Appendix

The Benefits of Trade Facilitation – A Modelling Exercise

Introduction

The benefits of improved global trade facilitation far exceed those available from further tariff reduction.¹ The authors estimate that an ambitious (but still incomplete) improvement in two key components of trade facilitation – border administration and transport and communications infrastructure – in which all countries raise their performance halfway to global best practice   would lead to an increase of approximately US$ 2.6 trillion (4.7%) in global GDP and US$ 1.6 trillion (14.5%) in global exports.² By contrast, the gains available from complete worldwide tariff elimination amount to only US$ 400 billion (0.7%) in global GDP and US$ 1.1 trillion (10.1%) in global exports. 

Even a more modest improvement in trade facilitation, in which all countries raised their performance halfway to regional best practice, would lead to increases of US$ 1.5 trillion (2.6%) in global GDP and US$ 1.0 trillion (9.4%) in global exports. These estimates are illustrative rather than precise, and are meant to provide only a broad indication of the potential impact of the policies being modelled. 

Figure 36: Reducing supply chain barriers has a larger effect than removing tariffs

While the increases in trade from tariff elimination are similar in magnitude to those associated with trade facilitation, the increases in GDP are many times greater. The reason is that the kinds of efficiencies brought about by improved trade facilitation are more powerful than those associated with tariff reduction. Reductions in supply chain trade barriers improve the efficiency of the movement of goods, in a manner analogous to an increase in transportation productivity, thereby recovering resources that are otherwise wasted. In contrast, tariff reductions primarily represent a reallocation of resources within an economy, while capturing only the more modest inefficiency created by the tax.³

Gains in GDP associated with trade facilitation would take place in all regions, though they would be concentrated in those with the greatest improvements. In the more ambitious scenario, these would include sub-Saharan Africa, South Asia, and parts of Central and West Asia (referred to as “Rest of Asia” in the tables) as well as other developing regions. The distribution of gains from barrier reductions is more even across countries than the gains associated with tariff elimination, which disproportionately accrue to specific countries, such as Russia and China. 

Trade facilitation leads to expansion of trade in a broader range of sectors than tariff elimination, so global exports would increase for most categories of goods.⁴ The trade-creating effects of tariff elimination are focused on products such as agriculture, processed foods, and textiles and apparel, which currently have tariff peaks. By contrast, a modest amount of trade facilitation would lead to trade expansion in a wider variety of manufactures, while ambitious trade facilitation is particularly helpful for trade expansion in technologically complex goods with long supply chains, such as transport equipment, machinery and electronics.

Figure 37: Impact per region varies under the ambitious scenario

Further gains are available if countries improve market access and the business environment. While these gains are not included in the above estimates, they are likely to be substantial. Improvements in market access which includes not only tariffs, but non-tariff measures, sanitary and phytosanitary and technical barrier to trade requirements, quotas, licenses, rules of origin and other issues   and improvements in the business environment, including the regulatory environment, investment policy, security and related issues, are important complements to improved trade facilitation. A change in market access and the business environment comparable to those modelled above could increase the overall economic gains by about 70%.⁵

Of course, reducing certain supply chain barriers   particularly those related to infrastructure requires upfront investment, whereas tariffs can be eliminated with the stroke of a pen. However, as this report shows, detailed analysis can enable policy-makers to prioritize investments that are most critical and cost-efficient.

The effect of cutting tariffs

As a benchmark, the report begins by presenting estimates of the potential economic effects of removing all tariffs globally, including tariff-rate quotas, as well as all export subsidies and taxes. These estimates are generated using a computable general equilibrium model, which simulates their removal. The estimates can be viewed as an upper bound for the economic gains that remain to be achieved from global tariff liberalization. The authors use the model of the Global Trade Analysis Project (GTAP), using a 2007 baseline.⁶ This has certain implications for the tariff results, as discussed below. The model is implemented using an aggregation of 23 regions, which are further aggregated into 15 larger regions for presentation purposes.

For the purposes of the simulation, the world’s countries are grouped into several regions, some of which consist of single countries and some of which consist of many countries. The estimates for each region reflect both their own tariff liberalization and that of their trading partners. They are summarized in table 1.

Table 1: Geographic impacts of global tariff liberalization

Note: Changes in trade are for merchandise trade, except for oil and gas, and are presented with respect to a 2007 baseline. The simulation removes all tariffs and tariff-rate quotas, export taxes and subsidies. All estimates are presented in volume terms, i.e. in terms of prices in the pre-liberalization baseline. Some regions in the simulation model have been aggregated for presentation purposes.

As a rule, the regions with the largest current tariffs gain the greatest benefits from their removal. Though trade liberalization does increase exports, countries generally achieve most of the gains from trade through cheaper imports. 

The results of the tariff liberalization should be considered as an upper bound, as they reflect the current 2007 baseline in the GTAP model. A large share of the estimated gains from tariff liberalization occurs in two regions: China/Hong Kong SAR/Taiwan and Russia and other former Soviet Union. The baseline tariffs in the model do not include tariff reductions from processing trade or other duty drawback programmes. Such programmes cover a large share of China’s trade, meaning that its actual applied duties are significantly lower than the most favoured nation duties appearing in the model. The model also does not reflect duty reductions implemented after 2007, such as those expected as a result of Russia’s recent World Trade Organization (WTO) accession. If both processing trade programmes and Russia’s WTO accession were fully reflected in these results, the remaining gains to be achieved by global tariff elimination would be estimated at less than the approximately US$ 400 billion shown here. 

The Effect of Trade Facilitation 

An ambitious scenario – improving all countries halfway to Singapore

Substantial literature suggests that improvements in trade facilitation are associated with increased trade flows. Following this literature, there is an estimated impact of a 1% improvement in the Enabling Trade Index (ETI) on trade between pairs of countries, using a standard econometric method for analysing trade flows known as gravity modelling. The ETI scores each country’s environment for trade on a scale of 1 to 7, with 7 being best. In particular, the focus is on the average of sub-index B (Border Administration) and sub-index C (Transport and Communications Infrastructure, hereinafter “Infrastructure”) as being the two components of the index most closely related to trade facilitation. These estimates, which take into account country size, purchasing power, geography and other determinants of trade, suggest that an improvement of 1% in an exporting country’s score in border administration and infrastructure increases exports by 1.2%, while a similar 1% increase in an importing country’s score increases imports by 0.6%. Countries benefit from each other’s improvements in trade facilitation, so that if both the exporter and the importer improve border administration and infrastructure by 1%, trade between the two countries should expand by 1.8% (1.2% + 0.6%). The details of the gravity model are presented at the end of this note.

Figure 38: Expected increases in trade will vary on the current Enabling Trade Index of a country

Examples for particular trade flows

As an illustration of the level of improvements envisioned in a halfway-to-global-best practice scenario, table 2 shows some examples of countries’ current scores for the ETI components representing border administration and infrastructure, along with comparison countries to represent the degree of improvement modelled in the scenario.

Table 2: Illustrative improvements in trade facilitation – ambitious scenario

Clearly this is a high degree of ambition, but it is useful to show the potential gains available from trade facilitation.

The next step is to translate the improvements in trade facilitation into bilateral increases in trade, aggregating these to the regions in the model. As noted above, countries benefit both from their own trade facilitation and from their trading partners’ improvements. Applying the benchmark improvement in trade facilitation to pairs of regions gives estimated increases in trade for each of the regions of the model. Some examples of the estimated increases in trade are given in table 3.

Table 3: Illustrative increases in trade in the ambitious scenario 
(improving trade facilitation half-way to global best practice)

These estimates are used as an input into the simulation model, which then shows the implications of the trade increases for GDP and for the sectoral distribution of trade. These implications include changes in relative prices, and changes in production and consumption levels in all regions of the world consistent with global clearing of markets and reflecting an expansion of trade approximating that implied by the econometric results shown above.⁷

Geographic impacts on global GDP and trade

Table 4 shows the estimated GDP effects from improving border administration and transport and communications infrastructure halfway to global best practice. There are gains in GDP and trade for all regions, with the largest gains for lower income regions, which would make the biggest improvements in trade facilitation in the ambitious scenario. The modelled gains in GDP range from 12.0% for sub-Saharan Africa to 2.0% for Japan. Countries and regions benefit both from their own trade facilitation and from their partners’ trade facilitation. Gains in GDP arise both from cheaper imports and from improved export opportunities.

Table 4: Estimated GDP effects from improving border administration and transport and communications infrastructure halfway to global best practice

Note: Changes in trade are for merchandise trade, except for oil and gas, and are presented with respect to a 2007 baseline. All estimates are presented in volume terms, i.e. in terms of prices in the pre-liberalization baseline. Some regions in the simulation model have been aggregated for presentation purposes. 

Extension to sub-indices A and D

The effects of improved border administration and transport and communications infrastructure, which account for approximately 59% of the cost increases due to supply chain barriers, have been estimated. The estimate of trade facilitation benefits does not include improved market access or business environment, but the Forum survey indicated that the costs associated with these two factors are about two-thirds of those associated with border administration and infrastructure.⁸

A modest scenario – improving all countries halfway to regional best practice

Here a more modest scenario is considered in which all countries improve trade facilitation not to global best practice (Singapore) but to regional best practice. This is done for two reasons. First, it may be difficult for countries to achieve the improvements in border administration and infrastructure envisioned in the ambitious scenario, so it is of interest to show the gains that may be achievable with a more modest effort. Second, the improvements in a regional best practice scenario are uneven, since the best practice is different in every region. Thus, the modest scenario reflects a case in which some countries’ efforts in trade facilitation lag behind their neighbours more than would be expected given their current performance. The consequences of such uneven improvements can be illustrated on a global scale.

As noted earlier, the simulation model consists of some single-country regions. In order to give all countries scope for improvement in the modest scenario, “regional best practice” is defined with respect to the seven regions used in World Bank statistics, six geographic regions for developing countries and a non-geographic “high income” region. The regional best practice countries in terms of border administration and infrastructure are Malaysia (East Asia and Pacific), Lithuania (Europe and Central Asia), Singapore (high income), Chile (Latin America and Caribbean), Tunisia (Middle East and North Africa), Sri Lanka (South Asia) and Mauritius (sub-Saharan Africa). Illustrations of the effect of improving each country’s Enabling Trade Index average score for border administration and infrastructure to regional best practice appear in table 5. 

Table 5: Illustrative improvements in trade facilitation – modest scenario

Several features of the modest scenario are worth noting. First, because the high-income countries are still improving halfway to global best practice, the same improvements are assumed in trade facilitation for these countries as in the ambitious scenario. For developing countries, smaller improvements are assumed. Second, among the World Bank regions, the standard for regional best practice is lowest in South Asia. This scenario thus assumes that Bangladesh and India do not make very large improvements in trade facilitation in order to go halfway to Sri Lanka. This feature of the scenario is illustrative; for example, a piece of sub-Saharan Africa, or some other grouping of countries, with lower regional best practice could have been constructed. But it shows what happens if any region lags in implementing trade facilitation.

In parallel with the ambitious scenario, table 6 illustrates the estimated increase in trade volumes between pairs of countries associated with the modest scenario. These are naturally smaller than the gains shown in table 3, except for trade among the high-income countries (illustrated by Singapore-Canada), which grows by the same amount as in the ambitious scenario.

As in the ambitious scenario, the increases in trade shown in the table are provided to the simulation model, which reconstructs a global equilibrium in which all markets clear and which reflects increases in trade approximating those illustrated in table 6.

Table 6: Illustrative increases in trade – modest scenario

Finally, table 7 illustrates the estimated GDP and trade effects from the modest scenario. These estimates highlight several points. First, even with more modest improvements in trade facilitation a significant portion of the gains in the ambitious scenario can be achieved (US$ 1.5 trillion as compared to US$ 2.6 trillion in the ambitious scenario). Second, the uneven implementation envisioned in the modest scenario is reflected in the economic outcomes. South and Central Asia, for instance, includes countries that do relatively little trade facilitation in the modest scenario. Its estimated gain in GDP is thus significantly smaller than in the ambitious scenario (1.8% versus 8.0%), its imports remain virtually unchanged, and its exports actually decline a little. By contrast, a larger share of the available gains in the modest scenario relative to the ambitious scenario accrues to higher income regions, which are assumed to engage in the same degree of trade facilitation in both scenarios. The greater trade imbalances in the modest scenario likely reflect the fact that countries achieving only modest improvements would find their exports to be relatively more expensive and less competitive than those from countries doing more to reduce supply chain barriers. 

Table 7: Estimated GDP effects from improving border administration and transport and communications infrastructure halfway to regional best practice 

Note: Changes in trade are for merchandise trade, except for oil and gas, and are presented with respect to a 2007 baseline. All estimates are presented in volume terms, i.e. in terms of prices in the pre-liberalization baseline. Some regions in the simulation model have been aggregated for presentation purposes. 

Details of gravity modelling 

The relationship between improvements in trade facilitation and increases in bilateral trade flows was estimated using a gravity model for all country pairs for which the Enabling Trade Index (ETI) is available. These increases in trade were then aggregated to the regions used in the GTAP model.

The gravity model estimated is of the following basic form, used for exploratory analysis:

(1) Where: 

IMP_ij represents the non-fuel imports of country i from country j. Measurement of import flows are likely to be more accurate because they represent a tax base. Where IMP_ji is missing, export data is used instead, if available. Non-fuel imports correspond to total imports minus imports in chapter 27 “Mineral fuels, oils, distillation products, etc.” of the Harmonized System (HS) 1996 classification. Data cover the period 2007–2010 and were retrieved from the United Nations Comtrade Database using the WITS interface on 10 October 2012. Trade data for Taiwan was retrieved from Taiwan’s Bureau of Foreign Trade on 11 October 2012. They cover total imports – not just non-fuel trade. Import flows from Taiwan are approximated by Taiwan’s export flows.

DIST_i,_j is the distance between i and j, expressed in kilometres. The distance is weighted to account for the geographic distribution of population within a country. Distance data come from the Centre d’Etudes Prospectives et d’Informations Internationales (CEPII).

GDP and GDPPC denote the gross domestic product (GDP) and the GDP per capita, respectively. Both measures are valued at purchasing power parity (PPP). Figures come from the April 2012 edition of the International Monetary Fund’s World Economic Outlook.

REM is a measure of economic remoteness. It corresponds to the sum of distances between a country and all the other countries weighted by their relative economic size. Formally for country i: 

ETI_i is country i’s overall score in the 2009, 2010 and 2012 editions of the Enabling Trade Index.⁹ For any given edition of the ETI most of the composing indicators are from two years before. Therefore the 2009 edition of the ETI needs to be recoded as 2007, the 2010 edition as 2008, and the 2012 edition as 2010. This ensures the best alignment possible between the indicators composing the ETI and the other covariates of the model.

D_k are a set of dummy variables that take the value 1 when a certain condition is met and zero otherwise. The source is CEPII unless mentioned otherwise. CONTIG_ij is 1 if i and j share a common border. COMLANG_ij is 1 if i and j have at least one common official language and/or if one language is spoken by at least 9% of the population in both countries. LLOCKED_ij is 1 if i, j, or both, are enclosed. Data on landlocked status were collected by the authors from various sources. 

Finally, Y2008 and Y2010 are two dummies controlling for time-specific effects.

The sample consists of 46,296 observations: 14,520 observations for 2007 (121 economies covered), 15,006 observations for 2008 (123 economies), and 16,770 for 2010 (130 economies covered).¹⁰

Using model (1), the hypothesis that the factors included in the ETI together play some role in explaining the variance of bilateral trade flows is tested. Drawing on the latest research, a Poisson pseudo-maximum likelihood estimation method is relied on.¹¹ In their influential article, Santos Silva and Tenreyro (2006)¹² show why Poisson is superior to other estimation techniques when it comes to constant elasticity models. Poisson has the major advantage of dealing with the heteroskedasticity issue, as well as with the presence of zeros in the dependent variable, two features of the gravity model that traditional log-linear methods do not handle properly, thus leading to biased estimates.¹³ With Poisson, the gravity model need not be log-linearized and can be estimated in its original multiplicative form.¹⁴  

 A number of alternate specifications of the model were tried. The most important choice regarding the specification regarded whether the ETI index should be used in its entirety, or whether it was alternately possible to identify separate effects for each of the ETI sub-indices, namely (a) market access, (b) border administration, (c) transportation and communications infrastructure, and (d) business environment. It proved unfeasible to identify the effects of the four components separately because of a high degree of multicollinearity, i.e. the four sub-indices are highly correlated with each other. On the other hand, using the entire ETI index led to unrealistically high estimates of improvements in the index, about three times larger than those used in the report.

The final specification chosen expands the ETI index into three subparts. The average of the sub-index scores for border administration and transportation/communications infrastructure is designated as “trade facilitation” and denoted as ETIbc in the final specification. The coefficients for ETIbc are used to calibrate the estimated trade expansion effects of trade facilitation. The remaining sub-indices, denoted as ETIa and ETId, are utilized as control variables. The final specification also utilizes country-pair fixed effects, which sweep out the effects of distance, contiguity, common language and landlocked status. This leaves a final specification of the following form:

This leads to the following estimates in table 8. In this table, recall that the subscript i denotes variables describing the importer and j denotes variables describing the exporter. The primary estimates of the impact of trade facilitation are in bold. The constant term is not shown. The standard errors shown are semi-robust.

Table 8: Poisson estimates of gravity equation

Overview of Barrier Quantification in Literature (elaborate)

Supply chain barriers to trade as discussed in literature have often been quantified, but usually for a specific region, sector or barrier. Most of these studies have been top-down reviews using macro models to identify the impact of these barriers on trade. A brief overview of the quantification studies and a summary table of a selection of quantifications follow.

The literature produced thus far has followed a trajectory from tariff-focused and geographically specific work to a broader understanding of barriers and cross-regional interactions. This move has been relatively recent – in the last 10 years – and the breadth of barriers discussed largely falls under four main areas of work.  

The most comprehensive and recent analysis of policy-induced barriers comes from Portugal-Perez and Wilson (2012),¹⁵ in which four main categories of barriers are discussed. These barriers include physical and technological infrastructure on the hard side as well as the softer border efficiency and business environment. Portugal-Perez and Wilson found that the bulk of the impact on trade volume, in ad valorem equivalent (AVE) tariff reduction, comes from the technology employed by the target country as well as the existing business environment. These factors contributed to a reduction in AVE, ranging from 5% to as high as 65%, assuming underperforming countries could improve halfway to the top performer in each region. 

This work stands at the front of a body of prior work that is mostly geographically focused and built from a top-down perspective. Iwanow and Kirkpatrick (2009)¹⁶ demonstrated that studying specific geographies (Africa in their case) can bias the quantification of the effects of policies, highlighting the need for a top-down case study-based perspective.  

Most previous work has focused on gravity models to estimate the impact of reductions in trade barriers. Wilson, Mann and Otsuki (2003)¹⁷ used a four-indicator structure that is similar to Portugal-Perez and Wilson’s (2009), focused on Asia-Pacific Economic Cooperation (APEC) countries, to determine that trade between APEC countries could increase by 21% if below-average countries raised their indicators halfway to average, largely driven by infrastructure. In 2005,¹⁸ the group looked at a wider selection of countries with a similar set of metrics and found that the same assumption of improvement in trade facilitation yielded a 10% increase in trade. 

Below are the main categories of barriers that are commonly identified in literature.

Category 1: Customs environment and non-tariff measures 

There is a healthy body of work focused on specific non-tariff measures (NTMs), such as technical barriers to trade like sanitary and phytosanitary measures. Kee, Olarreaga and Nicita (2006)¹⁹ developed indexes to measure the size of both specific duties and NTMs and found that, similarly, low-income countries impose and are subject to greater barriers to trade (and are thus best positioned to benefit from relaxing those barriers). Looking at total trade, their research found average AVE tariffs to be 10.7% for the sample studied, with a standard deviation of 23%. Otsuki, Wilson and Sewadeh (2001)²⁰ looked at sanitary and phytosanitary (SPS) regulations around aflatoxin standards on groundnuts using European and African trade data; they found that tightening standards by 10% led to an 11% reduction in trade. Disdier, Fontagné and Mimouni (2007)²¹ looked more broadly at different sectors to measure SPS impacts and found that SPS restrictions affect developing countries differentially when they export to developed countries. Moenius (2004)²² noted that SPS standards differ in their impact: They typically harm trade in agricultural goods, while they improve trade in manufacturing goods, because the information provided by those standards improves a product’s ability to sell.

Czubala, Shepherd and Wilson (2007)²³ showed that, although product standards are generally negatively correlated with trade growth, when customs standards are harmonized broadly, the correlation is significantly less negative. 

These works have been augmented by a recent World Trade Organization report,²⁴ which delves into the NTMs of the like described above. Since this paper focuses on measures beyond these, the review of the literature on this topic is capped here. 

In addition to those mentioned above, several research papers have examined rules of origin as an additional barrier to trade. Most notably, Cadot and de Melo (2007)²⁵ demonstrated that rules of origin generally impose rents on the developing partner that are collected by the developed partner – a form of non-explicit tariffs. They showed that a 10% decrease in the regional value content (the percentage of unit cost coming from a rule-of-origin country) leads to a 2% to 5% increase in utilization (a measure of the cost of preference for a rule-of-origin partner).

Category 2: Efficiency of ports

Hummels and Skiba (2004)²⁶ found that doubling trade quantities led to a 12% reduction in shipping costs when they looked at importer-exporter pairs to show that per-unit barriers differentially benefit higher quality goods. To quantify how much the size of a port directly impacts its efficiency, Arvis, Raballand and Marteau (2007)²⁷ demonstrated that goods in Africa are affected by the high cost of freight services and unpredictable transport times. This effect is enhanced for landlocked nations that are impacted by not only physical constraints but also high, widespread rent activities and poor transit systems. Limão and Venables (2001)²⁸ examined how poor port infrastructure significantly deters trade: moving from median levels of infrastructure (including port efficiency) to the 75th percentile raised transport costs 12% and reduced trade flows by 28%. 

Category 3: Infrastructure of the service sector (transportation, trucking, communications and others)

Portugal-Perez and Wilson (2009)²⁹ looked at the effects of improving logistics in Africa and found, for example, that when Ethiopia’s level of logistics was improved halfway to South Africa’s, that resulted in an average of 7.5% AVE tariff reduction in Ethiopia.

Behar, Manners, and Nelson (2012) similarly considered the effects of logistics, but for a large cross-section of countries. The authors found that a one-standard deviation increase in an international logistics index is associated with a 36% increase in exports for an average-sized country.³⁰

Focusing on transit delays, Freund and Rocha (2010)³¹ showed that a one-day reduction in inland travel time in Africa led to a 7% increase in exports, the equivalent of reducing tariffs by 1.5%. Djankov, Freund and Pham (2010)³² looked more broadly geographically and found that saving a day led to a 1.3% increase in trade; their study also potentially indicates that the effect of saving a day is more pronounced where it is more needed. Korinek and Sourdin (2009)³³ found that an extra day of delays at the border hurt trade by 4.5%, a result they reconfirmed in 2011.

Looking at a specific example, Buys, Deichmann and Wheeler (2006)³⁴ investigated the African Development Bank’s (ADB) proposed network of roads and found that the US$ 20 billion investment would yield US$ 250 billion in additional overland trade over 15 years. These excess returns seem less surprising in the wake of results from research such as the African Development Bank (2003), which showed that roadblocks and border controls contributed four to seven days to a three-day trip. Limão and Venables (2001) found that transport costs of the median landlocked country are 55% higher than the median costal economy due to poor infrastructure, and that a 10% drop in transport costs would increase trade flows by 25%. 

Category 4: Business environment or corruption

Other research looked at pieces of the supply chain. Abe and Wilson (2008)³⁵ looked specifically at corruption and transparency in APEC countries and found that reducing these negative factors to the group average would increase trade by 8% to 11% in the region.  

Anderson and Marcouiller (2002)³⁶ looked at corruption as a barrier to trade, finding that if indexes of Latin American institutions were raised to the level of institutions in the EU, Latin American trade could rise to an estimated 34%, the same amount by which trade would expand if Latin American tariffs were reduced to tariff levels in the US. Helble, Shepherd and Wilson (2007)³⁷ showed that raising below-average APEC nations up to the regional average would also raise trade flows in the region by 7.5%. Mauro (1996)³⁸ found a strong correlation between a decrease of one standard deviation in corruption based on a widely used index and a four-percentage-point increase in a country’s investment rate over the same time period. 

In addition, Freund and Weinhold (2004)³⁹ concluded that increasing an exporter’s Web presence may impact export growth – a 10% increase in Web volume was correlated with 0.2% increase in export growth.

Summative barrier: Enabling Trade Index 

Several papers have also looked at the Enabling Trade Index (ETI), developed by the World Economic Forum, and found reasonably similar results. Korinek and Sourdin (2011)⁴⁰ found that a 10% increase in an exporter’s ETI led to a 36% increase in trade, all else equal (for importers the percentage was a bit lower). This is similar to the Forum’s Global Enabling Trade Report 2009, which found that a 1% increase in the average ETI score of any given country pair led to a 4% increase in trade flow, using a gravity model. 

Relative importance of non-tariff barriers

A number of papers have examined the importance of certain non-tariff barriers relative to tariffs. Arvis, Duval, Shepherd, and Utoktham (2012) found, for example, that, under certain model specifications, the effect of improving logistics on trade costs is nearly 10 times greater than that of a reduction in tariffs.⁴¹ Looking at trade volume, Hoekman and Nicita (2011) estimated that if low-income countries converged to the average of middle income countries in logistics performance, exports could increase by 15.1%. This figure compared with a 10.6% increase associated with tariff convergence.⁴²

Estimating welfare gains 

As with this report, there is existing literature that considers welfare effects in addition to trade costs and trade volume. Adler, Brunel, Hufbauer and Schott (2009) estimated the potential increases in trade and income for a sample of 22 countries in response to improvements in trade facilitation. The authors applied the coefficients estimated by Wilson, Mann, and Otski (2005) to all 22 countries in the sample. To be conservative, the coefficient for Organisation for Economic Co-operation and Development (OECD) countries was applied to all the 22 countries in the sample, including developing countries. The authors found that improvements in trade facilitation could lead to a US$ 340 billion increase in exports and a US$ 385 billion increase in GDP.⁴³

Decreux and Fontagne (2009) analysed a scenario in which countries above the world median reduced clearance times for exports and imports by half, and found that GDP would increase by US$ 99 billion.⁴⁴

Table 9: Literature overview: Impact of change in metric on trade flows, except where specified differently

Table 10: Literature overview: Additional quantification of trade barriers

US–Mexico Competitiveness Agenda – The Urgency of Accelerating the Pace

Author: Beatriz Leycegui, Senior Fellow, Autonomous Technological Institute of Mexico and Undersecretary for Foreign Trade of Mexico (2006-2011)

To successfully integrate and compete in global value chains, the leaders of Canada, the United States and Mexico need to address with a greater sense of urgency the competitiveness agenda of the region. The agenda should include the following priorities: regulatory cooperation, 21st century borders and intellectual property. 

Regulatory cooperation

Costs in time and money associated with regulatory compliance are one of the main obstacles to international trade. Therefore, regulatory cooperation is central to the competitiveness agenda.  

To facilitate trade, Mexico and the United States need to continue working on adopting compatible measures, either through harmonization or through mutual equivalence recognition. (Despite being different, they provide the same degree of protection.) That way, producers do not have to differentiate their products depending on the rules or measures required in the target market. Also, mutual recognition of conformity assessment procedures represents savings to producers, as it helps them avoid going through certification procedures several times.

To facilitate communication, coordination and cooperation among the more than 20 agencies in charge of regulation in each country, the North American Free Trade Agreement (NAFTA) parties have complemented the institutional framework contemplated under the agreement creating bilateral high-level regulatory cooperation councils, which comprise senior-level regulatory, trade and foreign affairs officials. After extensive public consultations, the US-Mexico Council enacted a Working Plan focused on seven priority areas: food, transportation, nanotechnology, electronic medical records, oil and gas, and conformity assessment procedures. Equivalence has been implemented in measures applicable to electronics, medical devices, food safety, aircraft components, and telecommunications equipment.

21st-century border

A significant part of the benefits in competitiveness derived from tariff elimination under NAFTA have been offset by excessive delays at ports of entry between the US-Mexico and US-Canada, as shown in the accompanying graphs. In 2010, approximately 95% of the flow of goods and persons between the US and Mexico took place through land transportation. Addressing this situation has been a priority.

Despite efforts under NAFTA´s institutions and those put in place after 9/11, progress has not met the demand of this enormous challenge. The seamless and intelligent border became a thickened one. To improve coordination, in 2010 the US-Mexico Executive Steering Committee on Twenty-First Century Border Management was created. It comprises high level government officials from 17 offices of the United States and Mexico.⁴⁵ The committee focuses on improving infrastructure, establishing effective risk management procedures and facilitating pre-clearance of goods:

• Infrastructure After 10 years, three new ports of entry were inaugurated and the first new railroad in 100 years is being constructed, increasing the number of railroad border crossings from seven to eight. Railroad transportation is considered as a priority because of its greater efficiency, security and reduction of emissions.
• Risk management Mexico, together with its partners, is promoting effective risk management programmes to rigorously monitor high-risk trade while at the same time accelerating the flow of low-risk passengers and trade through the expansion of reliable traveller and safe load programmes.
• Pre-clearance Three pilot programmes are in the process of starting operations and, if successful, will become permanent and broader: Laredo, Texas Airport; Otay Mesa, Baja California for agriculture products; and Foxconn company facilities adjacent to Mexican customs in San Jerónimo, Chihuahua.
• Metrics The United States and Mexico have agreed to a common methodology to measure wait times at land ports of entry. As of August 2012, three of the seven research studies on wait times have been completed. It is necessary to carry out regular assessments to obtain complete information on the impact of implemented measures to compare the performance of both countries and, if convenient, to make the necessary adjustments to the adopted policies.

Figure 39: Each trade barrier can disrupt a company’s supply chain and lead to higher costs

Intellectual property 

To promote creativity and innovation, generate foreign and domestic investment and increase competitiveness, the NAFTA Parties subscribed to several commitments under the Agreement, followed by the Security and Prosperity Partnership, and most recently the signing of the Anti-Counterfeiting Trade Agreement.

Over the past years, the three countries have been working on these issues. Now, with US President Obama’s re-election and a new administration taking office in Mexico, there is an opportunity to adopt a common and comprehensive strategic vision that will hopefully accelerate the implementation of the former agenda. There is no more time to lose.