Measuring the value of public goods: A new approach with applications to recreational fishing and public utility pricing

Program evaluation (PE) techniques are adapted to measure the value of public good access. The premise is that interventions in the supply of a public good can be considered ‘programs’ where ‘use’ is tantamount to participation or ‘treatment’. Three chapters (Chapters 2 through 4) explore this premise.; Chapter 2 compares the ‘treatment effects’ approach (TEA) to conventional revealed preference (RP) methods for valuing public good access. Program evaluation techniques are adapted to derive access value from differences in related nonmarket activity expenditures between actual and potential public good users. Unlike methods such as the travel cost approach, this approach does not estimate a structural demand or utility model to derive welfare measures. Thus, the TEA avoids many of the widely recognized problems of endogeneity in RP models. A key insight is that alternative counterfactual assumptions can be used to condition estimates of the demand for a public good.; Chapter 3 applies the TEA to measure the recreational fishing value of Gulf of Mexico petroleum platforms. Those anglers who currently fish at these platforms are the treatment group, while those who fish elsewhere are the controls. An econometric model developed in Chapter 2 is used to obtain a measure of the expected value of platform access. The measure is relatively comprehensive because the TEA readily incorporates capital expenditures. Results from the TEA model are compared to those from a travel cost model.; Chapter 4 examines the conservation value of a program that informs public utility customers about the price of service. An analytical model of perceived price is developed that can be used to assess the value of price information. The corresponding empirical models are built around a dataset of Florida water customers that assigns households to treatment and control groups based on whether or not they know the price. The results from demand and treatment effects models are used to derive the expected value of price information for an uninformed household. Importantly, the notion of counterfactuals developed in Chapter 2 allows the welfare measures to be adjusted for the possibility that price elasticities change once a household learns the price.