條件評估法處理抗議性樣本及無法確定／不知道樣本之一般化架構

A General Framework for Protest and Uncertainty/Don't Know Bidders for Contingent Valuation Method

吳珮瑛(Pei-Ing Wu) ； 陳思豪(Si-Hao Chen) ； 劉哲良(Je-Liang Liou)

《農業與經濟》 53期 (2014/12) Pp. 1-54

https://doi.org/10.6181/agec.2014.53.01

摘要

本文主要建立一可處理條件評估法所有誘導支付方式下，產生之抗議性或無法確定／不知道樣本的一般化架構，所有誘導支付方式可歸納成「呈現WTP金額誘導支付」、「純選擇誘導支付」及「選擇加開放誘導支付」等三大類詢價模式。此一架構除了可以更有效、更簡易的估計方式處理抗議性樣本外，尚可合併處理抗議性樣本及過去極少涉及的無法確定／不知道樣本。而此一般化架構將以雙界二元選擇加開放條件評估法，所調查完成評估墾丁國家公園價值之資料為驗證，墾丁國家公園的抗議性與無法確定／不知道樣本佔整體樣本約37%，結果顯示本文所建立的一般化架構，在處理只涵蓋抗議性樣本或是同時處理抗議性及無法確定／不知道樣本之結果，比過去將抗議性樣本和無法確定／不知道樣本刪除，或視抗議性樣本為零時更有效。同時也讓過去將抗議性樣本及無法確定／不知道樣本刪除，估計而來的平均WTP之高估情形，或是將抗議性樣本視為零支付處理方式，所產生之平均WTP之低估情形得以避免。然結果則一致顯示，不論是傳統將抗議性樣本及無法確定／不知道樣本刪除，或以不當方式納入抗議性樣本的處理方式，均使得對應母體總數因而縮減或／及平均WTP低估，因此傳統處理方式均低估資源總價值，低估比例最高達142%、最少為31%。

關鍵字

雙元常態模型 ； Tobit模型 ； 雙界二元選擇加開放支付方式 ； 國家公園 ； 願付價值

並列摘要

The purpose of this study is to construct a general framework to treat the protest bidders and uncertainty/don't know responses in contingent valuation method (CVM). This framework is applicable to all kinds of elicitation methods in CVM. All types of elicitation methods are classified as three groups. They are ＂WTP revelation＂, ＂pure discrete choice,＂ and ＂discrete choice with open question follow-up.＂ Such framework is much more efficient and easier estimation in treating protest bidders and/or handling uncertainty/don't know responses which are rarely accounted for. The estimation results from this general framework will be compared with those traditional treatments. Traditionally, it is normally delete the protest and uncertainly/don't know responses or treat the protest bidders with zero willingness to pay. This framework is applied to the case of benefit evaluation of Kenting National Park conducted by double-bounded discrete with open question follow-up elicitation method. There is 37% of protest and uncertainty/don't know bidders in this survey. The results from this general framework show that the treatment of protest responses only or the treatment of protest and uncertainty/don't know responses have relatively high efficiency than those from the traditional treatments with such high percentage of protest and uncertainty/don't know responses for this sample. The estimated average willingness to pay (WTP) is over estimated while the protests and uncertainty/don't know responses are excluded traditionally. On the other hand, the estimated average WTP is under estimated while the protests are treated as zero WTP responses. However, total benefit of the National Park is under estimated in either treatment of protests and/or don't know bidders. The under estimation is caused either by the shrink of population due to the elimination of protests or by zero WTP responses for protesters. The total benefit is under estimated by 142% as the highest and at least by 31%.

並列關鍵字

Bivariate model ； Tobit model ； Double-Bounded with open question follow-up ； National park ； Willingness to pay ； Inverse mills ratio

參考文獻

吳珮瑛、林佳穎、蘇明達(2005)。抗議性樣本與答覆『無法確定』樣本之特質—這一群人對假設市場價值評估結果之影響。調查研究：方法與應用。17，66-107。

吳珮瑛、鄭琬方、蘇明達(2004)。複檻式決策過程模型之建構—條件評估法中抗議性答覆之處裡。農業與經濟。32，29-69。

行政院主計總處（2014），《臺灣地區家庭收支調查》，南投：行政院主計總處地方統計推展中心家庭收支科。http://win.dgbas.gov.tw/fies

Arrow, K., R. Solow, P. R. Portney, E. E. Leamer, R. Randner, and H. Schuman (1993), “Report of the NOAA Panel on Contingent Valuation,” Federal Register, 58, 4601-4614

Ahmed, S. U.,Gotoh, K.(2007).The Choice of Elicitation Method in CVM and Their Impact.Reports of Graduate School of Engineering, Nagasaki University.37(68),47-52.