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.
S-1 Freight Trip Generation and Land Use The main objective of NCHRP Project 08-80/NCFRP Project 25 was to study the relation- ship between freight trip generation (FTG) and land use and â. . . to develop a handbook that provides improved freight trip generation rates, or equivalent metrics, for different land use characteristics related to freight facilities and commercial operations to better inform state and local decision-making.â As part of that quest, the research: consolidated the avail- able FTG models in an electronic database (available at http://transp.rpi.edu/~NCFRP25/ FTG-Database.rar) to assist practitioners interested in using these models; undertook an in depth examination of the key concepts to identify the most appropriate approaches to develop and apply FTG models; and, used data previously collected by the team to estimate establishment-level FTG models for a number of case studies. This process led to the identi- fication of a number of premises considered to be central to the development of FTG models able to satisfy the needs of both transportation planning and traffic impact analyses. The most important of these premises is the need to make a distinction between FTG, i.e., the generation of vehicle trips, and freight generation (FG), i.e., the generation of the cargo that is transported by the vehicle trips. FG is an expression of economic activity per- formed at a business establishment by which input materials are processed and transformed generating an output that, in most cases, is transported elsewhere for further processing, storage, distribution, or consumption. FTG, on the other hand, is the result of the logistic decisions concerning how best to transport the FG in terms of shipment size, frequency of deliveries, and the vehicle/mode used. Of great importance is the shipperâs ability to change shipment size to minimize total logistic costs, as it allows shippers and carriers to increase the cargo transported (the FG) without proportionally increasing the corresponding FTG. As a result, FTG cannot be universally assumed to be proportional to business size because large establishments could receive larger amounts of cargo without concomitant increases in FTG. This has major implications for FTG modeling, as standard practices implicitly assume proportionality between FTG and business size variables (e.g., square footage, employment). Another important premise is that the accuracy of FG/FTG analyses depends on a number of key factors: (1) the adequacy of the classification system used to group commercial estab- lishments in a set of standardized classes; (2) the ability of the measure of business size used to capture the intensity of FG/FTG; (3) the validity of the statistical technique used to estimate the model; and, (4) the correctness of the aggregation procedure used to estimate aggregate values (if required). In addition to these FG/FTG specific factors, other basic prin- ciples hold true: the better the quality of the data, the better the results; and, that disaggregate models (establishment level) are generally better than aggregate models (zonal level). In order to ensure proper understanding and use of the terms, brief descriptions are provided. A classification system is a systematic way to group individual entities into pre-defined groupings or classes with which they share common features. An example is a simple land S U M M A R Y
S-2 use classification that considers, for instance, the three basic classes of âresidential,â âcom- mercial,â and âindustrial.â A measure of business size is a variable that tries to capture the scale of the operation at the establishment level. Examples include the square footage of the establishment as well as the total number of employees working there. The statistical tech- nique is the process used to compute the parameters of the models. Although there is a wide range of approaches that could be used, the research found two techniques to be particu- larly useful: ordinary least squares (OLS) (regression analysis), and multiple classification analysis (MCA). The aggregation procedure is the technique used to obtain aggregate values of FG/FTG from the establishment level estimates produced by a disaggregate model. This routinely overlooked aspect is at the core of many of the problems reported by practitioners when producing FG/FTG forecasts. The analyses revealed a number of aspects of great relevance for modeling purposes, including the following: â¢ It is important to use land use classification systems that lead to internally homogeneous classes, in terms of the determinants and patterns of FG/FTG activity. The heart of the issue is that there is a wide range of land use classification systems that exhibit various degrees of ability to capture the FG/FTG propensity of the business establishments in their classes. At one end of the spectrum one could find land use classification systems that only consider aggregate land use classes (e.g., commercial, industrial) are likely to group together a disparate set of economic activities. In such a case, the ability of busi- ness size variables, such as square footage, to be predictors of FTG is undermined by the internal heterogeneity of the FTG patterns in the land use class. In essence, the higher the degree of internal homogeneity, the greater the ability of a business size variable to predict FG/FTG. This implies that if land use classes are defined so that they repre- sent a homogenous set of economic activities, the corresponding business size variables would have a better chance of being good predictors of FG and FTG. At the other end, one finds land use classification systems that are based on a rather comprehensive set of formal descriptors that consider all key dimensions. The Land-Based Classification Stan- dards (LBCS), for instance, classifies land use using five dimensions: the activity (taking place at the establishment); the function (type of enterprise being served); structure type (building characteristics); site development character (physical description of the land); and ownership (e.g., public or private). A unique feature of the LBCS is that its activity dimension contains classes that are defined at a fine level of detail. Such a way of char- acterizing the activity with this strategy is likely to support proper modeling of FG/FTG, as the resulting classes are expected to be relatively internally homogeneous. The same would occur if a formal industry (economic) classification system [e.g., Standard Indus- trial Codes (SIC) or North-American Industry Classification System, (NAICS)] is used to classify land use, as the underlying classes meet the condition of internal homogeneity. â¢ It is important to use, as predictors of FG/FTG, variables that correctly measure the intensity of FG/FTG activity. The research clearly showed that commonly used variables, such as square footage and employment, have significantly different levels of explanatory power. The reason is related to their inherent ability to capture the intensity of the FG/ FTG. As an example, three establishments of exactly the same square footage will pro- duce different amounts of FG and FTG depending on the amount and type of economic activity being performed, and whether or not the establishments are empty, lightly used, or heavily used. Thus, variables such as employment are likely to be better explanatory variables because they are likely to rise and fall in concert with the level of economic activity and FG/FTG. Alternatively, if variables such as square footage are used, they should be complemented with an additional parameter that represents the percent of capacity being
S-3 used (e.g., full production, minimum production). This would help mitigate the lack of ability of square footage to capture the intensity of the FG/FTG activity. â¢ It is important to use the aggregation procedure that corresponds to the underlying disaggregate FG/FTG model. The research conclusively showed that not using the correct aggregation procedure leads to significant errors in the estimation of FG/FTG. Most notably, the research revealed that the widely used process of obtaining aggregate esti- mates of FTG by multiplying total employment by an FTG rate per employment is only valid if the underlying model is one in which FTG is directly proportional to employ- ment. In all other cases, different aggregation procedures must be used; in cases where the FTG is a constant that does not depend on business size, aggregate estimates must be found by multiplying the number of establishments in that industry segment by the average FTG for the industry segment. Alternatively, if the disaggregate model includes a constant and a term that depends on employment, the correct way to do the aggregation is to multiply the constant by the number of establishments in the industry segment and add the result to the multiplication of the industry segmentâs total employment by the FTG rate per employee (the modelâs second term). Not following these procedures could lead to significant estimation errors. The premises and conjectures discussed herein were tested using cases studies. To this effect, the research used FG/FTG data from: 362 receivers of supplies in Manhattan and Brooklyn, 339 carrier companies in Northern New Jersey and New York, a furniture store chain in Midwestern states, and, supermarkets in the Puget Sound region and Manhattan. In the cases where the data were most complete, the team had access to establishment-level data, including: employment, location, size, revenue, line of business, some trip data (e.g., number of truck trips per day/week, shipment sizes), and land use information. Using the data, the research compared the performance of FG/FTG models based on: (1) Industrial classification systems (i.e., SIC and NAICS); (2) Land use classification systems [i.e., LBCS and New York City Zoning Resolution (NYCZR)]; (3) the statistical technique used (e.g., ordinary least squares, multiple classification analyses); (4) the aggregation procedure used; and (5) the business size variable used as predictors of FG/FTG. The case studies led to the following findings: â¢ The case studies confirmed the superiority of economic classification systems over standard land use classification systems. The research revealed that using economic classification systems as the foundation for the estimation of FG/FTG models is signifi- cantly better than using standard land use classification systems such as the NYCZR, or land use classification systems that can be applied nationally such as LBCS. In cases where these standard systems were used, the vast majority of business size variables were found not to be significant. In contrast, when the economic classification systems were used they tended to produce models that were statistically stronger than the ones obtained using the standard land use classification systems. The best results were found when an economic measure of business size, e.g., employment, was used in combination with an economic classification system (i.e., two-digit SIC codes, or three-digit NAICS codes). In fact, the models using the NAICS codes produced better vehicle trip production models, while the SIC models produced better vehicle trip attraction models. It is important to mention that the results concerning the LBCS are not entirely conclusive due to lack of variability in the data, which suggests the need to conduct additional research with a larger dataset. The team would expect that using LBCS will produce better models than using the standard land use classification systems (such as NYCZR) in terms of its ability to support FG/FTG modeling. Moreover, if the activity codes in the LBCS are made
S-4 consistent with economic classification systems (e.g., SIC, NAICS), one could expect even more improvements in performance. â¢ The case studies confirmed that proportionality between FTG and business size only happens in a minority of industry segments. The research revealed that: in 51% of industry segments, the FTG is constant as it does not depend on business size; in 31% of cases, the FTG model is a function of a constant and a rate that multiplies the estab- lishmentâs employment; and in the remaining 18% of cases, the FTG model is propor- tional to employment and a constant FTG rate. The fact that the most commonly used approach (the constant FTG rate per employee) is correct in only a minority of cases, should be a concern. â¢ The case studies suggest that the models estimated at the establishment level are transferable, though more testing is needed to reach solid conclusions. As part of the research, the models estimated with New York City data were applied to supermarkets in the Seattle region. The models were found to produce very good estimates of FTG. This is a very encouraging result, though larger testing is needed to reach definitive conclusions. â¢ The case studies suggest that the NCHRP Project 08-80/NCFRP Project 25 models outperform both the Institute of Transportation Engineers (ITE), and some industry segments of the Quick Freight Response Manual (QFRM). The models were compared to the ones in ITEâs Trip Generation Manual, and the Quick Response Freight Manual. The results show that the land use based models estimated in NCHRP Project 08-80/ NCFRP Project 25 produce more accurate FTG estimates than the ITE rates. When com- paring with the Quick Response Manual, results show that models for most industry sectors have a similar performance, with the exception of models estimated for the âbuilding materialâ industry, which perform significantly better. â¢ The case studies indicated that Multiple Classification Analysis (MCA) performed better than Ordinary Least Squares (OLS) models. In conducting the case studies, two alternative estimation techniques were used: OLS (regression analysis) and MCA. The results indicate that for those industries with FTG dependent on employment, MCA performs slightly better than OLS. This was the case for both economic and land use classification based models. Although the work completed has primarily focused on FTG, the findings have significant implications for both freight transportation planning and traffic impact analyses. During the second phase of the projectâwhich will use the Commodity Flow Survey (CFS)âthe research will focus on the estimation of FG models. Ultimately, the entire set of findings will be synthesized in a set of guidelines for FG/FTG modeling.