WORLD TRADE REPORT 2015

C. The theory and

measurement of
trade facilitation

This section first provides a conceptual framework

for understanding the economic effects of trade
facilitation – how improving trade procedures
reduces trade costs, and how that in turn affects
the pattern and volume of trade, the allocation
of resources, and economic welfare. Given that
trade facilitation can, in principle, be implemented
unilaterally, this section examines the reasons why
countries would want to include trade facilitation in a
multilateral trade agreement. Finally, it examines the
indicators – from narrower customs-related ones to
broader regulatory and infrastructural areas – that
have been developed to measure trade facilitation,
and identifies what indicators can best be employed
to estimate the economic benefits of implementing
the WTO’s Trade Facilitation Agreement.

56
 II. SPEEDING UP TRADE: BENEFITS AND CHALLENGES OF IMPLEMENTING
THE WTO TRADE FACILITATION AGREEMENT

Contents
1. Trade facilitation in models of international trade 58
2. The economic rationale for an international trade facilitation agreement 62
3. Measuring trade facilitation 65
4. Conclusions 69

C. T
TRADE FACILITATION
MEASUREMENT OF
 HE THEORY AND
Some key facts and findings

•• Existing models of international trade, including recent ones that take into account the
ways in which trade costs are compounded and magnified along supply chains, can
be used to better understand the trade and economic effects of the Trade Facilitation
Agreement (TFA). For example, the “iceberg” model of trade cost draws an analogy
between the way trade costs reduce the value of goods to both exporters and
importers and the way an iceberg melts as it moves through the ocean.

•• If a country improves its trade procedures so that trade costs are reduced, importers
benefit from a lower price, while exporters receive a higher price for the traded good.
Thus, trade facilitation benefits both exporting and importing countries.

•• Incorporating trade facilitation in a multilateral agreement creates additional benefits

compared to what can be achieved unilaterally. It provides greater legal certainty,
helps reforming governments marshal support from domestic constituents, assists
with the adoption of similar trade procedures and coordinates the provision of donor
support for capacity-constrained developing countries.

•• A wide range of trade facilitation indicators has been developed by international

organizations and within academic literature. Among these, the Organisation for
Economic Co-operation and Development’s (OECD) Trade Facilitation Indicators (TFIs)
are well suited to analysing the trade and economic effects of implementing the TFA,
as these indicators are mapped to the provisions of the Agreement.
57
 WORLD TRADE REPORT 2015

1. Trade facilitation in models of because of demand for variety and increasing returns
to scale in production. Finally, a branch of more recent
international trade
models incorporates differences in the productivity
Trade facilitation aims to reduce trade costs, which of firms which result in only some of them being able
in their broadest definition include all costs, apart to overcome the fixed trade cost of entering export
from the cost of production, incurred in getting a markets (Melitz, 2003). A second branch focuses on
good from a producer to a final consumer (Anderson fragmented production and value chains and tells us
and van Wincoop, 2004). Among other constituents, that trade costs are particularly pernicious because
they include the costs of transportation, tariffs, non- they are cumulated and magnified along the supply
tariff measures and inefficient trade procedures. This chain (Yi, 2010).
section begins with a graphical analysis of the impact
of trade facilitation using a partial equilibrium supply- (a) A simple “iceberg” partial equilibrium
and-demand model. However, because the effects
of trade facilitation on a particular market may spill
model
over to other markets, the analysis is extended to a
The “iceberg” model by Samuelson (Samuelson, 1954)
general equilibrium setting using standard models of
is a useful device for analysing the effect of trade
international trade, from the classical models to the
costs, although it was originally designed to model
most recent models of global value chains.
transportation costs (see Box C.1). Inefficient trade
The early or classical trade models explain why procedures increase the cost of trade and drive a
trade emerges between dissimilar countries (inter- wedge between the price received by the producer
industry trade) based on differences in productivity of the good and the price paid by the consumer. This
(Ricardo, 1817) or endowment in factors of production represents a pure loss (“deadweight loss”) akin to the
(Heckscher, 1949; Ohlin, 1934). While these early part of the iceberg’s mass that is melted away as it
trade models do not bring trade costs explicitly into the moves through the ocean. In the iceberg model, trade
analysis, later trade models do. The new trade theory costs are proportional to the value of goods shipped,
(Krugman, 1979; 1980) explains why trade between but the main results will continue to hold even in cases
similar countries (intra-industry trade) takes place where trade costs are additive instead.1

Box C.1: The “iceberg” model

Figure C.1 gives a graphical illustration of the iceberg model for an imported good. For simplicity, it is assumed
that the good is not produced domestically. Domestic demand is given by the line D while foreign supply is
given by the line S. In the initial market equilibrium, trade costs are high, denoted by δ0 . Domestic consumers
pay a price of Pd 0 and foreign producers receive Ps 0 , which is lower by the trade cost δ0 while the total quantity
imported is equal to Q 0 .

Figure C.1: Iceberg partial equilibrium model

Price

Increase in consumer surplus due to trade facilitation

Pd0 A S

B
P* δ0

Ps0 C
D

Increase in producer surplus due to trade facilitation

Q0 Q* Quantity

58
 II. SPEEDING UP TRADE: BENEFITS AND CHALLENGES OF IMPLEMENTING
THE WTO TRADE FACILITATION AGREEMENT

Box C.1: The “iceberg” model (continued)

Assume that the country improves its trade procedures so that trade cost is reduced to zero. The quantity
of goods imported in equilibrium rises to Q *, domestic prices fall to P * and foreign prices rise to P * as well.
The price wedge caused by trade costs disappears. Both domestic consumer and foreign producer welfare
increase by the amounts indicated by the trapezoidal areas P d 0ABP * and P s 0 CBP * respectively. Observe that
trade facilitation improves the terms of trade of both countries because it simultaneously reduces the price
paid by domestic consumers for imports and increases the price received by foreign exporters. This terms-of-
trade improvement in both countries (a “win-win” outcome) as a result of trade facilitation is taken up again
in subsection C.3, which deals with the economic rationale for a multilateral agreement on trade facilitation.
The gains from trade facilitation will be smaller than those shown in Figure C.1 if inefficient trade procedures
create rents captured by some economic agents rather than pure deadweight losses (Dee, 2006). The analysis
has also not taken the cost of implementing trade facilitation reform into account, which would reduce the
gains shown in Figure C.1.

(b) Classical general equilibrium models land (Heckscher, 1949; Ohlin, 1934). In these models,
of trade countries specialize in goods in which they have a
comparative technological advantage relative to other
The analysis has focused on a single market so far, and countries or in goods that use their abundant factors

C. T
of production more intensively. They then import the

TRADE FACILITATION
MEASUREMENT OF
 HE THEORY AND
is therefore only partial in nature. It will be useful to
know whether these results are modified or additional other goods from their trade partners. These models
insights are obtained when the analysis is extended to provide a rationale for inter-industry trade (e.g. a
a general equilibrium setting. country exporting automobiles and importing wheat)
but not intra-industry trade (e.g. a country exporting
In classical models, gains from trade result because sports cars and importing sports utility vehicles).
countries are assumed to possess either different Box C.2 provides a more detailed discussion on the
relative productivities (Ricardo, 1817) or endowments effects of trade cost in classical models of trade.
of factors of production such as labour, capital and

Box C.2: The effects of trade costs in classical trade models

Classical trade theories explain trade in homogeneous goods under constant returns to scale and perfect
competition. Factors of production are assumed mobile across sectors within one country, but immobile across
countries. The basic versions of these models assume that two different final goods are produced.

The Ricardian model

The assumption motivating trade in the Ricardian model is that countries have different relative labour
productivities. This implies that under autarky, i.e. when countries do not trade at all with one another, the relative
price of one good expressed in terms of the other good differs between the countries.

In a hypothetical world without trade costs, this difference in relative prices opens up opportunities for welfare-
enhancing international trade at a world price lying between the two autarky prices, which is determined by
countries’ consumption preferences and relative sizes (Markusen et al., 1995). At least one country specializes
completely in the production of the good in which it has a comparative advantage.

Inefficient trade procedures result in trade costs that drive a wedge between the relative prices faced by the
two countries. They now face international prices closer to their respective autarky price. They may continue to
remain specialized but there will be less consumption and trade and hence lower economic welfare. If trade costs
become high enough, the international price faced by one country can become less favourable than its autarky
price and trade ceases altogether, returning both countries to their autarky equilibria. Relative country sizes
play a role in how likely this may happen. If one country is much larger, then the frictionless international price is
already close to its autarky price and trade ceases for smaller transaction costs.
59
 WORLD TRADE REPORT 2015

Box C.2: The effects of trade costs in classical trade models (continued)

The Heckscher-Ohlin model

In contrast to Ricardo, the Heckscher-Ohlin model assumes the same productivity in both countries. There
are two factors of production, capital and labour, and endowments of these factors of production vary across
countries, making one country labour-abundant and the other country capital-abundant. There are two sectors
producing two different goods; one sector, for instance automobiles, uses capital more intensively and the other
sector, for example textiles, uses labour more intensively.

In autarky, relative prices in the two countries will differ because of differences in their factor endowments.
The price of textiles relative to automobiles is lower in the labour-abundant country and higher in the capital-
abundant country. If trade is opened up and in the absence of trade costs, both countries produce more of
and export the commodity that uses their abundant factor intensively: i.e. the labour-abundant country exports
textiles and the capital-abundant country exports automobiles. But, unlike in the Ricardian model, complete
specialization is unlikely. They will trade at a world price lying between the two autarky prices, which means the
world price of textiles relative to automobiles is higher than the autarky price in the labour-abundant country
and lower than the autarky price in the capital-abundant country. Another important outcome of free trade is a
convergence of factor prices in the two countries (factor price equalization).

Trade costs drive a wedge between the relative prices faced by the two countries, creating a situation where
they both face international prices closer to their autarky price. Countries will be less specialized, and both
trade and consumption will be lower compared to a frictionless world. Again, economic welfare suffers as a
consequence. Furthermore, this wedge in the relative prices faced by the two countries also means a divergence in
factor prices.

Irrespective of their differences, trade costs work possible by a number of assumptions: consumers
through the same mechanism in these classical trade prefer variety in consumption, the market is populated
models. Inefficient trade procedures drive a wedge by firms selling different varieties of a good and there
between the relative prices faced by the two trading are increasing (internal) returns to scale in production,
countries. These relative prices move closer to the initial meaning that a firm’s average cost of production falls
autarky price, reducing the scope for specialization and as its volume of production increases.
trade. As a result, consumption possibilities are lower,
and so is economic welfare. The theory predicts that trade costs can have a
disproportionately adverse impact on small developing
One interesting result from the Heckscher-Ohlin economies. Typically, small developing economies have
model concerns how trade facilitation improves the large agricultural or natural resource sectors typified by
real income of the abundant factor of production. By constant returns to scale, and only a small manufacturing
reducing trade costs, it leads to greater specialization sector. In contrast, big developed economies have a
in the sector that uses the abundant factor more large manufacturing sector operating under increasing
intensively. This increases the demand for the returns to scale. In this setting, trade costs lead both
abundant factor and increases the real return to the to less trade and to a disproportionate relocation of
factor. If one of the countries involved is a labour- manufacturing to the big developed countries (the
abundant developing country, trade facilitation can “home market effect”). Meanwhile, small developing
make workers better off. countries become concentrated in the agricultural or
natural resource sector.
(c) The “New Trade Theory” – monopolistic
competition The key to explaining this result lies in the tension
created between the consumer’s love of variety and
In contrast to the classical theories, the “New Trade increasing returns to scale. With open trade and
Theory” (Krugman, 1979; 1980) explains why countries zero trade costs, consumers in the big developed
engage in intra-industry trade. This is a valuable result country will purchase both foreign and domestic
because the great bulk of global trade is intra-industry manufactured goods because of their preference for
rather than inter-industry in nature. The ability of the variety. All things being equal, love of variety leads to
theory to explain this feature of global trade is made more trade. On the other hand, increasing returns to
60
 II. SPEEDING UP TRADE: BENEFITS AND CHALLENGES OF IMPLEMENTING
THE WTO TRADE FACILITATION AGREEMENT

scale gives a cost advantage to manufacturing firms While it might be obvious that a reduction in trade
in the developed country because of the size of the costs will increase a country’s exports, this literature
market and the larger scale of production that could shows the need to distinguish between the two
be achieved by firms there. All things being the same, ways in which trade costs can be reduced and the
consumers in the developed country will prefer to different ways exports can increase as a consequence
purchase lower-cost domestic varieties than higher- (Chaney, 2006). Trade costs can be categorized as
cost foreign varieties. either variable or fixed. Variable trade costs are costs
that have to be paid on every unit of export. Tariffs
Inefficient trade procedures that lead to higher trade are a prominent example of variable trade costs,
costs upset this balance by making purchases (imports) as an importer needs to pay duty on every unit he
of foreign varieties more costly. As a consequence, imports. Fixed trade costs are costs that have to be
consumers in the developed country substitute away incurred independently of the volume of exports.
from foreign varieties towards domestic varieties. A firm deciding on whether to enter a particular market
This shift in demand towards domestic manufactured might have to incur a cost to learn about the trade
goods gives greater scope for what are already procedures in that country. These are costs incurred
powerful scale forces to operate. The manufacturing even before it ships a single product to the foreign
sector in the big developed country expands even market.
more while it shrinks in the small developing country.
This analysis suggests that small developing countries An increase in exports can take place along two
that want to diversify their economies have a strong dimensions or margins: the intensive and extensive
interest in lowering trade costs, as this reduces margins. The intensive margin refers to existing

C. T
incentives for manufacturing to concentrate in the exporters increasing the volume of their exports, while

TRADE FACILITATION
MEASUREMENT OF
 HE THEORY AND
biggest markets. the extensive margin refers to an increase in exports
achieved by new firms entering the export market.
(d) The “New New Trade Theory” –
A reduction in variable trade costs affects both the
heterogeneous firms
extensive and intensive margins of trade. It enables
existing exporters to capture a larger share of the
In the classical theories of trade, it is countries that are
export market and firms with a lower level of productivity
the objects of interest and analysis. In the last decade,
than incumbent exporters to enter the export market. A
new models of trade have emerged that have shifted
reduction in fixed trade costs only affects the extensive
this focus to firms – the so-called “heterogeneous
margin of trade. Trade facilitation will reduce both
firms” literature (Melitz, 2003). These models are
fixed and variable trade costs, making it possible for
motivated by empirical studies that reveal the striking
incumbent exporters to capture a larger share of the
diversity of firms in terms of size, productivity and
international market, and for firms that have never
participation in international trade (Bernard et al.,
exported before to begin to do so.
2007a; 2007b).

If trade facilitation reduces both fixed and variable

The studies find that only a small number of firms export,
trade costs, this analysis implies that one should see
and that the vast majority are only able to sell in the
trade expansion along both margins. Those enterprises
domestic market. The reason for this disparity is that
that are currently engaged in international trade as
firms differ in productivity: those with low productivity
exporters will most likely expand the volume of their
do not survive competition, more productive firms can
exports. In addition, firms that were shut out of foreign
compete but only in the domestic market, while only the
markets will now find it possible to enter these markets
most productive firms are able to enter and compete
and begin exporting. These new firms may be smaller
in the export market. There are two productivity
and less productive than current incumbents but the
thresholds: the minimum level needed for a firm to
survive, and the level at which a firm starts exporting reduction in trade cost now gives them an opportunity
part of its production. to participate in international trade.

The main result of the heterogeneous firms literature (e) Supply chain models
is that any reduction in trade costs brings the two
thresholds closer to each other, increasing the range of Supply chain models of trade emerge at around the
firms that are driven out by competition and the range same time as the heterogeneous firms literature. 2
of firms that enter the export market. This is beneficial While traditional trade theory assumes that each final
to the economy, as resources (capital and labour) are good is produced entirely within one country, supply
released from the least productive firms and reallocated chain models recognize that the parts and components
to the most productive firms. that make up complex final goods such as electronic
61
 WORLD TRADE REPORT 2015

products or motor vehicles are made in many different 2. The economic rationale for an
countries.
international trade facilitation
As a result of this way of organizing global production, agreement
trade costs become amplified (Yi, 2010). This occurs
through “cumulation” and “magnification” effects. Given the widespread benefits of trade facilitation,
Trade costs are cumulated through the different every country should have an incentive to undertake
stages of the value chain, as goods cross national reforms on its own. The questions, therefore, are:
borders multiple times while they are in process. They why is trade facilitation still on the agenda of many
are magnified because the trade costs at any stage countries; and why have these countries decided to
must be paid out of the share of value added in the proceed with the reforms by signing the TFA?
cost of production.
Evidence reviewed in this report suggests that trade
The existence of the cumulation and magnification facilitation can stimulate trade, promote diversification
effects mean that trade costs have a far greater and increase aggregate welfare. It also shows that
deterrent effect on global value chain-related trade trade facilitation benefits both the economy that
than on trade involving only final goods. The higher takes facilitating measures and its trading partners.
the trade costs, the less scope there is for supply The discussion so far suggests that governments
chain trade. In the extreme case where trade costs are would not need to cooperate to derive the benefits
very high, it is not worthwhile to divide up production from trade facilitation and that they could benefit
between different countries, and only final goods are from proceeding unilaterally with the reforms. Yet, the
traded. This means that trade facilitation is crucial to signature of the TFA suggests that there are reasons
the viability of global value chains, allowing for more why incorporating trade facilitation in an international
specialization in those production stages in which agreement creates additional benefits.
countries have a comparative advantage. Any reduction
in trade costs, such as what would be made possible Economists have identified several rationales for trade
by the TFA, also becomes amplified in the opposite agreements. The first one is that trade agreements may
direction. The cumulation and magnification effects serve as a means to escape from a terms-of-trade-
explained above take effect, but in a positive way, driven prisoners’ dilemma. 3 Countries with sufficient
thereby lowering barriers and allowing more developing market power have an incentive to impose tariffs which
countries to become involved in global value chains raise their terms of trade, i.e. the (untaxed) price of their
(GVCs). exports relative to the (untaxed) price of their imports,
but lower the terms of trade of their trading partners.
More complicated production arrangements in GVCs In the absence of cooperation, this may give rise to a
have been analysed by Baldwin and Venables (2013). trade war, that is, a prisoners' dilemma situation where
They distinguish between “snakes”, i.e. sequential countries set their tariffs too high, and the volume of
production processes with each operation adding trade is inefficiently low. A trade agreement, according
value in a predetermined order, and “spiders”, which to the terms of trade theory, allows countries to derive
combine different intermediate inputs in an assembly benefits from reciprocally reducing their tariffs, thereby
stage. Any GVC can be viewed as a combination of escaping the prisoners’ dilemma.
spiders and snakes.
This rationale may also play a role in explaining an
Given these differences in structure, the impact of agreement on trade facilitation. First, if customs
trade facilitation on GVCs and trade will be more procedures and practices can be manipulated to
complicated and vary depending on the structure generate rents and governments can be captured by
of these chains. Firms face a trade-off between private interests, countries may end up in a terms-
setting up manufacturing sites in different countries of-trade-driven prisoners’ dilemma similar to the one
to reduce production costs and keeping production just described. However, more interestingly, even if
in one country to limit trade costs. In the case of inefficiencies at the border generate costs rather than
snake-type GVCs, a fall in trade costs would lead to rents, a slightly modified version of the terms of trade
greater fragmentation and offshoring of production explanation may shed light on the rationale behind a
and expansion of trade, although the results are less trade facilitation agreement if the implementation of
straightforward in the case of spider-type GVCs. trade facilitation measures is costly (see Box C.3).

62
 II. SPEEDING UP TRADE: BENEFITS AND CHALLENGES OF IMPLEMENTING
THE WTO TRADE FACILITATION AGREEMENT

Box C.3: The effect of inefficient customs procedures on an economy

Consider first the effect of inefficient customs procedures. As shown in Figure C.2, such procedures raise a large
country’s trade costs and the price of its imports, lowering its terms of trade while at the same time they cause
the partner’s terms of trade to deteriorate.4 Inefficient procedures raise the domestic price in the importing
country to P w+c and reduce the demand for imports which, if the country is large enough, may push down the
world price – i.e. the price received by exporters – from P w to P w ’. While in the case of a tariff, this reduction of
the world price generates a terms of trade gain equal to the area of the orange rectangle, it generates a loss
equal to the same area in the case of inefficient customs procedures. Overall, for the importing country, the
welfare effect of the inefficiency is a large deadweight loss equal to the sum of the areas of the striped trapezoid
and the orange rectangle.

Consider now the effect of trade facilitation. Trade facilitation, by eliminating cost-raising inefficiencies, generates
a welfare gain for both the importing country and its supplier. At the same time, however, implementing trade
facilitation measures is costly. The importing country has an incentive to invest in trade facilitation inasmuch as
the gains exceed the implementation cost. However, as explained, eliminating inefficiencies also benefits the
exporting country, as this imparts a positive externality on foreign exporters. This externality provides a rationale
for international cooperation on trade facilitation. Without a trade facilitation agreement, (i.e., under unilateral
decisions about making efficiency-enhancing investments in customs procedures) this positive externality will
result in too little investment in improving customs procedures by large importing countries. A prisoners’ dilemma
type situation may arise where two large importing countries do not invest enough in trade facilitation, thereby

C. T
imposing costs on each other. A trade facilitation agreement can help countries to internalize these positive

TRADE FACILITATION
MEASUREMENT OF
 HE THEORY AND
(terms of trade) externalities and thereby lead to greater investments in efficient customs procedures.

Figure C.2: Impact of inefficient custom procedures on welfare

Price

Demand Supply

Pw+c
Pw’+c
Pw
Pw’

Quantity

The second rationale identified by economists is that for the WTO TFA, it may allow governments to tie their
trade agreements can help governments address a hands against anti-facilitation lobbies. In other words,
credibility problem. The idea is that governments value commitment may be one of the rationales behind
trade agreements as a way to tie their hands against, the TFA.
and thus resist pressure from, lobbies. 5 According to
Hoekman (2014), this theory does not help much in Another possible rationale is proposed by Hoekman
understanding the rationale behind a trade facilitation (2014), who argues that the TFA reflects international
agreement because trading partners would not be in coordination or collective action considerations. As
a position to enforce an agreement by threatening to already mentioned, implementing trade facilitation
withdraw concessions. It would, indeed, be difficult for measures unilaterally yields significant economic gains
a government to selectively “unwind” trade facilitation as customs procedures become more transparent,
measures to enforce a trade facilitation agreement. predictable and efficient. However, if countries use
If, however, the agreement foresees the possibility of different approaches and adopt different standards and
using other enforcement instruments, as is the case procedures, there will be redundancy in documentary
63
 WORLD TRADE REPORT 2015

requirement and control procedures at the borders. If A similar line of reasoning can be applied to the
procedures differ between countries, exporters and coordination problem related to asymmetries in
importers need to learn about multiple standards, implementation costs and capacity. Indeed, the TFA
which can create significant learning costs. The foresees that richer members will provide assistance
adoption of common procedures can reduce the time and support for capacity-building to developing and
and costs required to become familiar with customs least-developed countries to help them implement the
procedures in different countries as well as improve agreement. 6 Without the agreement, many countries
the efficiency and timeliness of the movement of goods might not have engaged in trade facilitation because
through customs worldwide. Coordination among WTO they might have preferred to allocate scarce resources
members in the context of the TFA and the adoption to other priorities, which would have resulted in a
of common approaches towards customs and related suboptimal situation for all members. Coordination
matters could further increase the gains from trade benefits may thus explain international cooperation
facilitation by harmonizing customs procedures on trade facilitation. However, this explanation may
worldwide. This international coordination problem has not be sufficient in itself to explain the TFA. This is
been conceptualized in a game theory framework by because if a trade facilitation agreement only serves a
Snidal (1985) (see Box C.4). coordination purpose, it would not need to be enforced
through dispute settlement procedures.

Box C.4: Coordination problems explained

Coordination problems are situations in which every individual gains from coordinating their actions with other
individuals. We face coordination problems in our everyday life. For example, imagine that Mike and his wife Lucy
both want to spend the night out. Mike would like to go to the cinema while Lucy wants to attend a play, but both
would rather spend the night together than alone. Their levels of satisfaction, depending on their actions, are
shown in Table C.1. In each cell of the table, the first number refers to Lucy’s level of satisfaction and the second
to Mike’s. If they do not coordinate, they will end up with lower levels of satisfaction. For example, if Mike goes
to the cinema and Lucy attends the play they will both get 1. This is lower than they would obtain if they went
together to either the cinema or the play. If they both go to the cinema Lucy’s satisfaction would be 3 and Mike’s
4 as he prefers the cinema and vice versa if they both went to the play which is Lucy’ preference. Therefore,
coordination and negotiation can lead to an outcome in which both Mike and Lucy are better off than if they had
not coordinated.

Table C.1: Coordination problem between Mike and Lucy

Mike
Evening Out
Cinema Play

Cinema 3;4 0;0

Lucy
Play 1;1 4;3

Snidal (1985) has conceptualized this coordination game in the context of international regimes. He underlines
the difference between a collective action problem and a coordination problem. The terms-of-trade-driven
prisoners’ dilemma discussed previously in this subsection is a good example of the former. In this case, once
a tariff agreement has been implemented, enforcement mechanisms will have to be put in place to prevent
countries from raising their tariffs again, as doing so would serve their short-term interests. In contrast, in the
case of a coordination problem both countries want to adopt the same behaviour and will have no incentive to
deviate once they have selected a given behaviour. In other words, it requires no more than communication and
common sense to achieve an outcome that is optimal both individually and collectively.

This coordination problem arises in the context of trade facilitation. Indeed, if Country 1 plans to implement trade
facilitation measure X and Country 2 trade facilitation measure Y, they will both experience gains. However, if
they manage to coordinate and both implement either X or Y, they will further the harmonization of customs
procedures worldwide and increase their gains from trade facilitation. Consequently, the TFA, by providing
a forum for negotiation and discussion on the best available approaches and standards, can help countries
coordinate and maximize the benefits stemming from trade facilitation. Table C.2 displays such a scenario.
64
 II. SPEEDING UP TRADE: BENEFITS AND CHALLENGES OF IMPLEMENTING
THE WTO TRADE FACILITATION AGREEMENT

Box C.4: Coordination problems explained (continued)

Table C.2: Coordination problem between Country 1 and Country 2

Trade Facilitation Country 1

Measures X Y

X 4;4 1;1
Country 2
Y 1;1 4;4

The only challenge comes from the fact that country 1 might prefer to standardize customs procedures with
method X whereas country 2 might go for method Y. However, this can readily be solved through negotiations as
both countries benefit from adopting common standards regardless of the method ultimately chosen.

3. Measuring trade facilitation Subsection B.3 described how the activities of a number
of international organizations in the trade facilitation
As discussed in Section A, there are varying definitions area complement the role of the WTO. Subsection

C. T
of trade facilitation which differ in whether they include C.4(a) will go on to describe the main indicators that

TRADE FACILITATION
MEASUREMENT OF
 HE THEORY AND
soft or hard infrastructure and whether they are have been developed by international organizations to
confined to border measures or also include behind measure trade facilitation, and subsection C.4(b) will
the border measures. As a result, numerous indicators identify which indicator best reflects the provisions of
of trade facilitation exist which reflect this variation in the TFA and which has been used as the basis for the
the scope of what is involved in the definition of trade estimation and simulations undertaken in the rest of
facilitation (see Box C.5 on what makes for a good this report.
indicator).

Box C.5: What is an indicator and what makes for a good indicator?

According to Walz (2000) and to Heink and Kowarik (2010), “[a]n indicator is a variable that describes the state
of a system”. An indicator allows benchmarks to be established, comparisons to be made across countries,
and monitoring of the state of a system by different agents. It can function as an early warning system and
alert actors on the need to make improvements to the state of the system (Mainguet and Baye, 2006). A good
indicator should be:
• Relevant from a policy point of view;
• Robust, that is, not sensitive to accidental fluctuations and suitable to be used in the long term;
• Connected with priorities and most significant issues;
• Coherent with other indicators on the same topic;
• Feasible, which requires the availability of its data sources;
• Accessible;
• Valid, which means that the indicator should be connected with the research question – this validity
is measured by the strength of the association between the indicator and the concept to analyse
(Pierce, 2008);
• Reliable, in that the measurement errors are reduced (Kimberlin and Winterstein, 2008);
• Accurately measured, in such a way that the indicator is close to the true value.

Indicators should be periodically updated, in order to incorporate new challenges, adapt to new issues and
improvements in the measurement techniques and data availability (Brown, 2009).
65
 WORLD TRADE REPORT 2015

(a) Measures of trade facilitation (i) The World Bank Group’s “Doing
Business” (DB) indicators
According to Orliac (2012), there are more than
twelve indicators of trade facilitation testifying to The “Doing Business” indicators measure the effect
the importance of trade facilitation, as well as to of business regulation and the protection of property
its complexity. It will not be possible in this report rights on businesses, especially small and medium-sized
to review all of these indicators. Instead, the focus domestic firms (World Bank, 2014). They are based on
will be on those that have been used frequently in surveys of “local experts”, including lawyers, business
the economic literature to determine the economic consultants, accountants, freight forwarders, government
impact of trade facilitation reform. They include the officials and other professionals routinely administering
World Bank Group’s “Doing Business” (DB) indicators, or advising on legal and regulatory requirements. The
particularly those related to trading across borders; surveys have been conducted annually since 2004
the World Bank’s Logistics Performance Index (LPI); and now cover 189 economies. For most of these, the
the Organisation for Economic Co-operation and collected data refer to businesses in the largest business
Development’s (OECD) Trade Facilitation Indicators city. The latest DB report contains 11 indicators which
(TFIs); and the World Economic Forum’s Enabling Trade measure the complexity of the regulatory process and in
Index (ETI). particular, through the indicator “trading across borders”,
the costs related to standardized import and export
It may be useful to distinguish between indicators activities. Table C.3 lists the indicators included in the
that measure policy inputs and those that track the DB, which are then summarized by two indices:
outcomes of policy. Policy-makers should obviously
(i) “Ease of Doing Business”, which ranks countries
be interested in both since they are complementary,
according to their relative performance (World
and should also be interested in understanding the
Bank, 2014);
outcomes of trade facilitation, as well as in identifying
policies that can achieve the desired outcomes. While (ii) The “Distance to Frontier” score, which refers to
this is not a perfect categorization, the DB indicators how distant, on average, an economy is at a given
measure outcomes, the OECD TFIs focus on policy time from the best practice, i.e. the best performing
inputs and the LPI and ETI are a mixture of both. economy.

Table C.3: List of indicators and indexes

Indicators Index

Doing Business 1) Starting a business; Two main indexes:

(DB)
2) Dealing with construction permits; 1) Distance to the Frontier.
3) Getting electricity; 2) Ease of Doing Business.
4) Registering property;
5) Paying taxes;
6) Trading across borders;
7) Getting credit;
8) Protecting minority investors;
9) Enforcing contracts;
10) Resolving insolvency;
11) Labour market regulation.

Logistics 1) Customs; The LPI is constructed from the six indicators using
Performance Index a Principal Component Analysis (PCA). The scores
2) Infrastructure;
(LPI) obtained are a weighted average of the six measures,
3) Ease of arranging shipments; with the weights being the components loading.
4) Quality of logistics services;
5) Tracking and tracing;
6) Timeliness.

66
 II. SPEEDING UP TRADE: BENEFITS AND CHALLENGES OF IMPLEMENTING
THE WTO TRADE FACILITATION AGREEMENT

Table C.3: List of indicators and indexes (continued)

Indicators Index

Trade Facilitation 1) Information availability (a); There are 16 indicators based on 97 variables. The
Indicators variables have been normalized using a “multiple binary”
(TFIs) 2) Involvement of the trade community (b); scoring system (see Moïsé et al. (2011) and Moïsé and
3) Advance rulings (c); Sorescu (2013)).

4) Appeal procedures (d);

5) Fees and charges (e);

6) Formalities – Documents (f);

7) Formalities – Automation (g);

8) Formalities – Procedures (h);

9) Cooperation – Internal (i);

10) Cooperation – External (j);

11) Consularization (k);

12) Governance and impartiality (l);

13) Transit fees and charges (m);

14) Transit formalities (n);

C. T
15) Transit guarantees (o);

TRADE FACILITATION
MEASUREMENT OF
 HE THEORY AND
16) Transit agreements and cooperation (p).

Enabling Trading Fifty-six indicators classified into seven pillars: ETI is computed as the unweighted average of the
Index (ETI) various indicators.
1) Domestic market access;

2) Foreign market access;

3) Efficiency and transparency of border administration;

4) Availability and quality of transport infrastructure;

5) Availability and quality of transport services;

6) Availability and use of ICTs;

7) Operating environment.

The seven pillars are then grouped into four areas or

subindexes:

1) Market areas;

2) Border administration;

3) Infrastructure;

4) Operating environment.

(ii) The World Bank Logistics Performance since 2007. In 2014, the data covered 160 countries.
Index (LPI) The survey is divided in two parts, an international
one and a domestic one. In the international part,
The LPI focuses on the logistics friendliness of respondents assess the logistics friendliness of a
a country and ranks countries according to six country in eight selected overseas markets. In the
dimensions: customs; infrastructure; ease of arranging domestic part, respondents provide qualitative and
shipments; quality of logistics services; tracking and quantitative data on the logistics environment of the
tracing; and timeliness. The LPI indicators can be country in which they operate (Arvis et al., 2014).
grouped according to whether they refer to inputs to
the supply chain (customs, infrastructure and services The six indicators are summarized into the LPI index by
quality) or to the outcomes (timeliness, international using a Principal Component Analysis (PCA), which is a
shipments and tracking and tracing).7 statistical technique used to reduce the dimensionality
of a dataset. The LPI is, then, a weighted average of
Data are collected through an online survey of operators the scores assigned to each indicator with the weights
in charge of moving and trading goods (Gogoneata, determined by the PCA. The index goes from 1 (worst
2008). The survey has been conducted every two years score) to 5 (best score).
67
 WORLD TRADE REPORT 2015

(iii) The OECD Trade Facilitation Indicators and their destinations (WEF, 2014). It contains data
on 79 indicators from 2010 to 2014 annually for
The OECD TFIs correspond to the main policy areas 138 countries. 9 Data on 56 of the indicators are collected
under negotiation at the WTO, enabling the indicators through information provided by different international
(there are about 97 variables grouped into 16 indicators) organizations, while data for the remaining indicators are
to be mapped to relevant provisions of the TFA (see collected from the WEF Executive Opinion Survey, which
Table C.4). The OECD database, launched in 2012 and survey CEOs and top business leaders. The seventy-nine
updated in 2015, contains information on 152 countries. variables are scored from 1 to 7, with 7 indicating the best
The information used for the TFIs is collected from possible outcome. These are grouped into seven pillars
questionnaires to governments and the private sector. which are then further consolidated into four areas:
market access; border administration; infrastructure; and
The variables seek not only to reflect the regulatory operating environment (see Table C.3). The ETI score is
framework in the concerned countries, but to delve, to computed as the arithmetic mean of the 79 indicators
the extent possible, into the state of implementation and therefore also ranges from 1 to 7.
of various trade facilitation measures. Each of the
variables follows a “multiple binary” scoring system, in (b) Choice of the trade facilitation indicator
which a score of 2 corresponds to the best performance,
0 corresponds to the worst performance and a score of As the subject of this report is the TFA, and the OECD
1 to performance that lies in-between. 8 TFIs were designed on the basis of that agreement, the
TFIs will be used as a measure of trade facilitation and
(iv) The World Economic Forum Enabling country performance. In particular, the OECD indicators
Trade Index (ETI) will be employed in Section D to estimate and simulate
the economic impact of implementing the WTO TFA.10
The ETI assess the extent to which economies have
in place institutions, policies, infrastructure and Based on the criteria discussed in Box C.5, the TFIs
services facilitating the flow of goods over borders satisfy many of the requirements for a good indicator.

Table C.4: TFIs and TFA articles

Trade Facilitation Indicator Trade Facilitation Agreement article

(a) Information availability Article 1: Publication and availability of information

(b) Involvement of the trade community Article 2: Opportunity to comment, information before the entry into force, and consultations

(c) Advance rulings Article 3: Advance rulings

(d) Appeal procedures Article 4: Procedures for appeal and review

Article 6: Disciplines on fees and charges imposed on or in connection with importation and
(e) Fees and charges
exportations and penalties

(f) Formalities – documents Article 10: Formalities connected with importation, exportation and transit

Article 7: Release and clearance of goods

(g) Formalities – automation
Article 10: Formalities connected with importation, exportation and transit

Article 7: Release and clearance of goods

(h) Formalities – procedures
Article 10: Formalities connected with importation, exportation and transit

(i) Cooperation- Internal Article 8: Border agency cooperation

(j) Cooperation – external Article 8: Border agency cooperation

(l) G overnance and impartiality Article 5: Other measures to enhance impartiality, non-discrimination and transparency

(m) Transit fees and charges Article 11: Freedom of transit

(n) Transit formalities Article 11: Freedom of transit

(o) Transit guarantees Article 11: Freedom of transit

(p) Transit agreements and cooperation Article 11: Freedom of transit

Note: The OECD TFI indicators include an item “(k) Consularization” which has no corresponding provision in the TFA.
68
 II. SPEEDING UP TRADE: BENEFITS AND CHALLENGES OF IMPLEMENTING
THE WTO TRADE FACILITATION AGREEMENT

The indicators are relevant from a policy point of view One can also compare how the different indexes score
precisely because they are based on the TFA, which the trade facilitation performance of countries to see
members have committed to implement. This also makes if major discrepancies emerge. Figure C.3 compares
it a useful indicator to monitor the implementation of three trade facilitation indexes: the TFIs average, LPI
the TFA. The statistical robustness of the TFIs has and ETI scores.11 It classifies countries according to
been improved through the study of the underlying links the WTO region classification, the level of development
of the dataset and tested with traditional indicators and whether they are landlocked developing countries
(Moïsé et al., 2011). The TFIs are also robust with or not. It should be noted that, when accounting for
regard to temporary fluctuations in economic activity the level of development and distinguishing between
as the indicators would only change as result of the landlocked/non-landlocked countries, the three
implementation efforts of each country. Furthermore, indexes score countries in the same general way.
the TFIs are consistent and correlated with the other Groups performing best on the TFI average also
widely used indicators of trade facilitation (despite perform best on the ETI and on the LPI. Among the
some indicators being measures of outcomes rather WTO regions, North America and Europe are the best
than policy inputs). Table C.5 shows the correlation performers in all the indexes.
between the TFIs, the DB trading across borders
components, LPI and ETI for the latest available year. When considering the level of development, developed
The TFI average score is positively correlated with the countries register the highest scores. Among developing
LPI and the ETI measures. As expected, the TFI average countries, those that are not landlocked obtain higher
is negatively correlated with the DB cost of export/ scores compared to landlocked developing countries,
although the differences between them are smaller if

C. T
import and number of days to export/import indicators.

TRADE FACILITATION
MEASUREMENT OF
 HE THEORY AND
The correlation coefficients are all significant at the 5 measured with the TFIs and larger if measured with the
per cent level. other indicators (DB, LPI or ETI). This result suggests
a double burden for landlocked developing countries:
apart from being isolated from global markets by having
Table C.5: Correlation between Doing
no access to the sea, they also have in place inefficient
Business Indicators, the Logistics
trade procedures that further hinder their trade.
Performance Index, the Enabling Trade Index
and the Trade Facilitation Indicators
Indicator TFI 4. Conclusions
Average
This section has shown that trade models of all
DB: Trading across borders – costs to export -0.25*
generations can be adapted to draw interesting and
DB: Trading across borders – costs to import -0.29* complementary conclusions regarding the impact of
DB: Trading across borders – number of days -0.42* trade facilitation. Yet, with the increased academic and
to export policy focus on trade facilitation, researchers should be
DB: Trading across borders – number of days -0.47* encouraged to develop more specific economic models
to import of trade facilitation that incorporate salient features
DB: Trading across borders – number of -0.47*
of how today’s international trade is conducted.
documents required to export For instance, none of the models discussed above
specifically consider the role of time in trade costs, but
DB: Trading across borders – number of -0.45*
documents required to import recent work suggests lengthy shipping times impose
significant costs on firms engaged in trade (Hummels
LPI Score 0.43*
and Schaur, 2013).
LPI Customs 0.41*

LPI Timeliness 0.42* Aside from the time question, there is also empirical
work on global value chains that indicates traders are
Enabling Trading Index 0.59*
concerned with the overall reliability of the supply chain
ETI Efficiency and transparency of border 0.51* and that hedging against uncertainty of delivery time
administration
makes up a significant part of logistics costs in many
ETI Customs transparency index 0.43* developing countries (Arvis et al., 2007a; 2007b). Work
by the WTO and the OECD on global value chains and
ETI Efficiency of the clearance process 0.36*
trade in value added has made researchers much more
ETI Irregular payments in import/export 0.47* aware of the role of trade in services. Might anything be
ETI Time predictability of import procedures 0.41* said about the relationship between trade facilitation
and trade in services? One hypothesis is that trade
*Significant at the 5 per cent level. facilitation should also increase services trade since
69
 WORLD TRADE REPORT 2015

Figure C.3: Average TFIs, Enabling Trade Index and Logistics Performance Index
(latest available year)

(a) (b)

2.00 2.00

1.50 1.50
Score

Score
1.00 1.00

0.50 0.50

0.00 0.00
the erica tral

cou loped
Asia

t Stah of
a

pe

ica

ibbe nd

cou loped

cou ping
(in Gntries
tes

Eas
Afric

an

s
)
ntrie

-20
Euro

ntrie

ntrie
en
a
mer
pen wealt

evelo
nd C

e
dle

ve
cou
Dev
th A

Car
Mid

t-de
den
Inde mon

Am
th a

er d
ing
Nor

s
Lea
Com

elop
Sou

Oth
Dev
WTO regions Level of development

(c)

1.50

1.00
Score

0.50

0.00
Landlocked Non-landlocked
Developing countries

Average TFIs ETI LPI

Note: ETI and LPI scores have been rescaled from 0 to 2 to make them comparable to the OECD TFIs.
Source: OECD TFIs, WEF ETI and World Bank LPI.

logistics and transport activity are likely to expand This section has also examined four major trade
along with merchandise goods trade. Alternatively, one facilitation indicators: the World Bank’s Doing Business
can imagine border delays increasing service trade indicators, the World Bank’s Logistics Performance
through more costly shipping and other transport costs. Index, World Economic Forum’s Enabling Trade Index
If so, trade facilitation will, in part, reduce service trade and the OECD’s Trade Facilitation Indicators. The
even as it expands trade in merchandise goods. main difference between them is the scope of trade
facilitation they take into account. This report will use
Future research could also distinguish between the the OECD TFIs as the indicator for the TFA because
impacts of different types of trade facilitation measures, they were constructed on the basis of the TFA, satisfy
consider the role of country circumstances along the the criteria of a good indicator, are correlated with
lines of Duval (2007), and examine the contribution of the other major indicators and, when accounting for
complementary policies in achieving success in trade the development and geographical characteristics of
facilitation reform (Borchert et al., 2012; Iwanow and countries, they are consistent in their ranking with the
Kirkpatrick, 2007; Francois and Hoekman, 2010). other indicators.
70
 II. SPEEDING UP TRADE: BENEFITS AND CHALLENGES OF IMPLEMENTING
THE WTO TRADE FACILITATION AGREEMENT

Endnotes
1 The reader is nevertheless encouraged to read Hummels thresholds can be identified. In these cases, if the variable
and Skiba (2004) and Hummels (2007), who examine in is numerical in nature, the score could be determined by
great detail how additive or non-proportional trade costs deviation from the sample mean or by its percentile rank.
affect the pattern of trade. See Orliac (2012).

2 Some recent contributions include Yi (2003; 2010) and 9 The country coverage has been increased in 2014.
Baldwin and Venables (2013). Before 2014, it covered 132 countries.

3 See Bagwell and Staiger (1999; 2002) and WTO (2012). 10 For the analysis in this subsection and the simulations in
Section D, we use the 2009 OECD TFI database, which
4 See also the discussion in subsection C.1.
has information on 133 countries, 26 of which are OECD
5 See Maggi and Rodriguez-Clare (1998; 2007), Matsuyama members, and 107 non-OECD members. Since previous
(1990), Staiger and Tabellini (1987), and WTO (2012). studies on the economic effects of trade facilitation that
have used the OECD TFIs have relied on the 2009 data,
6 See subsection E.4.
using the same data makes the analysis in this report
7 Arvis et al. (2014). comparable to those previous studies. All 26 OECD
members are also WTO members. Of the 107 non-OECD
8 A scoring system that assigns discrete numerical values
countries, 96 are WTO members and 11 are WTO observers.
according to some metric of performance requires
determining thresholds for what is best, worst or in between. 11 The “Ease of Doing Business” and/or the “Trading Across
Sometimes there are “natural” thresholds, as for example Borders” indicators have not been taken into account
for the variable “Establishment of a national Customs because they simply rank countries.
website”. Thus, a country without a customs website will be

C. T
assigned a score of 0; a country with a customs website

TRADE FACILITATION
MEASUREMENT OF
 HE THEORY AND
will be assigned 1; and a country with a customs website
which makes available a minimal set of information related
to import or export procedures in one of the official WTO
languages will be assigned a 2. In other cases, no natural

71