In this research, as in the aforementioned papers, WTP measures are found by taking ratios of marginal utility parameters from discrete choice models. These parameters are conditional on the choices made by respondents (that is, the estimates represent individual-specific behaviour). To illustrate, consider a simple model resulting in estimated preference parameters of 0.5 for transit time (measured in utils per minute) and 0.2 for travel cost (measured in utils per dollar). By taking the ratio of the former estimate to the latter, one identifies a ratio of 2.5 dollars per minute of transit time, or US$150 per hour of transit time. In this study, by dividing preference estimates for transit time, reliability and frequency (measured in utils per day of transit time, utils per percentage point of reliability and utils per trip per week, respectively) by preference estimates for the freight rate (measured in utils per dollar), we obtain WTP measures for each attribute (that is, dollars per day of transit time, dollars per percentage point of reliability and dollars per trip per week). These estimates, reported and discussed in the next section, provide critical insight into the market’s perceived valuation for level-of-service characteristics that could be influenced by changes in the cost, effectiveness and availability of services under carbon pricing and the availability of short sea shipping services. This is of particular relevance given both the unobserved nature of these sensitivities in the current market (that is, there may be no direct purchase option available for marginal changes in levels-of-service) and the as yet unobserved preference structure in the market under carbon pricing or the presence of short sea shipping alternatives.