Once a government agency has entered into a contract with a private service provider, it faces the challenge of verifying that the services delivered meet expected standards for implementation and quality. This entails the development and use of accurate performance measures, a task that the literature identifies as both particularly difficult and essential to successful privatization of social services (Kammerman and Kahn 1998; GAO 1997b; Sclar 2000). Creating measures of contractor performance and refining them to suit a specific program frequently require time and experimentation. It is especially difficult to identify and measure desired results for social services. This is because the objectives of many social services, including improved family and child well-being, are often difficult to define simply and clearly.
The challenge of contract monitoring extends not only to the definition of performance measures, but to the mechanics of the measurement process as well. Government agencies and contractors must decide who will conduct the monitoring, how data will be collected and reported, and how feedback will be communicated (Johnston and Romzek 2000). Without deliberate attention to these details, the value and impact of the monitoring system may be undermined. To ensure effective monitoring, many contracting agencies would need to invest heavily in staff training.
Although monitoring can serve as a beneficial link between government agencies and contractors, observers have noted that conflict occasionally arises between government demands for accountability and contractors' interest in maintaining autonomy (Frumkin 2001). This issue is especially salient for nonprofit organizations that see themselves as social innovators. Well-designed incentive structures have helped resolve this conflict in some instances. Under a "milestone payment system" used in Oklahoma, for example, contractors are compensated as they accomplish a set of predetermined program results, which were designed cooperatively by the service provider and the government agency (Frumkin 2001). Generally, the nature of the service being provided, and the risks that outsourcing poses to clients and the government, will affect decisions about the proper balance between promoting innovation through outsourcing and maintaining direct control over program operations (Blank 1999; Cohen 1999).