Linking Citizen Satisfaction Data to Performance Measures: A Preliminary Evaluation

A dvocates of citizen surveys assert that the idea of surveying citizens is not contrary to representative or a first step on the slippery slope toward direct democracy, but an exercise in deliberative democracy where elected representatives bring citizens into the decision-making process (Miller & Miller, 199 la, p. 7). This kind of direct input into the political process can enhance community participation in decision making, if not community control over service decisions. Contrary to the expectations of those who believe citizens dislike and distrust government, citizens give their local governments generally high marks for service delivery (Miller & Miller, 1992). It is not so much, it seems, that citizens do not trust government, but that many administrators do not trust citizens to render fair judgments about government. A recent study revealed that administrators in Atlanta municipal government believed that citizens would rate city services much lower than they actually did (Melkers & Thomas, 1998). Those opposed to the use of citizen surveys in public policy, many of whom are grounded in the public administration literature of bureaucratic professionalism, argue that government professionals are uniquely qualified to assess service quality and quantity. Sometimes characterized as technocrats (Lovrich & Taylor, 1976), this camp relies on objective internal measures of service quality like performance measures or benchmarks as opposed to citizen evaluations, which are unquestionably subjective. And so the two sides of the service evaluation debate are engaged. What is missing from the debate is a quantitative evaluation of how these two tools of service quality evaluation intersect. Proponents of performance monitoring assume that public service professionals know what constitutes good performance and can measure it accurately. Proponents of citizen satisfaction surveying assume that the public can evaluate service quality even if they have had no direct contact with the service provider. Advo-