THE WORLD TRADE ORGANIZATION ENLARGEMENT, TARIFFS AND GLOBAL SEAFOOD TRADE

FIPM/C1100 (En) FAO Fisheries and Aquaculture Circular ISSN 2070-6065 THE WORLD TRADE ORGANIZATION ENLARGEMENT, TARIFFS AND GLOBAL SEAFOOD TRADE F...
Author: Kristian Boone
16 downloads 0 Views 563KB Size
FIPM/C1100 (En)

FAO Fisheries and Aquaculture Circular ISSN 2070-6065

THE WORLD TRADE ORGANIZATION ENLARGEMENT, TARIFFS AND GLOBAL SEAFOOD TRADE

FAO Fisheries and Aquaculture Circular No. 1100

FIPM/C1100 (En)

THE WORLD TRADE ORGANIZATION ENLARGEMENT, TARIFFS AND GLOBAL SEAFOOD TRADE

Arne Melchior Senior Research Fellow Norwegian Institute of International Trade Oslo, Norway

FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS Rome, 2015

The designations employed and the presentation of material in this information product do not imply the expression of any opinion whatsoever on the part of the Food and Agriculture Organization of the United Nations (FAO) concerning the legal or development status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. The mention of specific companies or products of manufacturers, whether or not these have been patented, does not imply that these have been endorsed or recommended by FAO in preference to others of a similar nature that are not mentioned. The views expressed in this information product are those of the author(s) and do not necessarily reflect the views or policies of FAO. ISBN 978-92-5-108737-4 © FAO, 2015 FAO encourages the use, reproduction and dissemination of material in this information product. Except where otherwise indicated, material may be copied, downloaded and printed for private study, research and teaching purposes, or for use in non-commercial products or services, provided that appropriate acknowledgement of FAO as the source and copyright holder is given and that FAO’s endorsement of users’ views, products or services is not implied in any way. All requests for translation and adaptation rights, and for resale and other commercial use rights should be made via www.fao.org/contact-us/licence-request or addressed to [email protected]. FAO information products are available on the FAO website (www.fao.org/publications) and can be purchased through [email protected].

iii

PREPARATION OF THIS DOCUMENT The study contributes to FAO’s ongoing activities to analyse international seafood trade and trade regulation, focusing particularly on developments in the World Trade Organization (WTO) since its establishment in 1995. While the latest negotiation round of the WTO, the Doha Round, has been protracted and difficult, the WTO has had a major success by adding 31 new members. This publication studies how this has affected seafood tariffs for the new members, and it concludes that WTO accession has clearly contributed to more liberal trade. The impact of WTO membership on trade in seafood is analysed, distinguishing between the number of suppliers (the “extensive” margin of trade) and the change in import value. The results, some of which were presented at the FAO and Eurofish regional seminar “The WTO and fisheries”, 29–31 October 201, St. Petersburg, the Russian Federation, indicate that WTO membership has promoted trade, particularly at the extensive margin, by increasing the number of suppliers for seafood in each market. This trade-promoting effect has been uneven across product groups, and driven particularly by trade in processed seafood and crustaceans.

iv

Melchior, A. 2015. The World Trade Organization: enlargement, tariffs and global seafood trade. FAO Fisheries and Aquaculture Circular No. 1100. Rome, FAO. 24 pp. ABSTRACT The document shows that there has been a major success in the World Trade Organization (WTO) with the addition of 31 new members. This has contributed to lower tariffs for seafood trade. Among the founding members of the WTO, many have considerable “water in the tariffs” and the legal option of raising their most-favoured nation (MFN) applied tariffs. The analysis shows that most countries have not used this option; there has been no protectionist backlash for fisheries tariffs but, on average, a trend towards more liberal trade. The econometric analysis investigated whether WTO membership affected trade, for whatever tariff or non-tariff reason. The results are in line with recent contributions showing that WTO membership leads to increased trade. According to the results, WTO membership stimulates trade particularly at the “extensive margin”, by promoting new entry into seafood markets. This result is driven particularly by trade in crustaceans and processed seafood.

v

CONTENTS Preparation of this document

iii

List of tables

vi

List of figures

vi

Abbreviations and acronyms

vii

Acknowledgements

viii

1.

INTRODUCTION: WTO – A SECRET SUCCESS IN THE SHADOWS OF TRICKY NEGOTIATIONS

1

2.

SEAFOOD TARIFFS AND WTO DISCIPLINES: AN OVERVIEW

4

3.

WTO ENLARGEMENT AND SEAFOOD TARIFFS

7

4.

DID WTO MEMBERSHIP INCREASE TRADE?

9

5.

FIXED EFFECTS: A MEASURE OF NON-TARIFF BARRIERS?

13

6.

CONCLUDING COMMENTS

14

REFERENCES

15

APPENDIX

16

Tariffs for seafood for new WTO members, 1995–2013

16

Classification of seafood products

18

Fixed effects panel regressions: robustness checks, estimates for the WTO variable

18

Regressions for more detailed sub-groups of seafood

19

The 25 countries with the largest negative deviations for fixed effects

22

The 25 countries with the largest positive deviations for fixed effects

23

Seafood trade for new WTO members (USD millions)

24

vi

LIST OF TABLES 1. 2. 3.

Tariff binding rates and average tariffs for seafood Fixed effect panel regressions on the impact of WTO membership on seafood trade Regressions for more-detailed subgroups of seafood – estimates for the impact of WTO membership

4 11 12

LIST OF FIGURES 1. 2. 3. 4. 5. 6.

WTO members’ share of world GDP (current USD) and population, 1995–2013 Bound vs MFN applied seafood tariffs for 103 (129) WTO members, 2007–09 Changes in applied seafood tariffs for 88 WTO members, 2001–09 New bound tariff averages for seafood and the average of MFN applied tariffs before accession WTO binding vs real change in tariffs for seafood Number of countries supplying seafood to each market (median values for 111 countries), 1996–2012

2 5 6 7 8 10

vii

ABBREVIATIONS AND ACRONYMS CTS DDA GATT GDP MFN NAMA NTB RTA SPS TRAINS UNCTAD UR WITS WTO

Consolidated Tariff Schedules Doha Development Agenda General Agreement on Tariffs and Trade gross domestic product most-favoured nation Non-Agricultural Market Access (negotiations) non-tariff barrier regional trade agreement sanitary and phytosanitary Trade Analysis and Information System United Nations Conference on Trade and Development Uruguay Round World Integrated Trade Solution World Trade Organization

viii

ACKNOWLEDGEMENTS Thanks are extended to Audun Lem and Victoria Chomo, FAO Fisheries and Aquaculture Department and to Hege Medin, Senior Research Fellow, Norwegian Institute of International Affairs, for their useful comments on an earlier version of this circular. Appreciation is also extended to Gloria Loriente, FAO Fisheries and Aquaculture Department, for the layout design of this publication.

1

1. INTRODUCTION: WTO – A SECRET SUCCESS IN THE SHADOWS OF TRICKY NEGOTIATIONS “We have saved the WTO” was the relieved expression of Karel de Gucht, Trade Commissioner of the European Union (Member Organization) after the Bali meeting of the World Trade Organization (WTO) in December 2013,1 and he continued that it “marks the return of the WTO from the darkness of multilateral irrelevance into the light of multilateral action and success.” However, the Bali package left out most of the major results negotiated over 12 years for the fisheries sector, including substantial tariff reductions under the Non-Agricultural Market Access (NAMA) negotiations (Melchior and Johnsen, 2011). In 2008, the WTO almost achieved agreement on these issues, but the moment was lost and the WTO was set for five years of dwindling ambition and frustration.2 Some important elements were included in the Bali package, in particular, the new agreement on trade facilitation, reducing the scope for non-tariff barriers at the border. This is certainly of importance for fisheries trade. On NAMA, however, it is uncertain what will happen in the future; few observers expect that the 2008 drafts can be recovered or reintroduced at this stage.3 While it is symbolically important that the Bali meeting succeeded, the post-Bali WTO is a less dynamic organization – few participants would aim for grand design reforms in the near future. Notwithstanding this element of fatigue, this paper shows that there has actually been considerable progress and success in the WTO. Parallel to the protracted negotiations of the Doha Development Agenda (DDA), the number of WTO members has climbed steadily, and WTO disciplines now cover almost the whole world. The WTO was established in 1995 with 128 “founding members”.4 Since then, 31 new members have been added (country and year of entry in brackets): 

  

Fifteen countries from East and Southeast Europe (Bulgaria, 1996; Kyrgyzstan, 1998; Latvia, 1998; Estonia, 1999; Georgia, 2000; Albania, 2000; Croatia, 2000; Lithuania, 2001; Republic of Moldova, 2001; Armenia, 2003; The former Yugoslav Republic of Macedonia, 2003; Ukraine, 2008; Montenegro, 2012; the Russian Federation, 2012; Tajikistan, 2013). Seven countries from East and South Asia (Mongolia, 1997; China, 2001, Taiwan Province of China, 2002; Nepal, 2004; Cambodia, 2004; Viet Nam, 2007; the Lao People’s Democratic Republic, 2013). Three countries from the Near East (Jordan, 2000; Oman, 2000; Saudi Arabia, 2005). Two countries from Latin America and four other countries (Ecuador, 1996; Panama, 1997; Tonga, 2007; Cabo Verde, 2008; Samoa, 2012; Vanuatu, 2012).

As a result of this expansion, the WTO now covers almost the whole world. Figure 1 shows the cumulative share of world gross domestic product (GDP) and world population for the members of the WTO for 1995–2013.5 When the WTO was established in 1995, it covered 67 percent of the world population and 81 percent of world nominal GDP. In 2013, the share had risen to 93 percent of world population, and 97 percent of GDP. The most significant steps were the inclusion of China (2001) and the Russian Federation (2012).

1

http://trade.ec.europa.eu/doclib/docs/2013/december/tradoc_151951.pdf The Doha Development Round of the WTO was launched in Doha, Qatar, 2001, and ambitions were reduced after the contentious WTO Ministerial Meeting in Cancun, Mexico, 2003, where there were sharp disagreements over the new issues suggested in Singapore in 1996. 3 The Doha Development Round also addressed some other aspects of fisheries, including subsidies (Meliado, 2012). 4 Sixteen founding members actually entered into the WTO in 1996. 5 For this illustration, GDP and population in 2011 are used as a basis for measuring the shares. 2

2

Figure 1. WTO members’ share of world GDP (current USD) and population, 1995–2013

Source: World Development Indicators.

Many countries still remain outside the WTO. With Yemen about to become member number 160, there are 23 others involved in accession procedures. They includes important countries in East Europe (e.g. Kazakhstan, Belarus, Uzbekistan, Azerbaijan), and a number of countries in the Near East and North Africa (e.g. Algeria, Iran (Islamic Republic of), Iraq, the Syrian Arab Republic, Libya, Eritrea, Somalia, South Sudan, the Sudan). Does WTO membership lead to trade liberalization and the expansion of trade? While this has been the general belief among trade economists and practitioners for decades, a paper by Rose (2004) cast some doubt on this. He ran a standard “gravity regression” where the trade flows between two countries are expected to be a function of their economic sizes and the distance between them, plus other aspects that could influence trade, such as common language, common border, and past colonial ties. To this equation, he added variables (dummies) measuring whether one or both of the countries in each trading pair were members of the WTO / General Agreement on Tariffs and Trade (GATT). Using data for 1948–1999, Rose found that WTO members did not trade more than others – according to this, there is no significant trade increase if one becomes a WTO member. The paper by Rose ignited a new debate on WTO’s role. Subramanian and Wei (2007) criticized the econometric method used by of Rose.6 Distinguishing between the impact of WTO membership on imports vs total trade (imports plus exports), these authors used similar data to Rose and found that WTO membership led to a significant increase in imports, especially for developed countries. Balding (2010) also found that WTO membership affects imports and exports differently, and suggested this might be the underlying reason why Rose (2004) found no effect. His results indicated that WTO membership had a positive impact on trade (exports as well as imports), particularly for high-income countries. Dutt, Mihov and Van Zandt (2011) follow the more recent trade literature (see e.g. Melitz, 2003) by distinguishing between whether the WTO affects the intensive margin (changed trade volume for existing trade flows) or the extensive margin (trade increase due to new entry into markets). Measuring the extensive margin by trade in new products at a detailed level of classification, they find that WTO membership has a positive trade effect only at the extensive margin (+31 percent in their preferred specification), but a negligible or even negative effect at the intensive margin.

6

In particular, the authors show that results are changed when exporter and importer fixed effects are included, correcting for unobserved country characteristics that influence trade.

3

In the light of these findings, this paper examines the impact of the WTO on world seafood trade, with a particular focus on tariffs and the impact of WTO enlargement. While the impact of WTO accession may capture tariff as well as non-tariff effects, this document does not present specific data and analysis of non-tariff barriers. Given that rich countries are globally the largest importers of seafood, the development of new and stricter regimes for food safety control of imports in some countries is a potentially important issue that is not addressed explicitly here. According to WTO (2012), most trade concerns raised by WTO members about non-tariff barriers are for trade in the food sector. While the econometric analysis of WTO accession effects may also capture the impact of non-tariff barriers in seafood trade, a more extensive study would be warranted, but this is outside the scope of the present study. For a useful analysis of selected sanitary and phytosanitary (SPS) issues in world seafood trade, readers may refer to Ababouch, Gandini and Ryder (2005). The organization of the paper is as follows. Section 2 reviews the worldwide pattern of tariffs for seafood trade and how it is regulated by the WTO. Section 3 examines seafood tariffs for new WTO members and whether they have been affected by accession. Section 4 examines whether WTO accession has had a statistically measurable impact on seafood trade, distinguishing between the extensive and intensive margin of trade. Section 5 presents evidence on countries having particularly large or small imports of seafood. Section 6 concludes by discussing some limitations and possible extensions of the study.

4

2.

SEAFOOD TARIFFS AND WTO DISCIPLINES: AN OVERVIEW

Melchior (2006) presents an analysis of tariffs in world seafood trade, covering 169 countries with tariff observations later than 2000.7 Given that there have not been major global tariff reforms since 2006, many of these results for tariffs are still valid. The analysis covers bound as well as “most-favoured nation (MFN) applied” tariff rates. Bound rates are those written in the protocols of the WTO; i.e. upper bounds for the tariffs used by a member country for each tariff line/product. The “MFN applied” tariffs are the ordinary tariffs applied by a country for imports from all countries, unless they have special conditions owing to regional trade agreements or other trade preferences. As bound tariffs are upper bounds only, MFN applied tariffs – the tariffs applied in practice – may be lower. If the MFN applied tariffs are below the bound rates, there is said to be “water in the tariffs” or a “tariff overhang” or “binding overhang”. 8 An implication is that if bound tariffs are reduced because of WTO negotiations, it will not necessarily affect the MFN applied rates if there is water in the tariffs. For example, if the MFN applied rate is 10 percent and the bound rate 20 percent, the latter may be cut by half before it has to affect the applied rate. Owing to the Uruguay Round (UR) of trade negotiations, which led to the establishment of the WTO, the extent of tariff binding increased dramatically. However, many tariffs remained unbound also after the UR. Melchior (2006) showed that out of 119 WTO members, about half (59 members) had bound all the tariff lines for seafood at the WTO. Forty-five members had bound less than 20 percent of the tariff lines for seafood. It was particularly the low-income WTO members that did not bind seafood tariffs. While the binding average for all 119 members in the data was 61 percent, low-income countries had only bound 43 percent on average (Table 1). In the UR, the price to pay for more tariff binding was that many countries were allowed to bind their tariffs at rather high levels. Hence, a country could have a tariff average at 10 percent but bind tariffs at 80 percent in the more extreme cases. This was also the case for seafood. There was substantial water in the tariffs – while the median bound tariff average for seafood was 34 percent, the median MFN applied tariff average was 14 percent. Table 1 shows simple tariff averages for seafood for different country groups. There is wide variation across countries. For some countries, seafood is a “sensitive” sector with higher import tariffs; for others, trade is more liberal. On average, tariffs for seafood are slightly higher than for industrial goods (Melchior, 2006). Table 1. Tariff binding rates and average tariffs for seafood Share of six-digit items bound for seafood

Simple average of bound tariffs

79 60.7 71 43.1 60.5

(%) 12.4 42.9 31.6 51 34.2 70 WTO members

High income Upper middle income Lower middle income Low income All countries Countries covered by data:

119 WTO members

Simple average of MFN applied tariffs 5.2 18.7 16.3 17.8 15.6 140 countries, WTO members and non-members

Source: Melchior (2006), Table 5.

Hence, there is considerable “water in the tariffs” for all country groups. As a consequence, reductions in bound tariffs would not “bite” strongly in all cases. For example, a 40 percent

7

Seafood includes the following items in the Harmonized System (HS) classification: Chapter 3, sub-headings 1603-1605, and positions 051191, 150410-20, 230120. 8 All three expressions are now standard in the trade policy vocabulary. For an introduction see www.wto.org/english/tratop_E/markacc_e/nama_negotiations_e.htm

5

proportional cut in all bound seafood tariffs worldwide would only lead to a mandatory reduction in MFN applied tariffs of 9 percent (Melchior, 2006). Using more-recent data for 2007–09, Figure 2 shows that bound seafood tariffs are still higher than MFN applied rates in the majority of cases. Most of the data are from the Trade Analysis and Information System (TRAINS) tariff database of the United Nations Conference on Trade and Development (UNCTAD), accessed though the World Integrated Trade Solution (WITS), supplemented with data from WTO’s Consolidated Tariff Schedules (CTS) database (also through WITS), and in some cases other WTO online data. Figure 2. Bound vs MFN applied seafood tariffs for 103 (129) WTO members, 2007–09

   Sources: TRAINS, CTS, WTO.

If a data point is to the right of the 45 degree line, the bound average is highest. The MFN applied averages also include tariff lines that are not bound; this explains why some data points are above / to the left of the 45 degree line. In general, the data points are spread out, so there is a wide variation in the extent of water in the tariffs. Countries are legally free to change their MFN applied tariffs provided they are below bound rates. It is therefore of some interest to examine what happened. Did existing WTO members use the “water in the tariffs” to raise tariffs for seafood? Figure 3 compares MFN applied tariffs for seafood in 2001–02 with levels in 2007–09, for 88 founding WTO members. As tariff data are not available for every year/country, ranges of 2–3 years have been to increase data coverage. Most data points are close to the 45 degree line, showing that there was little change in many cases. Large deviations are more often to the right than to the left of the 45 degree line, showing that there was more liberalization than protectionism – a number of countries cut their MFN applied tariffs for seafood even though they did not have to. If one says that a change of more than one percentage point is an “increase” or “decrease”, then there was a decrease in 32 cases, an increase in 17 cases, and no change in 39 cases. Hence, on the whole, the trend was slightly in a liberal direction, with some exceptions.

6

Figure 3. Changes in applied seafood tariffs for 88 WTO members, 2001–09

Sources: TRAINS, CTS, WTO.

 

7

3.

WTO ENLARGEMENT AND SEAFOOD TARIFFS

With respect to seafood trade, there are data available for 29 of the 31 new WTO members (Appendix Table A7). About half of them were net exporters, and half net importers. The largest exporters were China, Viet Nam, Ecuador, the Russian Federation and Taiwan Province of China. The largest importers were China, the Russian Federation and Taiwan Province of China. While the Russian Federation had almost balanced trade, China was a significant net exporter of seafood. Here, seafood also includes processed seafood (see footnote 7 for classification). How did WTO accession affect tariffs for the new entrants? The newcomers had bound no tariffs before they became members. A first question is: Did they bind their tariffs? Or did they do like the old average, binding some but not all? The answer is straight and simple: all 31 new WTO members bound 100 percent of their tariff lines for seafood. This is good news: significantly raising the average binding rate to about 66 percent.9 At what level did the new entrants bind their tariffs? Did they bind them far above their MFN applied level or not? Appendix Table A1 shows the new bound rates, and MFN applied rates before and after accession. Figure 4 compares the new bound tariff averages for seafood with the average of MFN applied tariffs before accession. Figure 4. New bound tariff averages for seafood and the average of MFN applied tariffs before accession

   Sources: Data for 23 countries from TRAINS, CTS, WTO.

There is wide variation. Some countries bound seafood tariff at levels far below earlier MFN tariffs, while other countries bound tariffs higher. The most liberalizing country was Albania, having an MFN applied average for seafood at 22 percent before accession, and then binding tariffs at zero. At

9

This is not fully comparable with the average of 61 percent shown in Section 2, owing to slightly different data coverage.

8

the other end lies Vanuatu, with binding tariffs at 61 percent, far above the former applied level.10 On the liberal side are net exporters such as China, Viet Nam and Taiwan Province of China, but also net importers such as the Republic of Moldova, Georgia, The former Yugoslav Republic of Macedonia and Tajikistan. Hence, there is no clear correspondence between net export ratios and the extent of “water in the tariffs” that was introduced. For Albania and other countries below the 45 degree line, it is certain that WTO accession led to a reduction in seafood tariffs. This was the case for a number of countries, including important countries such as China, the Russian Federation, Taiwan Province of China and Viet Nam. Some countries bound tariffs at low levels but had already cut MFN applied tariffs to a low level before entry; a case in point is Ukraine. For countries that bound seafood tariffs at high levels, it would in principle be possible to raise tariffs after WTO accession. Using the data in Appendix Table A1, Figure 5 plots the newly acquired “water in the tariffs” (= bound average minus MFN applied average before accession) against the change in applied tariffs (= MFN applied tariffs after accession minus MFN applied tariffs before accession). Figure 5. WTO binding vs real change in tariffs for seafood

Source: Appendix Table A1, using data from TRAINS, CTS, WTO.

Countries to the right in Figure 5 have “water in the tariffs”, so they could in principle raise their tariffs. Countries to the left, on the other hand, were forced to cut tariffs. The figure shows that none of the countries raised its applied tariffs; even Vanuatu reduced its applied tariff level in spite of very high binding. The data points follow an inverse U-shaped pattern. There were two “clusters”: in the middle are many countries with little change in their tariffs; and to the left there is a liberalizing group where WTO accession led to a significant tariff cut. On average, WTO accession was accompanied by lower applied tariffs.

10

The Baltic States are included. However, they joined the European Union (Member Organization) in 2007 and their tariffs were then changed to the levels of that organization.

9

4.

DID WTO MEMBERSHIP INCREASE TRADE?

From the analysis in Sections 2 and 3 above, it is evident that tariffs were reduced for existing as well as new members. Hence, there is not a very sharp one-to-one relationship between applied tariffs and membership. This is perhaps one reason why one may not necessarily find a strong trade effect of accession. Membership of the WTO may also lead to many changes other than tariffs. Often, WTO accession may be part of a broader process of reform, and it may then be difficult to distinguish the impact of WTO from the impact of reform more generally. With respect to tariffs, this analysis has only examined MFN tariffs but not preferential tariffs within regional trade agreements (RTAs). In recent years, RTAs have become so prevalent that is little point in analysing the impact of MFN tariffs only, with no data for RTAs. However, obtaining accurate tariff data including preferences is an extensive data exercise that is beyond the scope of this study. Therefore, this analysis follows another route, more similar to the approach of Rose (2004) and later contributions, checking whether WTO membership as such affects trade. It uses dummy variables capturing WTO, and examines whether seafood trade is larger for WTO members. Following Dutt, Mihov and Van Zandt (2011), it checks whether the “intensive” vs the “extensive” margin of trade responds differently. Unlike Dutt, Mihov and Van Zandt (2011), data for aggregate seafood trade are used, i.e. not at the detailed product level. Instead of running a gravity regression with bilateral flows, two country-level variables are used: total import value for seafood (in current USD, using the definition in footnote 7); and the number of supplying countries for seafood in each market. Data from WITS/COMTRADE for 1996–2012 are used, and panel regressions are run. There are 17 years in this period and data are required for at least 13 years for each importing country (= each cross-section in the panel regression). With this criterion, there are data for 111 countries, and 1 735 observations (i.e. with data for 15.6 years on average). Among the 111 countries are 83 “founding members” of the WTO, 20 acceding countries and 8 countries that were non-members all the time. First, simple panel regressions are run of the form: Mit = α1 + α2Fi +α4 GDPit + β Dit + εit Here, Mit is the dependent variable, which can be seafood import value or the number of seafood suppliers (exporting countries), while Fi is a set of fixed effects that control for time-invariant country characteristics that affect trade for each importing country i, and GDPit is assumed to be a main and time-variant country determinant of imports, capturing market size. As bilateral trade data are not used directly in the regression, and there are not have variables for common borders, language etc., it is assumed that these characteristics are time-invariant over the period covered and captured by the fixed effects. Some regions and countries consume more fish than others, but this probably does not change too much during the period studied so it will be reflected in the fixed effects. As a check, a set of time effects, Tt, is included in some specifications. This is a set of time controls, capturing common shocks at each year t such as the financial crisis, or supply shocks for seafood. The key variable is Dit, taking the value 1 if the country is a WTO member and 0 if not. For the 83 founding members, it is 1 all the time, and for the 8 non-members it is always zero. For the 20 acceding countries, it varies over time from 0 to 1 from the year of accession and onwards. Import value is derived directly from the data, and the number of seafood-supplying countries in each country is calculated from the data.11 Changes in the number of suppliers must clearly reflect the extensive margin of trade, while import value is affected by the extensive as well as the intensive margin. For seafood, it is to be expected that fresh seafood is different from frozen or preserved

11

In order to generate the “number of suppliers” variable, the analysis uses a data set with bilateral trade and 181 000 observations. Deriving the number of suppliers variable, one for each importing country/year, yields a data set of 2 404 observations. This is cut further to have at least 13 years in each cross-section, giving the final sample of 1 735 observations.

10

seafood as fresh food often cannot be transported over long distances, and this limits the number of potential suppliers. Therefore, the data are split into fresh seafood and “other seafood”, and the regressions run separately, in addition to regressions for all seafood. Appendix Table A2 shows the classification of the two groups. Figure 6 shows the median number of suppliers for each crosssection in the data, for fresh and other seafood. Figure 6. Number of countries supplying seafood to each market (median values for 111 countries), 1996–2012

Source: Author’s calculations using WITS/COMTRADE data.

The typical number of supplying countries varies from a handful in some markets to more than 100 at the other end. The two curves seem to converge at very high levels, but for intermediate levels the relative difference between fresh and other is larger. In the panel regressions, the dependent variables and the GDP variable are expressed in logarithms. Table 2 shows the results for the whole sample and for the two subgroups. Using fixed effects, R2 is very high, as would be expected, at 0.93 and above. The GDP variable is always strongly significant, as expected. Many of the fixed effects are also highly significant; as shown by the bottom line in each section of the table, 77–95 out 111 fixed effects were significantly different from zero at the 1 percent level or better (more on this below). For the purpose of the analysis undertaken here, the most interesting variable is the WTO dummy. For seafood imports taken together, it is not significant in the “intensive margin” specification, with the value of imports as dependent variable. However, in the “extensive margin” specification, it is strongly significant – WTO membership leads to a 12 percent increase, on average, in the number of suppliers in a market. The subdivision into fresh and other seafood reveals that the extensive margin effect of WTO membership is driven by “other seafood”. The WTO effect on the extensive margin is not significant for fresh but only for “other” seafood. This is in line with the expectation – the feasible set of suppliers is more limited for fresh seafood. For fresh seafood, there is a weakly significant (at the 10 percent level) WTO effect on the intensive margin. However, this is not strong enough to create an intensive margin effect that is significant for seafood as a whole.

11

Table 2. Fixed effect panel regressions on the impact of WTO membership on seafood trade Dependent variable: Import value (in logs) All seafood Fresh seafood Other seafood Variable Est. P value Est. P value Est. P value Intercept –1.104