Elections and Policy By Richard Johnston, Department of Political Science, Penn

Elections matter for policy, but does policy matter for elections? Not always, and when it does the connection is pretty loose. Voters and candidates both do their part to make it so.

Start with voters. It is clear that most voters do not have much detailed policy knowledge, and can even be induced to pronounce on persons, institutions, and policies that do not exist. Where voters express opinions on policy, there is little resonance from debate among elites. Apparent positions on any given question are commonly unconnected to opinions on closely allied questions. Indeed an individual’s opinion at any given moment is only loosely related to the same person’s expressed opinion at another time, even where the survey questions are absolutely identical.

The picture is more reassuring when individuals are pooled, into groups or into the electorate as a whole. Where individuals’ opinions are not very stable, aggregate distributions change slowly, if at all. The change that does occur is interpretable in terms of sensible factors. In the short run, opinion responds to shifts in economic well-being; the current crisis is an example. In the long run, change reflects cultural evolution, as for example with gay rights. In the medium term, opinion on policy interacts with actual policy-making. A cyclical pattern is driven by party control of Congress. As that party enacts its program, it typically overshoots public opinion. This in turn shifts the electorate’s policy mood, such that the party is eventually called to account. The other party then proceeds to overshoot in the opposite direction. Opinion on defense spending is also moved by the sense of external threat. Perception of threat or of policy shift may be as important as reality, and perceptions can be manipulated. And the connection is highly general: what counts is a policy mood, rather than opinion in individual policies. Reward or punishment of individual Representatives or Senators bears little relationship to their policy positioning. In 2006 and 2008, for example, Republicans were punished for overshooting to the right, but within Republican ranks the victims have been the centrists. There are moments when individual Representatives are vulnerable for high-profile behavior, but almost always these are matters of malfeasance, not of substantive policy.

Candidates make their own contribution to the strength or weakness of the opinion-policy link. Candidates often obfuscate, although clarification of the choice is sometimes in the interest of at least one candidate. Negative advertising, notwithstanding its bad reputation, actually helps do this, as negative claims are commonly more truthful than positive ones, which tend to be self-congratulatory puffery. In many House races, however, challengers are so weak that they cannot afford to get their message out. In most Presidential elections, one or both candidates have only a weak national profile or none at all. The classic formulation has been to run for Washington by running against Washington. This makes it hard for even highly-informed voters to impute policy positions to the candidate. This problem is mitigated when an incumbent is running for re-election, and it helps that candidates carry party labels. By itself, the label fills lots of blanks, the more so where parties are polarized. But the label is a stereotype and not always a guide to a candidate’s position on a particular policy.

This year has been unusual in that both Presidential candidates are Senators, each with a national profile. Even so, the campaign was partly a contest over perception: How closely should John McCain be tied to an unpopular incumbent? How closely should Barack Obama be tied to the rest of the Democratic Party? Each candidate went to great lengths in the primaries to detach himself from his party. To the extent that each succeeded, they may have reduced the clarity the choice. Neither campaign really processed the economic crisis, other than to acknowledge that a crisis existed. But then the crisis may illustrate a quite common form of mandate, a mandate to do something, anything—if it works.

Richard Johnston is a Professor of Political Science at the University of Pennsylvania. His specialties include elections, public opinion, and representation, with a special interest in campaign dynamics and survey methods. View Professor Johnston’s webpage.

See the other side

Related posts:

1. Health Care Professionals Favor Single-Payer Insurance By Barry Slaff of the Penn Democrats
2. Spring Political Tickers
3. Consumers Favor More Healthcare Choices By Nick Smith-Wang of the Penn Republicans

Leave a Reply

You must be logged in to post a comment.