¹ Throughout this article we use the term “electoral-situation” to mean simply one general election for one office in one election jurisdiction. “Nominative-situation” means one primary for one office in one election jurisdiction; so for each electoral-situation there are two nominative-situations, i.e., one primary for each major party. In some cases the election jurisdiction is a county, in others two or more counties joined by law for the election of an office such as state senator, state representative, or prosecuting attorney.

² This of course is only one guide, and a crude one, to probable interest in a party's nominations. But it sets some outer limits and is justified, in the minds of the present authors, because it greatly narrows the area which must later be explored by more costly field inquiry into other relevant factors such as the availability of financial and organizational support.

³ “Primary Elections as the Alternative to Party Competition in ‘Safe’ Districts,” Journal of Politics, Vol. 15 (May 1953), pp. 197–210CrossRefGoogle Scholar.

⁴ Key, V. O., American State Politics: An Introduction (New York, 1956), p. 99Google Scholar; Ranney, Austin and Kendall, Willmoore, “The American Party Systems,” this Review, Vol. 48 (June 1954), pp. 477–485Google Scholar.

⁵ This seemed necessary because, conceivably, after a series of safe wins by the dominant party, a low percentage of the vote in the general election by the same party might be reflected in increased contesting in the subordinate party.

⁶ Two-Party Politics in Indiana (Ph.D. dissertation, Harvard University, 1955), pp. 6–7Google Scholar.

⁷ Key, op. cit., p. 172; Ranney and Kendall, op. cit.; Schlesinger, Joseph, “A Two-Dimensional Scheme for Classifying the States According to Degree of Inter-Party Competition,” this Review, Vol. 49 (December 1955), pp. 1120–1128Google Scholar; Hyneman, C. S., “Tenure and Turnover of the Indiana General Assembly, II,” this Review, Vol. 32 (April 1938), p. 317Google Scholar, footnote 26.

⁸ The 57 districts used in this analysis contained 67 counties. The 25 excluded districts contained 29 counties, mostly rural: fourteen were in the lowest quartile for density (that is, most rural), nine in the second quartile, four in the third, and two in the fourth (most urban). One would expect this rural bias in redistricting because counties with small populations are much more likely than large ones to be joined in multi-county circuits.

Since we could not apply our definitions of model safe and model competitive to redistricted counties for this office, we cannot classify the electoral-situations for prosecuting attorney in these counties. We can tell in an approximate way what we are excluding by looking at simultaneous legislative contests. One hundred eighty-five simultaneous house contests were Safe Republican, 58 were Safe Democratic and 15 were Competitive. Apparently, these excluded situations should not bias our study.

⁹ American State Politics, p. 172. Key based his generalization on study of elections for the state house of representatives in Missouri (1942–1950), Indiana (1948–1950), and Ohio (1948).

¹⁰ Thirteen jurisdictions moved from safe for one party to safe for the other party during the period 1926 to 1954. Eight were house districts; one was a senate district; four were judicial circuits electing prosecuting attorneys. Six of the shifts were from Safe Republican to Safe Democratic; seven were from Safe Democratic to Safe Republican. In every case at least five years intervened between the last electoral-situation that was safe for one party and the first electoral-situation that was safe for the other party.

¹¹ Op. cit., p. 181.

¹² We are indebted to Professor H. V. Thornton, Director of the Bureau of Government Research at the University of Oklahoma, for supplying us with the following table which illustrates a similar tendency for “bad times” to draw a larger number of primary contestants than prosperous times. Note that Thornton shows that the post-war period, 1947–54, did not draw more primary candidates than the depression. Our Figure 1 shows a pick-up in the number of primary contestants in the post-war years. One might explain this in terms of a “pent-up” demand for office on the part of ambitious veterans who were unable to seek office during the war; but this still did not equal the demand during the depression.