are Overrated

Public preferences and public policy are weakly correlated and marked by high and rising inequality, with the affluent wielding greater influence than the less well off no matter who controls the presidency and the Congress, Martin Gilens argues. Gilens’s research is always exemplary, so his conclusions deserve serious consideration. I want to raise some issues and offer some caveats, although my response should not be taken as suggesting Gilens’s findings are not worrisome.

Gilens relies on policy preferences as expressed in public opinion polls and then looks for congruence with policy outcomes. There are, however, a number of serious problems in assessing the match. The choices that policymakers face are often much more complex than those presented to poll respondents. And most Americans are not highly knowledgeable about public policy; after all, they have their lives to lead. A notorious example is the public’s gross overestimate of foreign aid spending and therefore the deficit-reducing impact of cutting it.

Furthermore, poll questions seldom present respondents with the trade-offs that policymakers face. Cutting taxes for everyone (à la the 2001 and 2003 Bush tax cuts) is an easy sell; the deficits that predictably resulted were also predictably unpopular. Cutting the deficit is broadly supported, but there are few government programs—other than foreign aid—that a majority of Americans favor cutting. Sometimes it is literally impossible to follow public opinion.

And what should decision makers do when public opinion seems to be based on too little knowledge? For instance, TARP was and continues to be highly unpopular. There may have been better ways of structuring it, but it’s unlikely that alternatives would have been met with much greater enthusiasm. Yet there was consensus among experts of all stripes that some action like the one the government eventually took was necessary.

Poling data shows us preferences, but interests matter, too, and they tell a different story.

Gilens’s finding that influence—or at least congruence—varies with income is not terribly surprising to those of us who study policymaking. Even aside from the influence derived from campaign contributions, the affluent are much more effective at making their views and interests known to policymakers. Most congressional scholars would probably agree that the influence of business interests is greatest on issues that are unlikely to make it into Gilens’s data set because they fly under the radar; the public never hears about them.

That raises another question about Gilens’s reliance on public opinion polls. When most of the public has no opinion on a matter, we cannot talk about the relationship between the public’s preferences and policymakers’ choices. The same is arguably true when the public’s opinion is based on insufficient or erroneous information.

Yet, even if uninformed, the public often has serious interests at stake. Political scientists, including Gilens, have shied away from imputing interests to people because, in addition to raising all sorts of methodological problems, it gets us into the normative realm. But any analysis of influence over public policy decisions across income groups will be incomplete if interests are ignored. And we can make some reasonable assumptions about the interests of the less affluent. Once we do, Gilens’s curious finding that the less well off have no more influence during Democratic administrations than Republican ones applies only to a group of issues for which preferences have been expressed in opinion polls. This limited agenda is not a stand-in for the interests of less well off Americans. In fact all sorts of federal government programs that aid the less well off are better funded when Democrats are in power, regardless of whether poll respondents know about them.

Gilens asks how representation in America can be made more equal and proposes a number of reforms based on his findings. If we were to focus on interests and not just preferences, then other reforms would be added to the list, and we would have to weigh them differently. Encouraging organizations that inform the less affluent and that speak for their interests would be a top priority. Labor unions have the best track record of doing just that. Unions have not declined so drastically in other industrialized countries and, with friendlier labor legislation, could well be revived in the United States as a powerful instrument of reform.