Combining non-market and market prices in cost-benefit analysis
The ultimate goal of assigning monetary values to non-market goods and services is to enable them to be examined in cost-benefit analyses alongside goods and services with market prices. Because the subjective well-being valuation technique is still in its infancy, and tends to produce wide-ranging estimates of value, the UK government Treasury’s current position (as articulated in the 2011 update to The Green Book: Appraisal and Evaluation in Central Government) is that whilst valuations based on life satisfaction might be a useful way to quantify the relative value of two or more non-market goods, they are not yet robust enough to allow comparisons with market prices. An alternative to this view would be to generate and test a series of cost-benefit analysis models - starting with a benchmark model based on market prices and other estimations widely regarded as robust, and extending this with valuations based on preference-based approaches on the one hand, and subjective well-being on the other.
One example of this multi-stage approach is the work of Gyarmati et al. (2008), who undertook a comprehensive quasi-experimental evaluation of the Community Employment Innovation Project in Canada in order to estimate the overall costs and benefits of the project to individuals, communities and governments. The project was designed to evaluate a long-term active re-employment strategy for unemployed individuals who volunteered to work on locally-developed community projects in return for wages (as an alternative to receiving state-funded income transfers). A benchmark cost-benefit analysis model was based on administrative costs, participant earnings and the market prices of fringe benefits, taxes, transfer payments and premiums, as well as market-based estimations of the value of volunteering. An extended model was then developed that included a valuation of foregone leisure (based on 20% of earnings), as well as a valuation based on subjective well-being (based on Helliwell and Huang, 2005) of the social networks that respondents built as a result of their work placements, and the reduction in perceived hardship experienced. In the benchmark model, each dollar in net cost to government was estimated to produce between $1.02 and $1.39 in net benefits to society.48 In the extended model, each dollar in net cost to government was estimated to bring between $1.21 and $1.61 of net benefits. Gyarmati et al. point out that one dollar direct cash transfer (one alternative to the employment project) has meanwhile been estimated to deliver only $0.85 in net benefits to the intended recipient.