On the gap between payment card and closed-ended CVM-answers

Abstract

The paper presents contingent valuation (CV) estimates of benefits provided by a proposed 'Kalkalpen' National Park in Austria. Although descriptive results of welfare measures are presented, the focus of the paper is on methodological questions concerning the analysis of CV answers. Evidence is given regarding the difference between payment card (PC) answers and closed-ended question formats. Based on different estimation models for CV questions substantial differences are found between closed-ended and payment card welfare measures. On average PC-willingness to pay (WTP) measures are below the closed-ended figures. Since the evaluation models are based on different premises in the calculation of WTP figures a more precise disclosure of the underlying evaluation methods is required if different question formats are compared to one another. Identical assumptions on the probability distributions have to be assumed whenever open- and closed-ended CV welfare measures are compared. Taking theoretical arguments into account the application of the closedended double-bounded Spike model that provides an average welfare measure is recommended.

Warm glow, free‐riding and vehicle neutrality in a health‐related contingent valuation study

“…Among the group of 1449 respondents who were not prepared to pay the lowest bid amount 1223 persons explicitly stated a zero willingness to pay. Since the spike model takes these zero answers into account a better approximation of the WTP-distribution function can be achieved (see [19]). …”

Willingness‐to‐pay for Sporting Success of Football Bundesliga Teams

“…The historical date value is typically the highest bond value issued. Although some researchers consider the payment card format problematic (Hackl and Pruckner 1999), others note its resurgence (Carson and Groves 2011). A payment card format will be more appropriate when there are more than two alternatives.…”

Framing the Willingness‐to‐pay Question: Impact on Response Patterns and Mean Willingness to Pay

“…OE and PC) than for the methods inferring WTP from choices. In addition a number of studies have shown that DC produces higher WTP estimates than OE (Brown et alRyan and Watson, 2009) as well as higher valuations than the PC approach (Holms and Kramer, 1995;Hackl and Pruckner, 1999). Further, several studies have compared OE and PC (Donaldson et al, 1997;Whynes et al, 2003;Lopez-Martin et al, 2008).…”