Much of our purpose in the past four chapters has revolved around reformulating and decomposing social decision problems so as to simplify analysis and interpretation; but at the same time we have been trying to reduce information-gathering requirements by constructing welfare measures from observable information (such as tax revenue, market demand, and the like). We have succeeded reasonably well in developing utility (and welfare) indicators from demand-and-supply information, but the associated methods are practical only when the relevant supply-and-demand functions can be observed without error. Even when transactions data exist (as for market goods), the necessary observations are rarely complete, and the econometrician is left with difficult obstacles; however, these difficulties seem minor compared to corresponding problems associated with collective goods for which the corresponding markets do not exist. Here, we will explore ways of identifying the demand for collective goods from sufficiently rich information on actual market transactions. Later, in Chapter 16, we will address the underlying econometric problems briefly.

How are we going to estimate the necessary “demand prices” for collective goods? We saw in Chapter 5 that attempts to elicit this information directly run into serious incentive compatibility problems. Although there is some hope for designing mechanisms that will avoid these problems, we are still a long way from having practical methods with general applicability.

However, perhaps we can identify demand for collective goods from observable behavior on markets for private goods.