Some Reflections On School Financial Referenda: a Response

Abstract

HE PRECEDING paper, . . A Critique of Cataldo and Holm, attempts to make a number of criticisms of our article on school financial elections (1983). While most of the points focus on issues of research design, our critic also attempts to promote social choice theory as the most fruitful approach to the study of school financial referenda. Before commenting on this alternative formulation, we will respond to the main technical issues our critic raises. 1. Our use of Pearson's r on nominal and ordinal data was inappropriate. We acknowledge that we did not always use pure textbook standards in choice of statistics. Preliminary analysis of our data showed that the direction and magnitude of the observed relationships were basically unaffected by the statistic employed. We chose Pearson's r since it is best known by our readers, and to fit our objectives of ranking the independent variables in order of importance and determining the unique contribution of the most important ones. 2. We did not mention sample size. Sample size was mentioned three times: on page 623 where a response rate of 88 percent was reported based on 523 completed interviews from a sample total of 602; in Table 2 which reported results of a regression analysis based on an N of 264; and in Table 3 which reported the results of a commonality analysis based on an N of 272. 3. Three of our variables - religion, desegregation and legitimacy -were not analyzed separately. These three variables were analyzed separately on three different occassions, with the results clearly reported in Tables 1, 3 and 4. 4. We overlooked the review of the relevant literature contained in Piele and Hall (1973). This publication was cited three times, most prominently in the very first sentence of the article where we noted that this book provides a useful summary of the socioeconomic and attitudinal literature on school financial referenda. 5. We improperly omitted the effects of political party and the press. Both political parties studiously avoided taking a position on the issue, and the nonpartisan school board barely campaigned for its own levy. In the absence of partisan political mobilization, it made little sense to test for party influence. Newspaper endorsements can influence elections for public office at all levels, but their ability to influence local referenda outcomes is unclear. In any event, neither Cleveland daily took a strong enough position on the levy to include this variable in our study. 6. We failed to study non-voting. We chose to study voting. 7. We used a modified version of the SRC political efficacy scale without justifying the modification. The modified version was pretested and