UNTANGLING FRENCH VOTING BEHAVIOR: Pierce Replies to Rosenthal

Rosenthal (1981) gets off to a bad start. The first sentence says "Pierce (1981) claims to have tested the Rosenthal-Sen (1973, 1977) spatial model of French voting behavior by using individual survey responses" (p. 363). Readers will look in vain for any such claim in my article. What I did, "by a different method than that employed by Rosenthal and Sen" (Pierce, 1981, p. 118), was to apply the same basic theory from which the Rosenthal-Sen spatial model was derived: the Davis, Hinich, and Ordeshook (1970) concept of the relationship between alienation and electoral participation. I presented evidence in support of three propositions: (1) Participation at the second ballot in France in forced-choice situations can be accounted for by partisan attitudes (as measured by thermometer scores) but not by perceptions of partisan locations on the Left-Right dimension; (2) the partisan choices of the voters in such situations can be better explained by partisan attitudes than by Left-Right perceptions; and (3) deliberate abstainers in such situations resemble ballot spoilers, while chronic abstention and random abstention are long-term phenomena. I. The first of these propositions is the most important, both substantively and in relation to the Rosenthal-Sen spatial model (1973). That model accounted for spoiled ballots (as opposed to participation) at the second round of voting in France in terms of distances between the available parties and the voter both when measures of partisan attitudes and Left-Right locations were employed as indicators of the distances involved. In other words, the Rosenthal-Sen (1973) work implied that Left-Right perceptions were as powerful a determinant of electoral participation as partisanship was. Now, I do not doubt Rosenthal for a moment when he writes that "Sen and I were mainly interested in testing the spatial approach, not in arguing for the Left-Right dimension" (1981, p. 368). But while I too had a