Respondents were excluded from the sample if their reasons for 

 responding "no" indicated they were protesting the market setup 

 rather than legitimately considering the question which was 

 asked. These "protest" responses included: 



* Respondents saying they didn't understand the questions, 



* Respondents indicating opposition to any increased taxes, 

 or fees. 



In total, 13 responses were excluded from the analysis because 

 their responses were deemed to be protests. 



Model Specification and Estimated Equations 



The net economic value for a waterfowl hunting trip was 

 determined by analyzing the respondents' yes or no answers to the 

 CVM questions. These responses were analyzed using logistic 

 regression. A more comprehensive discussion of the theory and 

 methods for analyzing these types of models is provided in 

 Duff ield and Patterson (1991) . 



Economic theory suggests that certain variables should influence 

 the response of an individual to a CVM question. In this study, 

 a bivariate form of the willingness to pay model was used which 

 regressed "yes" or "no" responses upon the bid amount asked. It 

 is expected that as bid increases, the probability of a "yes" 

 response will decrease. 



The following bivariate logit model was used for the CVM 

 questions : 



ln(-^ = a + pin{SJD) (2) 



where: P = Probability of a "yes" response 



BID = Dollar amount of increased trip costs the 

 respondent was asked to pay 



The estimated coefficients and standard errors for the model 

 shown in Equation 2 are shown in Tables 10-13 . In all cases, 

 the coefficients for the independent variable (In (BID)) had the 

 expected sign (negative) and added significantly to the model at 

 the .05 level. These results indicate the responses are logical 

 and consistent with economic theory. 



23 



