The paper includes a brief review of theoretical and empirical literature on the effort expended in making choices. It then describes an experiment, conducted in laboratory conditions with 123 subjects, in which subjects were presented with a series of road charging scenarios and were asked to estimate the charges payable, to indicate their confidence in that estimate and how difficult they had found it, and to indicate whether they would change their behaviour if the charges were introduced. The time taken by the subjects to answer each question was recorded and background data was collected on their personal characteristics and attitudes as well as on their own assessment of their decision-making style (need for cognition, need to evaluate, tolerance of ambiguity).

Analysis of the resulting data shows that the time taken to estimate the charge and decide on a response varied not only with the characteristics of the scenario (with more time taken to assess scenarios which were objectively more complicated) and with the order of presentation (indicating a learning, or fatigue, effect), but also with personal characteristics (notably age, gender, income, attitude to the policy inherent in the scenario, and self-reported decision-making style). Interestingly, although the time someone took to estimate a charge was not reflected in the degree of accuracy of the resulting estimate, it was significantly (and positively) related to the confidence they expressed in their estimate and to their assessment of the degree of difficulty they had experienced in understanding that scenario.

These findings, particularly on the existence of groups with very different levels of motivation to make careful assessment of costs and benefits, have important implications for our understanding of real-world decision-making, and hence for the way that we should collect choice data and model the decision-making process. They have particular implications for the interpretation of errors and for the appropriate segmentation of populations.