Government-to-Government Contracting: Stewardship, Agency, and Substitution

Abstract

ABSTRACT This paper provides empirical evidence detailing the distinctive nature of service delivery provided through contracts with other governments. The results of a survey of Ohio city and county managers both confirm and stand in contrast to implications derived from stewardship theory. Consistent with stewardship, our data demonstrate that contracts with public sector service partners generate less intensive monitoring by contracting governments than do services contracted with private entities. In contrast to stewardship theory, we find that contracting governments do not use other governments for services requiring intensive monitoring. In an era of accountability and results-oriented management, reliance on trust may not satisfy constituents who seek evidence of effective service delivery. The inability of the contracting government to affect another government's service delivery reduces the attractiveness of that government as a contracting partner. If the tools of stewardship prove to be inadequate, the imposition of carrots and sticks appropriate for a principal-agent relationship could undermine the trust central to stewardship. Given these tensions, it is not surprising that governments are contracting less with other governments. Notes a Percentages do not add to 100 because some governments employ multiple modes. Monitoring frequency measured on a scale of 1–4, corresponding to 1 = "Never," 2 = "Infrequently," 3 = "At least yearly," and 4 = "At least quarterly." a Percent of jurisdiction's population with a Bachelor's degree or higher. *Significant at the ten percent level **Significant at the five percent level ***Significant at the one percent level Our survey differs from the ICMA approach in several important ways. ICMA asked governments to describe their monitoring activity, only in general terms, for all services that were contracted out. The ICMA approach assumes that monitoring of external service delivery is independent of the external delivery method employed. Our survey, however, matched monitoring methods with types of service and modes of service delivery. We further differentiated service provision by asking respondents to indicate the "most relied upon" method of service delivery for a particular service. This allowed us to make explicit comparisons about how monitoring for a particular service varies across mode of provision. The ICMA survey provided a truncated inventory of potential service delivery providers, with only private providers, defined by ICMA as comprising "for-profit, non-profit organizations, and private industries" being of interest (ICMA 2002, 1). The ICMA survey did not inquire about monitoring for internally provided services or services delivered through public-public agreements. Our survey included monitoring for provision exclusively by one's own employees, in part by one's own employees, through agreements with other governments, for-profit firms, and non-profit organizations. Additionally, the ICMA survey offers respondents only a simple yes/no choice for a series of oversight methods. Our survey explored the frequency with which different monitoring activities are undertaken by the contracting government. To obtain our grouped indexes, we computed simple averages of the survey responses for each of the questions in the category. The questions were grouped as follows: TableDownload CSVDisplay Table Of the 137 respondents, 83 represented cities, 26 villages, 23 counties, and 5 townships. Our results parallel the rank ordering of providers reflected in the ICMA survey for three of the four services. For both surveys, other governments rank first, in terms of the percentage of governments employing this option, for public health and child welfare programs. For-profit firms and non-profit organizations rank second and third, respectively. In both surveys, non-profit organizations rank first, in terms of governments employing this option, for programs for the elderly, with other governments ranking second, and for-profit firms ranking third. Only for drug and alcohol programs do the two surveys reveal different rank orderings for external service provision. In our survey, for-profit firms rank second and other governments third. In the ICMA survey these rank orderings are reversed. In both surveys, non-profit organizations ranked first, in terms of the percentage of governments indicating utilization of that option. It is interesting to note the changes in the utilization of external providers over time in the ICMA survey. Comparing the results in 1997 (ICMA 1997) to those in 2002, one is struck by the decline in the percentages of governments reporting utilization of other governments for the four services used in our survey. Other governments declined from 31.6 to 26.2 percent for programs for the elderly from 48.8 to 44.6 percent for public health programs from 51.4 to 37.5 percent for child welfare programs and from 42.1 to 38.3 percent for drug and alcohol programs. Nevertheless, other governments remain an important provider of contracted services. The percentage of governments reporting for-profit service provision for the four services declined, as well, in three of the four cases; child welfare was the sole exception. For every service the percentage of governments indicating that they used nonprofit deliverers for the four services increased. Additional informationNotes on contributorsMary K. Marvel Mary K. Marvel is Associate Professor, John Glenn School of Public Affairs Howard P. Marvel Howard P. Marvel is Professor of Economics & Law, both at The Ohio State University. We are indebted to Matthew Potoski, Mildred Warner, participants in a session at the APPAM research conference in Madison, Wisconsin, and the referees for perceptive suggestions. All conclusions and/or errors are solely the responsibility of the authors.