Trade Impacts of the Terrorist Attacks of 11 September 2001: A Quantitative Assessment Peter Walkenhorst and Nora Dihel * May 2002 Paper prepared for the Workshop on "The Economic Consequences of Global Terrorism" DIW, Berlin, 14-15 June 2002 *) The authors are economists in the Trade Directorate, Organisation for Economic Co-operation and Development, 2 rue André-Pascal, 75775 Paris Cedex 16, France. The views expressed in the paper are not necessary those of the OECD or its Member countries. TRADE IMPACTS OF THE TERRORIST ATTACKS OF 11 SEPTEMBER 2001: A QUANTITATIVE ASSESSMENT Abstract This paper used estimates of changes in frictional trading costs and the Global Trade Analysis Project (GTAP) computable general equilibrium model to assess the effects of the terrorist attacks of 11 September on international trade. Following the 11 September events, regulatory measures were implemented to tighten security at air and seaports as well as land border crossings. As a result, trading and transport companies have been confronted with additional frictional trading costs, in particular costs relating to transport, handling, insurance, and customs. These additional costs make international trade more expensive and will tend to reduce trade flows. The results from scenario analysis show that regions with high trade to GDP ratios, such as South Asia or North Africa and the Middle East, suffer the largest trade and welfare losses in relative terms. Across sectors, trade in agri-food products, textiles, non-metallic minerals, and machinery is most severely affected by the higher frictional costs of trade. Keywords Terrorism, security concerns, non-tariff measures, frictional trading costs, CGE modeling. JEL classification National security and war (H560); Country and industry studies of trade (F140). 1. Background The nature and scale of the terrorist attacks of 11 September 2001 on New York City and Washington, D.C., meant that the political and economic impacts would not be locally or nationally confined, but would affect governments, businesses, and people world-wide. Consumer and investor confidence plummeted after the tragic events, and economic researchers downgraded their forecasts of short and medium-term economic growth for virtually all countries around the globe. Moreover, emergency measures were taken to tighten security at air and seaports as well as land border crossings. Some disruption of trade flows during the immediate aftermath of the attacks seemed almost inevitable, yet additional frictional trading costs due to tighter security, notably costs relating to transport, handling, insurance, and customs, have the potential to continue to affect trade in the medium to long term. Some have likened the higher frictional trading costs to additional taxes on business activity or increases in border tariffs. Yet, a comparison with business spending on mandatory pollution abatement equipment seems more appropriate, as the higher expenses for the private sector provide benefits to the general public (higher environmental quality and lower risk of terrorist attacks, respectively), but are generally not accompanied by additional tax or tariff revenues for governments. Some industry experts have estimated that the additional security-related costs amount to up to three per cent of the value of traded goods (Leonard, 2001), while other analysis suggests that the magnitude of the effects is more modest (OECD, 2002a, 2002b). Nevertheless, in an era of large-scale economic interdependance and vertical specialisation, even small changes in trading costs can have considerable economic effects. Empirically, the impacts of the higher transport and trading costs on trade flows are to a large extent masked through other economic developments. For example, following the 11 September events world oil prices fell, thereby reducing the fuel costs of transport operators and at least partly offsetting the security-related cost increases. Also, existing capacities of transport operators did not make it possible for them in the short run to fully pass on the increases in frictional costs to consumers, so that transport fees after the attacks did not necessarily reflect the changes in security-related costs. This paper uses estimates of changes in frictional trading costs and the Global Trade Analysis Project (GTAP) computable general equilibrium model to separate out and assess the effects of the terrorist attacks of 11 September on international trade. Combining data on frictional cost changes with GTAP's information on transaction costs and bilateral trade flows across a large number of sectors and countries makes it possible to analyse and quantify the trade effects of the 11 September events. Moreover, the general equilibrium framework enables an evaluation of the incidence of changes in the trading environment and an assessment of economic welfare effects. The remainder of the paper is organised in four parts. Section 2 reports on available empirical evidence of the terrorist attacks of 11 September on the business environment for transport and trading companies. Section 3 presents the quantitative approach to model the impact of frictional cost changes on international trade flows and economic welfare. The results of the modelling analysis are then discussed in section 4, and the final section 5 contains conclusions. 2. Impact of the terrorist attacks on frictional costs of trade Because of the fear of future terrorist attacks, the 11 September events have triggered a variety of regulatory measures and demands on transport companies that entail new costs and longer delays at border crossings. The prevalence and thoroughness of inspections at airports, seaports, and land borders has been increased, shipping companies have had to face expenditures for additional security equipment and personnel, and transport insurance rates have been augmented. These developments have been most pronounced in North America, but are similarly felt in other parts of the world. 2.1 Air transport Tightening security on airports and aircraft has been one of the foremost priorities of public authorities. Immediately after the attacks, all US airports were closed for three days to review and strengthen security procedures. Around the world, access to airports and aircraft was tightened, training programmes for ground and flight staff were initiated, and more sophisticated alert and airspace management systems were launched. Concerning passenger traffic, more thorough controls of passengers and their luggage have been implemented, the classification of objects in hand luggage that could potentially be used as weapons has been modified, and in-flight protection has been stepped up by blocking cockpit access and putting armed air marshals on flights. Air freight transport has been similarly affected. Since the 11 September events, x-ray machines to examine cargo boxes and check whether the contents match shipping labels have been used more extensively. Other precautions taken include earlier drop-off deadlines at airports, bans on shipments from unknown customers, and waiting periods before shipments are put on planes. Following the 11 September attacks, insurers and re-insurers modified their aviation plans considerably whenever a "war and allied perils clause" in existing contracts authorised such changes. Coverage was reduced while premiums went up. Premia for air and sea cargo shipments were raised by 2 between 0.027 per cent and 0.05 per cent of total insured fleet value.1 Additional premia of $1.25 to $1.85 per passenger have been charged in passenger transport. Also, insurers decided to limit the cover for third party damages caused by terrorist action to $50 million (instead of $1.5 billion). In order to avoid major disruptions of air traffic due to lack of insurance coverage, regulators in several countries enforced shortterm measures to support third party insurance for airline companies in case of a new terrorist attack with damages above $50 million. In response to increased security provisions and insurance costs, many airlines began to apply "security surcharges". Generally, fee increases amounted to up to $8 per passenger or fell within the range of $0.10-$0.15 per kg of cargo (OECD, 2002a). In addition to increases in security fees, cargo shipments have been affected by frequent hold-ups and detours. During the three-day shutdown of the US air transport system, many transportation providers had to shift from air transport to road or rail transport to maintain shipment commitments. Later on, many airlines in North America, Europe, and Asia dropped flights from schedules, because of the weakening of passenger confidence and demand for air transport. On a year on year basis, world-wide passenger traffic dropped by 23 per cent in October and freight traffic by 9 per cent.2 This decrease was far more pronounced than the reduction in air traffic during the Gulf War, for example. Yet, traffic volumes have been recovering towards normal levels by the end of 2001. The marked decline in passenger traffic in autumn 2001 caused delays in the movement of goods that are carried in the belly space of passenger planes. Also, several airlines closed some of their freight handling facilities as a result of decreasing cargo volumes. For example, American Airlines temporarily closed its freight handling operations at the international airports of Buffalo-Niagara, Greater Rochester, and Salt Lake City. As a result, cargo had to be routed through other transport hubs. 2.2 Maritime transport A large share of the world's cargo moves through seaports. Yet, in the past port facilities often lacked security equipment, such as x-ray devices, and had relatively lax controls of access to docks and ships. Only about 2 per cent of the 72 million containers that are moved internationally every year used to be inspected.3 After 11 September, more extensive security checks and use of surveillance cameras and cargo scanners have been a priority. The Port Authority of New York and New Jersey closed its operations for two days, and more intensive screening of cargo loads in US and Canadian harbours has caused substantial delays before cargo could be picked up. These hold-ups disturbed shipping timetables worldwide during the first weeks following the terrorist attacks. As ships to and from North America have experienced delays, subsequent routes have been affected by delays as well. Other precautionary measures in Canada and the USA have consisted of mandatory 96-hour advance arrival notices and more frequent onboard Coast Guard inspections of crews and cargo. Also, ships have had to travel at slow speeds inside US harbours, flanked on each side by a tugboat in order to protect, for example, bridge supports against abrupt changes in direction. Shipping companies have been charged $1 000 to $1 500 for the required tugboat escorts. 4 Over time, emergency measures taken immediately after the 11 September events have been relaxed, transport operators have adjusted to the new procedures, and transport and trading activities seemed to be running relatively smoothly again by the end of 2001. In the longer run, technological and procedural progress might make it possible to further reduce the post-11 September delays and procedural costs while providing the desired high level of security. Similar to developments in the air transport industry, war-risk insurance premia for maritime transport rose sharply after the terrorist attacks on New York City and Washington, D.C. In response, ship 3 owners and vessel operators have implemented war risk surcharges on ocean freight cargo transiting the Middle East, Red Sea and Eastern Mediterranean. Beginning in early October, ocean carriers operating in the Europe/Far East and certain Middle East trade lanes announced war risk surcharges for traffic to and from specific ports as well as cargo transiting the Suez Canal, regardless of the origin and destination ports. War risk surcharges range from $10 to $450 per Twenty Foot Equivalent Unit of full container loads, and from $5 to $12 per cubic metre of less than full container loads.5 The application of the surcharges meant that total insurance rates for shipments from India to Persian Gulf ports, for example, went up by about 50 per cent.6 2.3 Road and rail transport The reinforcement of US customs vigilance during the processing of commercial vehicles led to lengthy delays immediately after the terrorist attacks. The US-Mexican border was closed for a short period, but hold-ups at the US-Canadian border were at least equally substantial. In Detroit, Port Huron, and Buffalo waiting times amounted to 10 to 12 hours in mid-September. Later, delays were reduced as additional customs personnel took up duty and logistics were improved by posting wait times on the US customs internet-site. But by the end of September, truck operators still reported four-hour backups at the US-Canadian border while loads were searched by customs agents. Prior to 11 September, delays at the border had typically lasted only about 30 minutes (OECD, 2002a). Road transportation companies themselves have been taking a variety of new measures against terrorist attacks. Some are fencing unsecured freight yards and terminals, conducting background checks on drivers, and issuing identity badges to employees, while others have installed satellite-tracking systems to monitor the exact location of trucks and trailers or equipped their vehicles with sensors that can detect whether a cargo container has been opened before reaching its destination. Similarly, train operators have taken precautionary measures such as more frequent inspections of tracks, bridges and tunnels, strengthening critical buildings and communication facilities, and installing fibre optic cables along tracks to detect tampering. 2.4 Customs procedures Prior to the terrorist attacks, estimates of the cost of time delays, paperwork, and compliance related to border crossing ranged from 5 per cent to 13 per cent of the value of the goods involved, depending on the types of goods traded (OECD, 2002a). The increased security at borders after 11 September 2001 has further augmented these customs-related costs. However, in the longer term the attention and resources newly devoted to customs inspections might trigger efforts to develop more efficient security procedures and improve border management. Measures to facilitate trade could include the reduction of the number of in-transit cargo inspections, the electronic collection of customs duties, and improved information sharing between authorities. As a result, customs services could become more able to reconcile policing of borders with smooth and open trade flows. The benefits of customs harmonisation have recently been illustrated through research by Japan's Ministry of Economy, Trade, and Industry and the Mitsubishi Research Institute that shows that the introduction of automated customs would lower the direct costs of customs clearance by the equivalent of 0.2 per cent of the value of traded goods. If furthermore the indirect benefits of a reduction in customs related delays are taken into account, additional cost reductions of up to 1 per cent of merchandise value could be realised (Hertel, Walmsley and Ikatura, 2001). Some concrete results of trade facilitation efforts following the 11 September events have already materialised. For example, on 12 December 2001 Canada and the USA signed a "smart border 4 declaration" that outlines an 30-point action plan to collaborate in identifying and addressing security risks while expending the flow of people and goods back and forth across the Canada-USA border. The action plan aims to enhance security and strengthen cross-border commerce through improved technology, coordination, and information sharing.7 A similar initiative for the US-Mexico border was unveiled in March 2002. 3. Implications for international trade Most operators will try to pass along any new frictional trading costs to their clients, although excess capacities and resulting price competition might make it difficult to raise prices in the short term. For example, during the first weeks following the 11 September attacks, many airlines actually lowered their fares for passenger transport as demand collapsed. One major reason for the demand reduction was the substantial decline in travel for tourism purposes. People were concerned about potential further attacks by terrorists and postponed or cancelled their travel plans. Destinations world-wide were affected. Excess capacity has also been an issue in the air and maritime cargo industry, so that transport operators have not been able in the short run to pass on higher security-related costs. At the same time, developments for fuel prices have been favourable during autumn 2001, lowering overall operating costs for transport companies. Transport activities are fuel-intensive, and expenditure on propulsion fuels accounts for a large share of operating costs. Thus, increases in fuel costs have a substantial impact on transport operators. In past incidents of war and terror in the Middle East, oil prices tended to increase markedly, due to concern over a possible disruption of supplies. For example, after Iraq's invasion of Kuwait in 1990, oil prices shot up from less than $20 to more than $35, triggering a slowdown in global economic growth as well as a substantial increase in transport costs. Oil price developments after the 11 September events have in comparison been rather atypical. During the first month after the events, oil prices were on average about $4 per barrel lower than during the preceding month (Figure 1). Throughout the subsequent military action in Afghanistan, oil prices stayed below $20 per barrel. The corresponding decline in fuel costs facilitated the adjustments to new security arrangements for transport operators. Indeed, several North American and European airlines, including Northwest Airlines, British Airways, and Lufthansa, eliminated previously existing fuel surcharges that had been introduced earlier in response to then rising expenditures on fuel. (Figure 1) Industry experts have estimated that the total costs of extra security measures implemented after the 11 September events could amount to 1 to 3 per cent of the value of traded goods (Leonard, 2001). However, these estimates were made soon after the 11 September events and seem to reflect the major disruptions in the weeks following the terrorist attacks. Over time the costs of the new border security measures have been declining and might continue to do so as technical and procedural progress is realised. On the other hand, new terrorist attacks could give rise to renewed security concerns with implications for the level of frictional costs to international trade. Some researchers have empirically estimated the impact of changes in transport costs on trade. For example, a recent study of bilateral trade between 103 countries reported an elasticity of trade flows with respect to transport costs of about -3, implying that a transport cost increase of 1 per cent would lead to a reduction in trade volume of 3 per cent (Limao and Venables, 2001). Also, concerning delays at borders, researchers estimated that the daily cost of hold-ups equals on average a 0.5 per cent ad-valorem tariff, with lower values for raw commodities and higher values for intermediate and consumer goods (Hummels, 2001). 5 In the USA, transport and insurance costs amounted on average to 3.4 per cent of customs value in 2000 (OECD, 2002a). Yet, the cost shares ranged from about 1 per cent for pharmaceuticals to more than 23 per cent for crude fertilisers. Increases in transport and insurance costs will naturally have a stronger impact on those commodities for which frictional costs are more important. Hence, trade in commodities with a high ratio of weight and volume relative to value, like fertilisers, coal, and certain fruits and vegetables, will tend to be more impeded by higher frictional costs than imports and exports of, for example, transport equipment, pharmaceuticals, and natural gas. A second cross-commodity consideration concerns the prevailing transport mode. Air transport has been relatively most affected by the tightening of security after the 11 September events, so that commodities that are to a large extent shipped by aircraft will likely be subject to higher relative increases in transport costs than commodities that go normally by ship, train, or truck. Goods that are typically transported by air tend to be light, high-value products that need to be delivered quickly, as they are typically subject to frequent lifecycle or fashion changes. Examples include electronics equipment and apparel. However, as indicated in the preceding paragraph, frictional costs generally represent only a small percentage of the value of these products. Besides higher transport and insurance costs, longer delays at air and seaports, as well as land border crossings have represented another type of international trade impact. Such security-related delays may make producers of perishable products, such as vegetables or fish, reluctant to ship overseas. The risk of spoilage at airports may induce producers to limit their distribution to distances that can be served by truck or train, for example. Moreover, about half of all trade between the USA, Canada, and Mexico is in machinery and transportation equipment, in particular automotive products. Over the past years many automotive companies have adopted just-in-time inventory practices that depend on frequent and predictable shipments from suppliers, including those located in other countries. In this context, unforeseen delays at border crossings can have immediate, disruptive effects on production runs, leading in several cases to factory shutdowns after the 11 September events.8 The tightening of security measures and the impact on trade flows has been most pronounced in the countries and regions directly affected by the terrorist attacks and the subsequent "war on terror", i.e. North America, the Middle East, and Central Asia. Hence, trade with countries in these “sensitive” regions has been subject to relatively steep increases in frictional costs following 11 September. Naturally, all US trade was affected by the tightening of security at US borders, but as a large share of Canadian and Mexican exports go to the USA, these NAFTA-partners were almost as strongly exposed to the new measures as the USA itself. Yet, changes in frictional trade costs were not limited to North America, but occurred worldwide. Indeed, North America is a region with a very low trade to GDP ratio, implying that changes in trade costs tend to affect NAFTA countries to a lesser extent than other regions (Figure 2). In particular, South Asia as well as North Africa and the Middle East, but also Western and Eastern Europe, show a trade to GDP ratio that is more than twice as high as the North American one. (Figure 2) In order to determine the overall impact of higher frictional costs on different countries and regions, the importance of trade for particular countries, the mix of traded commodity and their mode of transport, as well as the exposure to "sensitive" locations have to be taken into account. Given the complexity of international trade relations, applied economic models that make it possible to derive estimates of the quantitative impacts of changes in commodity-specific trade costs are necessary tools for this type of analysis. Quantification also enables comparisons of the relative magnitude of the economic 6 effects of terrorism risk occuring through trade vis-à-vis those coming about through other economic activities, such as tourism, finance, and insurance. 4. Modelling the effects of increased security concerns on trade Earlier research on the economic effects of terrorism has covered a number of issues, including game-theoretic aspects of terrorist action (Lapan and Sandler, 1993), the impact of terrorism on investment and economic growth (Enders and Sandler, 1996; Poirson, 1998; Abadie and Gardeazabal, 2001), and the assessment of terrorism-related losses in the tourisms sector (Enders, Sandler, and Parise, 1992; Pizam and Smith, 2000). Concerning the trade impacts of tighter security measures, limited research has been carried out to date. Anderson and Marcouiller (1999) investigate the linkages between economic insecurity and trade and find that inadequate institutions to constrain corruption and contract negligence reduce trade flows to a larger extent than tariffs. Moreover, Bougheas, Demetriades, and Morgenroth (1999) report a positive relationship between the level of trade cost-reducing infrastructure and the volume of trade, and Hertel, Warmsley and Ikatura (2001) predict considerable gains from reducing frictional costs of trade through improved customs administration. The following analysis aims to complement existing research by deriving estimates of the post11 September impact of increased frictional costs on trade flows and economic welfare across geographical regions and economic sectors. Half a year after the terrorist attacks, it remained unclear to what extent the existing increases in frictional costs would persist, and how businesses and public authorities would adjust to the new security environment in the longer term. Evidence to assess these issues will only become available over time, so that a certain degree of uncertainty concerning the frictional cost effects of 11 September remains. Hence, any quantitative estimate should be seen as being primarily indicative. 4.1 Model and data The analysis is carried out by using the well-established GTAP database and model. The latter is a static, multi-region, computable general equilibrium model that operates under assumptions of perfect competition and constant returns to scale. The model reflects bilateral trade flows, international transport margins, and country and sector-specific rates of import protection. GTAP thereby makes it possible to determine changes in production, consumption, trade, and economic welfare from particular trade-related external shocks, such as changes in trade margins. A full description of the model can be found in Hertel (1997). The following investigation uses information from the GTAP-5 database to specify ten regions and ten sectors. Nine of the ten regions refer to semi-continental or continental groupings, and the tenth to an aggregate of small countries, for which no detailed information on economic structure is available in the GTAP database. Of the ten sectors, nine represent different types of goods, while services are aggregated into the tenth one. The correspondence of the regions and sectors modelled and their GTAP-5 components is given in Annex Tables 1 & 2. The GTAP trade data set contains information on transport margins, which are derived from actual transport costs data at the detailed commodity level (SITC-4/5) in the United States. Based on this US data, bilateral margins for other countries are estimated (Gehlhar and McDougall, 2002). Similarly, US data on the commodity-specific modes of transport (ground, sea, air) is extrapolated to other countries. As discussed earlier, the increase in frictional costs following the 11 September events results in higher trading costs, which are, unlike tariffs, not accompanied by additional revenue for the government. Instead, the benefits of the additional precautionary expenditures take the form of increased public security. Yet, the costs and benefits of providing this public security service are not as such covered in the 7 GTAP database. Therefore, the analysis of non-tariff shocks and their impact on trade flows requires some adjustments. The approach taken in this study is analogous to the analysis of non-tariff liberalisation carried out by Hertel et al. (2001). It rests on the notion of defining an “effective price” PMS*irs of commodity i, imported from country r at domestic prices in destination market s. This price is related to the observed price PMSirs as follows: (1) PMS*irs = PMSirs / AMSirs ; where the technical coefficient AMS is unobserved and equal to one in the initial equilibrium. Changes in the value of AMS capture the impact of changes in unobserved trading costs on the price of imports from a particular exporter. Thus a decrease in AMSirs triggers an increase in the effective domestic price of good i exported from r to s. In order to maintain balanced trade flows in the data set, a compensating quantity adjustment is required, so that the “effective quantity” of exports associated with this price is defined such as to ensure that the product of observed price and quantity equals the product of effective price and quantity: (2) QXS*irs = QXSirs • AMSirs ; Taking the total derivative of expressions (1) and (2), re-arranging terms, and expressing them in percentage terms (denoted by lower case letters) gives the equations for import demand (equation (3)) and composite import price (equation (4)): (3) qxsirs = - amsirs + qimis - σim • [pmsirs - amsirs - pimis] ; (4) pimis = Σkθiks • [pmsiks - amsiks] ; where: qxsirs: percentage change in bilateral imports of i of s from r; qimis : percentage change in total imports of i into s; σim : elasticity of substitution among imports of i; pmsirs: percentage change in price of imports of i from r in s; amsirs: percentage change in effective price of i from r in s due to change in unobserved trade costs; pimis: percentage change in average import price of i in s; θiks: share of imports of i from k in total imports of s; With the existing trade elasticities in GTAP, it can be expected that an increase in trade costs will result in a decrease of both observed expenditures on imports and the share of imports from the particular trading partner to which the increase in trade costs applies. 8 4.2 Simulation scenarios Several experiments are carried out to analyse the impact of post-11 September increases in frictional costs on trade flows and economic welfare. There is ample anecdotal evidence that frictional costs have risen since the terrorist attacks on New York and Washington. Yet, systematic statistical data on the magnitude and structure of the changes is not available to date. Hence, the simulated changes in trade cost have to be based on assumptions that generalise from the available empirical evidence. Four scenarios were evaluated that use differing data on the structure of frictional cost changes. The overall magnitude of the frictional cost change was taken to be one per cent of the value of transported goods. Given the average magnitude of transport and insurance costs of 3.4 per cent ad valorem and estimates of the average costs of time delays, paperwork, and compliance related to border crossing of 9 per cent, an increase in frictional trading costs by one per cent ad valorem appears to be an upper bound for the long-term impacts of the increased security concerns following the terrorist attacks on the World Trade Center and the Pentagon. Further terrorist action triggering additional tightening of border controls could lead to renewed increases of trade costs, and impact predictions made immediately after the 11 September events (Leonhard, 2001) put the increases of frictional costs as high as three per cent of the value of traded goods, but relatively flexible adjustments of businesses and customs authorities to the new security environment during the six months following the attacks suggest that the cost increases in the medium and long-term are much lower than during the days immediately after the tragic events. The lower bound of trade impacts from the 11 September events might be seen as no change in frictional costs and trade flows, implying that in the long run the increased attention devoted to optimisation of customs services and technical progress will offset any burden from security measures put into place after the attacks. This no-change situation is taken as the baseline against which four other scenarios are evaluated. The latter are outlined in the following: • Uniform increase in frictional costs: In this scenario, all frictional costs are increased by one per cent ad valorem. • Region-specific increase in frictional costs: In this scenario, frictional costs are on average increased by one per cent ad valorem. However, the increases are not uniform across regions. High risk regions (North America, North Africa & Middle East) and medium risk regions (Western Europe & South Asia) are assumed to experience increases in frictional costs that are two and oneand-a-half times as high, respectively, as the cost increases in low risk regions (Eastern Europe, Latin America, Sub-saharan Africa, Oceania, North Asia, Rest of World). Using expenditure on frictional costs as weights, the frictional cost increases amount to 0.7 per cent in low-risk regions, 1.05 per cent in medium-risk regions, and 1.4 per cent in high-risk regions. • Sector-specific increase in frictional costs: In this scenario, frictional costs are on average increased by one per cent ad valorem. However, the increases are not uniform across sectors. In particular, it is assumed that the change in frictional costs for commodities that are transported by air and sea are two and one-and-a-half times as high, respectively, as the cost increases in ground transport. In particular, when using expenditure on frictional costs as weights, the mode-specific frictional cost increases amount overall to 0.69 per cent for ground transport, 1.03 per cent for sea transport, and 1.38 per cent for air transport. With the information on the sector-specific composition of transport modes in the GTAP database, sector-specific estimates of frictional cost changes can be derived. • Region and sector-specific increase in frictional costs: In this scenario, frictional costs are on average increased by one per cent ad valorem, and a regional and sectoral differentiation as in the 9 "region-specific" and "sector-specific" scenarios is applied. The derived changes in frictional costs range from 0.52 per cent (products in a low-risk region that are largely transported by road or rail) to 1.78 per cent (products in a high-risk region that are largely transported by air). The changes in frictional costs for the selected sectors and regions were introduced into GTAP, and the model was solved using the Gregg 2-4-6 step solution procedure under standard closure. 4.3 Results The results from the simulations suggest considerable trade and welfare impacts of a one per cent ad valorem increase in frictional costs to trade. World welfare, measured in terms of equivalent variation, would decline by about $75 billion per year. The biggest welfare losses in absolute terms would occur in Western Europe, North America, and North Asia (Table 1). (Table 1) Yet, if the welfare losses are related to the size of the economies in the different regions, the ranking of most affected regions changes considerably (Figure 3). In relative terms, South Asia as well as North Africa and the Middle East are far more affected by higher frictional costs than, for example, North America or North Asia. These results are largely driven by the economies’ strong dependence on international trade (Figure 2). (Figure 3) Across sectors, the most marked impacts occur for agriculture and food products, textiles and leather, non-metallic minerals, and machinery. Trade in these products is projected to decrease by one per cent or more as a result of a one per cent ad valorem increase in frictional costs (Figure 4). In other words, import demand for these commodities is elastic, while for mining products, chemicals, and basic metals trade flows are expected to change only to a less than proportional extent. (Figure 4) Differences in the results across scenarios are apparent, but do not fundamentally change the assessments. Only in a few cases are the scenario outcomes significantly different, as, for example, for the region of North Africa and the Middle East that seems less affected than Western and Eastern Europe under the assumption of uniform increases of frictional costs, but suffers more marked welfare losses than the two European regions if it is taken to be a high-risk region that experiences higher than average increases in frictional costs. Hence, exposure to international trade and the elasticities of import demand appear to be more important parameters for determining the impact of higher frictional costs than regional or sectoral differences in the magnitude of the frictional cost increases. In general, the findings highlight the close economic relationships in the international economy. Small changes in trade costs can have considerable impacts on trade flows and economic welfare. Increases in frictional costs for imports into particular countries affect not only the importing and exporting countries, but also other countries through the diversion of trade flows and substitution between products. Hence, even countries that are not directly involved in a conflict or subject to terrorist attacks might nevertheless suffer losses in trade and welfare as a result of increased security concerns and higher frictional costs of trade. 10 5. Conclusions This paper used estimates of changes in frictional trading costs and the GTAP model to assess the effects of the terrorist attacks of 11 September on international trade. Following the 11 September events, regulatory measures were implemented to tighten security at air and seaports as well as land border crossings. As a result, trading and transport companies have been confronted with additional frictional trading costs, in particular costs relating to transport, handling, insurance, and customs. These additional costs make international trade more expensive and will tend to reduce imports and exports. Four scenarios are evaluated that quantify the trade and welfare impacts of a one per cent ad valorem increase in frictional costs of trade applying either uniformly across sectors and countries, or being sectorally or regionally differentiated according to presumed exposure to terrorism risk following the 11 September events. Given that after the initial disturbances following the terrorist attacks on New York and Washington international trade relations have returned towards normal again, a one per cent ad valorem increase of trade costs likely presents an upper bound for the impacts of 11 September. The global welfare losses of such an increase in frictional costs are estimated to amount to about $75 billion. The analysis shows that even though North America would face substantial losses from the increase in frictional costs, other regions that depend to a larger extent on international trade, such as South Asia or North Africa and the Middle East, are more severely affected in relative terms. Many of the countries in these regions are middle-income developing countries that have progressively opened their economies towards international markets. The above-average losses in trade and welfare from increased security measures following the 11 September events could raise popular concerns about further steps of trade liberalisation and thereby compromise the long-term development prospects of these countries. Agriculture and food products, as well as textiles and leather are among the product categories for which above average impacts on trade flows are predicted. Trade in these products is often politically sensitive and many countries retain substantial protection against imports of these products. In the Uruguay Round of multilateral trade negotiations some progress was made in reducing trade barriers for agri-food products and textiles, which would be partly offset through increases in frictional costs. 11 6. References Abadie, A., and J. Gardeazabal, 2001. "The economic costs of conflict: a case-control study for the Basque country." NBER Working Paper No. 8478, National Bureau of Economic Research, Cambridge/Mass. Anderson, J.E., and D. Marcouiller, 1999. "Trade, insecurity and home bias: an empirical investigation." NBER Working Paper No. 7000, National Bureau of Economic Research, Cambridge/Mass. Bourheas, S., P.O. Demetriades, and E.L.W. Morgenroth, 1999. "Infrastructure, transport costs, and trade." Journal of International Economics 47: 169-189. Enders, W., T. Sandler, and G.F. Parise, 1992. "An econometric analysis of the impact of terrorism on tourism." Kyklos 45: 531-554. Enders, W., and T. Sandler, 1996. "Terrorism and foreign direct investment in Spain and Greece." Kyklos 49: 331-352. Gehlhar, M., and R.A. McDougall, 2002. "Transport margins and modes." in Dimaranan, B.V. and R.A. McDougall (editors): Global Trade, Assistance, and Production: The GTAP 5 Data Base. Center for Global Trade Analysis, Purdue University, West Lafayette/Indiana. Hertel, T. (editor), 1997. Global Trade Analysis: Modeling and Applications. New York and Melbourne: Cambridge University Press. Hertel, T., T. Walmsley, and K. Ikatura, 2001. "Dynamic effects of the 'new age' free trade agreement between Japan and Singapore." Journal of Economic Integration 24: 1019-1049. Hummels, D., 2001. "Time as a trade barrier." Mimeo, Department of Economics, Purdue University, West Lafayette/Indiana. Lapan, H.E., and T. Sandler, 1993. "Terrorism and signalling." European Journal of Political Economy 9: 383-397. Leonard, J., 2001. "Impact of the September 11, 2001 terrorist attacks on North American trade flows." Manufacturers Alliance e-Alert, Arlington/Virginia. Limao, N., and A. Venables, 2001. "Infrastructure, geographical disadvantage, transport costs and trade." World Bank Economic Review 15: 451-479. OECD (Organisation for Economic Development and Co-operation), 2002a. "The impact of the terrorist attacks of 11 September 2001 on international trading and transport activities." Unclassified document TD/TC/WP(2002)9/final. Paris: OECD Publications. OECD (Organisation for Economic Development and Co-operation), 2002b (forthcoming). "The economic consequences of terrorism," in: OECD Economic Outlook #71. Paris: OECD Publications. Pizam, A., and G. Smith, 2000. "Tourism and terrorism: a quantitative analysis of major terrorist acts and their impact on tourism destinations." Tourism Economics 6: 123-138. Poirson, H., 1998. "Economic security, private investment, and growth in developing countries." IMF Working Paper WP/98/04, International Monetary Fund, Washington, D.C. 12 Notes 1. See "Survey of members: national measures in respect of war risk insurance," International Air Transport Association, 27 November 2001. 2 See "October traffic shows intensified decline." Press release no. 36, International Air Transport Association, 30 November 2001. 3 See "When trade and security clash." The Economist, 6 April 2002. 4. See "Freight-transportation system gets more expensive, slower." Wall Street Journal of 27 September 2001. 5. See "War risk surcharge summary," Fritz Transportation International, December 2001. 6. See "Ocean rates rising in Indian region." Journal of Commerce, 25 September 2001. 7. See "Canada and the United States sign smart border declaration." Government of Canada News Release No. 162, 12 December 2001. 8 See "The Canada-US border: an automotive case study." Center for Automation Research, Ann Arbor, Michigan, January 2002. 13 Figure 1: World oil prices (weekly data), Jan. 2000 - Dec. 2001 (US Dollar per barrel) 30 25 20 15 11-Sep-01 10 01-Jan-00 01-Apr-00 01-Jul-00 01-Oct-00 01-Jan-01 01-Apr-01 01-Jul-01 01-Oct-01 01-Jan-02 Source: OECD (2002a). Figure 2: Ratio of imports to gross domestic product in selected regions 50% 40% 30% 20% 10% 0% Western Europe Eastern Europe N. Africa & Sub-saharan Middle East Africa Oceania Source: GTAP-5 database. 14 North Asia South Asia North America Latin America Rest of the World Figure 3: Scenario results on welfare changes due to a 1% ad valorem increase in trade costs Unifo rm Regional diversity Sectoral diversity Regional & sectoral diversity 0.0% -0.1% -0.2% -0.3% -0.4% -0.5% -0.6% -0.7% W estern Europe E astern Europe N. Africa & M iddle East Subsaharan Africa O ceania North Asia South Asia No rth America Latin America Rest of the W orld Note: Welfare changes are calculated as equivalent variation divided by gross domestic product. Source: Authors. Figure 4: Scenario results on changes in imports due to a 1% ad valorem increase in trade costs U nifo rm R egio nal d iversity S ecto ral d iversity R egio nal & secto ral diversity 0.00 -0.20 -0.40 -0.60 -0.80 -1.00 -1.20 -1.40 -1.60 A gricu ltu re M ining Fo o d T extiles & leather Wood & p aper Source: Authors. 15 C hemicals N o nm etallic minerals B asic metals M achinery O ther Mining Food Textiles & leather Wood & paper Chemicals Nonmetallic minerals Basic metals Machinery Other Uniform increase Western Europe Eastern Europe N. Africa & Middle East Sub-saharan Africa Oceania North Asia South Asia North America Latin America Rest of the World Total Regional differentiation Western Europe Eastern Europe N. Africa & Middle East Sub-saharan Africa Oceania North Asia South Asia North America Latin America Rest of the World Total Sectoral differentiation Western Europe Eastern Europe N. Africa & Middle East Sub-saharan Africa Oceania North Asia South Asia North America Latin America Rest of the World Total Reg. & sect. differentiation Western Europe Eastern Europe N. Africa & Middle East Sub-saharan Africa Oceania North Asia South Asia North America Latin America Rest of the World Total Source: Authors. Agriculture Table 1: Change in imports and welfare as a result of a 1% ad valorem increase in trade costs Imports (per cent) -0.9 -1.0 -1.0 -1.5 -1.4 -0.6 -1.9 -1.0 -1.3 -1.4 -1.1 0.8 -0.8 -1.3 -0.6 -0.2 0.4 0.2 -0.3 -0.7 0.1 -0.6 -1.1 -0.9 -0.6 -1.1 -1.2 -0.7 -1.7 -1.0 -1.0 -1.0 -1.1 -1.0 -1.3 -1.0 -1.2 -1.1 -1.6 -2.0 -1.1 -1.8 -1.3 -1.4 -0.7 -0.7 -0.6 -0.7 -0.7 -0.6 -1.2 -0.9 -0.7 -0.6 -0.8 -0.2 -0.2 -0.2 -0.2 -0.1 -0.5 -0.3 -0.6 -0.3 -0.3 -0.3 -1.2 -1.0 -1.2 -1.0 -1.0 -1.2 -1.5 -1.1 -0.9 -0.8 -1.2 -0.3 -0.6 0.0 -0.4 -0.7 -0.9 -0.9 -0.9 -0.9 -0.5 -0.5 -1.2 -1.3 -0.8 -0.5 -0.8 -1.8 -0.6 -1.3 -0.8 -0.9 -1.3 -1.1 -1.1 -0.8 -0.7 -0.8 -0.7 -1.5 -0.6 -0.5 -0.7 -0.9 -31282 -3518 -2942 -1027 -972 -11845 -6380 -13434 -3105 -639 -75143 -0.9 -0.7 -1.5 -1.1 -1.1 -0.4 -2.0 -1.4 -1.0 -1.0 -1.1 0.8 -0.6 -1.8 -0.5 -0.1 0.2 0.2 -0.4 -0.5 0.1 -0.7 -1.1 -0.6 -1.0 -0.7 -0.8 -0.5 -1.8 -1.5 -0.7 -0.7 -1.1 -1.1 -1.0 -1.5 -0.9 -0.8 -1.1 -2.0 -1.5 -1.3 -1.0 -1.3 -0.7 -0.5 -0.9 -0.5 -0.5 -0.4 -1.3 -1.3 -0.5 -0.4 -0.9 -0.2 -0.2 -0.4 -0.2 -0.2 -0.4 -0.3 -0.8 -0.3 -0.3 -0.3 -1.2 -0.7 -1.8 -0.6 -0.6 -0.8 -1.6 -1.6 -0.6 -0.5 -1.2 -0.3 -0.4 0.0 -0.3 -0.4 -0.7 -0.9 -1.2 -0.6 -0.3 -0.5 -1.3 -0.9 -1.3 -0.3 -0.5 -1.3 -0.6 -1.8 -0.5 -0.6 -1.3 -1.2 -0.7 -1.3 -0.4 -0.4 -0.4 -1.5 -0.9 -0.2 -0.4 -1.0 -32798 -2413 -4564 -664 -684 -8470 -6672 -18602 -2166 -456 -77489 -0.7 -0.7 -1.0 -1.6 -1.5 -0.7 -2.0 -0.9 -1.2 -1.3 -1.0 0.7 -0.7 -1.3 -0.7 -0.2 0.4 0.3 -0.2 -0.7 0.0 -0.6 -0.9 -0.7 -0.6 -1.2 -1.2 -0.7 -1.7 -0.9 -0.9 -0.9 -1.0 -0.9 -1.0 -1.0 -1.3 -1.2 -1.6 -2.0 -1.0 -1.7 -1.3 -1.3 -0.6 -0.5 -0.6 -0.8 -0.8 -0.6 -1.3 -0.7 -0.7 -0.6 -0.7 -0.2 -0.1 -0.3 -0.3 -0.2 -0.5 -0.4 -0.6 -0.3 -0.3 -0.3 -1.0 -0.8 -1.3 -1.1 -1.1 -1.3 -1.6 -1.0 -1.0 -0.8 -1.1 -0.2 -0.4 0.0 -0.5 -0.7 -1.0 -1.1 -0.7 -0.8 -0.4 -0.5 -1.1 -1.1 -0.9 -0.6 -0.9 -2.0 -0.8 -1.3 -0.8 -1.0 -1.3 -1.1 -0.9 -0.9 -0.9 -0.9 -0.9 -1.8 -0.6 -0.6 -0.7 -1.0 -29494 -2993 -3144 -1096 -1063 -12772 -7008 -13451 -3214 -650 -74886 -0.8 -0.5 -1.6 -1.2 -1.1 -0.5 -2.1 -1.2 -0.9 -1.0 -1.1 0.8 -0.5 -1.8 -0.6 -0.1 0.2 0.3 -0.3 -0.5 0.1 -0.7 -0.9 -0.5 -1.0 -0.8 -0.9 -0.5 -1.8 -1.2 -0.7 -0.7 -1.0 -1.0 -0.7 -1.4 -0.9 -0.8 -1.2 -2.1 -1.4 -1.3 -1.0 -1.2 -0.7 -0.4 -1.0 -0.6 -0.5 -0.4 -1.4 -1.0 -0.5 -0.4 -0.7 -0.2 -0.1 -0.4 -0.3 -0.2 -0.4 -0.4 -0.7 -0.3 -0.3 -0.3 -1.1 -0.5 -1.9 -0.7 -0.7 -0.9 -1.7 -1.4 -0.7 -0.5 -1.1 -0.2 -0.3 0.1 -0.3 -0.5 -0.7 -1.2 -1.0 -0.6 -0.3 -0.5 -1.2 -0.7 -1.5 -0.4 -0.6 -1.5 -0.8 -1.7 -0.6 -0.7 -1.3 -1.1 -0.5 -1.5 -0.6 -0.6 -0.6 -1.9 -1.0 -0.3 -0.5 -1.0 -31160 -2010 -4835 -736 -763 -9390 -7394 -18401 -2302 -470 -77461 16 Welfare (mill. USD) Annex Table 1: Sectors defined by reference to the GTAP-5 Sectoral Classification Sectors Agriculture Mining Food Textiles and leather Wood and paper Chemicals Non-metallic minerals Basic metals Machinery Other manufacturing and services GTAP-number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 GTAP-sector Paddy rice Wheat Cereal grains not elsewhere covered Vegetables, fruit, nuts Oil seeds Sugar cane, sugar beet Plant-based fibers Crops not elsewhere covered Bovine cattle, sheep and goats, horses Animal products not elsewhere covered Raw milk Wool, silk-worm cocoons Forestry Fishing Coal Oil Gas Minerals not elsewhere covered Bovine meat products Meat products not elsewhere covered Vegetable oils and fats Dairy products Processed rice Sugar Food products not elsewhere covered Beverages and tobacco products Textiles Wearing apparel Leather products Wood products Paper products, publishing Petroleum, coal products Chemical, rubber, plastic products Mineral products not elsewhere covered Ferrous metals Metals not elsewhere covered Metal products Motor vehicles and parts Transport equipment not elsewhere covered Electronic equipment Machinery and equipment not elsewhere covered Manufactures not elsewhere covered Electricity Gas manufacture, distribution Water Construction Trade Transport not elsewhere covered Water transport Air transport Communication Financial services not elsewhere covered Insurance Business services not elsewhere covered Recreational and other services Public Administration, Defense, Education, Health Dwellings Source: Authors. 17 Annex table 2: Regions defined by reference to the GTAP-5 regions Model region GTAP-Code AUT Western Europe BEL DNK FIN FRA DEU GBR GRC IRL ITA LUX NLD PRT ESP SWE CHE XEF HUN Eastern Europe POL XCE XSU TUR N. Africa & Middle East XME MAR XNF BWA Sub-Saharan Africa XSC MWI MOZ TZA ZMB ZWE XSF UGA XSS GTAP-region Austria Belgium Denmark Finland France Germany United Kingdom Greece Ireland Italy Luxembourg Netherlands Portugal Spain Sweden Switzerland Rest of EFTA Hungary Poland Rest of Central Europe Former Soviet Union Turkey Rest of Middle East Morocco Rest of North Africa Botswana Rest of SA Customs Union Malawi Mozambique Tanzania Zambia Zimbabwe Other Southern Africa Uganda Rest of Sub Saharan Africa Source: Authors. 18 Model region GTAP-Code AUS Oceania NZL CHN North Asia HKG JPN KOR TWN IDN South Asia MYS PHL SGP THA VNM BGD IND LKA XSA CAN North America USA MEX XCM Latin America COL PER VEN XAP ARG BRA CHL URY XSM XRW Rest of World GTAP-region Australia New Zealand China Hong Kong Japan Korea Taiwan Indonesia Malaysia Philippines Singapore Thailand Vietnam Bangladesh India Sri Lanka Rest of South Asia Canada United States Mexico Central America Colombia Peru Venezuela Rest of Andean Pact Argentina Brazil Chile Uruguay Rest of South America Rest of World