sound. They should be nonthreatening. Formats 

 that threaten the respondent with a welfare shock 

 that he may view as unfair should be avoided. 



(6) The method of payment (called payment vehi- 

 cles) should be carefully pretested. At the pretest 

 stage, always include a neutral vehicle, e.g., "The 

 money collected will be placed in a trust fund and 

 devoted entirely to providing (the good)." 



(7) The respondent should be given price or 

 value information and asked, "Would you buy?" 

 with the clear understanding that "if no, you would 

 go without." The wording "Would you be willing to 

 pay * * * ? " should be avoided be- 

 cause some respondents may interpret it as an 

 appeal for voluntary contributions. The question 

 must be worded to suggest the pragmatic "take it, 

 or leave it" atmosphere of the marketplace. 



(8) Depending on the "yes" or "no" answer, the 

 price or value is varied iteratively and the question 

 repeated until the respondent's point of indifference 

 between the money and the good is identified. 

 Early iterations may change the price widely until 

 the enumerator senses that he is approaching the 

 respondent's indifference point; then iterative price 

 variations will become finer. 



(9) The starting price quote (called "starting 

 point") will vary across respondents. The particular 

 starting point assigned to a given respondent will 

 be chosen randomly. 



(10) The payment vehicle should be specified. 

 Payment vehicles that may generate an emotional 

 reaction should be avoided because they might in- 

 troduce a confusing element into the bid data. Vehi- 

 cles based on increments in taxes, utility bills, and 

 hunting or fishing license fees may generate such 

 reactions. 



(11) General formats for iterative bidding ques- 

 tions are presented below, followed by specific ex- 

 amples. The questions must be specific to the par- 

 ticular measure of value to be elicited from the re- 

 spondent. WTP formats should always be used; 

 they may be incremental (willingness to pay for an 

 increment in a desired recreation opportunity) or 

 decremental (willingness to pay to avoid a threat- 

 ened decrement in a desired recreation opportuni- 

 ty). The incremental format has two major advan- 

 tages: it is the theoretically correct measure and, 

 since it offers the respondent the (hypothetical) 

 chance to pay for a desired good, it is unlikely to 

 provoke an offended reaction. The decremental 

 format, which asks the respondent how much he 

 would pay to avoid a change he does not want, 

 may seem unfair or morally offensive to some, and 

 thus may elicit biased or otherwise unreliable value 

 estimates. The incremental version is preferred 

 wherever it is credible. 



(12) The incremental version may not be credible 

 if the real world experience is typically one of de- 

 crements rather than increments. For example, the 

 question "if a new, unspoiled natural recreation en- 

 vironment could be created and the right to use it 

 would cost $ , would you buy?" may be re- 

 jected as fantasy by some respondents in a world 

 in which "unspoiled natural recreation environ- 

 ments" are fast disappearing. In such circum- 

 stances, it may be necessary to resort to decre- 

 mental formats. However, since reasonable doubts 

 can be raised, a priori, about the efficiency of WTP 

 decremental formats, the following precautions are 

 essential: The format designed must be the most 

 consistent and plausible and least offensive possi- 

 ble; and at least two different formats must be pre- 

 tested to permit statistical testing for differences in 

 their performance. 



(13) General examples of the WTP formats are: 



WTP incremental: "If you had the opportunity to obtain [describe 

 an increment in recreation facilities, hypothetical market rules, 

 and payment vehicle], would you pay [starting price]? Yes (pay) 

 — . Or would you refuse to pay, and do without [the incre- 

 ment]? No (pay) ." Reiterate with new prices until the 



highest price eliciting a "yes" response is identified. 



WTP decremental (example 1); "[Describe a decrement in recre- 

 ation facilities] will occur unless [describe market rules and pay- 

 ment vehicle]. Would you pay [starting price] to avoid [the decre- 

 ment]? Yes (pay) . Or would you refuse to pay, and thus 



permit [the decrement]? No (pay) ." 



WTP decremental (example 2): "[Describe a recreation facility 

 currently available to respondent] Is currently available [describe 

 existing market rules, existing payment vehicle, and existing 

 price). Unless [the existing price] is increased, [describe a decre- 

 ment] will occur. Would you pay [starting price, which Is some 

 Increment over the existing pnce] in order to prevent [the decre- 

 ment]? Yes (pay) . Or would you refuse to pay, and thus 



permit [the decrement]? No (pay) " Reiterate .... 



(14) Since some respondents may bid only zero 

 amounts to WTP questions, it is important to identi- 

 fy which zero bids represent true zero valuations 

 and which, if any, represent a protest against the 

 market rules or payment vehicles in the bidding 

 format. Check questions should always be used to 

 probe "zero" responses to WTP formats, e.g., "Did 

 you bid zero because (check one): 



a. You believe [the stated increment] would be 

 worth nothing to you? 



b. You believe [the payment vehicle] is already 

 too high? 



c. You believe [the stated increment] would be of 

 value, but you do not think it is fair to expect (the 

 respondent's class of citizen, e.g., hunting license 

 holders, utility customers) to pay for it? 



(15) Answers (b) and (c) above are "protest" re- 

 sponses, addressed not to the value of the good 

 but to some element of the question format. Protest 

 bids should be recorded but eliminated from calcu- 



80 



