Suchergebnisse

Voting on two rapid transit referenda held in Atlanta, Ga, in 1968 & 1971 is analyzed using a framework based on the rational voter hypothesis, focusing on three variables influencing transit demand -- quantity & quality of transport services, & the tax/price of transport. The basic differences in the two proposals -- including property tax vs sales tax subsidization & the relative importance of bus transit in the two referenda -- allow several different hypotheses to be drawn. These hypotheses are tested using logit regression analysis of precinct-level voting data from the two referenda together with SE data from the 1970 census. In general, the results from each year separately, as well as between the two years, strongly support the hypothesis that voters act in their own economic self-interest. Current use of transit, importance of the central business district (CBD) as a work site, distance from the CBD, & distance from the proposed train stations all significantly influenced voting outcome in individual precincts. The relative importance of bus commutes was extremely instrumental in determination of the outcome of the 1971 referendum, which emphasized the role of bus service in the overall transit package. 3 Tables. AA.

The question of rounding of income in self-reported income surveys is explored. Using data from the 1966 Survey of Economic Opportunity it was found that R's were more likely to report income in even thousands than would be suggested by simple probabilities. This "bunching" of income means that the definition of the income class interval will affect the class X. For example, the X income for a class interval of 4,000 - 4,999 is 4,460, but for 4,001 - 5,000 the X is 4,536. In addition, the effect that using alternative class intervals have when using weighted regression is explored. 4 Tables. AA.

"January 1985." ; Includes bibliographical references. ; Mode of access: Internet.