Stated preferences over job characteristics: A panel study

When making choices over jobs with different characteristics, what trade‐offs are decision‐makers willing to make? Such a question is difficult to address using typical household surveys that provide a limited amount of information on the attributes of the jobs. To address this question, a small but growing number of studies have turned to the use of stated preference experiments; but the extent to which stated choices by respondents reflect systematic trade‐offs across job characteristics remains an open question. We use two popular types of experiments (profile case best–worst scaling and multi‐profile case best–worst scaling) to elicit job preferences of nursing students and junior nurses in Australia. Each person participated in both types of experiments twice, within a span of about 15 months. Using a novel joint likelihood approach that links a decision‐maker's preferences across the two types of experiments and over time, we find that the decision‐makers make similar trade‐offs across job characteristics in both types of experiments and in both time periods, except for the trade‐off between salary and other attributes. The valuation of salary falls significantly over time relative to other job attributes for both types of experiments. Also, within each period, salary is less valued in the profile case compared to the more traditional multi‐profile case.