International Choice Modelling Conference, International Choice Modelling Conference 2015

Font Size: 
Do varying attributes in a DCE form a context influencing people’s WTP?
Priska Johanna Weller, Julian Sagebiel, Jette Bredahl Jacobsen, Jürgen Meyerhoff

Last modified: 11 May 2015


Carrying out an environmental valuation study typically involves valuation in a specific context, and a long debated question is how strongly results are influenced by context. Economic theory assumes that individuals have stable and independent preferences. However, this may be questioned, particularly in environmental economics where identifying relevant attributes may be non-trivial. Many studies have revealed that changes in the context of the choice situation have an impact on stated choices. One way to investigate the contextual influence is by varying context variables, e.g., the purpose of a journey (Molin and Timmermans 2010) or the situation in which the good in question is consumed (Jaeger and Rose 2008). Tinch et al. (2014) investigate the differences between on-site vs. off-site and ex-ante vs. ex-post surveys. Another context that could be crucial in a discrete choice experiment (DCE) is whether the attributes themselves are building a frame that influences people’s willingness to pay (WTP). We took up this point to explore to what extent the WTP for certain attributes in a DCE is dependent on the context formed by the other attributes. Are people willing to pay the same amount of money for a certain change (e.g., an increase in the share of forest cover in their vicinity) when this attribute appears among other attributes that also focus on forests or when it appears among attributes that focus on agriculture? Apart from the context that is formed in a questionnaire, preferences change if the context changes in which they experience the good. For example, Marsh et al. (2011) investigate the importance of status quo perception for valuation. One underlying reason would be loss aversion (Booij and van de Kuilen 2009). Another reason would be diminishing marginal utility as economic theory would predict.


The data used for the analysis come from a survey that dealt with land use changes in Germany and comprised six independent samples. The salient feature of this survey is that two non-price attributes - share of forest cover and field size in the vicinity of the respondents’ place of residence - and one price attribute were identical across all six samples. Additionally, each of the six samples included three non-price attributes that were different across the six samples. These non-price attributes, 18 altogether, refer to three landscape types (forest landscape, agricultural landscape and general landscape) and cover a variety of characteristics; e.g., protection of biodiversity, amount of invasive tree species or surface water quality. This set up allows us to compare the marginal WTP for the two fixed attributes share of forest cover and field size across samples.

The data consists of 8786 answers to an online interview conducted in March and April 2013 across Germany. Prior to the choice experiment we recorded people’s perception of the present situation of the landscape attributes, and thus we are able to control for the influence of the perceived status quo on the strength of the context effect. Each respondent received nine choice sets and was asked to choose between policies changing the appearance of the landscape within a radius of about 15 kilometres around their place of residence. We geo-coded respondents’ place of residence to gain basic information about the region respondents live in.


In the paper we analyse whether there are differences between the WTP for share of forest cover and field size in the different contexts and landscape types. We do so by using a mixed logit model with scale adjustment for the different samples, and interaction terms between sample dummies and the two fixed attributes. Hence, we can directly test for the difference in WTP across samples. Furthermore, in the models it may be advised to also control for the influence of the perceived status quo of the attributes on WTP. The paper contributes to the literature on context effects by testing attribute context dependency. Here, both the context people are presented with and their surrounding landscapes play a role. Furthermore, it may contribute to the use of valuation studies for benefit transfer and for validity check.



Booij, A.S., van de Kuilen, G., 2009. A parameter-free analysis of the utility of money for the general population under prospect theory. Journal of Economic Psychology 30 (4), 651-666.

Jaeger, S.R., Rose, J.M., 2008. Stated choice experimentation, contextual influences and food choice: A case study. Food Quality and Preference 19, 529 - 564.

Marsh, D., Mkwara, L., Scarpa, R., 2011. Does Respondents’ Perception of the Status Quo Matter in Non-Market Valuation with Choice Experiments? An Application to New Zealand Freshwater Streams. Sustainability 3, 1593 - 1615.

Molin, E.J.E., Timmermans, H.J.P., 2010. Context dependent stated choice experiments: The Case of Train Egress Mode Choice. Journal of Choice Modelling 3(3), 39 - 56.

Tinch, D., Colombo, S., Hanley, N., 2014. The impacts of elicitation context on stated preferences for agricultural landscapes. Journal of Agricultural Economics (online first).

Conference registration is required in order to view papers.