Being able to understand and disseminate information about the behavioural underpinnings of freight transportation demand requires at least three stages. The collection of data that are able to provide insight into these underpinnings; appropriate analysis of the data; and a method by which to communicate the findings.

One of the main reasons that the link between behavioural underpinnings has until recently been relatively unexplored is the difficulty in obtaining the necessary information. Understanding the behavioural underpinnings of freight transportation demand amounts to understanding shipper preferences for transportation services. Understanding these preferences requires understanding how shippers make their choices of carriers. This type of information is normally collected through surveys.

The problem, however, is that it can be difficult to get this information. The primary reason for this is that the transportation industry is very competitive. Moreover, freight data are generally private and companies can be reluctant to provide information to researchers that they think might compromise their competitive position. As a result, asking shippers about decisions they actually make (or asking them about their Revealed Preferences) can be very difficult.

One method by which these challenges have been overcome is through the use of Stated Choice (or Stated Preference) methods. Whereas revealed preference techniques ask respondents about their actual choices, Stated Choice techniques involve asking respondents to choose between hypothetical (albeit realistic) alternatives, designed to simulate the actual choice environment. The benefit of the Stated Choice approach is that it can overcome some of the drawbacks of revealed preference techniques.

In particular, the fact that respondents make choices between alternatives for which information has been provided means that the respondent does not have to reveal any information of a competitive nature that might discourage participation in the survey.

It goes without saying that the development of Stated Preference datasets, however, is quite onerous. Luckily, this project will take advantage of the fact that such a dataset has recently been developed, and is now available to be used to understand the behavioural underpinnings of freight transportation demand. In fact, this dataset was developed by the proponents of this project.

The dataset to be used in this project comes from a Stated Preference survey of shippers in the Quebec City to Windsor Corridor. The survey obtained responses from close to 400 shippers (manufacturers, wholesalers, retailers and third party logistics companies (3PLs)) in this important corridor about their choice of carriers. It produced a very rich data set with over 7,000 shipper choices of carriers. The dataset includes a great deal of information on the shippers themselves, the characteristics of the carriers chosen, as well as shipment types for which these choices were made. Of particular importance, the survey was designed explicitly to understand shipper preferences for the use of rail for the transport of their shipments.

As a result, this project focuses on: the analysis of this dataset, and the dissemination of the results of this analysis in a format that can be readily used by transportation planners.

The analysis of the data ranges from the relatively straightforward to the somewhat more analytically complicated. It includes basic analyses of the results including the number of respondents, the number of responses as well as the number of responses of different types (shipments going by truck or by rail). It further includes an analysis of the types of shipments that were considered, the overall sample population and how the realized sample compared with the population.

The analysis includes the number rail vs. road shipments tabulated across different response subsets, i.e. by shipment type (high value vs. low value shipments, fragile vs. non-fragile, etc.), as well as by shipper type (e.g. 3PL vs. manufacturer). The idea is to understand the effect that these different factors have on carrier choice, and in particular on the likelihood that rail is chosen as a shipping option. Apart from providing information on the factors that affect carrier choice, these cross-tabulations could be used from a planning perspective to develop estimates for the proportion of shipments likely to be shipped by different modes.

We also present a more involved econometric discrete choice analysis. This analysis involves the development of a series of econometric models that predict freight mode share for intercity and interregional freight movements. This analysis should be seen as an extension and more refined version of the results presented in the cross-tabulation analyses. As with the cross-tabulations, the idea is to produce models that could be used by transportation planners to estimate freight mode share between cities and regions of the country. The overall purpose of this is to allow planners to estimate infrastructure requirements for future transportation demand.

The most important aspect of this project is the dissemination stage. This dataset and the resulting analysis is only useful to the extent that it is actually used by planners and decision makers. As a result, all of this analysis has been made to be accessible, easily interpreted and its resulting models easily applied.