Modelling welfare estimates in discrete choice experiments for seaweed-based renewable energy

Area of Science:

• Environmental Economics
• Energy Policy
• Econometrics

Background:

• Discrete choice experiments (DCEs) are crucial for valuing environmental goods and services, particularly for renewable energy programs.
• Commonly used DCE models include the random parameter logit (RPL) with correlated parameters, RPL with uncorrelated parameters, and the hybrid choice model.
• Understanding the implications of model choice is vital for reliable preference measurement.

Purpose of the Study:

• To evaluate the gains and losses for researchers using three distinct models for DCE data analysis.
• To assess the suitability of these models for preference measurement in renewable energy contexts, specifically seaweed for biogas production.
• To highlight the importance of testing model assumptions before application in environmental valuation.

Main Methods:

• Analysis of three datasets using discrete choice experiment data.
• Comparison of three models: random parameter logit (RPL) with correlated parameters, RPL with uncorrelated parameters, and the hybrid choice model.
• Focus on median willingness-to-pay (WTP) values and underlying model assumptions.

Main Results:

• All three models can converge to similar median willingness-to-pay (WTP) values.
• Despite median WTP similarity, the models are not interchangeable due to differing assumptions.
• Standard sample sizes in environmental valuation are often insufficient to detect differences between these models.

Conclusions:

• Researchers must not treat the random parameter logit (RPL) and hybrid choice models as indistinguishable in discrete choice experiment analysis.
• Model assumptions underlying each discrete choice model must be explicitly tested prior to application.
• The inability of standard sample sizes to differentiate model outcomes does not justify their indiscriminate use.