modest data requirements for the estimation of net benefits conferred by the 

 exogenous change or improvement. Moreover, the TCM will only capture changes 

 in on-site benefits changes; off-site benefits generated by on-site improvements 

 will not be estimated by a TCM study (Loomis 1987a). 



The contingent value method is an important tool for estimating the 

 benefits conferred by qualitative improvements at a wetland site because the data 

 requirements are relatively modest. If the change in the wetland induces a 

 habitat improvement for an economically significant, naturally reproducing 

 species, only a CVM can estimate the increase in off-site (existence) benefits 

 induced by the increase in habitat values (Loomis 1987a). A contingent value 

 method estimate of benefits conferred by a publicly owned wetland recreation site 

 uses survey instruments to provide Hicksian aggregate willingness-to-pay 

 estimates of benefits conferred by estimating the aggregate bid for the site. 



The sums of money that a recreationist would exchange for a natural 

 resource amenity are known as bids; the individual bids usually vary by fixed 

 incremental sums. If there is a fixed maximum bid, the instrument is said to 

 be a closed-end bidding game; if there is no predetermined maximum bid, the 

 instrument is said to be an open-end bidding game. There are various techniques 

 for summing over the individual bids to estimate the aggregate net benefits 

 conferred. 



Both closed-end and open-end bidding games are fixed-sum bidding games. 

 Roughly, the aggregate bid is the sum of the individual bids. Regression 

 analysis is often used to refine the technique so that the total bid is the 

 product of the "true" average bid (given the socio-economic characteristics of 

 the populace that is making the aggregate bid) times the size of the population. 

 Regression analysis can be used to add elegance and precision to CVM estimates 

 of benefits conferred by qualitative improvements in a wetland site. An 

 independent variable representing the quality change can be introduced into the 

 regression model to estimate the marginal or discrete change in total benefits 

 conferred caused by the quality improvement. Also, the statistical significance 

 of the quality variable provides a valid means of testing the basic premise that 

 aggregate benefits are linked to on-site quality. 



Dichotomous bidding games elicit information as to the aggregate bid in 

 the form of a set of probabilities. Thus a respondent is asked to reply with 

 a yes or no to each of a sequence of questions about his willingness to add a 

 certain fixed sum amount to the initial bid. For each incremental sum, there 

 is a probability of eliciting a yes that can be calculated from the responses 

 of the respondents. These probabilities can be used to directly calculate the 

 aggregate bid for the incremental sums. Correction for the socio-economic 

 characteristics of the population is made possible through the use of a 

 qualitative response regression analysis called a logit regression, in which the 

 response probabilities are the dependent variables and the socio-economic 

 characteristics of the population are the independent variables. 



Contingent value methods can be tailored to conform to the Hicksian 

 measures of consumer surplus. However, bidding game techniques can only simulate 

 market responses (McKean and Walsh 1986) whereas the Hicksian measures give the 

 true changes in benefits conferred by real changes in prices, quantities, and 



