Analysis of Agri-Food Trade Structures
To Promote Agri-Food Trade Networks
In the Islamic Countries
25
Table 4: Centrality Scores for Fish Products, by Region, 2016
Out Degree Centrality Eigenvector Centrality
East Asia & Pacific 41.083 0.058
Europe & Central Asia 48.700 0.077
Latin America & Caribbean 22.949 0.043
Middle East & North Africa 22.857 0.062
North America 92.333 0.095
South Asia 36.125 0.054
Sub-Saharan Africa 14.000 0.039
Source: Authors’ calculations.
Network analysis shows that world trade in agricultural products tends to be relatively
dispersed, in the sense that although trade within regions plays an important role, inter-regional
trade links are also important. Of course, geography means that trade costs are lower between
countries in the same geographical region, and that is a factor that promotes intra-regional
trade. But geography is also an important determinant of comparative advantage in agricultural
trade: relative availability of arable land, as well as weather conditions, determine the types of
agricultural products countries can specialize in, as well as patterns of comparative advantage
and comparative disadvantage across countries. Whereas trade in manufactured goods is
characterized by two-way trade in similar goods, results show that trade in agricultural
products is closer to what would be expected from a simple factor proportions model, where
countries trade dissimilar goods. This factor explains why inter-regional trade links are
relatively important: countries within the same region have relatively similar geographical and
climatic characteristics, which leads them to specialize in similar products; as a result, there is
less scope than might be expected for commercial exchanges across borders with close
neighbors as compared with distant countries. Policy also plays a role in this distribution of
agricultural trade, a factor that the next subsection examines.
2.4. Agricultural Trade Policies
Trade in agricultural products is affected by different types of trade policies in all countries. To
analyze the course of trade policies over the last two decades, the focus is on changes in ad
valorem tariffs over time. Under WTO law, three types of tariffs are of relevance. Bound tariffs
are the legal ceilings countries agree to in WTO negotiating rounds, and which are contained in
their schedules of commitments. They cannot exceed these rates without the possibility of legal
action by trading partners. However, they can apply lower tariffs if they wish. Applied MFN
tariffs capture the tariffs actually applied by WTO members on a most-favored nation basis, i.e.
a basis of non-discrimination across trading partners. These rates are typically substantially
lower than bound rates. Finally, many countries have signed free trade agreements that allow
them to impose lower rates with partner countries. These preferential rates, as well as
preferences granted unilaterally by developed countries to developing and least developed
countries, are captured in the concept of effectively applied rates.
Figures 12 through 14 present applied MFN and effectively applied tariff rates for the three titles
identified in the Annex 1 classification. The analysis is undertaken at this level because tariff
rates can vary substantially at more disaggregated levels, so it is less feasible to identify
wholesale trends as opposed to momentary movements.
First, both types of applied tariff rates are substantially less than WTO bound rates, as would be
expected. For agri-food products, bound rates were nearly four times as high as applied MFN