Identification and Evaluation of Freight Demand Factors (2012)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

5 EXECUTIVE SUMMARY Many national freight forecasts have been predicting continued increases in the demand for freight transportation, and it has been generally accepted that the United States will have to invest many billions of dollars in new infrastructure. Many approaches to estimating freight growth have been used over time, for example Gross Domestic Product (GDP) or population growth. O ver the past 30 years, various structural changes in the United States and even the world’s economy require new approaches to understanding the key factors that influence freight demand. The deregulation of most freight transportation markets in the early 1980s led to a new era of flexibility in the arrangements that providers could offer to their customers. A long with consolidation by providers, real prices decreased and demand increased as shippers, suppliers and retailers began to remake their production and distribution systems to take advantage of lower transportation costs and improved service. Economic expansion led to growth in freight transportation demand that was not necessarily evenly distributed and did not always move smoothly with GDP. Free trade agreements and the revolution of intermodal transportation led to longer supply chains and a shift in the utilization of transportation in the global marketplace. The primary objective of this research effort was “to describe and analyze various types of demand drivers that shape the volume and movement of freight in North America.” The goal was to identify a set of regularly generated, well-documented, easily-obtainable variables with high statistical significance in explaining the variability of different kinds of freight demand. Moreover, a contribution would be made if the research could identify one or more influencing variables that predicted changes in transportation demand during a subsequent period. By determining the significance and usefulness of these predictor variables, infrastructure investors and capital planners would be better able to predict freight transportation demands and make better decisions. Although it is tempting to extend the relationship between exogenous economic and demographic variables to freight transportation demand on a regional basis, this effort was directed towards a national level. It was concluded that it was important to lay a firm foundation with regards to modeling freight demand using well-regarded, consistent and accessible economic data alongside reliable summaries of freight activity before trying to break the analysis down into sub-regions. In essence, the data as well as the desire for a strong foundation push the analysis toward national form. W hether the same relationships between the independent variables and freight demand that exist on a countrywide level also exist on a regional or state/local level is not demonstrated in this research, although in many cases it is an intuitively attractive hypothesis worthy of testing if appropriate data can be secured. The research took the following steps as part of this analysis: • Beginning with a review of NCFRP-01 “Review and Analysis of Freight Transportation Markets and Relationships,” more than a hundred different studies and papers that examined the influencers of freight were reviewed so that the selection of independent variables would

6 reflect state-of-the-industry and academic thinking regarding the question “What generates transportation?” • It was determined that only those data that were readily and freely obtainable from 1980 through 2007 – beginning with the transitional event of US freight industry deregulation - would be considered. This assured a relatively long time series and results that others could replicate and build upon. • Nine key measures of transportation demand, via trucking, rail and inland waterways -- 95% of domestic freight tonnage -- were summarized over 28 years sampled. • A variety of statistical analyses – correlation, regression (with lagged/leading and dummy “shock” variables), and Principal Component Analysis – were used to derive multiple models that provided very high R2 and low standard errors in respect to these measures. • “Back casting” was performed to test and validate the quality of several of these models. In brief, production factors such as GDP and Industrial Production measures are good predictors of “pure” transportation demand – namely total tonnage or volume of goods transported. Consumption factors such as Housing Starts and Imports, because of their associated lengths-of- haul, provide good predictive value to Ton-Mileage transported. Trucking is more sensitive to consumer factors while rail is affected by a broader set of economic data including industrial production. Water freight grows with Total Capacity Utilization, Grain and Coal production, while water ton-miles have decreased as railroads have become more competitive at capturing share of long-hauls of heavy freight (such as export grain). Exogenous political factors like the deregulation of highway/railroad transportation during the early 1980’s and the North American Free Trade Agreement (NAFTA) in 1993 led to significant, one-time increases in freight as well. Besides the results of the models, major conclusions and takeaways include: • The availability of quality data on economic activities and freight transportation demand is an inherent limitation that guides this analysis. • While several of the economic variables or measures provide much of the explanatory variation regarding freight demand, other factors that might appear to be less significant serve to improve the various models. • Early warning factors were identified by testing how well fluctuations in independent factors’ values in one time period predicted changes in freight demand during the following time period by “lagging” the variables. Both the Purchasing Manager’s Index and the Number of Households had significant effects on subsequent freight transportation demands, with Fuel Prices also having some effect. Independent of the other values, NAFTA led to a substantial increase in freight transportation as well. • Principal Component Analysis provides a potentially helpful method to combine the explanatory benefit of multiple, similar independent economic factors on resulting freight transportation demand. While it is somewhat less intuitive and difficult to deconstruct, the method has been proven in other areas and appears to have some value at estimating freight demand from a large set of correlated independent factors.

7 • The backcasting analysis indicates strength and robustness of the test models as the calculated results shows good trending vs. the actual measures of economic data. A greater availability of quarterly data would have provided a richer detail to this end. • The comparison of correlations and the effect of NAFTA and some of the other variables indicates that the relation of freight transportation demand to exogenous factors is not static. Since 1980, t here have been significant shifts in the way the same economic and demographic data affects demand for freight transportation. Future changes in these relationships as economies and countries transform are expected. Ideas for additional research include extending the data capture through the 2008-2011 time period to determine how well the relationships between influencing variables and freight demand persist during and past the recent economic downturn. Also, if sufficient freight demand measures and independent economic variables exist on a multi-state regional basis, would similar analysis confirm that the same relationships that were found on a national level still apply?