Methodological issues in meta-analysis, benefit transfer and environmental valuation

Abstract

This thesis consists of four papers on methodological issues in meta-analysis (MA), benefit transfer (BT) and environmental valuation. The first paper presents a MA of stated preference studies valuing non-timber benefits in Norway, Sweden and Finland over the last 20 years. It investigates using different meta-regression models to what extent willingness to pay (WTP) estimates conform with standard expectations, tests a number of novel hypotheses and identifies gaps in the literature. Papers 2-4 then each pick up an important research theme following from the first paper. The second paper utilises the same data to investigate the precision in using MA for international BT, as compared with simpler and more common BT techniques. The third paper investigates, using a different and more extensive dataset of biodiversity and nature conservation values from Asia and Oceania, how the heterogeneity or scope of the MA data influences the results of different meta-regression models and their precision when used for BT. The fourth paper tests using a primary contingent valuation data set of WTP for forest protection in Norway, whether people state different WTP if asked as individuals or on behalf of their household. This paper tests in a more controlled way the question also investigated in the first paper. Results from the four papers are encouraging in contributing to our understanding of people’s preferences for complex environmental goods. However, more research is required to determine the conditions under which MA may be reliably used for BT.