Bureaucratic Ambition: Careers, Motives, and the Innovative Administrator

Appendix A: Survey Methodology

Appendix A: Survey Methodology This appendix details the survey sampling and administrative procedures that I used to gather the data on police chiefs and water utility managers. Sampling Frame The survey sampling frame includes municipal police chiefs and water utility managers serving local governments in the United States. The frame was defined for each profession using U.S. federal government catalogues of local government agencies. Police agencies listed in the U.S. Justice Department’s Bureau of Justice Statistics 2000 Census of State and Local Law Enforcement Agencies (CSLLEA) composed the police frame (U.S. Dept. of Justice 2003).∞ The CSLLEA includes all public law enforcement agencies in the United States that employ at least one sworn o≈cer. All agencies coded as ‘‘municipal police’’ departments serving cities, towns, villages, or townships were included in the frame. State, county, and university law enforcement agencies were excluded from the frame, as were agencies serving U.S. territories. Water utilities owned by local governments and listed in the U.S. Environmental Protection Agency’s (EPA) 2004 Safe Drinking Water Information System (SDWIS) composed the water utility frame.≤ Systems owned by general purpose municipal governments and special districts were included in the frame. Water systems serving U.S. territories were eliminated from the frame. The CSLLEA and SDWIS include size categories for police departments and water utilities, respectively. Very small police departments (with fewer than three full-time sworn o≈cers) were excluded from the frame. Small and very small water utilities (serving populations below 3,300) were excluded, too. While these small and very small agencies compose a substantial share of the total agencies in each profession, collectively they serve a very low proportion of the total U.S. population. Moreover, many very small water utilities are nonprofessional, in some cases operated entirely by volunteers. Heads of such agencies have very limited administrative and policymaking responsibilities. The final sampling frame included 9,784 police agencies and 15,973 water utilities. Tables A1 and A2 summarize the sampling frame by agency size. 184 Appendix A Table A1 Police agency survey sampling frame Municipal police agencies Population served Size group (number of sworn o≈cers) Agencies % Population % Very small (7≥)* ≤,∏≤≥ ≤∞.∞ ≥,≠≤∞,∫≠∞ ∞.∏ Small (≥–∞≤) ∑,∞Ω≤ ∂∞.∫ ∞Ω,∑≤≤,∫Ω∏ ∞≠.∂ Medium (∞≥–∑≠) ≥,≥∂π ≤π.≠ ∂≥,∑∑≥,∑∞∞ ≤≥.≥ Large (∑∞–≤∑≠) ∞,≠∫∏ ∫.∫ ∑∏,≥∑≠,∫∫∫ ≥≠.∞ Very large (8≤∑≠) ∞∑Ω ∞.≥ ∏∂,∑≠≠,∫≠Ω ≥∂.∑ Total ∞≤,∂≠π ∞≠≠.≠ ∞∫∏,Ω∂Ω,Ω≠∑ ∞≠≠.≠ Source: U.S. Justice Department, Bureau of Justice Statistics 2000 Census of State and Local Law Enforcement Agencies (CSLLEA). *Very small police agencies (fewer than three sworn o≈cers) excluded from the frame. Table A2 Water utility survey sampling frame Utilities Population served Size group (population served) Utilities % Utilities % Very small (7∑≠≠)* ∏,∑∞∑ ≤Ω.≠ ∞,∑∑π,∑∏∫ ≠.π Small (∑≠∞–≥,≥≠≠)* Ω,≤π≠ ∂∞.≤ ∞≥,π≠∞,ππ∂ ∏.∂ Medium (≥,≥≠∞–∞≠,≠≠≠) ≥,∏≠≥ ∞∏.≠ ≤∞,≠∑∞,∑∞∞ Ω.∫ Large (∞≠,≠≠∞–∞≠≠,≠≠≠) ≤,πΩ∑ ∞≤.∂ πΩ,≠π∫,πΩ∏ ≥∏.∫ Very large (8∞≠≠,≠≠≠) ≥≠∑ ∞.∂ ΩΩ,≥∞≠,∞≤∂ ∂∏.≥ Total ≤≤,∂∫∫ ∞≠≠.≠ ≤∞∂,∏ΩΩ,ππ≥ ∞≠≠.≠ Source: U.S. Environmental Protection Agency, 2004 Safe Drinking Water Information System (SDWIS). *Very small and small water utilities (serving fewer than 3,300 pop.) excluded from the frame. Sample Once the frame was established, I employed a stratified sampling method to draw representative data from agencies of many sizes. As seen in tables A1 and A2, great majority of agencies are very small and serve small proportions of the total U.S. population. For this reason, a simple random sample would likely o√er little data on large and medium-sized governments. Since large and medium-sized governments serve the majority of the U.S. population, stratifying to ensure their inclusion in the sample is important for drawing conclusions about the nation as a whole Appendix A 185 (Dziegielewski and Opitz 2004). Stratification also ensured that data are gathered from agencies occupying every stratum of these professions, which is important since career progression is an important element of this study. A random sample of agencies within each size stratum shown in tables A2 and A3 composed the final survey sample: thirty-eight per stratum for police and fifty per stratum for water utilities. Two police agencies were eliminated at random to result in exactly 150 sampled agencies for each kind of agency. A few selected governments were removed from the sample. In two cases the top administrative positions were vacant in the governments selected for the sample. One police agency drawn from the frame was removed from the sample because its government had been dissolved since the 2000 CSLLEA. Contact information in the databases for a handful of agencies proved to be inaccurate. Administration To gain maximum generalizability with limited resources this survey emphasized careful instrument design and response rate rather than a large sample size. To encourage their participation, sampled administrators...