The massive use of government contracting to provide public goods and services and achieve policy priorities has become a common and desirable government practice at all levels of governments. This governance mode has fundamentally reshaped the features and businesses of the public sectors (Kettl, 1993). Accordingly, "third-party government" (Salamon, 1981), "government by proxy" (Kettl, 1988), "hollow state" (Milward & Provan, 2000), and many other labels have gradually been attached to the public management narrative.
Government contracting itself is a complex process, roughly involving contracting decision-making and contracting management. The prosperity of government contracting, naturally, has aroused a huge number of studies on these two categories of topics (e.g., Cooper, 2003; Donahue, 1989; Kettl, 1993; Savas, 1987; Sclar, 2000). Particularly, the motivations behind government contracting decisions, however, are not well understood. In view of the potential benefits and costs of contracting, what factors motivate or hinder public managers to put contracting onto their agendas?
Current research fails to demonstrate a consistent answer. The main bifurcation lies in the wrestling of pragmatism and politics as the dominating driving forces in make-or-buy decisions. At first, empirical evidences diverge at the local level. Early studies in the 1980s and 1990s showed political factors, like public service constituency group and governance structure, as significant determinants of local government contracting (e.g., Ferris, 1986; Ferris & Graddy, 1986; Morgan, Hirlinger & England, 1988). However, Warner and Hebdon (2001), based on evidence from 201 local governments in New York, concluded government contracting was more pragmatic-oriented by information, monitoring, and service quality. Ironically, Fernandez, Ryu, and Brudney (2008) argued politics, such as political ideology and public employee opposition, still mattered. Even up to the state level, the debate continues. Using a large sample of 1,175 state agencies in fifty American states, Brudney, Fernandez, Ryu, and Wright (2005) showed that American state contracting decisions were immune to political intervention. However, Ni and Bretschneider (2007) and Price and Riccucci (2005) demonstrated that political rationales played a major role in contracting out state E-Government services and prisons, respectively. A possible source for these discrepancies comes from the different types of goods and services these researches examine.
This study continues exploring the driving forces in government contracting decisions, specifically in human services, aiming to add knowledge to contracting motivation literature. Human service contracting is an important and distinct component of government contracting. In human services, the majority of services are provided through publicly funded contracts, mostly with nonprofits. Moreover, human service contracting, not fitting well with the traditional privatization theory, is more like a negotiated and cooperative practice. Given these characteristics, the whole human service contracting process has always been considered as a political practice (Bernstein, 1991; DeHoog, 1984; Johnston & Romzek, 1999; Saidel, 1991). This sheds light on human service contracting research, but very little particularly on contracting decision motivation. However, because of the specificity of human service contracting, any knowledge of contracting may not be complete without human service element.
Thus, the research question of this paper is: Do political factors dominate government human service contracting decisions? Rather than using conceptual or case study methods like much of the current researches, this paper relies on a cross-sectional quantitative analysis. The unit of analysis here is state government, or, more precisely, state human service agency. The paper begins with a brief description of government human service contracting. After that, the concept of politics is operationalized and hypotheses are pointed out accordingly. The third section describes the data and model specification. In section four, the empirical results are discussed. Finally, limitation and future research agenda are mentioned in the conclusion.
AN ERA OF GOVERNANCE BY CONTRACTING
American has an ingrained tradition of using private power, rather than public administrative capacity, to solve public problems (Kingdon, 1999). Government contracting, the most common type of privatization, is so widely and durably used as a government tool that it has become a remarkable feature of the governance system. Although normative opposition to privatization are always present (e.g., Sullivan, 1987; Adams & Balfour, 2010), there is no evidence to see the ebb of contracting. At the federal level, for example, more than one third (37.63%) of federal spending is used in the form of contracts and grants from fiscal years 2000 to 2010 on average (1). Again, at the local level, the scope of contracting is equally prominent. Approximately 45.5% of local government services are delivered through contracting in 2007 (Warner & Hefetz, 2009).
The field of human services is an indispensable component of government contracting. In fact, the use of government-nonprofit contracting in human services is nothing new in the United States. The historic roots can be dated back to the colonial period (Cooper, 2003; Smith & Lipsky, 1995). Today, governments at every level do very little directly by themselves in human service provision. Rather, they fund third-party implementers to delivery services (Salamon, 1995). Among them, nonprofits deliver a large share of government funded human services. In the United States, for example, 56.3% of homeless shelters, 35.9% of drug and alcohol treatment programs, and 32.8% of day care facilities are run by nonprofits in local communities (Warner & Hefetz, 2009). In 2009, governments at all levels contracted with 33,000 human service nonprofits for approximately 200,000 contracts and grants worth over $100 billion (Boris, de Leon, Roeger, & Nikolova, 2010). Such an extensive government-nonprofit partnership is a widespread feature of the U.S. human service delivery system.
The significance of human service contracting lies in not only its scope, but in the distinctiveness. Human service contracting does not follow closely with the traditional market competition model, as the standard privatization theory suggests (Savas, 1987). Unlike many "hard services" such as garbage collection and transportation, in the human service field, competition is limited, performance is difficult to measure, and contract monitoring cost is high (DeHoog, 1984; Schlesinger, Dorwart, & Pulice, 1986; Van Slyke, 2003). These peculiarities cause human service contracting often to be made through negotiation instead of competitive bidding. In many ways, as Lipsky and Smith (1989) noted, "government purchase of services from nonprofit agencies is a substantially political process undertaken against a backdrop in which a market model is said, wrongly, to apply" (643). Thus, human service contracting, according to Dehoog (1990)'s terminology on contracting models, falls into the negotiation model, rather than the ideal competition model. Based on this, current research consistently views politics as a significant part of the contracting process (e.g., Bernstein, 1991; Saidel, 1991). For example, after examining Michigan's Title XX and employment and training service contracting, DeHoog (1984) found various political motivations in human service contracting decisions, such as political pressures from legislators elected officials, service providers and constituencies. Similarly, Johnston and Romzek (1999) also observed politicization of the service contracting process in state medical aid. All these indicate human services as a good field of inquiry to examine the motivations behind contracting decisions.
POLITICS IN GOVERNMENT CONTRACTING DECISIONS
Public management is not equal to generic management with scientific management tradition--politics matters. Public management, as Overman (1984) observed, incorporates "the tensions between rational-instrumental orientations, on the one hand, and political-policy orientations, on the other" (278). Correspondingly, all the contracting decisions are roughly made up of two types of concerns: the policy question and the administration question (Domberger & Jensen, 1997; Kettl, 2005). The policy question specifies the role of government, accountability, and other normative and political judgments. Thus, it has its roots in politics and needs political operation to achieve consensus. The administration problem, on the other side, has underlying logic in economics and management. The core concern is efficiency and effectiveness, i.e., paying less while doing more. As Ferris and Graddy (1986) argued, governments need to minimize the cost of the public service delivery within the political and legal constraints (2). Thus, both administration and policy motivations stand behind contracting decisions, making the decisions both pragmatic and political.
The political dimension of government contracting is in no sense a new research topic. The complexity of politics in public administration has been studied since the inception of public...