International Choice Modelling Conference, International Choice Modelling Conference 2015

Font Size: 
Disentangling the Status Quo and Zero Price Effect in Stated Choice Experiments
Ulf Liebe, Jürgen Meyerhoff, Andreas Kontoleon

Last modified: 11 May 2015


It is common practice to include a so-called status quo alternative in valuation studies using stated choice experiments. Typically, the status-quo alternative - generally described to respondents as maintaining the current situation - is presented with no additional costs for the respondents. This makes this alternative different from all other alternatives in the choice set which contain a positive price when willingness-to-pay estimates are the aim of the study. Therefore, the question arises whether respondents' choices of the status-quo alternative reflect a true preference for maintaining the current situation or whether the status quo alternative is chosen because it does not include any costs for the respondent, that is, the price for the status-quo alternative is zero.  In addition, in a forced choice design status-quo choices might also signal protest responses by respondents who do not accept the valuation scenario.

This is to the best of our knowledge the first study that tries to disentangle different reasons for respondents' status-quo choices. In order to do so, we applied an experimental design comprising four treatments, that is, differently designed choice sets. In a choice experiment on environmental valuation regarding mixed orchards each respondent was randomly assigned to one of the four treatments. Across treatments the number of choice sets did not differ requesting respondents to answer six choice sets. Moreover, apart from the design of the choice sets the questionnaire was identical. In the first treatment (1) choice sets comprised two alternatives with a positive price and a status-quo alternative with a zero price. In the second treatment (2) choice sets had the same two alternatives with a positive price. Additionally, choice sets in this treatment had an alternative that explicitly offered to maintain the current situation. Choosing this alternative implies that respondents had to pay a positive price but the price was lower than any of the prices of the alternatives that would increase environmental quality. The choice set also included a fourth alternative in which the quality of the environmental good would decrease without additional payments (zero-price alternative). In the third treatment (3) choice sets consisted of three costly alternatives and a status-quo alternative with zero price. Finally, in the fourth treatment (4) choice sets consisted of two alternatives with a positive price, a status-quo alternative with a zero price and an opt-out option.  

Comparisons across treatments should reveal to what extent the differently designed choice sets affect respondents' choices in view of our research interest. Firstly, comparing the effects of treatment (1) and (2) should indicate to what extent respondents are interested in maintaining the status quo of the good in question or to what extent the price of zero drives choices accepting a decreasing quality. Secondly, comparing treatments (1) and (3) should reveal whether an increased complexity or increased preference matching effect occurs due to a third alternative on the choice set with a positive price for increasing the quality of the good in question. Thirdly, comparing treatments (1) and (4) should indicate to what extent protest responses affect choices; it is assumed that protesters would opt-out if they have the possibility to do so.

The data was collected in a web survey in the South of Germany, in the federal state Baden-Württemberg, in November and December 2013. A professional survey organization carried out the sampling based on its access panel. Mixed orchards are typical for Baden-Württemberg but the number and quality (e.g., usage of orchards) has considerably decreased in the last decades. In the survey respondents were asked to choose between different alternatives to maintain or improve the quality of the orchards. Attributes used were area of orchards, the predominant fruit species, preservation of old fruit species, usage of orchards as a meadow or not, and price specified as payments to a fund. Respondents were randomly assigned to one of the four treatment groups described above and each treatment group comprised at least 250 respondents (i.e. in total more than 1,000 respondents). 

In the baseline treatment (1) 22 percent of all choices were directed towards the status quo alternative with a zero price. This share decreases to 9 percent in treatment (2), in which the zero-price alternative is associated with a loss in environmental quality. At the same time 26 percent of all choices account for the alternative in which respondents had to pay a positive price for maintaining the status quo. Note that this price was always lower than in the alternatives that offered an increase in environmental quality. This shows that among those respondents who chose the status quo alternative in a "regular" choice set with two hypothetical alternatives and a status quo alternative many have indeed a preference for maintaining the status quo. Respondents switched significantly from choosing the status quo alternative or choosing a positive price alternative in treatment (1) to the status quo alternative with a positive price in treatment (2). Treatments (3) and (4) suggest that the share of status-quo choices when the price for this alternative is zero is rather constant (20 versus 22 percent) if we add another alternative to the choice set, and that 16 percent of all choices are opt-out choices if the corresponding alternative is available. It seems that respondents mainly switch from alternatives with a positive price to the opt-out alternative (12 percent of all choices) rather than from the status quo alternative with a zero price to opt out (4 percent). These conclusions are supported in conditional logit and random parameter logit models estimated separately for each treatment and for a pooled data set.

A conclusion from our results is thus that the common design of status quo alternatives might result in biased estimates. On the one hand, given no other option, respondents might choose the status quo alternative albeit they are not interested in maintaining the good at hand. In this case status-quo choices do not represent ‘true' preferences. On the other hand, and somewhat unexpected, opt-out choices, if available, lower the frequency of choices directed towards alternatives with a positive price and not so much the number of status quo choices. Protesting might therefore not only emerge as a choice of the status quo alternative but also as a choice of alternatives with a positive price, a finding that is supported by previous research on protesting in contingent valuation and choice experiments.

Conference registration is required in order to view papers.