Governing by Polls

The study of policy responsiveness to public opinion is blossoming and propagating. Work published over the last two years includes the 2010 book by Stuart Soroka and Chris Wlezien (Canada, US and the UK), this paper by Sattler, Brandt, and Freeeman on the UK,  this paper on Denmark, my own article on the EU, Roberts and Kim’s work on post-Communist Europe, etc.  The latest edition to the literature is this article by Jeffrey Lax and Justin Phillips from Columbia University (forthcoming in AJPS).

“The Democratic Deficit in the States” takes a cross-sectional rather than a dynamic (time series) perspective and analyzes both responsiveness  (correlation)  and congruence between policy outcomes and public opinion in the US states for eight policies. In short, there is a high degree of responsiveness but far from perfect congruence between majority opinion and policy. More salient policies fair better, and having powerful interest groups on your side helps. Altogether, this is an interesting and important study that adds yet another piece to our understanding of policy responsiveness.

What starts to worry me, however, is that the normative implications of the policy responsiveness literature are too often taken for granted. Lax and Phillips seem to equate the lack of correspondence between public opinion and policy to democratic deficit(similarly, Sattler, Brandt and Freeman speak of ‘democratic accountability’). But there is quite a gap between the fact the a policy contradicts the majority of public opinion and the pronouncement of democratic failure. And we need to start unpacking the normative implications of the (lack of) policy responsiveness. 

Of course, at a very general level no political system can be democratic unless there is dynamic responsiveness and broad correspondence between the wishes of the public and what government does. But can we equate congruence of policy with public opinion with democracy? I don’t think so. Precise responsiveness and congruence are neither necessary nor sufficient for democratic policy making. Why?

First and foremost, public opinion as such does not exist. One doesn’t need to embrace a radical post-modern position to admit that the numbers we love to crunch in studies of policy responsiveness are, at best, imperfect (snapshot) estimates of a fluid social construct. It is not only that estimates of aggregate public opinion are subject to the usual measurement problems. It has been shown times and again that the answers we get from public surveys are sensitive to the precise wording, form, and  context of the questions (see George Bishop’s ‘The Illusion of Public Opinion’ for an overview). The questions themselves are often vague and imprecise. Polls will elicit responses even when the people have no meaningful opinion towards the issue (opinions will be regularly given even on fictitious issues). The availability bias is often a problem, especially in surveys of the ‘most important problem’ (open vs. close forms of the question).

A second problem is that public opinion as portrayed by mass surveys need not be the same as the opinion of a group of people after they (1) have been given relevant information about the issue, (2) have been allowed ample time to think about it, and (3) have had the opportunity to deliberate about it (on deliberative polls which come with their own set of problems see James Fishkin). People know astoundingly little about current policies even when they are personally affected by them (here). Do we expect congruence and responsiveness between policy and public opinion as given over the telephone after a modicum of brain activity, or policy and public opinion as it would have been if people made informed decisions with the common good in mind?

The third problem is that public opinion is expressed on various issues presented in isolation. I can very well support an increase in spending on defense, education, and health, and a decrease in the overall state budget at the very same time.  My opinion and preferences need not be consistent but policies need to be. The problem is compounded by the possibility of preference cycles in aggregate public opinion. Even if individual opinions are rational and well-behaved, preference cycles in aggregate public opinion cannot be ruled out.

There is some unintended irony in Stimson et al. designating the aggregate of attitudes and opinions they construct the ‘policy mood’ of the public. Normatively speaking, do we really expect policy to respond to the mood of the public with all the irrationality, instability and caprice that a mood implies? All in all, the lack of perfect temporal and spatial correspondence between public opinion and policy cannot be interpreted directly as a sign of democratic deficit and failure. Political institutions translating mass preferences into policy exist for a reason (well, a number of reasons, including preference aggregation, deliberation, and inducing stability).

The other side of the same coin is that responsiveness is not sufficient for democracy. The fact that a government follows closely majority opinion as expressed in the polls and adjusts policy accordingly cannot be a substitute for a democratic policy making process. This is especially clear in my own analysis of the EU: although I find that aggregate legislative production closely follows the ebbs and flows in public support for the EU during the 1970s, 1980s, and early 1990s this cannot dispel our misgivings about the democratic deficit of the EU during this period – the polls are not a substitute for elections, representation, and accountability.

The lack of sufficient reflection on the democratic implications of the (lack of ) policy responsiveness is especially worrying in view of the tendency (identified on the basis of my subjective reading of the political process in several European states) of more and more reference to and reliance on ‘instant’ polls in making policy. The increased availability and speed of delivery of ‘representative’ public opinion polls lures politicians into dancing to the tune of public opinion on every occasion. Sensible policies are abandoned if the poll numbers are not right (e.g. second hand smoking restrictions, see here), and retrogressive policies are enacted if the percentage of public support is high enough. But government by polls is only one step removed from the government by mobs. Politicians should sometimes have different policy opinions than the public and they should have the courage to pursue these opinions in the face of (temporary and latent) opposition by the citizens. Meanwhile, social science has the important task to uncover when and how policy responsivness and congruence works. But I see no need to inflate and oversell the normative implications of the research.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s