The Journal of Politics


The Sampling Problem in Measuring Policy Mood: An Alternative Solution

Samuel J. Besta1

a1 University of Notre Dame


Measuring policy mood poses difficulties for researchers because a random sample of salient policy issues necessary to derive this concept cannot be generated. Consequently, efforts must be taken to weight the contribution of existing survey items by the salience of their topics. Unfortunately, issue salience varies by political context, and no consensus exists on which setting should be used to assess salience. Stimson (1991) assumes in his seminal work that the frequency with which survey organizations tap opinions on various issues mirrors the salience of these issues in national politics. However, he provides no evidence that this is the case. Rather than assuming pollsters capture political salience, I explicitly measure salience by the prominence given to issues by the public. These alternative specifications of salience identify different structures underlying aggregate domestic policy preferences. These different forms of mood are not trivial, but can affect how mood shapes other political phenomena.

(Accepted July 01 1996)

(Received July 07 1998)

Samuel J. Best is assistant professor of government and international studies, University of Notre Dame, Notre Dame, IN 46556.