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.
13 2.1 Overview This chapter reviews the literature that covers both the fundamental concepts associated with, and the methods of assessing, economic impacts of supply chain disruptions. A general framework reflecting a hierarchy of possible approaches that are progressively more complex is used to organize the literature review. As shown in Table 2-1, there are two broad classes of tools that can be used to model economic impacts of a transportation network disruptionâ supply chain response models and economic impact models. These may be regarded as sequentially linked. First, supply chain response models consider the impact of a disruption on supply chains (e.g., is cargo diverted to another route? is it held back for a time? does sourcing adjust?). Following this, and based on how supply chains are thought to adjust, economic impact methodologies help to assess the impact of the disruption on the economy. The economic modeling component essentially asks: if the supply chain adjusts in a certain way, what are the economic losses associated with these adjustments? This may involve estimating the sum total of impacts on businesses, jobs, households, and people of a disruption at various scales, including local, regional, national, and even international levels. 2.1.1 Supply Chain Response Models There are several common approaches or categories of supply chain response models described in the literature. These include ⢠Network-based models â Simple cargo diversion models that are often based on a least-cost path algorithm of vary- ing degrees of complexity â Freight network simulation models ⢠Industry supply chain response models â Business supply chain optimization models â Dynamic supply chain simulation models Network-Based Models Simple Cargo Diversion Models: Diversion models are often quite simple and can be based on the assumption that, in the case of a transportation network disruption, cargo is diverted to the least-cost alternative route. This diversion leads to some direct impacts in terms of increased transportation costs (e.g., fuel, operator salaries, operations and maintenance, etc.) and certain indirect impacts (such as increased inventory costs imposed by the relative uncertainty of deliveries through the detour, lost shipper profit, etc.). The sum total of these costs can be C h a p t e r 2 Methodologies to Measure Direct and Indirect Economic Impacts of Disruptions to the Goods Movement System
14 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System considered the immediate, or first order, economic impact of the disruption. Inter-industry linkages or residual economic and societal dislocations may or may not be considered in these models, depending on the economic impact tool used. These models tend to take a short-term view of disruption impacts. They are relatively less complex and less data or skill intensive and hence can be easier to use than full supply chain modeling tools. An example of this type of model includes the Freidman et al. (2006) Disruption Impact Estimating Tool-Transportation (DIETT) Model. Although diversion models tend to be simple when employed at a large scale of analysis, they can be complex when applied to specific transportation facilities. For example, a number of very detailed port disruption models simulate potential operational (and even policy) responses to disruptions to the port itself, or to port intermodal connections. These models tend to be operational simulations of greater or lesser complexity. Examples include IOCGâs Maritime Security Risk Analysis Model (MSRAM), which is capable of performing scenario analysis as well as assessing the impacts of a disruption event. Several studies were found in the literature where no formal or rigorous attempt was made to actually model diversions based on cost parameters, but rather assumptions were put forth about what might happen to goods movement in the face of a disruption, and these were tested as alternative cases. See, for example, Arnold et al. (2006). Freight Network Simulation Models: These models are complicated versions of simple diversion models. Generally GIS-based platforms that replicate the complex network of routes through which freight flows occur, they either represent movement on a single mode (such as trucks over the road network) or over multiple modes. As with the simple cargo diversion mod- els, a least-cost (including time cost) or shortest route approach is applied to analyze freight flows in various scenarios. The models usually have the ability to analyze scenarios in which certain links or nodes in a network are rendered dysfunctional. Examples of this type of modeling tool include the Oak Ridge National Laboratoryâs (ORNLâs) Transportation Routing Analysis Geographic System (TRAGIS), which has the ability to conduct both single-route and multiple-route analysis. Other commercially available network simulation Table 2-1. Classification of available disruption impact assessment methodologies. Increasing Levels of Complexity SUPPLY CHAIN RESPONSE MODELS Network-Based Models Industry Supply Chain Models Assume freight is diverted and estimate transport cost and inventory value impacts. Includes: -Simple Cargo Diversion Models -Freight Network Simulation Models Are used to optimize business operations and address industry choices/decisions regarding sourcing, inventory levels, and route choice. Includes: -Business Supply Chain Optimization Models -Dynamic Supply Chain Simulation Models ECONOMIC IMPACT METHODOLOGIES Static/Input-Output- Based Models Dynamic Economic Simulation Models Assume declines in industry final demand and calculate direct, indirect, and induced impacts across all industries Assume changes in supply, demand, output, prices, or other direct economic impacts, and simulates overall economic impact via dynamic modeling
Methodologies to Measure Direct and Indirect economic Impacts of Disruptions to the Goods Movement System 15 models such as those included in the TransCAD GIS-based modeling software and VISUM also fall in this category of model. Industry Supply Chain Models Business Supply Chain Optimization Models: These models aim at optimizing business operations including the flow of goods (raw materials, intermediate and finished products) through a specific industry supply chain. Business and industry decisions regarding sourcing, inventory levels, and the transportation network or route choice form key components of the supply chain and are all vulnerable to impacts of major disruptions. These models help to re-examine key business decisions as the business situation changes. Port disruption models such as the Port Disruption Recovery Model (PDRM) are an example of this category of models. Wilson (2007), Lewis et al. (2006), and Dauelsberg and Outkin (2005) have demonstrated applications of these models to hypothetical cases of supply chain disruption. Dynamic Supply Chain Simulation Models: These models aim to capture supply chain responses dynamically and allow for ongoing assessment of raw material sources, factory locations and processes, distribution centers, transportation links, outsourcing, inventory and related costs, and constraints. These new modeling tools can be categorized as complete dynamic supply chain models. Some of these models are highly complex formulations that simulate supply chain networks spatially as well as temporally. Nagurney et al. (2002), for example, modeled the physical, information, and financial networks involved in commodity markets. These models are generally very demanding in terms of data requirements and skill levels of the modelers and are as yet in the early stages of development as far as their use in common practice. 2.1.2 Economic Impact Models Economic impact models can be classified into the following two broad categories: ⢠Static/Input-Output (I-O) Models ⢠Dynamic Economic Simulation Models Static/Input-Output Models These are models of varying complexity that focus on transport costs (the monetary value of time and other resource costs). The most widely used tools are I-O modeling tools. Within the context of disruption modeling, I-O modeling tools can be applied primarily to analyze the following two situations: ⢠Economic damages to freight transportation infrastructureâThese models focus on the costs to transportation facilities themselves and the damages incurred and can also include the eco- nomic costs of mitigation (supply chain disruption response measures) going into the future. A good example of such a model is the MARAD Port Economic Impact Kit, which focuses on the costs incurred directly within a port (e.g., when dock workers are laid off), and the spinoff impacts to industries and enterprises that supply goods and services to the port. ⢠Economic damages associated with supply chain disruptions and cargo diversions or bottle- necks that disrupt freight flows. I-O models capture the inter-industry linkages of a disruption event on a regional economy by using matrices that relate the outputs of one industry to inputs of all other industries. That is, the goods and services demanded by one industry constructed as inputs in the form of goods and services provided by other industries. Multipliers (employment, output, earnings, and value added), which are derived from these inter-industry linkages, are the key input into estimating long-term economic growth and development. Standard economic multipliers produced by input- output models estimate both direct impacts and two kinds of secondary impacts resulting from
16 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System these direct changes to an economyâindirect impacts and induced impacts. Direct changes to an economy usually are represented by employment, sales, or purchases (spending), which would result from an increase/decrease in final demand for a given industry sector or for households. Indirect impacts, sometimes referred to as backward linkages, result from the intermediate purchases necessary to operate a business. To the extent that local firms buy more/less from local suppliers, the indirect impact will be larger/smaller. Induced effects, sometimes referred to as forward linkages, stem from the re-spending of household earnings by workers who benefit from direct and indirect activity. In other words, if a new firm is attracted to the local area, the employees of that firm will spend some proportion of their earnings at local shops, restaurants, etc. Ham et al. (2005), for example, demonstrates the estimation of interregional impacts of major transportation disruptions using such an I-O approach. I-O models may be especially effective when there is a zonal-based freight network flow or travel demand model that can capture the change in freight flows on the network due to a disruption. In the absence of such a demand model there are âshort cutâ methods that can be used to simulate the disruptionâs effect on freight flows. Dynamic Economic Simulation Models These models are statistically estimated, simultaneous equation representations of the aggregate workings of an economy. They have forecasting capabilities that can be helpful in determining the potential impacts of a future event or in distinguishing the actual activity of an economy from what it would have been like in the absence of a shock. Econometric models have their own established set of criteria for model validation that are more rigorous than those for I-O models. Uncertainty is inherently incorporated in the models by virtue of their stochastic estimation. The downside of these models is their high level of complexity and user skill requirement. If available as off-the-shelf software, most aspects of the model, underlying assumptions as well as equations, are generally opaque to the user. Furthermore, these models require historical time series data and tend to be data intensive. Examples of dynamic simulation modeling tools that have been (or could be used) for disruption analysis include the Regional Economic Model Inc. (REMI) and University of Marylandâs Long-term Inter-industry Forecasting Tool (LIFT). Studies that have utilized this approach include Rose et al. (2007), Tsuchiya et al. (2007), and Arnold et al. (2006). 2.2 Research on Economic Impacts of Disruptions to the Goods Movement System This section reviews papers on the economic analysis of supply chain disruptions, primarily viewed from a conceptual perspective. The research represented in these papers provides some key concepts regarding supply chain disruptions that helped form the theoretical basis for the development of an impact assessment methodology. Rose (2009) presents a schematic framework for assessing direct impacts (such as property dam- age, site-specific business disruption) and extended impacts (supply chain interdependencies among different types of infrastructure) of major disruptions such as a terrorist attack (see Figure 2-1). The framework can be used for identifying and classifying direct, microeconomic, and macro- economic impacts to form a basis for disaster loss estimation. The paper emphasizes the concept of resilience, broken down into micro resilience (i.e., substitution, conservation, rescheduling) and macro resilience (i.e., price system and market strengthening). The paper also contends that behavioral linkages (e.g., fear of flying after an event like 9/11) or social amplification of risk is a major driver of the long-term impacts of disaster-occasioned disruptions that are not often quantified when estimating the economic impacts of such events.
Methodologies to Measure Direct and Indirect Economic Impacts of Disruptions to the Goods Movement System 17 The paper proposes an operational measure of direct economic resilience (DER), defined as the extent to which the estimated direct output reduction deviates from the likely maximum potential reduction given an external shock. Although there are several economic models that are used to estimate economic impacts of disruptions (such as I-O, computable general equilibrium, macro-econometric models), Rose contends that these estimates will likely not vary by more than 50 percent between methods. Resilience and behavioral linkages can, however, result in 10-fold differences in loss estimates. Benefit-cost analysis (by governments and businesses alike) should be broadened to also consider both mitigation (pre-event) and resilience (post-event) options. TranSystems (2008) proposes a framework to develop a tool to analyze potential transportation security incidents (TSI) and the resulting effects. The tool aims to address the interdependencies among port, terminal, vessel, rail line, and local and regional cargo flows. The tool provides a method of assessing service impacts (capacity, cost, time, etc.) through an operatIonal framework that dynamically mimics multiple freight flows across and through various diverted transportation modes, corridors, and facilities. Chopra et al. (2007) demonstrated the importance of decoupling recurrent supply risk and disruption supply risk. In terms of supply chains, disruptions are defined as interruption of supply due to an unforeseen circumstance, whereas recurrent supply risks are delays resulting from inherent (in the supply chain) or recurrent uncertainties. This classification of risk as recurrent or disruption risk can help inform the decision regarding an appropriate supplier. Bundling together two types of uncertainties would potentially underutilize a reliable source while over- utilizing a cheaper, but less-reliable source. The paper concludes that if most of the supply risk comes from recurrent uncertainty, then increasing quantity from a cheaper, but less reliable source is a better mitigation strategy. However, if most of the supply risk growth comes from an increase in disruption probability, then the firm should order more from a more reliable source and less from a cheaper, but less reliable one. McKinnon (2006) presented the result of a macro-level assessment of the impact of a failure of the United Kingdomâs road transport system due to the effects of the 2000 fuel protest and previous freight road crises on major supply chains. A qualitative extrapolation of the impacts relied on telephone interviews and on expert judgment from brainstorming sessions. The research maps out a timeline of the âcreepingâ effects of a hypothetical case of a truck stoppage in an entire country. Both immediate, defined as a 5-day impact, and long-term impacts are discussed in some detail. Source: Rose (2009). Figure 2-1. Framework for assessing direct and extended impacts of disruptions.
18 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System McKinnon concludes that much of the risk management research focuses on vulnerabilities of individual entities, such as companies or supply chains and the impact of localized disasters such as fire, bad weather, etc., but overlooks the wider effects of systemic collapse/catastrophic failure of critical infrastructure systems. Standard risk mitigation measures such as increasing safety stock, diversifying supply base, and building redundancy into logistical systems are unlikely to afford much protection in such systemic failures while also increasing costs. McKinnon argues that it is mainly the role of the government to work closely with private industry to prepare for such eventualities. RAND, Inc. (2005) drew attention to supply chain performance measures and how security mea- sures affect supply chains. The study provides broad recommendations for various stakeholders within the supply chain for balancing the need to increase efficiency, while at the same time providing for fault tolerance and resilience through supply chain redundancy. The paper presents a three-tiered structure for the international goods supply chain (see Figure 2-2), consisting of the following tiers: 1. Logistics layer (that includes physical transportation systems and entities such as truck carriers, rail carriers, ocean carriers, etc.); 2. Transaction layer (that procures and distributes goods and is primarily driven by information flow, e.g., customer, retailer, foreign supplier) and; 3. Oversight layer (that provides the policy framework and enforces rules of behavior through standards, fines, and duties). Source: Rand (2005). Figure 2-2. Multi-layered international goods supply chain.
Methodologies to Measure Direct and Indirect economic Impacts of Disruptions to the Goods Movement System 19 The study uses this structure to analyze the role of different stakeholders in improving supply chain security. The paper concludes that as a first step the public sector needs to take the lead in bolstering fault tolerance and resilience in the system by creating an appropriate policy environment. Second, security efforts need to address vulnerabilities along the supply chain network trade lanes and not just at port facilities. Finally, the paper recommends that research and development should focus on new technology for low-cost, high-volume remote sensing and scanning that does not compromise efficiency (by imposing delays) and is a reliable measure for ensuring supply chain security. Nagurney et al. (2002) proposed a multi-level network for conceptualizing supply chain dynamics. The multi-level network consists of a logistics network, an information network, and a finance network and has the ability to model a supply chain response in a competitive environment. Using this model formulation, commodity shipments can be modeled dynamically with the ability to adjust prices over space and time. Commodity prices and flows are brought into a spatial equilibrium. Rose and Lim (2002) presented several refinements in a hazard loss estimation methodology and applied it to measuring business interruption losses from utility lifeline disruptions following the 1994 Northridge earthquake. They used a combination of primary data tabulation collected through questionnaires, interviews, and telephone surveys and deterministic simulation, which involves the use of mathematical models, such as IâO or linear programming to examine the impact of electricity outage on all sectors in Los Angeles after the Northridge earthquake. The authors critique current methods of loss estimation. In particular, they point out that the application of standard I-O analysis to the estimation of indirect losses exaggerates the levels of losses due to the multiplier effect. Assumptions inherent in the modeling methodology are key to determining the order of magnitude of loss estimation. Consideration of bottleneck effects on other sectors of the economy other than the one being studied can push the estimates upward, whereas consideration of resiliency, or the ability of individual entities and communities to cushion losses, can push estimates downward. Boarnet (1998) presented the results of a survey of businesses in the Los Angeles-Orange County region following the Northridge earthquake. The survey was conducted with a view to developing insight into the business impact of the infrastructure disruptions as a result of the earthquake. Survey responses from 559 firms in the Los Angeles and Orange County area were collected to provide information on the extent and magnitude of the business losses that could be attributed to transportation system disruptions. It was found that 17 percent of the business losses among the surveyed firms were due to transportation damage. Boarnet concluded that transportation damage is an important factor in economic losses following an earthquake. Interestingly, the research also found there to be no correlation between the distance from the damage due to earth- quake and the incidence of transportation losses. This suggests that the transportation network in urban areas is highly interlinked and that firms located some distance from a damaged highway can also be affected by it. Weinstein and Clower (1998) estimated losses to the Texas economy due to disruption in Union Pacificâs (UP) rail service and to safety problems faced by UP in the latter part of 1997. They estimated the impact of these service disruptions in terms of lost sales, reduced output, and higher shipping charges. The study estimated impacts by surveying businesses in six key sectorsâchemical industry, agriculture, paper and forest products, building materials, electric utilities, and retail trade. The study included analysis of impacts such as higher prices of inter- mediate goods that might reduce profits or would cause consumer prices to increase, substitution of more expensive intermediate production inputs for less expensive ones, etc.
20 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System 2.2.1 Key Conclusions from Research on Disruption and Impact Assessment ⢠The spatial or geographic scale of disruption will likely have a direct bearing on the magnitude of the economic impact. Thus, for example, the closing of a major port or key links in a land transportation network could have negative impacts throughout the supply chain, assuming little resiliency in moving goods on alternate paths. From a national perspective, however, the economic impacts might be very local (assuming that shippers and carriers can adjust the transportation component of the supply chain to serve the final consumers and thus still provide national benefits). See, for example, Cohen (2002) and Hall (2004). ⢠A disruption could affect the entire freight system of an area or affect a specific mode. In the situation where a single mode is disrupted, shipments may be transferable to alternative modes. The question then becomes whether or not the alternative modes are suitable for the type of delivery service and cargo carried, they have the capacity to handle the shipments, and they have the required approvals to transport such an unexpected demand in a timely and cost-effective manner. ⢠The temporal nature of disruption will also have potentially important economic consequences. A short disruption, that is, one lasting a day or up to a week, might cause some temporary or short-term economic loss, but overall would have minimal economic impacts. However, one lasting a much longer time could have severe consequences (see, for example, Chopra et al. 2007; Ito and Lee 2005), depending on how industries and supply chains adjust. It should be noted that temporally short disruptions could also have long-term effects. Thus, network resiliency in placing back into service the necessary facilities and services to move cargo once more becomes an important consideration in assessing overall economic impact. ⢠The economic impact of severe bottlenecks resulting from disruptions, or from the rerouting of traffic around such disruptions, could affect a wide range of supply chain participants, not just the ocean carriers, truckers, railroads, and shippers that are using the network to transport the goods. Others affected include local labor unions, local retailers, warehousing and distribution providers, and potentially a significant number of consumers and economic organizations throughout the region and/or nation (see, for example, Boarnet 1996; Lewis et al. 2006; Tierney undated). The geographic incidence of impacts can shift over time as well. For example, the 2002 West Coast port labor strike led several major retail chains to shift to using multiple ports, thus altering international cargo flows. The 2011 tsunami in Japan and massive flooding in Thailand have caused many automobile parts makers to rethink the location of their manufacturing plants. ⢠Different types of disruption could have a range of direct and indirect economic impacts. For example, the removal of one critical link in a network could create a bottleneck that might or might not have important consequences to goods flow. On the other hand, widespread net- work disruption (for example, due to floods and hurricanes) could have a multitude of over- lapping and connected economic impacts (see Abt 2003; Baade et al. 2007; Suarez et al. 2005). ⢠Network resiliency, whether provided via redundancy in asset supply or by business flexibility in the responsiveness to changed supply and/or demand circumstances, is a very important characteristic of economic impact. If, for example, goods flowing through a particular bottle- neck can be rerouted without significant economic costs, the overall economic impact of a bottleneck could be minimal, ignoring for a moment the economic consequences to the local economy (see, for example, Sheffi 2007; Chu and Levinson 2008; and MIT 2008 on freight resiliency planning at the statewide level). ⢠Assessing economic effects has to take into account the nature of the methodologies being used. For example, a disruption that shifts shipments from rail to truck may require that far more truck drivers be used. Some economic models would see this as a positive impactâmore workers are being employed. However, from a user perspective, the system has become potentially less
Methodologies to Measure Direct and Indirect economic Impacts of Disruptions to the Goods Movement System 21 efficient and not enough drivers and/or highway capacity may exist to handle the increased shipments. In other cases, the methodologies might rely on tools such as scientific surveys (see, Washington State DOT 2008b). Accordingly, the analysis of disruptions may require more of a âtool kitâ rather than a one-size-fits-all model, organized within a consistent methodological framework. ⢠The global goods movement supply chain is a multi-tiered system with various entities, stake- holders, networks, and modes involved that span a huge physical space and by their very nature are susceptible to natural and human-caused disruptions. International supply chains are also intricately interconnected. ⢠In case of a major event, such as a terrorist attack or an earthquake, standard risk mitigation measures, such as increasing safety stock, diversifying supply base, and building redundancy into logistical systems, may or may not afford much protection from damage. At the same time, the probability of occurrence of such events is small. ⢠Disruption resistance (security) and tolerance (resilience or recovery) are both important measures in disruption management. These measures have to be balanced with concerns regarding productivity while promising to provide sufficient benefits to justify costs. These measures could include increased electronic surveillance, such as in the case of global container movement, or having back-up suppliers, as in the case of a business that procures raw material, so if a disruption event affects one supplier, there is a fallback. The assumptions regarding resistance and resilience are important in understanding estimates of disruption impacts. ⢠Various types of economic models can help estimate the potential for losses due to a disruption in the supply chain. These range from simple logical frameworks to complicated dynamic economic simulations. However, it is important to understand the underlying assumptions of these models, particularly with regard to resiliency of, and interdependencies among, businesses and infrastructures. The order of magnitude of loss estimation can be largely affected by these assumptions. 2.3 Available Modeling Tools that Estimate Economic Impacts of Disruptions to the Goods Movement System 2.3.1 Least Cost or Diversion Modeling Tools Disruption Impact Estimating Tool-Transportation (DIETT) This tool is designed to generate net national economic impacts as a function of commercial shipments by truck, rail, and waterways. The values estimated by the model include ⢠Increased cost of freight movement associated with detours and ⢠Increased inventory costs imposed by the relative uncertainty of deliveries through the detours. The DIETT Model identifies and prioritizes state-specific transportation choke points (TCPs) according to their potential economic impact on U.S. commerce. The tool can assist state departments of transportation (DOTs) and other state security organizations in identifying and protecting high-value TCPs. The model is executed in MS Access and MS Excel environments. The primary inputs to the model are length of the detour around the TCP, level of congestion (on the original route and detour), and unit cost of shipment. The output is developed through an interconnected and semiautomatic set of functions designed to estimate the least-cost alter- native route in case of a disruption and to compute incremental costs. High-value TCPs are located along major commercial transportation routes and are critical points that can affect flow of commodities should a disruptive event occur at these points. Decisionmakers can use DIETTâs prioritized set of outputs, along with risk information, to better focus capital resources, security, and emergency preparedness planning. Net economic and societal dislocations are not considered in the model.
22 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System Freight Performance Measures (FPM) Data The FHWA Office of Freight Management and Operations, through a research partnership with the American Transportation Research Institute (ATRI 2010), has developed numerous performance measures for the nationâs highway system. Known as the Freight Performance Measures (FPM) Initiative, one element is a data processing tool that determines average oper- ating speeds for trucks that travel on interstate highways. These averages are calculated using GPS technology to collect confidential on-board data from several hundred thousand trucks (https://www.freightperformance.org/fpmweb/user_login.aspx). Using FPM data, changes in truck speeds can be obtained and diversion analysis can be performed in the case of a disruption to the highway network where more than one interstate corridor is a viable option. With expansion of FPM truck speed data to cover major non-interstate routing alternatives, the data obtained from the FPM Web interface could also assist with visualization and quantification of the level of system disruption resulting from network closures. However, not enough trucks are sampled currently to allow for robust estimates of corridor-specific truck volumes. Freight Analysis Framework (FAF) This is a comprehensive multi-sourced and partially modeled database maintained by FHWA and in the public domain (http://ops.fhwa.dot.gov/freight/freight_analysis/faf/index.htm). It identifies the nationâs major highway, railway, waterway, and air freight corridors and how different classes of commodity are moved between broad regional origins and destinations (O-Ds). Truck traffic volumes based on these O-D flows are assigned to a detailed representation of the national highway network, and this database can be used to construct scenario analysis by disabling specific links (highway segments, bridges) and nodes (interchanges, intersections). Such scenario analysis can provide key insights such as where the critical nodes are and the number of affected vehicles upstream and downstream of a disruption, the tonnage and dollar values of the commodities affected, and the additional travel time required. Rail and water freight traffic route assignments are not currently available from the FAF, but can be created from the public domain versions of the Railcar Waybills dataset provided on the Surface Transportation Board (STB) Website (http://www.stb.dot.gov/stb/industry/econ_waybill.html), and on the U.S. Army Corps of Engineersâ waterborne commerce Website (http://www.ndc.iwr.usace.army.mil/data/ datawcus.htm). Mapping these rail and inland barge flows can be accomplished using GIS network representations maintained for the purpose by the Federal Railroad Administration (FRA) and Army Corps of Engineers respectively and found at the Bureau of Transportation Statistics (BTS) National Transportation Atlas Database Website (http://www.bts.gov/publications/ national_transportation_atlas_database/2011/). Both a single-mode and a multimodal version (that allows for intermodal connections) of a traffic-routable national freight transportation network, including trans-oceanic linkages, can also be found at the Center for Transportation Analysisâ Oak Ridge National Laboratory (CTA/ORNL) Website (http://cta.ornl.gov/transnet/). Table 2-2 presents a summary of the characteristics of the least-cost or diversion modeling tools that were reviewed for this research. 2.3.2 Industry Supply Chain Response Modeling Tools Supply Chain Operations Reference Model (SCOR) SCOR is an off-the-shelf process management tool including three main featuresâprocess modeling, performance measurements, and best practices. The SCOR Model has been developed to describe the business activities associated with all phases of satisfying a customerâs demand. It can be used for capturing the configuration of a supply chain, measuring its performance, and re-aligning supply chain processes and best practices to fulfill unachieved or changing business objectives.
Methodologies to Measure Direct and Indirect economic Impacts of Disruptions to the Goods Movement System 23 iThink/STELLA (isee systems Software) This tool provides a graphically oriented front end for the development of system dynamics models. This off-the-shelf software guides a user (typically a business) through the creation of models that simulate business processes and scenarios, pointing out the impacts of a new pro- cedure or policy, and offers an opportunity to fix undesirable outcomes. The stock and flow diagrams used in the system dynamics literature are directly supported in this interface with a series of tools supporting model development. The tool also provides the user with the ability to write equations through dialog boxes accessible from the stock and flow diagrams. The tool is designed to answer strategic âwhat ifâ questions (e.g., what if you increased sales and marketing effort without adding network bandwidth?). The tool has the ability to model a businessâ entire operations system, using mapping and modeling processes, running simulations and analyses, and communicating through readable dashboards. Critical Infrastructure Protection/Decision Support System (CIP/DSS) This project is an integrated joint effort of three Department of Energy (DOE) laboratories, sponsored by the Science and Technology Directorate of the Department of Homeland Security (DHS). The CIP/DSS is a decision support system that can be used by decisionmakers to analyze all 17 of the nationâs critical infrastructures and their interdependencies. The model answers questions regarding critical infrastructure stability in times of disruption through the use of âconsequence models.â Linking critical infrastructures through their strongest inter dependencies Table 2-2. Summary of least-cost or diversion modeling tools reviewed. Tool DIETT FPM FAF Type of disruption Any disruption along a goods movement supply chain Truck flow disruption due to interstate closures Commodity flow disruption by truck, rail, water, air, pipeline, intermodal, and other transportation modes Economic impacts measured Increased cost of freight movement through diversion and increased inventory costs as a result of uncertainty of disruption Does not explicitly measure economic impacts; however, data can be used as an input to economic impact estimation Cost of diverted cargo and the delay cost of diversion Data requirements State-level TCP databases containing relevant data on mountain passes, tunnels, and bridges; specific datasets (e.g., identifiers, bridge span, detour length, and number of vehicles) None None Typical outputs List of state-level TCPs, prioritized by net national economic impacts Truck speeds Nature and dollar values of affected commodities, congestion, and additional travel time due to disruption Short- vs. long-term responses Changes in transportation route or facility Changes in sourcing of goods Changes in Yes Yes No No No Yes No No No Yes No Nologistics practices
24 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System allows for determination of choke points and vulnerabilities within the nationâs systems. The outputs of the consequence models are captured in a consequence database from which âdecision metricsâ tuned to particular decisionmaker profiles are computed (Bush et al. 2005). Vensim Originally developed in the mid-1980s for use in a consulting project, Vensim was made commercially available in 1992. It is an integrated environment for the development and analysis of system dynamics models. Vensim runs on Windows and Macintosh computers and is used for developing, analyzing, and packaging high-quality dynamic feedback models. Features include dynamic functions, subscripting (arrays), Monte Carlo sensitivity analysis, optimization, data handling, application interfaces, and more. Table 2-3 summaries the characteristics of the industry supply chain response tools reviewed for this project. 2.3.3 Static/I-O Modeling Tools Transportation Economic Development Impact System (TREDIS) TREDIS has an economic impact analysis module with an I-O model as its core computation tool. It can be used for all modes, including all forms of passenger and freight transport using road, rail, aviation, and marine facilities (http://www.tredis.com). PB Regional Impact Scenario Model (PRISM) PRISM has an I-O model as its core computational tool. PRISM utilizes outputs from full freight network freight flows models to estimate impacts of transportation accessibility improvements across the freight network on industry productivity and output. I-O equations are applied to the Tool SCOR iThink/STELLA CIP/DSS Type of disruption Supply chain disruption Supply chain disruption Disruption to 1 of 17 of the nationâs critical infrastructures Economic impacts measured Cost to business of disruption and alternate scenarios Cost to business of disruption and alternate scenarios No explicit economic impacts measured Data requirements Existing supply chain operations of the business Existing supply chain operations of the business Existing supply chain operations Typical outputs Business decisions (supply chain response) to changes in business environment Business decisions (supply chain response) to changes in business environment âDecision matrixâ to determine alternative infrastructure protection strategies Short- vs. long-term responses Changes in transportation route or facility Changes in sourcing of goods Changes in Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes logistics practices Table 2-3. Summary of industry supply chain response modeling tools reviewed.
Methodologies to Measure Direct and Indirect economic Impacts of Disruptions to the Goods Movement System 25 increases in first order output to derive full regional or state economic impacts. Supply chain cost savings are applied to shippers and consignees based on market conditions, and output is assumed to increase based on underlying demand and constant returns to scale assumptions. MARAD Port Kit The MARAD Port Kit estimates the economic value of port activities in easily understandable termsâjobs, income, and taxes generatedâwhich is critical to educating public officials and the general population about the value of the port industry. The kit enables deep-draft ports and other organizations to assess the economic impacts of maritime-related construction and ongoing activities at the national, state, and local levels. The kit ⢠Quantifies the economic value of deep-draft port activities in readily understandable terms such as employment, income, and tax revenues generated; ⢠Shows how a deep-draft port is linked to other industries; ⢠Can be used to investigate âwhat if â policy simulations (such as shifting trade patterns and dredging policies); and ⢠Assesses the economic implications of potential investments and changes in business activity. The results of an economic impact assessment undertaken with the MARAD Port Kit not only show the direct port industry impacts of an investment, cargo flows, or passenger flows, but also identify the total effect on the local region, state, or nation (A. Strauss-Wieder Inc. and Rutgers University 2000). Table 2-4 summaries the characteristics of the I-O modeling tools reviewed for this project. TREDIS, for example, makes use of the popular IMPLAN I-O data from MIG, Inc. (http:// implan.com/V4/Index.php). IMPLAN I-O data is sold at the state, county, and zip code levels for use in a wide range of economic impact analyses. For agencies that have sufficient in-house freight modeling software and expertise, the use of IMPLAN, REMI, or other sources of I-O data as a supplement to their impact studies may also be a cost-effective option to explore. IMPLAN I-O data has similarly been used in the University of Southern Californiaâs NIEMO (National Interstate Economic Model), which has been used in recent years to study a number of transportation network disruptions (http://create.usc.edu/2007/05/economic_impact_modeling_ and_a_4.html). Tool TREDIS PRISM MARAD Port Kit Type of disruption Disruption that affects the flow of goods Disruption that affects the flow of goods Disruption to a maritime port facility Economic impacts measured IO â Multi-regional, state, and county IO â Multi-regional, state, county, and TAZ Port facility, state, regional, IO Data requirements Good travel demand model or substitute Good travel demand model or substitute Information on port industry (investments, cargo flows, etc.) and regional definitions Typical outputs Change in employment, personal income, GDP, and value added Change in employment, personal income, GDP, and value added Impact on economyâ jobs, output, income, and taxes Short- vs. long-term responses Changes in transportation route or facility Changes in sourcing of goods Changes in Yes No No No Yes No No No Yes Yes Yes Yeslogistics practices Table 2-4. Summary of input-output modeling tools reviewed.
26 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System 2.3.4 Dynamic Economic Simulation Modeling Tools Long-term Inter-industry Forecasting Tool (LIFT) LIFT is a full macroeconomic model of over 800 variables. It combines an inter-industry (I-O) formulation with extensive use of regression analysis and employs a âbottom-upâ approach to macroeconomic modeling. Its macroeconomic âsuperstructureâ contains key functions for household savings behavior, interest rates, exchange rates, unemployment, taxes, government spending, and current account balances. The toolâs demand/production block uses econometric equations to predict the behavior of real final demand (consumption, investment, imports, exports, government) at a detailed level. Then, the detailed predictions for demand are used in an I-O production identity to generate gross output (total revenue adjusted for inflation). LIFTâs approach to projecting industry prices is similar, involving behavioral equations that estimate each value-added component (e.g., compensation, profits, interest, rent, indirect taxes) for each industry. Regional Economic Models, Inc. (REMI) Model REMI is an integrated modeling tool that combines I-O modeling methodology with an econometric model, where underlying equations and responses are estimated using statistical techniques. The estimates are used to quantify the structural relationships in the model. The overall structure is that of a macroeconomic model with features of a market equilibrium model for the labor sector, as well as features of the new economic geography related to inter-regional competitiveness. It is empirically calibrated on the basis of region-specific data. Table 2-5 summarizes the characteristics of the econometric/simulation modeling tools reviewed for this project. 2.4 Empirical Estimations of Economic Impacts of Disruptions to the Goods Movement System 2.4.1 Least-Cost or Diversion Model Applications ATRI (2010) presented an analysis of FPM data for the impact on truck traffic of the closure of I-40 near the North Carolina-Tennessee border in late 2009 and early 2010. Three methods of analysis were used for analyzing truck flows during two, 10-day study periods (before and after the closure) as follows: Tool LIFT REMI Type of disruption Disruption to container shipment at an international port of entry Disruption that affects the flow of goods Economic impacts measured Overall impact on the IO economy US: multi-regional, state, county Data requirements Data from federal govt. on container imports arriving at individual ports nationwide, all modes Good travel demand model or substitute Typical outputs Reduction in GDP Change in employment, personal income GDP, and value added Short- vs. long-term responses Changes in transportation route or facility Changes in sourcing of goods Changes in Yes Yes Yes Yes Yes No No Yeslogistics practices Table 2-5. Summary of econometric/simulation modeling tools reviewed.
Methodologies to Measure Direct and Indirect economic Impacts of Disruptions to the Goods Movement System 27 ⢠Average travel speed of trucks and truck demand for I-40, ⢠Spot speed and roadway utilization that identifies specific points where trucks are operating, and ⢠Freight flow information that identifies trends in truck movement. The research presents, by means of maps and graphics, several variables such as truck flows, system speeds, spot speed analysis to assess alternative routing during the closure of the I-40 segment, and freight flow diagrams that contrast truck flows before and after the incident. This report acts only as a demonstration tool that highlights capabilities and applications of FPM data and analysis. To determine delay costs, researchers could calculate total hours of delay and increases in operational costs for those trucks that use the I-40 corridor as part of regular operations using the FPM program. This method can be applied to understand how well prepared a particular facility is to respond to a disruption. 2.4.2 Industry Supply Chain Response Model Applications Wilson (2007) investigated how a transportation disruption affects the supply chain perfor- mance of a traditional supply chain and a vendor-managed inventory (VMI) system. The author also discusses how individual supply chain risks are connected and suggests strategies for mitigating the risk from transportation disruption. The model uses a system dynamics simula- tion software, iThink, to estimate the outcomes and impacts of various types of disruptions. Disruptions are classified by the location in the supply chain echelon (e.g., between warehouse and retailer, between supplier and warehouse, etc.). The simulation compares the impacts on a traditional supply chain versus those on a VMI. The author models a period of 600 days with a 10-day disruption occurring at day 200. The model utilizes data such as customer demand, inventory policy, processing and transport capacity, operational details, and type of disruption. The metrics used to evaluate the performance of the supply chain are unfilled retail customer orders, maximum number of goods in transit, and maximum and average inventory levels. The study finds that a disruption between a Tier 1 supplier and the warehouse or distributor creates the most problems within a traditional supply chain structure. Using a VMI reduces some of the impact. A VMI involves sharing customer-demand information and retail and warehouse inventory positions with the Tier 1 supplier. Hence, the benefit of disruption impact mitigation would have to be balanced with risk of dissemination of intellectual property. Another strategy for mitigation of this risk is carrying additional inventory or having a redundant supplier, although some of these strategies translate into additional cost and may have limited impact in cases where the redundant supplier is also impacted by the disruption. Lewis et al. (2006) proposed a Markov decision model to solve the inventory management problem faced by a firm operating an international supply chain that utilizes a seaport prone to unexpected closures. The problem of unexpected seaport closures results in a need to opti- mize overall inventory management costs and the costs of unfulfilled demand, both of which are impacted by variability in product delivery lead times resulting from unexpected seaport closures. The model can be used to determine both the probability of seaport closures and the expected duration of closures. Furthermore, the study examines how these cost impacts may be mitigated by the availability of additional, post-disruption emergency processing capacity. The model assumes that port closures lead to ships waiting to offload containers (instead of rerouting) and uses a simple deterministic queuing approach to model port freight processing dynamics. The model uses firm-specific data such as supply chain lead time, value of contents of a 40-foot container, holding cost, penalty for unfilled demand, units of work/day, and assumptions about seaport closure and reopening probabilities. The study finds that the length of a seaport closure affects a firmâs supply chain more negatively than does the probability of closure. Hence, contingency planning (from the firmâs perspective) and minimizing the duration of a closure, in the event of one (from the governmentâs or port
28 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System authorityâs perspective) are critical. The paper also concludes that holding and penalty costs increase more than linearly with port utilization, indicating a need for increasing government investments in processing capabilities and contingency planning (such as rerouting of freight to other ports of entry) for highly utilized ports. Dauelsberg and Outkin (2005) present a model of economic impacts arising from disruptions to critical infrastructures. Disruptions and their economic consequences are modeled as non- equilibrium events where the interdependent nature of various infrastructures allows event and disruption propagation from one infrastructure to another. The Critical Infrastructure Metropolitan Model is a set of critical infrastructure subsectors modeled in a system dynamics framework. Disruptions and their economic impacts are modeled as non-equilibrium events, where the independent nature of different infrastructures allows event disruption propagation from one infrastructure to another. In addition to lost sales, the model computes lost value-added by computing per capita gross state product and factoring it by the number of days of worker absence due to death, illness, or quarantine as a result of the disruption. This approach is deemed superior to a traditional I-O approach where equilibrium conditions are implied and are often calibrated to annual data in order to capture long-term trends and permanent change only, smoothing out the short-term dynamics. 2.4.3 Static/I-O Model Applications Ham et al. (2005) proposed a model to estimate inter-regional economic impacts of disrup- tions caused by major events (such as earthquakes) to intermodal transportation hubs (such as the Midwest). The model takes into account inter-regional commodity flows by sector and mode on U.S. inter-regional transportation networks. It computes changes in mean shipment length as a result of disruptions in transportation networks caused by the disaster event and estimates the modified value of inter-regional shipments as a result of increased shipping costs. The analysis uses a transportation network model (a modification of the traditional four- step transportation planning model) to estimate the impact of route changes in the event of a disruption to key links in the transportation network. A simple least-cost routing method is used to estimate alternate paths. The model then goes on to estimate the change in the value of inter-regional commodity flows as a result of the disruption. It assumes that the net loss in the inter-regional commodity flows is converted to intra-regional commodity flows due to increased shipment costs as a result of disruption to the network. It also assumes a shift of commodities from highways to railways. An application of the model to a hypothetical case of an earthquake in the New Madrid Seismic Zone is presented in the paper. The model uses commodity flow data collected by the U.S. Census Bureau in cooperation with the Bureau of Transportation Statistics. Data for the transportation network was based on the National Transportation Atlas Database. A simplified version of the existing railway network was used in the model. The transportation network model may be applied to identify critical sections of the network and analyze post-event reconstruction strategies. The economic impacts estimation method may be used to estimate the indirect impacts of a catastrophic event on inter-regional commodity flows. Changes in demand after an event also need to be taken into consideration in the event of a catastrophic event, but have not been included in this model. 2.4.4 Dynamic Economic Simulation Model Applications Rose et al. (2007) estimated the economic impacts on the U.S. economy of a 1-year halt in all imports from the rest of the world in response to an external threat to the United States. The
Methodologies to Measure Direct and Indirect economic Impacts of Disruptions to the Goods Movement System 29 analysis uses the REMI Model with an I-O model at its core. Different data and refinements are needed for various types of closuresâshutdown of imports, exports, international travel, and immigration. Policy variables that represent the direct impacts of simulated events are determined. Industry sales/international exports for each sector, data on levels of tourism, and total expenditure by international tourists within the U.S. regional breakdown of expenditure statistics are other inputs to the model. The net loss of GDP resulting from a complete shutdown of U.S. borders to people and goods for a period of 1 year is estimated to be close to $1.4 trillion measured in 2006 dollars. Arnold et al. (2006) summarized the structure and economics of the U.S. container port industry and its significance to U.S. merchandise trade. They estimated economic losses of disruptions in that traffic by investigating two hypothetical scenarios involving closure of the ports of Los Angeles and Long Beach. The analysis of economic losses resulting from disruptions in container traffic was performed using the University of Marylandâs LIFT Model. Estimates of economic losses for the 2002 closure of the Los Angeles and Long Beach ports were performed and compared to other estimates for that event. The LIFT Model is described as being superior to other estimating techniques, because it takes an inter-industry view of disruption rather than a business-centric view, and hence assumed that the economy as a whole makes adjustments in response to adverse supply shocks, thus helping contain losses. In the case of a supply disruption, businesses adjust in various ways. In the case of U.S. imports, the blocked imports could ⢠Enter the country from other open ports with temporary adjustments and capacity increases at those facilities, ⢠Be replaced, in part, by U.S. production of those goods (with U.S. producers responding to the increase in price by increasing production), ⢠Be compensated for by inventory draw downs. Models estimating economic losses should take into consideration these adjustments in order to make a more accurate assessment of the impact. However, there is uncertainty in this approach as well. For instance, it is difficult to know the ability of importers to find alternative routes for bringing imports into the United States. Industries and companies that use just-in-time inventory management could be disproportionately upset by disruption to imports. Little data is available on which industries use this method of inventory management. Lastly, the outcome for GDP will depend in part on the speed with which increases in import prices are reflected in the prices of the final goods and on the Federal Reserveâs response, both of which are uncertain. Tsuchiya et al. (2007) propose an analytical framework to estimate the indirect economic impacts of disasters on a multi-regional scale. The model considers both highway and rail networks and includes freight as well as passenger movements. The model is integrated with a transportation network model. The inter-regional spillovers of direct damage due to disruption are estimated using a spatial computable general equilibrium (SCGE) model. In order to be used/applied, the model requires an inter-regional I-O table containing data on trade flows between regions, as well as railroad and highway information. It also utilizes other inputs such as production capacity rate, rate-based transportation cost, etc. Additionally, a transportation network model is required to estimate delays on the system after a major incident. Delays and hence loss estimation can be measured in the short term as well as in the long term. The model was applied to the Niigata-Chuetsu earthquake of 2004. Earthquake-related damage to transportation networks was considerable and widespread. Countermeasures were needed to reduce negative spillover effects to regions apparently unaffected by the disaster (i.e., those that sustained no physical damage as a result of the earthquake).
30 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System 2.5 Construction of Freight Cost Matrices For the purposes of simulating both modal and market competition, a consistent treatment of transportation costs is required. This means that the costs computed for each mode, and for each source-market pair, should include the same generically defined set of cost elements. For the most part, freight costs are obtained from the following three sources: 1. Freight rates based on averaging over a large number of individual shipping contracts, 2. Statistically based freight costing or freight rate models, and 3. Component-by-component constructed engineering cost models. Costs are usually derived on a commodity- and mode-specific basis. Cost and rates/fees are not the same. Rates and fees charged by freight providers generally include a profit margin, which varies by load, commodity, client size, and shipment characteristics, along with the competitive situation. However, rate and fee information can be easier to obtain than cost information. For example, published tariffs include some of this information. Statistical Cost Models These costs are usually based on regression modeling. The models use either cross-sectional or time-series data for calibration purposes and typically include travel distance or distance-based average operating cost, travel time, and one or more measures of service quality, notably measures of service reliability (such as on-time arrival percentage, or standard deviation of delivery times). There are many examples of regression-based freight costing models. Commercially available products include Global Insightâs COSTLINE© family of rail and barge costing models (http:// www.ihsglobalinsight.com/gcpath/Costline.pdf) and Commonwealth Logistics Rail Costing System© (http://commonwealthlogistics.com/). Benson, Vachal, and Byberg (1999) use historical data to estimate container rates for agricultural commodity cargos. Rate calculators based on this work can be found on the IODAâs Ocean Rate Bulletin Website (see http://www.ams.IOda.gov/ tmd/Ocean/calculatIons.htm). Engineering Cost Models These models often take the form of a detailed spreadsheet model summing costs over a modeâs principal cost components and producing a total dollar valued cost per ton or per ton-mile. Examples include ITIC-IM, the U.S.DOTâs Intermodal Transportation and Inventory Cost Model (FRA 2005), ORNIM, the U.S. Army Corps of Engineers Ohio River Navigation Investment Model (ORNL 2001), STBâs Uniform Railroad Costing System (URCS) software, and the Truck Load Analysis Model developed by Berwick and Faroog (2003). Each of these models includes a wide range of logistics as well as pure âtransportationâ costs, and each offers a means of reconciling fixed and variable costs for economic analysis purposes. An ongoing TRB project, NCFRP Project 26, is looking in depth into freight costs. 2.6 Summary Table 2-6 summarizes the supply chain and economic impact models reviewed for this project. Numerous important issues for economic impact modeling surfaced from this literature review, including the following: ⢠Duration of disruption (e.g., day[s], week[s], month[s], year[s])âThe extent and nature of the response to such disruptions will vary depending on how long the disruption lasts and how widespread the scope of the disruption. It is important to recognize that disruptions may have both short- and long-term impacts. When possible, it is best to assess and articulate both of these impacts using a common methodological framework.
Methodologies to Measure Direct and Indirect economic Impacts of Disruptions to the Goods Movement System 31 ⢠ModeâCargo can sometimes shift to alternative modes when disruptions occur. The extent and nature of this shift will depend on many factors, such as whether the alternative mode serves the same geographic markets, the degree of redundancy in alternative mode, the flex- ibility in a shipperâs business practices, the economics of goods movement for different modes (e.g., low value-added bulk goods moving by barge will not likely shift to truck, but rather to rail as the next best alternative), and the available capacity and suitable cargo carrying approvals on alternative modes. ⢠Value of commodities being shippedâWhether goods can be shipped economically via other modes depends, in addition to the availability of service, on the value and nature of the cargo itself. High-value commodities or commodities that are otherwise time sensitive, such as air cargo, may not economically be shifted to slower modes. This can have major negative Supply Chain Models Economic Impact Models Characteristics Least Cost/Diversion Models Logistics and Industry Response Models Static/Input-Output Models Dynamic/Econometric Models Representative Tools DIETT, FPM, FAF PDRM, CIP/DSS, SCOR, iThink/STELLA TREDIS, PRISM, MARAD Port Kit REMI, LIFT Empirical Applications ATRI (2010), Freidman (2005) Wilson (2007), Lewis et al. (2006), Dauelsberg et al. (2005) Ham et al. (2005) Rose et al. (2007), Arnold et al. (2006), Tsuchiya et al. (2007) Data Requirements Readily available by mode from published sources Quantitative or qualitative; quantitative analysis can utilize statistical or econometric results Requires commodity- and industry-specific freight data to be fully effective Requires substantial data and multiple iterations of analysis â time series, cross-section, or panel data can be used Economic Variables Unit freight and inventory costs (e.g., per ton mile) Changes in supply chain response (inventory management, sourcing, routing, etc.) Changes in final demand (final industry sales or output, capital expenditures, operating expenditures) Explanatory variables include direct impacts, control variables, length of disruption, geographic location, etc. Direct Impact Representation Yes Yes Yes Yes Indirect Impact Representation No Yes Yes Yes Typical Outputs Unit freight and inventory costs (e.g., per ton mile) Changes in output; industry location response variables Output, value added, inter-industry sales, employment, income Output, value added, inter-industry sales, employment, income, others as specified Representation of Uncertainty No Varies No Yes Cost of Use Low Can be high High if customized; low if off the shelf High because labor intensive Application to System Level Analysis and Specific Modes Yes Yes Not mode specific, but commodity specific Yes, if specified Application to Region Level Analysis Usually average unit costs for U.S. as a whole Yes Yes Yes Availability of Model (Public Domain/Private) Public, private if customized Public or private Usually public domain Usually public or needs to be developed Ease of Use Moderate Variable Moderate Low Table 2-6. Summary of models reviewed.
32 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System effects on some parts of the country such as the Pacific Northwest, where high value-added electronics and computer parts manufacturing occurs. ⢠Geographic area involvedâThe spatial level can affect the degree of impact, as well as the types of models used in the assessment. The time dimension is especially important and merits some extra discussion. In the short run, impacts may be smaller under ânormalâ supply chain conditions, because there will be inventory to fall back upon. In the long run (i.e., for very long-term disruptions), the impacts may also be relatively small in an absolute sense, as both supply chains and industries may adjust. However, the spatial and distributional impacts may be significant to the extent that adjustments become permanent. For example, industries might permanently change their production locations or supply chain paths and routing may per- manently change (shippers may shift to other modes, other ports of entry, etc.). These would have major regional economic impacts, for example, by shifting production offshore or shifting some warehousing and distribution activities from one coast to another. In the medium run, impacts might actually be greatest if long-run adjustments are not yet established but inventories are drawn down and real supply chain disruptions begin to occur.
