International Choice Modelling Conference, International Choice Modelling Conference 2015

Font Size: 
Uncovering Complexity-Induced Status Quo Effects in Discrete Choice Experiments for Environmental Valuation
Malte Oehlmann, Priska Weller, Jürgen Meyerhoff

Last modified: 18 May 2015




When individuals have to make choices among different alternatives, they have been observed to disproportionally often remain at the status quo (SQ) option. Samuelson and Zeckhauser (1988) found this phenomenon to be present in a series of decision-making experiments and termed it the SQ bias or SQ effect. The propensity to stay at the SQ has also been observed in the context of discrete choice experiments (DCEs) for environmental valuation, which usually include a SQ, current situation, or default option (e.g. Adamowicz et al. 1998, Scarpa et al. 2005).

In this study we use data from an extensive stated preference web survey on land use changes in Germany to explore possible sources of the SQ effect. The survey included a DCE considering attributes regarding the share of forest in the landscape, land conversion, and biodiversity in different landscape types. In total, 1684 completed interviews were obtained.

We focus on SQ effects induced by the complexity of the DCE. We define choice task complexity in two different ways. On the one hand, we focus on the quantity of information in the DCE expressed through the design dimensionality. We follow a Design of Designs (DoD) approach which was originally introduced by Hensher (2004) in the context of transportation research. Following the DoD approach, we use 16 different split samples or treatments systematically varying the number of choice sets (6, 12, 18, 24), the number of alternatives (3 to 5), the number of attributes (4 to 7) and their levels (2 to 4) as well as the level range (narrow, base, wide). On the other hand, we make use of three complexity measures that capture the nature of the information in a DCE and how it is configured (DeShazo and Fermo 2002. Entropy, and cumulative entropy, respectively, were introduced into the discrete choice literature by Swait and Adamowicz (2001). They may be used as a measure for the similarity of alternatives in a choice set (Zhang and Adamowicz 2011). A measure which can be seen as a proxy for the number of trade-offs to be made by respondents is the number of attribute level changes within a choice set (DeShazo and Fermo 2002). In addition to the complexity of the DCE we investigate systematic preferences for the SQ resulting from different perceptions of the current situation, which was stated by respondents on a four point Likert scale prior to the DCE. It has been found that different perceptions of the current situation may explain the propensity to choose the SQ (Marsh et al. 2011), and that not accounting for respondents’ beliefs about the current situation can bias welfare estimates (Kataria et al. 2012).

The main contribution of this paper is that it investigates the SQ effect simultaneously accounting for different possible sources. In particular, this is the first attempt to analyze SQ choices within a DoD approach systematically varying five design dimensions. Being aware of the sources of the SQ effect may be especially important for the researcher in the design stage of the DCE where, among other things, the design dimensions, and the SQ have to be defined. Since not accounting for SQ effects may introduce significant bias into choice models and subsequent welfare estimates (Adamowicz et al. 1998, Scarpa et al. 2005), the results of this study may also have implications for the estimation of discrete choice models.

A modelling approach that has little been used to address the reasons of choosing the SQ, see Marsh et al. 2011 for an exceptions, is the Error Component Mixed Logit (MXL-EC) framework with alternative-specific constant of the status quo (ASCSQ). It allows the researcher to simultaneously identify the systematic and stochastic component of the SQ effect and has been observed to outperform other approaches (Scarpa et al. 2005). In this study we specify a series of MXL-EC models each time interacting the ASCSQ with different variables. In total we use four sets of covariates: The design dimensionality of the DCE, the complexity measuring that capture the nature and configuration of the information in a DCE, and the perceived SQ. We further control for socio-demographic characteristics of each respondent. In the next step we compare willingness to pay (WTP) estimates across model specifications.

We find the probability of choosing the SQ to increase with higher levels of entropy, and cumulative entropy as well as with more choice sets. The current situation is less likely to be chosen in designs with five alternatives, three attribute levels, and split samples with more than four attributes and a narrow level range. Effects in the same direction are observed when more attribute level changes are presented on a choice set. The propensity to stay at the SQ is also strongly related to participant’s perception of the current situation, and the socio-demographic variables gender and education. Highly significant error components for alternatives different from the SQ indicate that the hypothetical alternatives share a common error structure. Moreover, there appears to be a high degree of unobserved heterogeneity among the non-SQ options which is not captured through random attribute coefficients. We further find that WTP estimates do not differ much across model specifications.



Adamowicz, W., Boxall, P. and M. Williams, 1998. Stated Preference Approaches for Measuring Passive Use Values: Choice Experiments and Contingent Valuation. American Journal of Agricultural Economics, 80(1), 64–75.

DeShazo, J.R. and G. Fermo, 2002. Designing Choice Sets for Stated Preference Methods: The Effects of Complexity on Choice Consistency. Journal of Environmental Economics and Management, 44, 123–143.

Marsh, D., Mkware, L. and R. Scarpa, 2011. Do Respondents’ Perceptions of the Status Quo Matter in Non-Market Valuation with Choice Experiments? An Application to New Zealand Freshwater Streams. Sustainability 2011, 3(9), 1593–1615.

Hensher, D.A., 2004. Identifying the Influence of Stated Choice Design Dimensionality on Willingness to Pay for Travel Time Savings. Journal of Transport Economics and Policy 38: 425-446.

Kataria, M., Bateman, I., Christensen, T., Hasler, B., Hime, S., Ladenburg, J., Levin, G., Martinsen, L. and C. Nissen, 2012. Scenario Realism and Welfare Estimates in Choice Experiments – A Non-Market Valuation Study on the European Water Framework Directive. Journal of Environmental Management, 94(1), 25–33.

Samuelson, W. and R. Zeckhauser, 1988. Status Quo Bias in Decision Making. Journal of Risk and Uncertainty, 1, 7–59.

Scarpa, R., Ferrini, S. and K.G. Willis, 2005. Performance of Error Component Models for Status-Quo Affects in Choice Experiments. In Applications of Simulation Methods in Environmental and Resource Economics, ed. Scarpa, R. and A. Alberini. Dordrecht, Netherlands: Springer.

Swait, J. and W.L. Adamowicz, 2001. Choice Environment, Market Complexity, and Consumer Behavior: A Theoretical and Empirical Approach for Incorporating Decision Complexity into Models of Consumer Choice. Organizational Behavior and Human Decision Processes, 86(2), 141–167.

Zhang, J. and W.L., Adamowicz, 2011. Unraveling the Choice Format Effect: A Context-Dependent Random Utility Model, Land Economics, 87(4), 730–743.





Conference registration is required in order to view papers.