Methodologies to Estimate the Economic Impacts of Disruptions to the Goods Movement System (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.

69 5.1 Overview of High-Level Methodology The following discussion and methodology template (Table 5-1) describe a high-level methodology that can be applied by transportation officials to estimate the economic impacts of transportation system disruption. The purpose of the high-level methodology is to use rules of thumb for estimating the economic costs associated with different compo- nents of a system disruption. Based on the relevant literature, case studies, and the research team’s experience in freight, logistics, and economic impact analysis, it was determined that a general methodology could be developed that grouped impacts according to common characteristics and that simplified the analysis in such a way that a high proportion of the economic effects could be readily captured. At the same time, it was clearly apparent that even a high-level methodology, while involving simplification, is not necessarily “simplistic.” Rather, the methodology must reflect important distinctions across a number of key variables that would determine economic outcomes and also dictate analytic methods. The high-level methodology is intended to capture the most likely general set of supply chain and economic impact outcomes. The methodology discussion features economic impact rules of thumb (i.e., impact factors that can be applied for short-term economic impact planning). Applying these rules of thumb factors requires that key assumptions be made beforehand, as follows: • What are the likely supply chain responses to disruptions (do shippers halt production, do they draw down inventory, or do they bid up prices for intermediate products)? • To what extent can freight realistically be diverted to other freight routings, facilities, or modes (are there other freight routing options at reasonable cost and capacity)? • What are the likely distances and time differentials for freight diverted to other routes or modes? As a pre-step to applying the high-level methodology, a DOT or MPO will need to have made an appropriate set of assumptions in advance of an event occurring. After describing the methodology, this chapter will illustrate the methodology by applying it to the Northridge earthquake case study, one of the case studies conducted as part of this study, which was presented earlier. As described in that case study, the January 1994 temblor centered in the San Fernando Valley caused major damage to a widespread area; in addition to buildings and utilities, the quake caused major damage to four freeways in Southern California and included a number of bridge and interchange collapses. Significant volumes of truck traffic were disrupted until most repairs could be completed. C h a p t e r 5 High-Level Methodology

Supply Chain Response Geographic Scope Economic Impact Economic Metric Economic Impact Economic Metric local/regional increased in-region shipper costs - transporta�on costs for in-region shipments; based on FAF data direct regional transportation cost increases - apply modal cost factors to change in ton miles and hours, to in- region TM/TH increase - apply ton to truck conversion factors as needed inventory cost increases - cost of increased �me in transpor�ng freight to market and/or holding freight in inventory �me value of freight - calculated as change in regional ton hours x cost of capital per hour for private sector state increased in-state shipper costs - transport costs for in-state Same as above applied to in-state TM/TH increase Same as above for state Same as above for state na�onal increased U.S. shipper costs - transport cost for total U.S. Same as above applied to U.S. TM/TH increase Same as above for U.S. Same as above for U.S. local/regional in-region shippers suspend produc�on/sales: industry output reduced for dura�on of suspension; based on FAF data reduced regional output, employment, earnings and value added for industry sectors affected - output of affected industries derived from BEA or other county or regional level data sources direct output losses expanded for multiplier effects - output mul�pliers from RIMS or IMPLAN used to es�mate total regional economic losses indirect and total economic losses for industry sectors affected - total impacts are sum of direct and indirect output losses; mul�pliers used to derive total impacts for output, employment, earnings, value added state Same as above for in-state shippers Same as above for in-state output, employment, earnings and value added Same as above for state Same as above for state na�onal Same as above for all U.S. shippers Same as above for total U.S. output, employment, earnings and value added Same as above for U.S. Same as above for U.S. local/regional increased in-region shipper costs - transporta�on costs for in-region shipments; based on FAF data direct regional transportation cost increases - apply modal cost factors to change in ton miles and hours, to in- region TM/TH increase - apply ton to truck conversion factors as needed inventory cost increases - cost of increased �me in transpor�ng freight to market and/or holding freight in inventory �me value of freight - calculated as change in regional ton hours x cost of capital per hour for private sector state Same as above for in-state shippers Same as above applied to in-state TM/TH increase Same as above for state Same as above for state na�onal Same as above for all U.S. shippers Same as above applied to U.S. TM/TH increase Same as above for U.S. Same as above for U.S. Impact # 1 - Direct Shipper Cost Impacts Impact # 2 - Inventory Cost Increases freight diverted to next nearest or least cost/time transport path/facility - same mode (e.g., highway to highway diversion) freight diverted to another mode Impact # 1 - Direct Economic Output Losses Impact # 2 -Indirect Economic Losses freight shipments disrupted/suspended Impact # 1 - Direct Shipper Cost Impacts Impact # 2 - Inventory Cost Increases Table 5-1. High-level economic evaluation methodology template: Table 5-1A. Major inland freight disruptions (highway/bridge, Class I or short line railroad (RR), waterway &/or pipeline links) — short term impacts (<90 days).

Supply Chain Response Geographic Scope Economic Impact Economic Metric local/regional persistent transport and inventory cost increases reduce business produc�vity and output - shippers affected reduce output as a func�on of increased produc�on costs total reduction in economic output, employment and earnings: output elas�ci�es applied to shipper cost increases, by industry; IMPLAN or other mul�pliers used to es�mate total direct, indirect and induced economic losses for earnings, employment, output and value added state Same as above for state Same as above for state na�onal Same as above for U.S. Same as above for U.S. local/regional direct loss in industry output and sales - industry output from BEA or other data total reduction in economic output, employment and earnings: output mul�pliers used to es�mate total direct, indirect and induced economic losses for earnings, employment, output and value added state Same as above for state Same as above for state na�onal Same as above for U.S. Same as above for U.S. local/regional persistent transport and inventory cost increases reduce business produc�vity and output - shippers affected reduce output as a func�on of increased produc�on costs total reduction in economic output, employment and earnings: output elas�ci�es applied to shipper cost increases, by industry; IMPLAN or other mul�pliers used to es�mate total direct, indirect and induced economic losses for earnings, employment, output and value added state Same as above for state Same as above for state na�onal Same as above for U.S. Same as above for U.S. freight diverted to next nearest or least cost/time transport path/facility - same mode (e.g., highway to highway diversion) freight shipments disrupted/suspended freight diverted to another mode Long Term Impact # 1 - Small to Moderate Reduc�ons in Economic Output, Employment, Earnings, GDP Long Term Impact # 1 - Moderate to Major Reduc�ons in Economic Output, Employment, Earnings, GDP Long Term Impact # 1 - Small to Moderate Reduc�ons in Economic Output, Employment, Earnings, GDP Table 5-1B. Major inland freight disruptions (highway/bridge, Class I or short line RR, waterway &/or pipeline links) — long term impacts (>120 days).

Supply Chain Response Geographic Scope Economic Impact Economic Metric Economic Impact Economic Metric local/regional increased in-region shipper costs - transporta�on costs for in-region shipments; based on FAF data direct regional transporta�on cost increases - apply truck mile and truck hour unit cost factors to in-region TM/TH increase - apply ton to truck conversion factors as needed inventory cost increases - cost of increased �me in transpor�ng freight to market and/or holding freight in inventory �me value of freight - calculated as change in regional ton hours x cost of capital per hour for private sector state Same as above for state Same as above applied to in-state TM/TH increase Same as above for state Same as above for state na�onal Same as above for U.S. Same as above applied to U.S. TM/TH increase Same as above for U.S. Same as above for U.S. local/regional increased in-region shipper costs - transporta�on costs for in-region shipments; based on FAF data direct regional transporta�on cost increases - apply truck mile and truck hour unit cost factors to in-region TM/TH increase - apply ton to truck conversion factors as needed inventory cost increases - cost of increased �me in transpor�ng freight to market and/or holding freight in inventory �me value of freight - calculated as change in regional ton hours x cost of capital per hour for private sector state Same as above for state Same as above applied to in-state TM/TH increase Same as above for state Same as above for state na�onal Same as above for U.S. Same as above applied to U.S. TM/TH increase Same as above for U.S. Same as above for U.S. local/regional in-region shippers suspend produc�on/sales: industry output reduced for dura�on of suspension; based on FAF data reduced regional output, employment, earnings and value added for industry sectors affected - output of affected industries derived from BEA or other county or regional level data sources direct output losses expanded for multiplier effects - output mul�pliers from RIMS or IMPLAN used to es�mate total regional economic losses indirect and total economic losses for industry sectors affected - total impacts are sum of direct and indirect output losses; mul�pliers used to derive total impacts for output, employment, earnings, value added state in-state shippers suspend produc�on/sales: industry output reduced for dura�on of suspension; based on FAF data reduced state output, employment, earnings and value added for industry sectors affected - output of affected industries derived from BEA or other state level data sources Same as above for state Same as above for state na�onal US shippers suspend produc�on/sales: industry output reduced for dura�on of suspension; based on FAF data reduced na�onal output, employment, earnings and value added for industry sectors affected - output of affected industries derived from BEA or other na�onal data sources Same as above for U.S. Same as above for U.S. Impact # 1 - Direct Shipper Cost Impacts Impact # 2 - Inventory Cost Increases freight diverted to next nearest or least cost US Port Impact # 1 - Direct Shipper Cost Impacts freight diverted to next nearest or least cost non-US Port freight shipment disrup�on/suspension Impact # 1 - Direct Economic Output Losses Impact # 2 -Indirect Economic Losses Impact # 2 - Inventory Cost Increases Table 5-1C. Ocean/air cargo port disruption (port out of service or major direct intermodal connections to port disrupted) — short term impacts (<90 days).

Supply Chain Response Geographic Scope Economic Impact Economic Metric Economic Impact Economic Metric Economic Impact Economic Metric local/regional persistent transport and inventory cost increases reduce business produc�vity and output - shippers affected reduce output as a func�on of increased produc�on costs employment and earnings: output elas�ci�es applied to shipper cost increases, by industry; IMPLAN or other multipliers used to es�mate total direct, indirect and induced economic losses for earnings, employment, output and value added affected port worker hours reduced; local/ regional port support businesses lose sales - es�mated based on current port cost information, earlier port economic impact studies; IO-based models total regional output, employment, earnings and GDP reductions - MARAD Port Economic Impact Kit or other IO-based models used to es�mate multiplier effects cargo throughput suspended - all in-port revenue genera�ng ac�vi�es lost for dura�on of disrup�on revenues loss - es�mated based on current revenues or revenue per container or per freight ton throughput state Same as above for state Same as above for state affected port worker hours reduced; local /regional port support businesses lose sales - same as above only if freight diverted to out-of-state port Same as above for state cargo throughput suspended - all in-port revenue genera�ng ac�vi�es lost for dura�on of disrup�on only if freight diverted to out-of-state port revenues loss - es�mated based on current revenues or revenue per container or per freight ton throughput na�onal Same as above for U.S. Same as above for U.S. No major net na�onal impacts where freight diverted to other US port No major net na�onal impact No major net na�onal impacts where freight diverted to other US port No major net na�onal impact local/regional persistent transport and inventory cost increases reduce business produc�vity and output - shippers affected reduce output as a func�on of increased produc�on costs total reduc�on in economic output, employment and earnings: output elas�ci�es applied to shipper cost increases, by industry; IMPLAN or other multipliers used to es�mate total direct, indirect and induced economic losses for earnings, employment, output and value added Affected port worker hours reduced; local/regional port support businesses lose sales - es�mated based on current port cost information, earlier port economic impact studies; IO-based models Total regional output, employment, earnings and GDP reductions - MARAD Port Economic Impact Kit or other IO-based models used to es�mate multiplier effects Cargo throughput suspended - all in-port revenue genera�ng ac�vi�es lost for dura�on of disrup�on revenues loss - es�mated based on current revenues or revenue per container or per freight ton throughput state Same as above for state Same as above for state Same as above for state Same as above for state Cargo throughput suspended - all in-port revenue genera�ng ac�vi�es lost for dura�on of disrup�on only if freight diverted to out-of-state port revenues loss - es�mated based on current revenues or revenue per container or per freight ton throughput na�onal Same as above for U.S. Same as above for U.S. Same as above for U.S. Same as above for U.S. Cargo throughput suspended - all in-port revenue genera�ng ac�vi�es lost for dura�on of disrup�on only if freight diverted to a non-U.S. port revenues loss - es�mated based on current revenues or revenue per container or per freight ton throughput local/regional Direct loss in industry output and sales as firms suspend produc�on - industry output from BEA or other data total reduction in economic output, employment and earnings: output mul�pliers used to es�mate total direct, indirect and induced economic losses for earnings, employment, output and value added Multiple Long Term Responses - 1) Shipper firms relocate out of region or 2) Shipper firms remain, permanently shif�ng supply chain to next nearest port Lost economic output, employment, wage earnings, GDP: In case of: 1) reloca�on: firm output permanently lost to region, and total regional economic ac�vity falls; or 2) next best port permanently adopted: regional industry output falls due to higher transportation and inventory costs state Same as above for state Same as above for state Same as above for state Same as above for state na�onal Same as above for U.S. Same as above for U.S. Same as above for U.S. Same as above for U.S. Long Term Impact # 2 - Direct port losses - reduced port employment, wage earnings, GDP Long Term Impact # 3 - Direct port losses - lost port revenues from shipping Long Term Impact # 2 - Major Impacts on Shippers from Sustained Shipment Disrup�ons, Inventory Depleted and Supply Chain Permanently Reorganized Long Term Impact # 1 - Small to Moderate Reductions in Economic Output, Employment, Earnings, GDP Long Term Impact # 2 - Direct port losses - reduced port employment, wage earnings, GDP Long Term Impact # 1 - Moderate to Major Reductions in Economic Output, Employment, Earnings, GDP Long Term Impact # 3 - Direct port losses - lost port revenues from shipping Long Term Impact # 1 - Small to Moderate Reductions in Shipper Output and Secondary Economic Impacts freight shipment disrup�on/ suspension freight diverted to next nearest or least cost non-US Port freight diverted to next nearest or least cost US Port Table 5-1D. Ocean/air cargo port disruption (port out of service or major direct intermodal connections to port disrupted) — long term impacts (>120 days).

Su ppl y Chai n Re sp on se Ge ogra phi c Sc op e Ec on om ic Im pact Ec on om ic Me tr ic Ec on om ic Im pact Ec on om ic Me tr ic air fr ei ght suspensi ons fo r a ffe ct ed US air carg o mo veme nt s na ti on al US sh i ppe rs su sp en d pr o ducti on /s al es : indus tr y output re du ce d fo r dur at ion of su sp en si on re duce d st at e ou tput , em pl oy me nt , ea rn in gs a nd va lu e a dde d for i ndus tr y se ct ors affe ct ed - output of a ffected indus tr ie s de ri ved from BEA or other state- level data so urce s di re ct ou tput lo ss es ex pande d fo r mu lt ip lier e ffe ct s - output mu ltipli er s from RI MS or IMPL AN us ed to es ti ma te total st at e econom ic lo sse s in di re ct and to ta l ec on om ic losses for in dus tr y se ct or s affe ct ed - to ta l im pact s ar e su m of di rect and indirect output lo sse s; mu ltipli er s us ed to derive tota l im pact s for output , em ploy me nt , ea rnin g s, value adde d air fr ei ght suspensi ons fo r a ffe ct ed US air carg o ex por ts na ti on al US sh i ppe rs su sp en d pr o ducti on /s al es : indus tr y output re du ce d fo r dur at ion of su sp en si on; US sh ippe rs ba se d on FA F data re duce d nati on al ou t put , em pl oy me nt , ea rn in gs a nd va lu e a dde d for i ndus tr y se ct ors affe ct ed - output of a ffected indus tr ie s de ri ved from BEA or othe r national data so urce s di re ct ou tput lo ss es ex pande d fo r mu lt ip lier e ffe ct s - output mu ltipli er s from RI MS or IMPL AN us ed to es ti ma te total national econom ic lo sse s in di re ct and to ta l ec on om ic losses for in dus tr y se ct or s affe ct ed - to ta l im pact s ar e su m of di rect and indirect output lo sse s; mu ltipli er s us ed to derive tota l im pact s for output , em ploy me nt , ea rnin g s, value adde d ma jo r re sour ci ng of US air fr ei gh t im po rt s fr om ot he r co unt ri es na ti on al in cr ea se d US sh ip pe r co st s - tran sp or t co st for US sh ip me nt s; US sh ippe rs ba se d on FA F data di re ct nati on al tr an sp or tati on co st in cr ea se s - apply mo da l co st fa ct or s to ch ange in ton mi le s a nd hour s, to US TM/T H in crea se - apply ton to tr uck co nv er si on fa ct or s as needed in ve nt or y co st in cr ea se s - co st of increa se d ti me in tr an sp or ti ng frei gh t to ma rket and/ or holdin g frei gh t in inventor y ti me va lu e of freigh t - ca lc ul at ed as ch ange in re gi on al to n hour s x co st of ca pital per hour for private se ct or Im pact # 1 - Di re ct Sh i ppe r Co st Im pact s - Ai r ca rgo sh ip me nt s rein st at ed from ot he r s uppl ie r co unt ries or region s Im pact # 2 -I ndi re ct Ec on om ic Lo ss es Im pact # 2 - In ve nt ory Co st In cr ea se s Im pact # 1 - Di re ct Ec on om ic Output Lo sses Im pact # 1 - Di re ct Ec on om ic Output Lo sses Im pact # 2 -I ndi re ct Ec on om ic Lo ss es Table 5-1E. Major air cargo disruptions — short term impacts (<90 days).

Supply Chain Response Geographic Scope Economic Impact Economic Metric Economic Impact Economic Metric air freight suspensions for affected US air cargo movements na�onal direct loss in industry output and sales - industry output from BEA or other data total reduction in economic output, employment and earnings: output mul�pliers used to es�mate total direct, indirect and induced economic losses for earnings, employment, output and value added multiple long term responses - 1) Shipper firms relocate out of US or 2) Shipper firms remain, permanently shif�ng supply chain to next nearest port lost economic output, employment, wage earnings, GDP: In case of: 1) reloca�on: firm output permanently lost to US and total na�onal economic ac�vity falls; or 2) next best port permanently adopted: na�onal industry output falls due to higher transporta�on and inventory costs air freight suspensions for affected US air cargo exports na�onal direct loss in industry output and sales - industry output from BEA or other data total reduction in economic output, employment and earnings: output mul�pliers used to es�mate total direct, indirect and induced economic losses for earnings, employment, output and value added multiple long term responses- 1) Shipper firms relocate out of US or 2) Shipper firms remain, permanently shif�ng supply chain to next nearest port lost economic output, employment, wage earnings, GDP: In case of: 1) reloca�on: firm output permanently lost to US and total na�onal economic ac�vity falls; or 2) next best port permanently adopted: na�onal industry output falls due to higher transporta�on and inventory costs major resourcing of US air freight imports from other countries na�onal persistent transport and inventory cost increases reduce business produc�vity and output - shippers affected reduce output as a func�on of increased produc�on costs total reduction in economic output, employment and earnings: output elas�ci�es applied to shipper cost increases, by industry; IMPLAN or other mul�pliers used to es�mate total direct, indirect and induced economic losses for earnings, employment, output and value added Long Term Impact # 1 - Moderate to Major Reductions in Economic Output, Employment, Earnings, GDP Long Term Impact # 2 - Major Impacts on Shippers from Sustained Shipment Disrup�ons, Inventory Depleted and Supply Chain Permanently Reorganized Long Term Impact # 2 - Major Impacts on Shippers from Sustained Shipment Disrup�ons, Inventory Depleted and Supply Chain Permanently Reorganized Long Term Impact # 1 - Moderate to Major Reductions in Economic Output, Employment, Earnings, GDP Long Term Impact # 1 - Small to Moderate Reductions in Economic Output, Employment, Earnings, GDP: Table 5-1F. Air cargo disruptions — long term impacts.

76 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System 5.2 Key Parameters 5.2.1 Type of Disruption Although many types of freight transport/supply chain disruptions can occur and vary in details, the high-level methodology has combined these into three broad types that capture most of the possible incidents and also would be analyzed similarly. These are as follows: • Major land freight transport link disruption (highway/bridge, Class I railroad, short-line railroad, inland waterway, pipeline, land-based international border crossing)—a partial or complete disruption to a major land freight transport link. Examples include the Minneapolis bridge collapse and the Baltimore rail tunnel fire. • Major ocean/air cargo port disruption—major service disruptions to an ocean port or airport, or major disruption of direct intermodal connection(s) to the port. An example of this is the closure of the Ports of Los Angeles/Long Beach (LA/LB) due to labor disputes. • Regional air cargo disruption—major disruption to air/air cargo operations over large areas or regions. Examples include the shutdown of European airspace as a result of the volcanic eruptions in Iceland. These three disruption types correspond to the major row categories in Table 5-1. Not included in Table 5-1 are large-scale national or regional disasters that cripple indus- trial output and normal social functioning for a significant period of time. In this case, major supply chain impacts can occur “at the source” of production. Impacts of this type can, and have, been shown to have very profound global economic impacts due to supply chain effects (other economic effects can also arise, such as disruptions to financial markets; reduced national income, consumption, and trade; loss of life; and other humanitarian needs). Prime examples include the March 2011 earthquake and tsunami in Japan and major flooding in Thailand in the fall of 2011. In cases such as these, key production sectors around the world can be significantly impacted. In the case of the Thai floods, computer manufacturers and assemblers around the world were rapidly affected. Thailand is a major global manufacturer of external computer disk drives; the disruptions to this component industry were quickly felt around the world. One estimate is that prices for these components were bid up by about 10 percent as a result of the disruption and resulting parts shortages (New York Times, Nov. 6, “Thailand Flooding Cripples Hard Drive Suppliers”). 5.2.2 Supply Chain Response As noted earlier, supply chains can be modified or will modify themselves in response to disruptions to the transportation system based on many factors, and the nature of the responses can be complex. Responses can also vary over time, so that for the shortest duration disruption, it is possible that a response of any kind is neither practical nor needed. Rather, shippers and carriers can simply wait out the disruption—shippers and producers, for example, can draw down inventories until the disruption is fixed and goods begin moving again across normal supply chain channels. For the high-level methodology, a simplified typology of responses is developed. As highlighted in Table 5-1, these include • Shifting freight transport to the next shortest/least-cost routing or facility—for example, if a major truck route is closed, trucks can detour to another route; if a port is shut down, import and export cargoes can be shifted to a reasonably nearby (next least cost or next lowest trans- port time) port, provided costs are not prohibitive and capacity is available. Port options are more complex—some import containers, for example, may be headed for interior U.S. land

high-Level Methodology 77 destinations, and other ports can pick up the slack. An example of this is when the Ports of Seattle and Tacoma served as an alternative port for some containers during the Port of LA/LB shutdown. However, in other cases, such as the transport of in-region cargo, no other port can be used because distances from the alternative port to the destination can become too great for cost-effective shipping. • Freight may, in some cases, be shifted to other modes—for example, rail shipments can be transported by truck, although this becomes less economical as the line-haul distance increases. • Goods may stop moving for the duration of the disruption—where no reasonable alternative routings or facilities are available, goods may simply not move, or in the case of disruptions to port intermodal linkages, may languish in port or near port storage and warehouse facilities. • Supply chain responses can be different in the case of major regional or international port or air cargo hub disruptions—here it is possible that the disruption is of such scale and geographic scope that firms may have to quickly change their regional sourcing of goods—something seen quickly, for example, after the 2011 Japanese earthquake and tsunamis. In this case, until supply chains are able to adjust by resourcing major products or inputs, severe global shortages may occur. Industrial output may be suspended or, alternatively, prices for disrupted products may increase substantially. A mix of responses is not only possible, but likely, depending on the mix of cargo types and their origins and destinations. For this reason, it is difficult to treat “freight” through a point as a homogenous flow; rather, some differentiation of freight flows by commodity and origin-destination is highly desirable from a methodological standpoint, assuming the data are available. 5.2.3 Impact Geography Economic impacts can vary significantly when viewed from different geographic/jurisdictional perspectives. For example, freight that originates or terminates outside of a given region (i.e., that is not shipped by producers within the region, but simply moves through) may be of little eco- nomic consequence to regional or local authorities or planners. At the same time, where a port that employs thousands of people in a region is shut, a long disruption can significantly affect the regional economy, both due to impacts on regional shippers, as well from lost port employee wages, business losses to port suppliers, and to the port itself in terms of lost cargo handling fees and tariffs. On the other hand, where freight shifts fairly seamlessly from a given route or facility to a reasonable alternative path, the impacts from a national perspective can be negligible. The high-level methodology considers economic impacts from the local/regional, state, and national perspectives. 5.2.4 Short-Term vs. Long-Term Timeframe Many, even the most apparently serious disruptions, will be of relatively short duration. For example, as described in its respective case study, Baltimore’s Howard Street Tunnel rail fire caused rail cargo disruptions/diversions of only several days (although street-level disruptions to local traffic, utilities, and commercial business activity lasted considerably longer). However, a highway bridge collapse can take months or even years to remedy. Supply chain responses can be very different over time—in the very short run, goods can simply be held back as inventory is drawn down. A less favorable short-run case is where goods cannot flow, and manufacturing or other economic production may cease for a time. Over time, new routings or modal shifts can be established and built into supply chains as a more normal practice. Higher transport and possibly inventory costs can be borne by shippers for a time, as these shifts can effect a “bandage” over the disruption. For long persisting disruptions, however, firms and shippers may be forced

78 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System to completely overhaul their supply chain networks, as the persisting increases in transport and inventory costs mount. In this case, some firms may seek to change their sourcing of inputs, rely on alternative warehousing locations, or, in the worst case, might consider relocating all or some operations. 5.2.5 Economic Impact Type A simplified typology of economic impacts, including both impacts and metrics, is detailed under Columns 4 and 5 in Table 5-1. Although specific impacts can vary in some details depend- ing on the type of disruption, the geographic perspective, and the timeframe, the high-level methodology groups impact the following categories: 1. Direct transportation (shipper) cost increases—These are the direct cost increases of moving freight across an alternative path, through another intermodal or port facility, or via another mode. Direct transport cost differentials can be derived, assuming some information about the origins, destinations, modes, and location of next-best alternative. Costs can be estimated as a function of changes in freight ton hours, miles, or both. 2. Inventory cost increases for commodities delayed—For commodities that have a high value and/or are time sensitive, the costs of slower transit times can be real. Freight economics identifies several reasons for these costs, including the costs of extra storage and delays in getting time-sensitive goods to market. Typically, such costs are measured against the market value of freight as a function of the delay, measured in freight ton-hours. Inventory costs can be estimated by multiplying the additional dollar value of freight delayed by the duration of the delay and by the opportunity cost of capital (discount rate or prevailing private-sector interest rates) prorated for the period of disruption. 3. Reduced economic output as goods cannot get to market—Industry output in an area will fall when freight disruptions compel freight users (shippers) to suspend production or when retail goods cannot be brought to stores for final sale. These effects may sometimes be referred to as general equilibrium impacts, although true impacts of this type would consider supply, demand, productivity, output, and price adjustments over time. Over a period of time, a region may see significant reductions in economic output and associated employment and earnings. A shutdown in production or significant cut back in sales will have a direct impact on those sectors affected. As output (supply) falls, there can also be price effects—prices can rise—but the output reductions would have to be large enough to have significant macro- economic effects nationally or regionally. Price effects are not considered in the high-level methodology. 4. Reduced economic output due to persistent, higher supply chain costs—Significant increases in transportation and inventory costs, when sustained for long periods of time, can result in reductions in producer firms’ economic output. Resources effectively become less efficient as they cost more, and reductions in those resources can be expected to lead to fewer inputs and less output. In the short to middle run, other things cannot be “substituted” for transportation (e.g., more inventories) so the increases in costs reduce the overall output of shipper firms. These impacts will be smaller than when firms must altogether shut down or greatly reduce production due to a complete break in their supply chain. 5. Multiplier effects—Whenever economic output in a region falls, either because of temporary or longer term production cut backs, or because of a contraction of output due to higher producer costs (due to higher transport and inventory costs), that reduction will carry over to other firms, which sell to the directly affected business, or due to reductions in household earnings, which affect consumption across many other economic sectors. These “multiplier” effects (indirect and induced impacts in the nomenclature of I-O analysis) are typically used in economic impact assessment.

high-Level Methodology 79 Multipliers are derived from several possible sources, but all are based on an I-O model of the regional, state, or national economies. Reduced economic output on the part of producers, as well as reduced final sales for retailers and wholesalers, are represented in I-O modeling as reductions in final demand. 6. Total employment, earnings, and gross domestic product (GDP) impacts—Reductions in final demand can be further used to estimate reductions in employment, wage earnings, and value added. Value added represents final sales for all producers in an area minus the costs of intermediate goods purchases used in production. The sum of value added by all sectors in an area is one way to measure gross product, the best measure of economic activity in an area. By summing changes in value added, changes in gross regional, state, or national product can be derived. Existing economic impact models such as the PRISM and TREDIS models are both effective in estimating these impacts. For example, PRISM utilizes the output of freight demand models to estimate changes in freight costs due to changes in the freight network (including possible reroutings), assigns these costs to industries, and utilizes researched elasticity assumptions to “translate” freight transport cost changes into industry output, employment, earnings, and value added effects, both for the industries directly affected and for other economic sectors. This is done utilizing IMPLAN-derived economic multipliers. 7. Port, airport, and other intermodal facility impacts—The high-level methodology includes regional and state economic impacts unique to ports and airports. These facilities, and perhaps major intermodal rail facilities, are regional economic engines in their own right. When a major port such as LA/LB is closed for a length of time, direct economic losses can accrue to a region as workers may (depending on actual labor practices) be furloughed and as port supplier firms lose business sales. In addition, these impacts have an effect on household earnings, which in turn, generates its own set of induced multiplier effects. Existing economic impact models such as the MARAD Port Economic Impact Kit are available to estimate such impacts. 5.3 Key Data Inputs and Major Assumptions To implement the high-level methodology, data must be obtained—or assumed—that capture the types of commodities being slowed or halted by the disruption, their approximate origins and destinations, and modes of transport. These must then be related to the industry structure of an area, so commodity flows can reasonably be associated with the regional, state, or national industries that are shipping the commodities and that are therefore impacted by disruptions. Gathering this information can be a daunting task and, clearly, for the methodology to work, it will almost always be true that some simplifying assumptions will be required. At a minimum, simple data sets readily accessible will be required. Data or assumptions required will include the following: • Commodity flows affected—Requires information about commodity origins and destinations, types of commodities, value or value class, and transport mode. FAF3 data may be sufficient in most cases when applying the high-level methodology (but see discussion with respect to Table 5-2). • Shipper industries affected—As noted, commodities must ultimately be related to the economy of an area, because the economic impacts relate to the industry structure. Conversion tables, which relate commodities to industries, can be developed for this purpose. I-O make-use tables can also be used, since these relate commodity sales to industries purchasing. • Commodity value—Time value of freight (inventory value) is a function of commodity value and can differ significantly for different commodities. As an alternative to a detailed commodity/

80 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System value analysis, it may be sufficient to break freight into rough time value classes—high, low and medium time sensitivity. For example, bulk commodities moving by barge are of very low unit value and are not typically time sensitive; by contrast, perishable food or high-value-added electronics (e.g., parts made to assemble an iPhone) would be rated high in time value. • Alternative transport paths—Without an elaborate freight transport model, assignment of disrupted freight to alternative paths can usually be made based on knowledge of the surrounding transport system and other factors. Educated guesses are a reasonable option in this case, including potential for mode shift. Mode shift potential (e.g., shift from rail to truck) can be informed by knowledge of origins and destinations of the cargo and the distance of transport. There are many commodity flow data options, including freight flows data, transport cost data, and other requirements. Table 5-2 summarizes data options that might be typically available for freight flows in a particular application of the high-level methodology. In addition to the above, ocean port cargo data can usually be obtained directly from the port, the U.S. Army Corps of Engineers, or other maritime data sources. In addition to its 123 (in 2007) region-to-regional national commodity flows by mode matrix, the public domain FAF database also supplies a set of route-specific highway (truck) volumes Network Flows Data Sources Source of information Available by mode? Trip distance information Available by commodity NAICS? Link specific flows Forecast data available? Behavioral sensitive to network changes Truck traffic counts by highway segment and possibly seaport MPO or state DOT supplied Truck only Usually no No Yes Usually no Yes Customized link node Travel Demand Model Usually MPO or state DOT where statewide TDM model available Usually some truck data for highway network Yes Usually no Yes, for truck Yes Yes Freight Analysis Framework (FAF) From FHWA Yes Yes Yes Yes for truck Yes No TranSearch Purchased from private vendor Yes Yes Yes No Yes No Commodity Flow Survey Data From BTS/Census Yes Yes Yes No No No Table 5-2. Summary of potential freight flows data sources, by model-relevant key characteristics (Step 2).

high-Level Methodology 81 for the nation’s interstates and other major intercity truck routes that can be useful to corridor- level analysis. Where rail or inland waterway routes are involved, STB’s railcar waybill data and the U.S. Army Corps of Engineers’ Waterborne Commerce data sets (cf Section 2.3.1) can be used to develop route-specific flows. Where a more detailed connection between O-D and commodity-specific flows and modal-corridor-specific flows are needed, further disaggregation of FAF inter-regional O-D flows may be required. While more detailed spatial and sectoral associations between industrial activity are more appropriate to the in-depth level of analysis and require additional resources, IHS Global Insight’s inter-county commodity flows matrix, known as its Transearch database, offers readily obtained, if proprietary, data sources (http:// www.ihs.com/products/global-insight/industry-analysis/commerce-transport/database.aspx). The IMPLAN Website also lists a recently developed doubly constrained gravity-model-based inter-county trade flows matrix. Such proprietary options trade off development effort and time against availability, transparency, and some flexibility in adaptation of county-level trade- cum-freight flow data. A third option is for a state DOT to generate its own O-D matrices using Census and other non-proprietary data sources. At the time of writing, NCFRP Project 20, “Guidebook for Developing Sub-National Commodity Flow Data,” is underway, looking at methods for combining FAF version3 region-to-region flows with other generally available data sources in order to produce more detailed corridor-level freight flows (http://apps.trb.org/cmsfeed/ TRBNetProjectDisplay.asp?ProjectID=2663). 5.4 Economic Impact Rules of Thumb Once a reasonable set of assumptions has been made about the likely supply chain outcomes and responses (e.g., freight diverted to another route or mode, freight does not move and shippers reduce output or suspend production, producers re-source inputs from other locations or move their production to other locations, etc.), the economic impact of those outcomes can be estimated. Figure 5-1 shows the basic concept that underlies the rules-of-thumb approach to estimating economic impact from disruptions. The economic impact of any particular disruption would depend primarily on the commodity characteristics, the characteristics of the disruption, and the costs associated with different elements of the cost structure (e.g., transport/logistics costs, inventory costs, productivity, and output losses). Commodity Characteristics • Low/high value • Low/high time sensitivity • Low/high volumes Disruption Characteristics • Short/long duration • Small/large geographic scale • Small/large number of transport alternatives available • Significant supply chain disruptions within a given industry sector = Economic Impact Default Values for: • Transport costs • Inventory costs • Lost industry productivity • Output Figure 5-1. Basic concept for “rules of thumb” analysis.

82 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System The basic economic impacts estimated in the high-level methodology are as follows: • Transport/logistics costs—These are primarily cost increases incident to shippers/beneficial cargo owners (BCOs), or transportation providers. They primarily include direct freight transport costs and inventory costs and are typically a short- to middle-term impact. Figure 5-2 represents a current estimate of the shares of total logistics costs attributable to direct transportation and inventory. • Regional economic impacts—These comprise wider economic impacts reflecting losses in regional output, sales, employment, wages, and GDP. They comprise any or all of the following impacts: – Lost economic output, employment, earnings, and GDP resulting from suspension of production or final sales due to breaks in freight shipping such as could occur when a major ocean port is shut down, the freight rail network is cut at some location, or a major shipping oil pipeline is cut (short run). – Reduced economic activity (output, employment, earnings, and GDP) from persistently higher transport costs (mid to long run). – Impacts on specific facilities such as sea ports or air cargo ports and multiplier effects within the region. Default rule-of-thumb values for estimating impacts are provided below and may be consid- ered as reasonable approximations for immediate estimates of potential economic impacts of disruptions, where results are needed quickly or the time and costs of collecting detailed data or performing economic modeling are unavailable. The rules of thumb cover the following areas: • Logistics costs – Direct unit transport cost values, by mode (e.g., transport cost per ton-mile and per vehicle- mile by truck)—Unit default costs for the rules of thumb are derived from a range of sources including those from ATRI, Association of American Railroads (AAR), FHWA, Vachal et al. (2005), the Arkansas River Historical Society (inland barges), and Bitzan et al. (2002) (on short line-railroads). – Unit inventory value of commodities, by commodity and value type (e.g., inventory cost per ton-hour of high-value manufactured products moved by truck)—Inventory costs can Figure 5-2. Logistics costs components. Source: http://ops.fhwa.dot.gov/freight/freight_analysis/econ_methods/lcdp_rep/index.htm#Toc112735358

high-Level Methodology 83 be viewed as composed of a number of cost elements. As shown in Figure 5-3, inventory costs include direct costs of warehousing (which, in turn, includes items such as real estate carrying costs, maintenance of warehouse structures, etc.); interest on the value of goods delayed in getting to market; and depreciation, taxes, insurance, pilferage, and loss, etc. Inventory costs are estimated based on the estimates reported by Macrosys (2005) and FHWA (2006). • Regional economic impacts – Direct output losses from disrupted production or sales—Total impacts in the high-level methodology are estimated per $1 million in direct output loss. Where specific industry output for disrupted industries in a region can be identified, the total economic losses in output, employment, wage earnings, and GDP can be estimated. Basic economic statistics such as U.S. Bureau of Economic Analysis regional and state GDP by industry data and U.S. Bureau of Labor Statistics (BLS) industry output data can be used to estimate direct output losses by industry in a given region or state on an annual basis. For given regions and areas, output by detailed NAICS industry class can be estimated based on national output to employment ratios, multiplied by the area’s employment in that industry. Employment data by NAICS are readily available for states, regions, and metropolitan areas through BLS or state economic data bases. Default impact multipliers would be applied to estimated output losses. – Indirect output losses from sustained increases in transportation cost and time—Total regional impacts on employment, output, earnings, GDP, and by industry class resulting from long-term increases in transport and inventory costs. Impacts are estimated in the high- level methodology as a function of logistics cost increases for a given commodity/industry group. The rate at which increased transport and inventory costs result in changes (reduc- tions) in the output of those industries is based on transportation “elasticity factors” that relate industry output (final demand) to logistics costs. Models such as Parsons Brinckerhoff ’s PRISM Model can be used to estimate the full economic impacts of short-term industry transport and logistics cost increases on commodities and industries within a region. PRISM utilizes IMPLAN I-O data for this purpose. – Total regional impacts on employment, output, earnings, and GDP from reduced or disrupted port throughput and revenues—Typically, short-term economic losses when large maritime Figure 5-3. Components of inventory-carrying costs. Source: http://ops.fhwa.dot.gov/freight/freight_analysis/econ_methods/lcdp_rep/index.htm#Toc112735358

84 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System ports (or other large intermodal facilities such as major intermodal rail hubs or interior ports) slow or become inactive, cargo is not handled, and port revenues cease to be earned. Port economic multiplier (or spinoff) impacts are also lost. Direct port impacts, including reduced container or other cargo throughput volumes can be estimated utilizing basic throughput data from the port itself, from Waterborne Commerce data, or from American Association of Ports data. Table 5-3 shows the impact multipliers that constitute the rules of thumb as derived from this research. The principal sources of the parameter values used are noted below this table. In applying these rules of thumb and parameter values, certain guidelines and limitations apply as follows: • Direct transportation costs—Average unit transport costs for modes must be treated as very approximate. Unit costs vary significantly within the modes by type of service, distance, speeds, equipment types, cargo type, geographic location, and multiple other factors. For example, truck costs will be higher in the New York metropolitan area than for the nation as a whole, since wage and fuel costs and taxes are higher there. Costs in metropolitan areas will also be higher than national averages, since congestion will affect driving times, rates of fuel consumption, etc. NCFRP Project 26, currently underway, is expected to provide detailed costing information and, when completed, can be used as a reference for analysts seeking to apply the high-level methodology. • Inventory costs—Costs can vary significantly for specific commodities, and grouping commodities into rough value classes, as in Table 5-3, should be considered very general approximations. Variations in inventory costs by mode may be considered preliminary in nature here and arise because of the simplified methodology used for deriving default values, which are based on a percentage of direct transport costs. Variations in inventory costs per ton hour of delay in freight transport will be due primarily to cargo value, rather than mode. • Regional economic impacts—Judgment must be applied to the question of whether disruptions will result in a stoppage of production (or sales for retailers), inventory draw down, changes in sourcing of inputs, or other market- and production-sector responses. These judgments should be based on the following considerations: – Timing—impacts are likely to be significant only if disruptions extend beyond several weeks, which would represent the period of inventory turnover for an average producer, retailer, or wholesaler (see below). – Severity and geographic scope of the disruption: widespread disruptions covering a large area and many industries and shippers and affecting numerous links along the supply chain will generate the most significant general equilibrium impacts. – Rate of inventory turnover (for assessing the likelihood of existing inventory draw down)— as noted, inventory turnover will typically be several weeks to a month, but different industries will have different inventory draw-down rates. – Other factors specific to the circumstance. • Direct facility impacts—Direct facility impacts relate to potential losses in regional eco- nomic activity resulting from a “stand down” of the facility itself. When a major port such as Los Angeles/Long Beach is closed, regional economic activity will be affected to the extent that port workers are laid off or furloughed, and as support businesses outside the port itself (e.g., drayage services, ship maintenance, bunkering services) lose sales. Port revenues will also be directly affected, although those losses would be partially offset by reduced marginal expenditures by the port itself. To illustrate, if a port handling 1,000 TEU per day (a mid- to big-size container terminal) is shut down for 10 days, then there would be 6,700 person days of lost employment, $1.07M in associated lost wage earnings, and state/regional GDP would fall by $1.60M.

high-Level Methodology 85 • Inventory Costs. Macrosys (2005), FHWA (2006) • Regional Economic Impacts – Direct Output Reductions. IMPLAN; US Bureau of Economic Analysis, RIMS II National Type I Direct Multipliers, 2008 updated to 2009 dollars – Higher Freight Transport and Inventory Cost impacts. Parsons Brinckerhoff PRISM model; IMPLAN • Direct Port Facility Impacts. Parsons Brinckerhoff port impact analyses (various); Le-Griffin, Hahn-Le and Melissa Murphy, “Container Terminal Productivity. Experiences at the Port of Los Angeles and Long Beach, Feb. 2006, University of Southern California, Dept. of Civil Engineering; ILA annual average earnings estimates http://www.ehow.com/info 8592745 longshoreman-salary.html Output (millions) Employment (job years, 000s) Earnings (millions) (millions) Ocean port $412,600 $670 $107,200 $162,400 Sources: Direct Transport Costs. For truck costs. ATRI, 2010, for Class 1 railroads AAR (2011), for short line railroads Bizman et al, 2002, for inland waterways the Arkansas River Historical Society, 2010 and Vachal et al., 2005. Truck Class I RR Short Line RR Inland Waterway/Barge Per ton-mile $0.07 $0.03 $0.04 $0.01 Per ton-hour $2.63 $0.52 $1.07 $0.06 Per vehicle-mile $1.39 $1.90 $4.07 $14.55 Per vehicle-hour $59.03 $39.56 $101.70 $87.30 Truck Class I RR Short Line RR Inland Waterway/Barge High-value manufacturing $1.05 $0.22 $0.45 n.a. Low- to moderate-value manufacturing $0.92 $0.19 $0.38 n.a. Low-value bulk commodities $0.74 $0.14 $0.29 $0.02 Perishable agricultural $1.19 $0.23 $0.47 n.a. Total Output (millions) Total Employment (job years, 000s) Total Wage Earnings (millions) Total GDP (millions) Direct Output Reductions (per $ million direct output loss, 2009$) Manufacturing $1.89 9.1 $0.42 $0.75 Services $1.59 12.6 $0.51 $0.92 Retail and wholesale $1.47 13.1 $0.49 $0.92 Agricultural, natural resources $1.69 9.8 $0.41 $0.86 Higher Freight Transport and Inventory Costs (per $ million transport cost increase, 2009$) Manufacturing $1.27 4.3 40.33 $0.50 Services $1.24 10.5 $0.58 $0.71 Retail and wholesale $1.14 8.7 $0.50 $0.77 Agricultural, natural resources $1.19 8.2 $0.48 $0.60 per day of impact Direct Transport Costs (per mile, hour) Inventory Costs (per ton-hour) Regional Economic Impacts Direct Facility Impacts (per 1,000 TEU Total Total Total Wage Total GDP Table 5-3. Economic impact rules of thumb—default values.

86 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System To estimate impacts in greater detail and to refine the “default” values presented above, various sources can be consulted. These include the following: 1. Freight transportation cost models—For the purposes of estimating transportation cost impacts, a consistent treatment of transportation costs is required. This means that the costs computed for each mode and each source-market pair should include the same generically defined set of cost elements. For the most part, freight costs are obtained from three sources (as freight rates based on averaging over a large number of individual shipping contracts, from statistically based freight costing or freight rate models, and from component-by-component constructed engineering cost models). Costs are usually derived on a commodity- and mode- specific basis. • Statistical cost models: These costs are usually based on regression modeling, using either cross-sectional or time-series data for model 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 rate models. Commercially available products include Global Insight’s COSTLINE© family of rail and barge costing models (http://www.ihsglobalinsight.com/ gcpath/Costline.pdf), and Commonwealth Logistics Rail Costing System© (http://common wealthlogistics.com/). Benson et al. (1999) used historical data to estimate container rates for agricultural commodity cargos. Rate calculators based on this work can be found on the IODA’s Ocean Rate Bulletin Website. (http://www.ams.IOda.gov/tmd/Ocean/calculatIons. htm). • Engineering cost models: These models now often take the form of a detailed spreadsheet model summing costs over a mode’s principal cost components and producing a total dollar valued cost per ton or per ton-mile. Examples include ITIC-IM, the U.S.DOT’s Intermodal Transportation and Inventory Cost Model Cost Model (FRA 2005), ORNIM, the U.S. Army Corps of Engineers Ohio River Navigation Investment Model (ORNL 2001), the Surface Transportation Board’s Uniform Railroad Costing System (URCS) software, and the Truck Load Analysis Model developed by Berwick and Faroog (2003). There are also rate-based commercially available software packages such as U.S. Rail Desktop that, through sampled rate data, provides carrier costs and margins and benchmark rail rates for carload, trainload, and intermodal shippers (http://www.IOraildesktop.com/). Each of these models includes a wide range of logistics as well as pure “transportation” costs, and each offers a means of reconciling fixed and variable costs for economic analysis purposes. NCFRP Project 26, is looking in depth into this freight cost modeling literature. 2. Economic impact models—These models relate direct reductions in industry output to overall regional or state economic activity and are derived from input/output model- ing sources, such as IMPLAN. More advanced models, which relate increased transportation costs to economic behavior and output of firms, would include models such as PRISM and TREDIS. The MARAD Port Economic Impact Kit is a useful and available tool for estimating overall economic impacts of port closures or disruptions, and that model can likely be extended to airports as well. Other airport economic impact modeling tools are described in research monographs such as ACRP Synthesis 7: Airport Economic Impact Methods and Models (http://onlinepubs.trb.org/onlinepubs/acrp/acrp_syn_007.pdf). Additional rules of thumb were inferred from the case studies. The case studies identified three initial considerations in framing the short- and longer term impacts associated with supply chain disruptions. All three of the following factor into the economic impact assessment: • Geographic scale of the disruption—The larger the geographic area affected by the disruption, the greater the impact on the supply chain. This includes the ability of other supply chain modes

high-Level Methodology 87 and routes to provide alternative capacity. For example, 9/11 temporarily shut all airports, seaports, and U.S. borders. This caused disruptions to those industries that kept minimal inventory on-hand and relied on international movements (e.g., automotive assembly lines). Hurricane Katrina involved substantial damage to miles of railroad infrastructure, which had to be rebuilt. In contrast, the Howard Street Tunnel fire affected the major north-south route of a Class I railroad, but was limited in geographic scope. Accordingly, time-sensitive rail moves could be shifted to alternate routes. • Length of time required for infrastructure repair/reopening facilities—The longer the time required to repair the infrastructure, the greater the level of supply chain disruption and related economic impacts. In the case of the Howard Street Tunnel fire, the route was reopened within days, minimizing the short-term disruption on rail movements. Similarly, the Port of New York and New Jersey was reopened quickly after 9/11, minimizing the impact on vessel operations. • Forced re-thinking of supply chain practices—In addition to the above, disruptions or additional costs to the supply chain can result in a re-thinking of practice, which in turn can have economic impacts. For example, the 9/11 closure of JFK International Airport (JFK) forced air cargo customers to try alternate airports. Some of these customers found the alternate airports less expensive or better than the “status quo” use of JFK. As a result, JFK’s air cargo customer base eroded. Similarly, the 2002 West Coast port strike resulted in major retail customers adopting a port diversification strategy, meaning companies switched to using multiple ports along several U.S. coasts for the receipt of cargo rather than relying largely on one port. The port diversification strategy resulted in increased traffic and economic value for some ports. 5.5 Illustrated Application of Methodology to Representative Case Study—Northridge Earthquake A case study narrative of the Northridge earthquake was presented in Section 4.4. In this section, the high-level methodology is applied in simplified form to illustrate its application, the types of assumptions required, and the outputs estimated. There is adequate information in the case study to apply the methodology in a credible manner, supplemented by information from other sources. Importantly, the case study includes estimates of economic impacts generated by other, more detailed studies of the earthquake’s economic effects. This provides information against which the high-level methodology can be evaluated. A more comprehensive analysis of the traffic and supply chain effects of the Northridge earth- quake can be found at http://ntl.bts.gov/lib/jpodocs/repts_te/13775.html#_Toc25646695. Input Assumptions Average Annual Daily Truck Traffic (AADTT) = approximately 20,000 trucks (http://www. goldenstategateway.org/IOer_content/Interchange_Spring_2011_WEB.pdf). • Duration of disruption = 4–6 months. • Truck (and other I-5) traffic detoured to longer and lower speed routes: – Caltrans and other surveys indicated major drops in traffic on I-5 leading to the affected sections. – Rerouted traffic traveled significant additional distances to use alternative regional highway corridors such as I-15 and I-110. – SR 14 bypass lanes (the primary detour route) accepted 75 percent to 80 percent of pre-quake I-5 trips. – Parallel routes all operated under congested conditions for the 4–6 month duration of the disruption.

88 Methodologies to estimate the economic Impacts of Disruptions to the Goods Movement System – By the fourth month, most major shippers using I-5 (96 percent) had returned to pre- disruption supply chain and routing practices. – Surveys indicated that 81 percent of trucks were rerouted and 69 percent were rescheduled, and driver overtime increased 55 percent. Other strategies included consolidated loads (29 percent) and reduced pick-up and delivery frequencies (38 percent). • Average trucking detour distance assumed to be 25 miles. • Average increase in truck delivery time of .75 hours. • Post-event surveys indicated business disruptions, with about 5 percent surveyed indicating significant interrupted supplies. Of that, assume 1 percent output loss. • Average truck freight assumed to be of moderate to high value. • Average truck payload = 10 tons. • In the absence of detailed truck commodity and industry data, industry sectors are assumed to be a mix (an average) across manufacturing, services, retail and wholesale, and agriculture and natural resources. • Because I-5 was restored in stages extending over 4 to 6 months at various locations, the average truck impact is assumed to be 5 months. Simplified Calculations Based on the above, the following simplified calculations of economic loss are made based on the high-level methodology. Losses are expressed in 2009 dollars. • Direct transport cost increases – 20,000 truck trips × $1.39/truck mile × 25 mile detour = @$700K per day for detour distance – 20,000 truck trips × $59.03/truck hour × .75 hours = @$900K per day for delay – Total for average of 5-month duration = $105M for distance + $135M for delay = @$240M • Inventory cost increases – 20,000 truck trips × 10 tons × .75 hours/truck × $1.00 inventory cost/ton hour = $150K/day – Total for average of 5-month duration = @$22.5M • Wider regional economic impacts—increased supply chain costs – Total direct supply chain (logistics cost) increase = $265M (from above) – Total lost output = $265M × 1.21 = $320M – Net regional economic loss (multiplier only) = @$55M • Wider regional economic impacts - disrupted production/output/sales – 20,000 daily truck trips × 150 days × $8,000 (FHWA, Value per Ton of IO Freight Shipments by Mode, 2002) average truck payload value × 1 percent lost output/final sales = $240M – Total lost output = $240M × 1.66 = @$400M • Summary impacts – Logistics cost increase = @$265M – Net regional economic impact n Due to supply chain cost increases = $55M n Due to disrupted production/output, sales = $400M Total Logistics Cost Increases and Lost Regional Output = @$720M The combined figures above, approximately $265M in direct logistics cost increases and another $455M in net lost regional economic output, compares favorably to estimates made by Gordon et al. (cited in the Northridge case study) of $769 million in output losses attributable to business logistics cost increases.