Freight Trip Generation and Land Use (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.

76 This innovation plan introduces a set of comprehensive and practical improvements to FTG modeling practices. There are three interlocking components: the land use and freight systems; the FTG models; and data that planners use or have available to understand the freight impacts of land use decisions. Future improvements to FTG modeling include the greater use of economic models based on employment. These models can capture the underlying activity of freight, as well as the use of correct spatial aggregation procedures when estimating disaggregate models. Improvements related to data include the use of a standardized data collection instrument which will homogenize the data collection process, and the need to further explore the use of CFS micro-data. More research is needed to quantify service trips generated by commercial establishments, as not much is known about them. In the case of land use based FG/FTG model estimation, the use of LBCS will strengthen the connection between land use definitions and FG/FTG, which will be an improvement on the use of typical land use classification systems that may not capture the underlying economic activity. Finally, along with the previous improvements, the team encourages the use of the database created as a part of NCFRP Project 25, and provides it as a platform wherein the transporta- tion community can locate information—whether models or literature—on FG/FTG. Users will also be able to improve on the database by editing existing models and adding new ones. The database will also allow for the sharing of data collected with the use of the standardized instrument; this will lead to significant advancements in the area of FG and FTG analysis. Enhance Freight Trip Generation (FTG) Models Database The database, created as a part of NCFRP Project 25, houses a comprehensive library of FTG models and publica- tions. This database constitutes a living document, an envi- ronment that practitioners and researchers can consult for easy access to the models in the literature, to enhance both practice and research on FG and FTG modeling and analy- sis. An added recommendation would be to allow the com- munity of practitioners and researchers to enhance the data- base by identifying gaps in the current models and helping to improve them by editing the existing models or adding new models. However, this would require a moderator to manage the information added to the database for quality assurance. Conduct Research on Service Trips Commercial trips are comprised of two types of trips, freight and service. While NCFRP Project 25 focused on freight trips, there is a need for exploration in the area of ser- vice trips. Service trips are an increasingly important compo- nent of the traffic generated by commercial establishments. Surprisingly, not much is known about how many service trips are produced by commercial establishments in urban areas. These trips are produced by households as well as businesses, freight-related and non-freight-related establish- ments alike. Currently, these trips are not being accounted for in trip generation models. There needs to be some under- standing about these trips, as this knowledge may have sig- nificant planning implications. Therefore, more research should be done in the area of service trips to/from commer- cial establishments so that the findings are in the database. One possibility is to include service trips as a part of FTG models, which is done implicitly when traffic data is used to estimate FTG models. Use Standardized Instruments for FTG Data Collection The findings of the literature review performed in Task 3 of NCFRP Project 25 revealed that there is a need for better primary freight data and to remove inconsistencies within C h a p t e r 7 Innovation Plan

77 collected data. Use of a standardized instrument (survey) in the data collection process will address both of these issues, by providing an instrument that will achieve better consis- tency within the data. The survey design should capture the key aspects of conceptual validity, practicality, and accuracy. Transportation professionals need access to a basic survey design that they can tailor to their specific needs. The uni- formity of the data will facilitate data pooling, thus enabling practitioners and researchers to add data to the centralized database. Resulting models from use of this data in FG/FTG analysis may also be uploaded to the database. The standardized survey instrument along with coding instructions on how to use it will provide more consistent data for model estimation that will allow better compari- sons between different cities. Further instructions on how to apply the use of LBCS in FTG analysis will further strengthen the transferability of the models. There also needs to be con- sistency in vehicle classes and time periods of data collection to achieve comparability of results. Therefore, a standard- ized instrument should be promoted for use in the data col- lection process, along with the use of LBCS as the preferred land use classification system in FG/FTG. Use Commodity Flow Survey (CFS) Micro-Data for Freight Generation (FG) Analysis The CFS is the most important source of freight demand data in the United States, and one of the oldest data collec- tion programs in transportation. The CFS collects data on the movement of goods in the 50 states and the District of Columbia. It provides information on commodities shipped, their value, weight, and mode of transportation origins and destinations of shipments. The main focus is on shipments sent by domestic establishments in: manufacturing, whole- sale, mining, and selected other industries (Fowler 2001; Bureau of Transportation Statistics 2008). As a result, the micro-data file contains about 2.5 million individual records collected for 100,000 establishments. The CFS data does have some limitations: (1) it only col- lects data about the outflow of cargo from establishments; and (2) the data collected is for freight generation only, not for FTG. Nevertheless, these limitations can be overcome by developing procedures to estimate the inflow of cargo as a function of the outflow, and also by estimating FTG as a function of the FG. Accordingly, CFS micro-data should be promoted for use in FG analysis. Further explorations that may be carried out include the use of CFS data to convert FG models to FTG models. The depth of the CFS data provides an opportunity to estimate FTG models at various levels of geographic detail, reflecting regional differences which will assist in the mapping of industry sector models to different land use classifications. Use Economic Models Based on Industrial Classification Systems Freight generation (FG) and FTG result from derived demand, they are a result of economic transactions involv- ing cargo. Therefore, it is important to account for the eco- nomic activity performed by the establishments—both those that produce and consume freight—as different economic activities may have different FTG patterns. This is typically achieved through the use of industrial classification systems in FTG modeling. The two most important industrial classifica- tion systems that are used in FTG modeling are the SIC system and the NAICS. [For more information on these systems see (Pearce, 1957) and (U.S. Census Bureau, 2010c).] FTG analysis from the case studies showed that the use of industrial classification systems estimated slightly more efficient models than those based on land use classification systems, because land use can only serve as a proxy for the underlying economic activity and industrial classification sys- tems. Results also revealed that for estimating freight produc- tion trips, NAICS generated better models than those derived using SIC. In terms of freight attraction, on the other hand, overall SIC produced more efficient models than NAICS, but only marginally so. Therefore, the use of economic models should be based on employment to create FTG models, and in doing so NAICS should be promoted as the preferred clas- sification systems to use in FTG analysis. Ensure a Better Connection Between Land Use Definitions and FG/FTG Connections between freight and land use consider two separate aspects: (1) how land use at the establishment level influences FTG; and (2) how freight activity and land use interact with each other at the system level (Holguín-Veras et al. 2011). As mentioned previously, it is necessary to account for the underlying economic activity when generat- ing freight models. In most cases, land use is only a constraint to the production process, not an input factor; therefore, at most, land use is a proxy to the underlying economic activ- ity being conducted by the businesses or more typically, an aggregation of economic sectors. As a result, the adequacy of land use attributes as explanatory variables depends on how well the land use class matches the FTG patterns of the indus- try segments that have been included under it. In cases where there is a good match, land use variables could be good pre- dictors. In contrast, if the economic/land use class groups use disparate economic activities, independent variables cannot be expected to work well (Holguín-Veras et al. 2011). Estimation of consistent FTG models based on land use will be valuable to transportation researchers and practitioners and urban planners. Those in charge of zoning regulations

78 will particularly benefit from the availability of information related to FTG per land use. FTG estimates based on land use are useful in assessing the FTG effects of planned develop- ments, where the size and footprint are given. There are various local land use classification systems such as the City of New York Zoning Resolution, which was used in NCFRP Project 25. It categorizes land uses for a specific region. These local land use systems need to be analyzed to assess their capabilities in accounting for freight activity because some land use systems are structured in a way that fully captures freight activity. There is also the question of transferability, how well models generated using a specific local land use clas- sification system are able to be applied to another geographic region. Several studies on FTG have shown that FTG estimates based on land use can produce consistent results. The LBCS is a national land use system that classifies land use based on the following dimensions: activity, func- tion, structure type, site development character, and owner- ship, according to the APA (see http://www.planning.org/ lbcs/standards). The flexibility of these classification systems enables FTG modeling to account for underlying freight activ- ity and also addresses the issue of transferability of models. The flexibility of the LBCS makes it adaptable for any city, and provides a uniform classification system that will estimate models that may be transferable to other locations. Therefore, LBCS should be promoted as the lands use clas- sification system that should be employed in FG/FTG analysis. Use of Appropriate Aggregation Procedures Spatial aggregation is the process by which estimates at the zonal level are created from a disaggregate model. Although aggregation procedure should be consistent with the math- ematical structure of the model at the core of the aggregation, this step is often overlooked. Unlike passenger transportation, where this a minor issue, the variations in FG/FTG patterns require that this issue be addressed on a case-by-case basis. The aggregation formulas for three key cases are listed below. Though employment is being used in the cases below, the results apply to any other variable as long as the structure is similar (Holguín-Veras et al. 2011). To understand the following cases, the following formu- lation is essential. The aggregated FTG, F, is equal to the summation of the FTGs for the different establishments: F fi i n = = ∑ 1 8( ) where F = Aggregate freight trip generation FTG Ei = employment at establishment i fi = FTG for establishment i The first case addresses when the FTG is a function of employment only. Hence, FTG for the establishment is pro- portional to employment (FTG rate per employee). The for- mulation for this is shown in Equation 9, where b is a constant FTG rate per employee. f Ei i= β ( )9 Substituting Equation 9 in 8 and taking b out of the sum- mation will result in the formulation shown in equation 10 for determining the aggregate FTG (F). F E E Ei i i n i n = = = == ∑∑β β β * ( ) 11 10 Therefore, in cases where the underlying FTG pattern is directly proportional to employment, total FTG is obtained by the product of the FTG rate (b) and total employment (E*). This estimation process is commonly employed by practitioners in determining zonal estimates of FTG. The second case formulation addresses the situation when FTG (fi) is a constant per establishment. The mathematical for- mulation is expressed in Equation 11, where a is a constant. fi = α ( )11 Substitution of Equation 11 in Equation 9 and taking a out of the summation results in the formulation shown in Equation 12. F n i n = = = ∑α α 1 12( ) Therefore, in cases where the FTG at the establishment level is constant, the correct estimation process for aggregate FTG is the product of the unit FTG (a) and the number of establishments (n). The final case is when FTG at the establishment level is determined by a constant and a term that is dependent on employment. The mathematical formulation is expressed in Equation 13, where a is a constant and b is a constant depen- dent on employment. f Ei i= +α β ( )13 Substitution of Equation 13 in to Equation 8 results in the formulation expressed in Equation 14. F E n E n Ei i i n i n = +( ) = + = + == ∑∑ α β α β α β * ( ) 11 14 Using this method, the total FTG will be obtained from the product of the total number of establishments and the con- stant (a) added to the product of the total employment and the FTG rate (b). As can be seen this case is a combination of

79 the two previous cases (Holguín-Veras et al. 2011). Table 64a shows the three cases and their correct spatial aggregation procedures. Therefore, the correct spatial aggregation procedure should be used in FG/FTG analysis. Use of Synthetic Correction Methodology to Improve Accuracy of Existing Models The objective of the synthetic correction methodology is to correct existing models to account for the differences in FTG patterns for small establishments and large establish- ments. In fact, the empirical evidence from the FTG models estimated with establishment-based data indicates that FTG rates depend on business size. In essence, small establish- ments tend to generate proportionally more trips than large establishments. This leads to a situation in which a constant trip rate underestimates the FTG of small establishments, and overestimates those for large businesses. This poses a prob- lem, because several FTG models reported in the literature are in the form of constant trip rates. The synthetic correction methodology takes advantage of the mathematical properties of OLS (regression) mod- els and of the case studies developed in Task 11. In essence, this methodology: (1) uses the intercept of the models devel- oped in the case studies to estimate the number of trips pro- duced by small business in the same industry sector; and (2) computes a new slope to account for the effect of employ- ment on large businesses. The key element is that, although an approximation, even a suboptimal assumption of the intercept is bound to perform better than the constant trip rate model. Case No. Model Type Aggregation Procedures 1 FTG rate per employee 2 FTG constant per establishment 3 FTG is a combination of a constant and a term that depends on employment level * 11 )( EnEnEF n i i n i i βαβαβα +=+=+= ∑∑ == * 11 EEEF n i i n i i βββ === ∑∑ == αα nF n i ==∑ =1 F = Aggregate freight trip generation FTG Ei = Employment at establishment i E* = Total employment β = Constant FTG rate per employee α = Constant n = Number of establishments Table 64a. Spatial aggregation procedures for disaggregated models.