Accountability in Democratic Systems: Applications of the Model

In this section, we demonstrate the applications of the model to the analysis of accountability, reporting, and monitoring in democratic systems. We refer to the three relationships presented in Figure 2.1: between senior administrators and the public, between senior administrators and politicians, and between senior administrators and mid-level bureaucrats. We first discuss how the model helps shape our understanding of the differences between the public and private sector in the areas of performance management and then investigate the application of the model in the three relationships in the public sphere mentioned earlier.

Public Sector Versus Private Sector: Implications for the Model

Managers in the public sector usually face different challenges and dilemmas than those in the private sector. While the New Public Management movement tried to apply the management techniques used in the private sector to the public sector (Barzelay 2001; Hood 1991), private organizations are usually smaller in size, have fewer responsibilities, and engage in fewer activities than governments and public organizations (Miller 2005; Wood 2010). They have more flexibility to design and enforce explicit contracts, including the terms aimed at motivating their employees. Managerial autonomy is greater than in the public sector and so is the mobility and flexibility of the workforce. Furthermore, while the principals in the private sector are managers who are also involved in nested principal-agent relations, they usually have more clearly defined interests than the principals in the public sector, namely, the public and politicians. These are collective players who suffer from the problems of collective action, conflicting interests, and the lack of a specific identity and set of interests that will allow them to function effectively as principals. As a result, there is less information asymmetry in the private sector than in the public sector, and controlling for gaming is easier (Miller 2005).

Another important difference between the sectors concerns the type of performance indicators used. While private organizations usually utilize simple quantitative performance indicators, the public sector uses three types of performance indicators: systemic, organizational, and program performance.

According to our model, the conditions needed to minimize gaming are less favorable in the public sector than in the private sector. To support this argument, let us see when the condition Y > Ca > Cp is most likely to be fulfilled. This condition applies to most scenarios developed in our model. Looking at the right side of the inequality, we want to determine when Ca is large enough to exceed Cp. In the public sector, the agents, that is, the senior administrators or mid-level bureaucrats, have much more expertise and control than do the principals (in the case of the politicians or the public) and the strong support ofthe labor unions and their contracts (Miller 2005;Wood 2010). Under such structural conditions, the principals in the public sector find it difficult to penalize the agents for gaming the system, especially because in order to be effective, these costs should be higher than the cost that the principals invest in monitoring the gaming. Therefore, in the public sector it is less likely that this part of the inequality will be fulfilled than it is in the private sector.

A similar rationale applies to the left side of the inequality, which requires that the principal’s benefit from the measured output be large enough, that is, larger than the cost that the agent pays if the gaming is discovered (Y > Ca). Generally, in the public sector, the principal’s benefit from improvements in measured output tends to be vague, because it is difficult to evaluate the impact of such improvements on the principal’s utility function. This is evident when the principal is the public, but it is also true when the principals are politicians or senior administrators. The vague nature of this benefit applies to all three types of performance indicators, meaning that they are not value adding. Therefore, in the public sector it is less likely that the second condition for optimization will be fulfilled than it is in the private sector. Nevertheless, given the distinction we made above among the various performance indicators, we will present several scenarios illustrating the players’ cost-benefit calculations in each interaction.

Moreover, when we consider the scenario of asymmetric information, we consider the likelihood that the condition 2Ca will be fulfilled in order to evaluate whether there is a possibility that certain types of agents will not game the system. Due to the conditions mentioned earlier, in the public sector the rewards for agents for their measured performance (X) are very limited and so are the sanctions imposed on those who game the system. At the same time, the cost of the effort invested in real performance is usually extremely high, because it includes overcoming numerous bureaucratic and political obstacles. Therefore, it is less likely that certain types of agents will not game the system in the public sector than it is in the private sector.

These insights can shed light on the empirical findings in the literature. As explained earlier, Hvidman and Andersen (2013) were unable to determine what dimension of publicness - ownership, funding, or source of social control - mattered most to the effects of management. Our theoretical model implies that the source of social control is a key component in determining the effectiveness of performance management systems. This insight goes beyond this specific empirical study because the model is context free.

However, these insights do not necessarily mean that the public sector should try to adopt more a business-like management orientation and emulate the private sector. On the contrary, our model demonstrates the limitations of performance indicators and performance measurement as a major managerial tool in a public sector that has already moved toward a business-like management paradigm. In other words, given that certain structural aspects of public democratic systems are here to stay, the application of performance management measures beyond a certain level becomes ineffectual, because these mechanisms are distorted through gaming manipulations. We now demonstrate the validity of this argument by considering three principal-agent interactions in the public sector. In the next chapter we will develop the argument that performance management mechanisms primarily influence and improve managerial quality. As a result, such mechanisms are more important tools for organizational learning and transforming organizational culture than for promoting accountability.

< Prev   CONTENTS   Source   Next >