As Behn (2003) has pointed out, systems for measuring performance in the public sector can be designed to further a variety of goals.  These purposes can be scaled conceptually in terms of the degree of accountability they impose on public managers.  For example, the use of performance measures to celebrate exemplary workers or to promote departmental activities does little directly to increase accountability relative to utilization of performance measures for resource allocation.  The development of performance measurement systems often involves movement from simple measurement toward incorporation of higher-level accountability regimes.  The implicit assumption is that the net benefits of performance measurement increase monotonically with increased levels of accountability.

This paper examines this assumption in greater detail, illustrating some of the tradeoffs in performance measurement design with case analysis from historical and current trials of performance budgeting in California.  In theory, increasing the stakes connected to performance measures should improve effectiveness, increase efficiency, and reduce financial waste.  At the same time the costs of the performance measurement are also likely to increase.  For example, incentives for the strategic manipulation of information asymmetries increase when there are higher stakes, which in turn increases costs for data verification and auditing.  Also, possible gains from coordinated or cooperative behavior may be lost when stakes are high, encouraging organizational sub-units to maximize their own performance at the cost of collaboration.  The question is whether the net benefits of performance measurement are maximized at a stringent level at which budgets are directly tied to performance measures. 

The State of California has conducted several experiments with performance budgeting, most recently, an initiative of the Governor’s Office to run several pilot performance budgeting trials.  As earlier work has discussed, recently a number of departments in the State have been independently developing performance measurement regimes, most of which have not instituted high level accountability provisions.  This case discussion will include a comparative assessment of the state’s experiences with differing approaches to high- and low-stakes performance management approaches.

Behn, R. D. (2003). "Why measure performance? Different purposes require different measures." Public Administration Review 63(5): 586-606.