International Choice Modelling Conference, International Choice Modelling Conference 2017

Font Size: 
Analyzing Attribute Cut-offs towards Characteristics of Renewable Energy Production
Malte Oehlmann, Jürgen Meyerhoff

Last modified: 28 March 2017


Strongly supported by the German Renewable Energy Act, the number of renewable energy facilities in Germany has been grown rapidly over the last years. This trend is expected to continue in the future. Despite general public support of renewable energy technologies in Germany, the construction of new facilities as well as new transmission lines often faces substantial opposition at the local level. For instance, people living in the vicinity of wind farms may experience significant negative external effects from changes in the landscape, noise emission or shadow cast. To facilitate the successful and efficient expansion of renewable energies, it is therefore crucial to understand public preferences towards renewable energy alternatives.

Discrete choice experiments (DCE) have been found to be particularly suited to achieve these goals. As usual, studies applying DCEs assume that individuals have compensatory decision strategies such that they make trade-offs among all attributes. However, it has been shown in several areas of the choice literature such as attribute non-attendance (e.g. Hensher 2006) or choice set formation (Swait and Ben-Akiva, 1987) that individuals may adopt non-compensatory decision strategies to, for instance, manage information acquisition and processing cost. One such decision-making process is the use of thresholds or attribute cut-offs (Leong and Hensher, 2012), which are defined as the minimum (maximum) acceptable level an individual sets for an attribute (Huber and Klein, 1991).

This paper investigates the use of attribute cut-offs in a DCE on the expansion of renewable energies and high-voltage transmission lines in Germany. The alternatives in the DCE were labeled as “electricity from wind power”, “electricity from solar power” and “electricity from biogas”. In addition, a status quo could be chosen. The DCE was part of a large-scale nation-wide online survey conducted in 2013. The DCE comprised six choice sets. Each choice set was composed of six attributes including the distance of renewable energy facilities to residential areas (Distance) or the type of high-voltage transmission line (overhead or underground).

Prior to the DCE and before introducing the choice attributes, participants were asked to state their maximum (minimum) acceptable levels for four out of the six attributes. For two of these attributes cut-offs were elicited alternative-specific. Cut-offs levels were presented on a card, and respondents were asked to select the category closest to their threshold. Attribute cut-offs were only requested from a split sample of around 50% of the participants. Respondents were randomly assigned to one of the two treatments. In total, 3211 usable interviews were collected of which 1613 included the cut-off questions.

First, we investigate whether preferences are sensitive to prior cut-off elicitation by comparing the choices made in the sample with to the sample without cut-off questions. By applying the Poe et al. (2005) test in a mixed logit framework, we find that willingness to pay estimates are significantly higher for those respondents who were asked toward their thresholds prior to the DCE. This result applies to two out of six attributes. Moreover, it is shown that choice consistency is higher in the split sample in which cut-offs were stated.

Second, descriptive statistics on attribute cut-offs are analyzed systematically using four criteria: occurrence of cut-offs, cut-off average, number of cut-off violations and extent of cut-off violation. Among other things, it is observed that the vast majority of participants do have attribute cut-offs (e.g. 85% regarding Distance). At the same time respondents do violate their stated minimum (maximum) levels within the DCE. For the attribute Distance, e.g., in more than 40% of the choices participants have been found to choose an alternative with a Distance that is lower than their minimum threshold.

Third, attribute cut-offs are incorporated into the discrete choice framework. Here, the Swait (2001) model is used. It incorporates attribute cut-offs into a linear compensation discrete choice model allowing for non-linearities in the utility function. To our knowledge there is only one attempt where this model has been employed in an environmental economics context (Bush et al., 2009).

Overall, we find that the model fit improves significantly when attribute cut-offs are considered. Moreover, willingness to pay estimates are affected significantly.

In sum, this research provides evidence the participants in a discrete choice survey do indeed have non-compensatory preferences. At the same time, however, participants are willing to violate their cut-offs to some extent when having to make trade-offs in the DCE.


Bush, G., Colombo, S., Hanley, N., 2009. Should all Choices Count? Using the Cut-Offs Approach to Edit Responses in a Choice Experiment. Journal of Environmental and Resource Economics, 44(3), 397 – 414.

Hensher, D.A., 2006. How do respondents process stated choice experiments? Attribute consideration under varying information load. Journal of Applied Econometrics, 21, 861–878.

Huber, J., Klein, N.M., 1991. Adapting Cut-offs to the Choice Environment: The Effects of Attribute Correlation and Reliability. Journal of Consumer Researhc, 18(3), 346–357.

Leong, W. Hensher, D.A., 2012 Embedding decision heuristics in discrete choice models: areview. Transport Reviews: A Transnational Transdisciplinary Journal, 32(3), 313–331.

Poe, G.L., Giraud, K.L., Loomis, J.B., 2005. Computational Methods for Measuring the Difference of Empirical Distributions. American Journal of Agricultural Economics, 87(2), 353-365.

Swait, J., Ben-Akiva, M., 1987. Incorporating random constraints in discrete choice models of choice set generation. Transportation Research B 21(2), 91-102.

Swait, J., 2001. A non-compensatory choice model incorporating attribute cut-offs. Transportation Research Part B: Methodigical, 35(10), 903–928.

Conference registration is required in order to view papers.