Decentralizing Land Use Decisions

Abstract

primed to give unduly consistent or instrument-induced responses clearly suggested by the rather substantial variation among service categories shown in Table 2 above. 11. Joel D. Aberbach and Jack L. Walker, Attitudes of Blacks and Whites Toward City Services: Implications for Public in J. Crecine (ed.), Financing the Metropolis (Beverly Hills: Sage, 1970), pp. 519-537; Angus Campbell and Howard Schuman, Racial Attitudes in Fifteen American Cities, in Supplemental Studies for the National Advisory Commission on Civil Disorders (New York: Praeger, 1968), pp. 3-67; Floyd J. Fowler, Jr., op. cit.; John P. Pelissero, Citizen Evaluations of Cotmmnunity Services in Oklahoma (Norman, Okla.: Bureau of Government and Research, 1978). is Race an Important Factor? in H. Hahn (ed.), People and Politics in Urban Societv (Beverly Hills: Sage, 1972), pp. 369-392; Francis X. Tannian with Kevin Dawson, Views of Neighborhoods, Commuters, and Businesses Concerning the Productivity of City Services (Newark: University of Delaware, College of Urban Affairs and Public Policy, 1977); Thomas J. Williams, Citizzen Evaluations of Local Government Services in a Southern Community (Washington, D.C.: National Association of Schools of Public Affairs and Administration, 1977). 12. In interpreting these correlations we limit ourselves to considering two variables related only when their coefficient of sufficient magnitude to be statistically significant at the .05 level of probability; i.e., there only a 5 percent or less probability that the variables are actually unrelated in the population as a whole. An important preliminary consideration in the interpretation of any multiple regression analysis the possibility that the results are compromised because of the problem of multicollinearity; i.e., very high correlations between two or more of the predictor (independent) variables in the equation. Very high correlations in this context have been considered by H. F. Weisberg and B. D. Bowen, An Introduction to Survev Research and Data Analvsis (San Francisco: Freeman, 1975) to be those greater than .70 (p. 214). K. B. Rai and J. C. Blydenburgh, Political Science Statistics (Boston: Holbrook, 1973), p. 232, consider multicollinearity present when coefficients greater than .80 are found between independent variables. In either case, none of the simple correlations between the independent variables in this analysis approach such a level. 13. See Advisory Commission on Intergovernmental Relations (ACIR), 1978: The Year of the New Populism, Intergovernmnental Perspective, Vol. 5 (Winter 1979), pp. 4-5. 14. As quoted in ACIR, States Tackle Tough Fiscal Issues, Intergovernmental Perspectmve, Vol. 5 (Winter 1979), p. 7.