Prepared for 10 th Annual ERF Conference (December 16-18, 2003, Morocco)

U.S. Trade Relations with Middle East (Has the United States lost market share to Europe because of its Middle East policy?) (October 2003) Nader Habibi* Prepared for 10 th Annual ERF Conference (December 16-18, 2003, Morocco) *Senior Economist Middle East and Africa Service Global Insight Inc. Nader.habibi@globalinsight.com

Abstract U.S. Trade Relations with Middle East (Has the United States lost market share to Europe because of its Middle East policy?) (October 2003) Nader Habibi* This article investigate the impact of the United States, Middle East policy on U.S. exports to Middle East. In particular it asks whether differences in U.S. and European policies toward the Arab-Israeli conflict and other regional crises have enabled Europe to increase its export share in Middle East at the expense of the U.S. Using a time series regression model with annual and monthly data I show that U.S. market share in these areas has been sensitive to some of the major political events. The Oslo peace negotiations (1994-97) are associated with a higher U.S. market share. The 9-11 attacks and the second Palestinian uprising, on the other hand, have led to an increase in European market share at the expense of the United States. 2

1) Introduction Bilateral trade between nations is primarily determined by economic factors such as the relative price of imports, trade barriers and economic growth. However, political and diplomatic considerations can have an impact on international trade under certain circumstances. Some countries, such as US routinely use trade sanctions against others as an instrument of foreign policy. Even when a government does not formally impose trade sanctions on a trading partner, popular sentiment could lead to the boycott of that country s products. In early 2003, French and German opposition to the U.S. plan for removal of Saddam Hussein caused an emotional backlash in the U.S. and many people called for boycott of French and German products. At the same time a negative campaign against popular American brand names such as Coca Cola and McDonalds was launched in Europe. Similarly, popular anger in the Arab world towards the U.S. for its support of Israel, led to a large scale grass roots movement in many Arab countries for boycott of American products in 2001-02. 1 The purpose of this study is to investigate whether the U.S. foreign policy towards the Arab world in general and the U.S. involvement in the Arab-Israeli conflict in particular have had any impact on export of U.S. products to the Arab world. A more specific objective of the study is to see how differences in US and European policies towards the Arab-Israeli conflict have influenced their relative share of the Arab world s import market. Over the past two decades the European countries have shown more sympathy for the Palestinian cause whereas the United States has been more supportive of Israel. In 1982, the European Union temporarily suspended its 1975 Trade and Cooperation agreement with Israel to protest its invasion of Lebanon. In another show of concern for Palestinians EU delayed the approval of three Trade and Financial Agreements with Israel in 1987 during the first Palestinian uprising. A year later EU approved these three agreements after Israel made certain trade concessions to Palestinians. In mid-1990s EU rewarded Israel for its participation in the Oslo peace negotiations by signing several trade and economic cooperation agreements that had been delayed for several years. In recent years, particularly after the second Palestinian uprising, EU has been highly critical of Israel. It has openly opposed the occupation of West bank and Gaza strip and has demanded an end to the expansion of Israeli settlements in these areas. In 2002 European political leaders became highly vocal in their criticism of Israel and reluctance of U.S. to put more pressure on Israel. 2 The European public opinion has also been far more sympathetic to the Palestinian cause than the U.S. The growing difference in U.S. and EU policies towards the Arab-Israeli conflict has led some analyst to claim that Europe could enjoy a sympathy dividend in the Arab world because its stronger support for Palestinians. At the same time the Arab world recognizes that since Israel is heavily dependent on American military and economic support, United States is the only country that can put pressure on Israel. This fact has lead to two conflicting views of how the Arab countries should deal with the United States. A moderate vision that is more popular in the GCC countries believes that Arab countries should expand their trade and economic ties with the United States and use their economic influence to lobby for the Palestinian cause in Washington. This view was echoed in 2002 by Prince al-walid of Saudi Arabia in response to the growing demands for boycott of American products (Bahrain Tribune, May 1, 2002). If an Arab country follows this recommendation then we will not observe any anti-us bias in its import demand behavior. To the contrary, such a country might even show a preference for U.S. products in hope of strengthening the Arab lobby in Washington. A more radical perspective in the Arab world rejects the notion that buying more from U.S. will influence U.S. policy in favor of Palestinians. Instead, it believes that Arabs should express their displeasure with U.S. support of Israel by shunning away from American goods and services. According to this view if American firms realize that they are loosing market share in the Arab world because of the perceived bias in America s Middle East policy, they will lobby the U.S. government for a more balanced approach to the conflict. This view gained widespread popularity soon after the second Palestinian uprising (Intifadha), which began in September of 2000. For Arab countries that might have been influenced by this perspective we expect to observe a decline in their imports of American products during periods of heightened Israeli aggression against Palestinians (such as the 1982 invasion of Lebanon). And by the same logic we expect an increase in demand for U.S. products during periods of progress towards peace and a just resolution of the conflict (such as 1993-96 Oslo negotiations). 3

However, one must be cautious about attaching too much weight to the Arab-Israeli conflict in US-Arab relations. Aside from this conflict, U.S. has developed close security and economic ties with various Arab countries particularly in the Persian Gulf region. The small oil exporting countries of the Gulf welcomed the U.S. presence in the region during Iran-Iraq war, during which US helped protect the shipping lanes and assure the safe navigation of oil tankers. The GCC countries developed close political and military ties with the United States after Iran s 1979 Islamic revolution. These ties grew even stronger after the United States ended the Iraqi occupation of Kuwait. Hence it won t be unreasonable to expect the GCC trade relations with the United States to have been influenced by the special role of the US in political and military conflicts of the Persian Gulf region. The Kuwaiti government for example pursued a formal policy of favoring American products and contractors as a show of appreciation for the U.S. role in liberating Kuwait from Iraqi occupation. The approach of GCC countries to using trade and investment to influence U.S. policy towards the Arab-Israeli conflict is, to a large extent influenced by their own military and political relations with the United States. To summarize the above discussion, we expect the Arab-Israeli conflict to have had an influence on Arab desire for trade and investment with the United States. However, the direction of this influence is not clear and it could vary from one country to another. In the remaining sections of this article I will use bilateral trade data between a select number of Arab countries and their major trade partners to investigate the impact of the factors that were discussed above on the U.S. share of these countries import markets. The sources of data and method of analysis will be discussed in section 2 followed by a descriptive analysis of recent trends in Arab imports from U.S. and Europe in section 3. I will then test several hypothesis about the role of political factors in US share of Arab markets in the context of a multivariate regression model in sections four and five. A summary of findings and conclusions will appear in section six. 4

2) Method of Analysis The empirical section of this analysis is dedicated to comparing U.S. and European shares of the Arab import markets. To realize this goal I have constructed a U.S. trade share variable as follows: S i = US i /(US i +X i ) Si = Share of US in the combined exports of US and another exporting country (X) to the Arab country i. USi = US exports to the Arab country/region i. Xi = Country X exports to the Arab country/region i. The fluctuations of S i over time will show whether U.S. has lost or gained trade share to X. An increase in S would indicate that U.S. has gained trade share at the expense of X while a decline would imply that U.S. has lost trade share. Figures 2-a to 2-e shows US export share in a select number of Middle Eastern countries or sub-regions during 1980-2002. For each of the selected Middle Eastern markets I compare U.S. trade share against major European countries; France, Germany, the United Kingdom, and Italy. The primary emphasis however, is on France and Germany, which are both key European exporters and whose policy towards the Arab-Israeli conflict diverges significantly from the United States. In addition to looking at U.S. market share against France and Germany individually, I have also analyze the U.S. market share against a) the combined exports of France and Germany and b) the total exports of the UK, France, Germany and Italy. Data: I have used IMF s Direction of trade Statistics to extract annual and monthly bilateral trade data (aggregate imports of commodities) for Arab countries. The analysis is based on annual data for 1980-2002. The monthly data is also used to observe the impact of the more recent events such as the September 11 attacks against U.S. targets. The fact that the data only covers the merchandise trade is a weakness that limits the scope of the conclusions. Both U.S. and Europe export sizable amounts of services to the Arab world. However, unfortunately comparable data for bilateral trade in services is not yet available. The monthly data is also used to observe the impact of the more recent events such as the September 11 attacks against U.S. targets. 5

3) Recent Trends in US and European exports to the Middle East Graphs 1-a to 1-e show U.S. merchandise trade with the (Arab) Middle East, the Gulf Cooperation Council (GCC: Saudi Arabia, Kuwait, UAE, Oman, Bahrain, Qatar), Saudi Arabia, Egypt and Israel. We notice that while U.S. imports from Middle East are highly volatile, most of this volatility is caused by the instability of oil prices (1-e). The U.S. exports to the Middle East, on the other hand enjoyed a steady and continuous growth between 1986 and 1998. The value of U.S. exports to the region grew from a low of $12.3 billion in 1985 to a high of more than $27 billion in 1998. The trend was revered after that and the U.S. lost most of its trade gains for 1990s during 1998-2002. The decline second half 1998 and 1999 could partly be attributed to the declining oil revenues of Arab oil exporters, which imposed a constraint on their purchasing power. However, the decline in 2000-2002 U.S. exports was associated with a period of relatively high oil revenues in the region. Graph 1-e also shows that after running a growing trade deficit in late 1980s the U.S. trade imbalance diminished significantly in mid-90s thanks to the moderate cost of U.S. oil imports. U.S. even registered a noticeable trade surplus in 1997-99. It must however be kept in mind that while the value of U.S. exports to Middle East has increased overtime, the share of Middle East in total U.S. exports declined from 0.037% in 1992 to 0.02% in 2000. Furthermore, because of proximity and a long history of economic relations, Western Europe is the largest trade partner of Middle East. The trade volume of four major European countries (The UK, France, Germany and Italy) with the Middle East exceeded $101 billion dollars in 2000. The U.S.-Middle East trade volume during the same year was $61.7 billion. Graphs 1-a and 1-b show that U.S. exports to GCC countries, particularly Saudi Arabia, followed the same trend as the entire Middle East by growing steadily between 1985 and 1998 and shifting into a downward trend after that. The growth of U.S. exports to GCC countries during 1985-98 was steeper than the rest of the Arab world. At the same time we also notice that the decline in U.S. exports to GCC countries after 1998 has also been steeper that rest of the Middle East. These observations lead us to conclude that U.S. trade relations with GCC have been more volatile than other sub-regions of Middle East. In contrast, graph 1-c shows that U.S. trade with Egypt has been highly stable and U.S. has enjoyed a trade surplus, which has been financed partially by its continuous economic aid to Egypt for the past two decades. 6

20000 US-GCC Merchandise Trade (US$ Millions) 16000 14000 U.S. Merchandise Trade with Saudi Arabia (US$ Million) 15000 12000 10000 10000 8000 6000 5000 4000 2000 0 80 81 8283 84 85 86 8788 89 90 91 9293 94 95 96 9798 99 00 0102 0 80 81 8283 84 85 86 8788 89 90 91 9293 94 95 96 9798 99 00 0102 U.S. Impoerts from GCC Exports to GCC Exports to Saudi Arabia Imports from Saudi Arabia 1-a 1-b 4000 U.S. Merchandise Trade with Egypt (US$ Million) 14000 12000 US-Israel Merchandise Trade (US$ Millions) 3000 10000 2000 8000 6000 1000 4000 0 2000-1000 0 80 81 82 83 84 85 86 8788 89 90 91 92 93 94 95 96 97 98 9900 01 02 80 81 8283 84 85 86 8788 89 90 91 9293 94 95 96 9798 99 00 0102 Exports to Egypt Imports from Eg Imports from Israel Exports to Israel 1-c 1-d 50000 40000 30000 20000 10000 0 The U.S. Merchandise Trade with the Middle East (US $ M illio n ) 80 81 8283 84 85 86 8788 89 90 9192 93 94 9596 97 98 99 0001 02 Exports to Middle East Imports from Middle East P ric e o f O il $/b 1-e 30 25 20 15 10 7

Graphs 2-a to 2-e below show how U.S. has faired against Western European countries in Middle East markets. Looking at the U.S. market share in all of Middle East (graph 2-e), we notice that contrary to some expectations U.S. has not lost market share to the four major European exporters (Germany, France, The UK and Italy) combined. To the contrary, the data shows a modest gain in U.S. market share against these four countries between 1990 and 1995 followed by a slow downward trend, which has still left U.S. share in 2002 larger than 1980s. Another general observation worth mentioning is that the U.S. market share against the combined exports of these four countries to the Middle East has been remarkably stable. It has hovered around within a range of 26%-34% over this 23-year period. However, when we look at the U.S. market share against France and Germany we observe more volatility. We also observe that during 1996-2002 U.S. has lost more market share to France than Germany. The U.S. share of the combined exports of U.S. and France to Middle East fell from 75% in 1996 to 64% in 2002. Turning to GCC countries (graphs 2-a, 2-b), we see that U.S. was able to preserve its market share against European countries in this sub-region up until 1997. During 1994-97 U.S. was actively involved in the Oslo peace accord. The gain in U.S. market share in this period could reflect the GCC countries favorite reaction to U.S. engagement in this process. Similarly, the declining U.S. market share in 1998-2002 is associated with the failure of the Oslo negotiations and the deterioration of U.S.-Arab negotiations after September 11 attacks. After 1997, U.S. has steadily lost market share to both France and Germany. In addition to political factors the exchange rate fluctuations (Dollar versus Euro) have also had an impact on the market shares of U.S. and Europe as will be demonstrated in the regression analysis. Within GCC we observe a significant difference in U.S. market share in Saudi Arabia and the remaining five members of GCC. The U.S. market share in Saudi Arabia (graph 2-c) suffered a steep decline in 1999-2002, which was particularly severe in 2002. This decline is mostly due to the deterioration of U.S.-Saudi relation after the 9/11 attacks. In the remaining five GCC countries, the U.S. market share regained some of its losses in 2002 after the sharp decline of 1998-2001. The 9-11 events had a less severe impact on U.S. relations with these small oil-exporting countries. The 2002 upswing in U.S. market share in GCC countries is visible against all four European countries in our sample. In case of Egypt, the U.S. market share has suffered from short-term cyclical fluctuations but at the same time it has followed a modest upward trend since 1983. In a sharp contrast to GCC countries, the U.S. gained market share in Egypt at the expense of Europe in 1998-2001. As mentioned earlier U.S. has offered a generous annual aid package to Egypt ever since the 1978 Egypt-Israel peace agreement. The minimum U.S. content conditions of aid package could have had a positive impact on U.S. exports to Egypt. Graphs 3-a to 3-e show U.S. market share in various Middle East sub-regions after 1990. Their only difference with graphs 2-a to 2-e is that instead of annual data they represent the 12-month moving averages of monthly data and they include the first few months of 2003. These two sets of graphs are consistent and the monthly data graphs show that the decline in U.S. market share in Saudi Arabia continued in 2003 but it appeared to have stabilized in the broader Middle East market. 8

0.90 0.80 0.70 0.60 0.50 0.40 0.30 GCC Imports from US and Europe Annual data: 1980-2002 0.20 8081 8283848586878889 909192 9394959697989900 0102 US vs. (Fr+Gf+Uk+It) US vs. (Fr+Gf) US vs. France US vs. Germany 0.90 0.80 0.70 0.60 0.50 0.40 0.80 0.70 0.60 0.50 0.40 0.30 GCC5 Imports from US and Europe Annual data: 1980-2002 (Saudi Arabia excluded) 0.20 8081 828384 8586 878889 909192 939495 9697 989900 0102 US vs. (Fr+Gf+Uk+It) US vs. (Fr+Gf) US vs. France US vs. Germany 2-a 2-b U.S. Share of Saudi Arabia's Imports Annual data: 1980-2002 0.30 8081 828384 8586 878889 909192 939495 9697 989900 0102 US vs. (Fr+Gf+Uk+It) US vs. (Fr+Gf) US vs. France US vs. Germany 0.90 0.80 0.70 0.60 0.50 0.40 0.30 U.S. Share of Egypt's Imports of Goods Annual data: 1980-2002 0.20 8081 828384 8586 878889 909192 939495 9697 989900 0102 US vs. (Fr+Gf+Uk+It) US vs. (Fr+Gf) US vs. France US vs. Germany 2-c 2-d 0.80 U.S. Share of Middle East Merchandise Imports Annual data: 1980-2002 0.70 0.60 0.50 0.40 0.30 0.20 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 US vs. (Fr+Gf+Uk+It) US vs. (Fr+Gf) US vs. France US vs. Germany 2-e 9

0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 90 91 92 93 94 95 Israel Middle East U.S. Share of (U.S. + EU4) Exports (12-month moving average, monthly data) 96 97 98 EU4=UK+It+Fr+Gf 99 00 01 02 03 0.70 0.60 0.50 0.40 0.30 GCC Imports and Exports to U.S. vs. (Europe: Fr+Gf+Uk+It) (12-month moving average, monthly data) 0.20 90 91 92 93 94 95 96 97 98 99 00 01 02 U.S. share in GC imports U.S. share in GC exports 3-a 3-b 03 0.50 U.S. Share of (U.S. + EU4) Exports (12-month moving average, monthly data) 0.22 0.20 Share of U.S. in total commodity imports (12-month moving average, monthly data) 0.45 0.18 0.40 0.16 0.35 0.30 0.14 0.12 0.10 0.25 90 91 92 93 Saudi Arabia 94 95 96 97 98 99 00 01 Gulf Cooperation Council (GCC) 02 03 0.08 90 91 92 93 94 95 Egypt Middle East 96 97 98 99 00 01 02 03 3-c 3-d Middle East Imports and Exports to U.S. vs. Europe: Fr+Gf+Uk+It (U.S. share) (12-month moving average, monthly data) 0.60 1.20 0.50 1.10 0.40 1.00 0.90 0.30 0.80 0.20 90 91 92 93 94 95 96 97 U.S. share of imports US$-euro exchange rate (right) 98 99 00 01 02 03 U.S. share of exports 0.70 3-e 10

4) Regression Analysis The trend analysis above suggested several hypotheses with respect to the role of non-economic factors in the U.S. share of Arab import markets that deserve a more systematic investigation. In this section I will use regression analysis to test these hypotheses. The conventional models of import demand in the international trade literature regress the exports of country A to country B on the relative price of imports, (which amounts to exchange rate 3 in case of aggregate imports) and B s GDP (which acts as a proxy for purchasing power). logmt= a*logpt+b*logyt+c (1) logm t = a*logp t +b*logy t +logm t-1 +c (2) M=Total Imports P= Relative Price of Imports Y= GDP Most studies of import demand do not pay any attention to non-economic factors. However, when the country under consideration heavily regulates its external sector (which is the case in most Middle East countries,) and government exerts a large amount of direct and indirect control over foreign purchases, the role of political factors cannot be ignored. Under such circumstances bilateral trade with other countries will be influenced by foreign policy considerations. A recent empirical analysis of import demand behavior in North Korea (Kim and Lim, 2002) clearly supports this argument. Kim and Lim add a set of political dummy variables to equation (1) to see how North Korea s relations with communist rivals, China and the former Soviet Union, affected imports from these two countries. Their analyses revealed that while aggregate imports from China and former Soviet Union were not sensitive to political factors, imports of some basic commodities such as fuels were sensitive to these factors. In my regression models (equations (3) and (4) below,) I use the U.S. trade share that was described in section two as dependent variable. This variable is regressed on the dollar/euro exchange rate, the national income of the importing country, and a series of political dummy variables. Notice that I have used the Euro/$ exchange rate instead of the importing country s exchange rate versus US$ and Euro. The reason behind this choice is that since my dependent variable measures the U.S. market share versus a specific European country or sub-region the price factor that matters to the importing country is the relative import prices of American and European products. This is best captured by the $/Euro exchange rate. Furthermore, most of the Arab countries that appear in this analysis maintain a fixed exchange rate against U.S. dollar and hence the only exchange rate factor that will influence their import decisions is the $/Euro exchange rate. logs i = a*logp US/X + b*logy i + λ 1 *D 1 + λ 2 *D 2 + +λ n *D n + C i (3) logs i = a*logp US/X + λ 1 *D 1 + λ 2 *D 2 + +λ n *D n + C i (4) P US/X = Exchange Rate between U.S. and X The political dummy variables (D 1, D 2, D n) that I have included in my model are the following: 1) The Lebanon War 1982-84: U.S. was viewed in the Arab war as supporting the invasion of Lebanon). 2) Iran-Iraq war 1984-87: During the last four years of this eight-year long conflict U.S. offered protection to oil tankers navigating through Persian Gulf against attacks by Iranian forces. U.S. also showed a commitment to preventing an Iranian victory against its Arab rival. 3) The first Intifadha (the Palestinian uprising I) 1987-90. 4) Gulf war I 1991-92: The U.S. role in liberation of Kuwait led to a significant improvement in US-GCC relations. 5) Oslo Peace negotiations 1993-96: During these negotiations U.S. image in the Arab world improved significantly. 11

6) Second Intifadha (Palestinian Uprising) and the 9-11 attacks: 2001-02. Failure of the Oslo agreement led to the second Palestinian uprising and a grassroots boycott of U.S. products in the Arab world. I estimate equation (3) using time series annual data with correction for serial autocorrelation AR(1). I then repeat some of the regressions with monthly data to see how the change of data set affects the results. Regression results appear in tables 1 to 5. We estimate the model for the entire Arab Middle East, GCC countries, Saudi Arabia, and Egypt. 5) Regression Results The model estimations for U.S exports to the entire Arab Middle East appear in Table 1. We notice that the Euro/$ exchange rate does not have a significant coefficient in any of the estimations in this table. On the other hand the U.S. trade share has a positive and significant correlation with the region s GDP. Among the political dummy variables the 9-11 variable has the strongest impact on U.S. market share. The coefficient of this variable is negative and significant for all regression estimates in table 1. The only other political variable with a significant coefficient is Oslo peace negotiations, which has a positive coefficient for U.S. trade share versus all four European countries. However only in the case of U.S. versus French exports to the region (equation 4) this coefficient is statistically significant. It implies that during Oslo peace negotiations, 1994-97, U.S. gained market share in the Middle East at the expense of France. The fact that other political factors did not have any significant coefficients indicates that up until recent years the U.S. policy towards the Arab-Isreali conflict did not cause U.S. to loose market share to European exporters in Middle East. In table 2 we see the results of the same regression models for U.S. and European exports to GCC countries. For GCC block the region s GDP is not significantly correlated with U.S. export share. But it is weakly correlated with the level of U.S. exports to the region. In other words, these regressions suggest that an increase in GCC income leads to more imports from U.S. but it does not necessarily increase the U.S. market share at the expense of European countries. The Euro/$ exchange rate is significant and (as expected) negative for U.S. export share versus the combined exports of four European countries. In here, again the most significant dummy variable is the 9-11, which also represents the recent Arab Boycott of U.S. products. The coefficient is significant for all four export share regressions. The only other dummy variable with a significant coefficient is Iran-Iraq war. However, the sign of this variable is surprisingly negative, and it is only significant for the U.S. versus France regression. Initially I expected this variable to have a positive sign. Another unexpected result of these regressions is that the dummy variable for the first Gulf war did not have a significant coefficient in any of the regressions. Perhaps one possible explanation for these unexpected results is that in both, the first Gulf war and the Iran-Iraq war, the European countries adopted policies similar to the United States. Both helped prevent an Iranian victory over Iraq, which could have threatened the political stability of GCC countries. They also both participated in military operations for liberation of Kuwait in 1991. Hence the GCC nations had no political motive to favor U.S. exports at the expense of Europeans. The regression results for Saudi Arabia are reported in table 3. Among the political dummy variables the 9-11 variable has the most significant coefficient in all four regressions. Another significant variable is the Oslo peace negotiations. The coefficient of this variable is positive and significant for U.S. versus France and U.S. versus all four European exporters. It suggests that U.S. gained trade share in Saudi Arabia during the Oslo negotiations. We also observe a weakly significant positive coefficient for the first Gulf war variable in case of U.S. versus France. Comparing the regression results of Saudi Arabia and GCC we notice that the U.S. export share in Saudi Arabia is more sensitive to political variables than in other GCC countries. This could be due to the fact that Saudi Arabia is a more politicized society compared to other GCC countries. Table 4 reveals that U.S. market share in Egypt has been more sensitive to political variables in comparison to GCC countries. All of the political dummy variables have statistically significant coefficients. Contrary to GCC countries, the 9-11 variable has a positive coefficient in this table and indicates an increase in U.S. market share at the expense of European exporters in 2001-02. In light of the strong campaign against American products in the Arab world this result is rather surprising. It shows that the boycott campaign not only failed to reduce the total volume of U.S. exports to Egypt (equation 1), but it also failed to shift the market demand in favor of European exporters (equations 2 and 3). We also notice that the Oslo peace negotiations and the first Gulf war 12

were also associated with higher U.S. market share in Egypt. In line with our initial expectations, the Lebanon war and the first Palestinian uprising were both associated with lower U.S. market shares in Egypt. In table five I have repeated the regressions with monthly trade data for 1990-2002. The results show that the only significant political variable is the 9-11 factor, which has had an adverse effect on U.S. market share in Saudi Arabia and through Saudi Arabia in GCC countries. It has also had a weakly significant negative impact on U.S. market share in the entire Middle East region. 13

1) 2) 3) 4) 5) Middle East imports from the U.S. (Log) t-stat U.S. share of exports to ME EU4)* (Log) t-stat U.S. share of exports to ME (US vs. EU2)* (Log) t-stat U.S. share of exports to ME France) (Log) t-stat U.S. share of exports to ME (US vs. Germany) (Log) Dependent Variable ---> t-stat euro/$ exchange rate (Log) -0.111 0.49-0.09 0.99-0.05 0.49-0.03 0.53-0.033 0.35 Middle East GDP$(Log) 1.062 5.38 0.52 6.8 0.45 5.02 0.09 1.86 0.45 5.96 9-11 (Dummy Var.) -0.44 0.48-0.13 3.43-0.15 3.41-0.08 3.46-0.11 2.82 Lebanon War (1982-84) -0.1 1.06-0.05 1.34-0.05 1.24-0.02 0.79-0.05 1.29 Oslo Agreement (93-96) -0.017 0.27 0.01 0.73 0.03 1.04 0.04 1.96 0.009 0.35 Intifadha 1 (1987-90) -0.2 2.96-0.02 0.82-0.007 0.22 0.006 0.356-0.01 0.5 Iraq-Iran war (84-87) -0.29 3.96 0.013 0.82 0.03 1.11 0.0095 0.48 0.039 1.26 Kuwait War (91-92) 0.03 0.42-0.02 0.59-0.01 0.37 0.04 1.82-0.055 1.44 C 3.13 4.2-4.6 9.3-3.65 6.34-0.97 2.97-3.62 7.01 F-Statistic 22.19 (9,12) 9.56 (9,12) 5.5 (9,12) 4.78 (9,12) 7.41 (9,12) AR(1) -0.216 0.88-0.4 1.45-0.25 0.82-0.21 0.61-0.27 0.96 Observations 22 81-2002 22 81-2002 22 81-2002 22 81-2002 22 81-2002 * EU4=UK+Italy+France+Germany, EU2=France+Germany Source: Direction of Trade Statistics (IMF) Trade Share is defined as US/(US+X) where US= US exports to the target country or region. Table 1: Middle East 14

1) 2) 3) 4) 5) U.S. share of exports U.S. share of exports U.S. share of exports U.S. share of exports to GCC to GCC to GCC to GCC GCC imports from the EU4)* EU2)* France) Germany) Dependent Variable ---> U.S. (Log) t-stat (Log) t-stat (Log) t-stat (Log) t-stat (Log) t-stat euro/$ exchange rate (Log) -0.17 0.49-0.23 1.859* -0.15 1.45-0.11 1.11-0.13 1.32 GCC GDP$(Log) 0.4 1.55-0.02 0.31-0.03 0.42-0.11 1.56 0.04 0.609 9-11 (Dummy Var.) 0.009 0.06-0.11 2.01* -0.14 2.81* -0.09 2.1* -0.09 2.35* Lebanon War (1982-84) 0.056 0.33 *** 0.035 0.59-0.05 1.18 0.013 0.241 Oslo Agreement (93-96) -0.03 0.36-0.03 0.72-0.001 0.03-0.01 0.47-0.001 0.043 Intifadha 1 (1987-90) -0.27 2.08* -0.06 1.32 0.001 0.02-0.017 0.45 0.003 0.08 Iraq-Iran war (84-87) -0.36 2.11* -0.09 1.78-0.06 1.18-0.07 1.65-0.022 0.48 Kuwait War (91-92) 0.06 0.39 0.022 0.355 0.04 0.83-0.003 0.08 0.03 0.69 C 7.01 4.88-0.85 1.74-0.41 0.87 0.36 0.87-0.65 1.66 F-Statistic 8.84 (9,12) 3.55 (8,13) 4.76 (8,14) 3.54 (9,12) 2.41 (9,12) AR(1) -0.08-0.22-0.13 0.38-0.06 0.19-0.27 0.79-0.1 0.29 Observations 22 81-2002 22 81-2002 22 81-2002 22 81-2002 22 81-2002 * EU4=UK+Italy+France+Germany, EU2=France+Germany GCC= Gulf Cooperation Council: Saudi Arabia, Oman, UAE, Bahrain, Qatar, Kuwait. Source: Direction of Trade Statistics (IMF) Trade Share is defined as US/(US+X) where US= US exports to the target country or region. Table 2: Gulf Cooperation Council 15

1) 2) 3) 4) 5) U.S. share of exports U.S. share of exports U.S. share of exports U.S. share of exports Saudi to Saudi to Saudi to Saudi to Saudi Arabia imports from the EU4)* EU2)* France) Germany) Dependent Variable ---> U.S. (Log) t-stat (Log) t-stat (Log) t-stat (Log) t-stat (Log) t-stat euro/$ exchange rate (Log) 0.27 0.5 0.04 0.39-0.01 0.102 0.05 0.79-0.07 0.48 Saudi Arabia GDP$(Log) 0.39 1.19 0.253 2.45* 0.133 1.142 0.06 1.15 0.027 0.29 9-11 (Dummy Var.) *** *** -0.207 3.74* -0.18 2.90* -0.1 3.23* -0.103 2.82* Lebanon War (1982-84) 0.045 0.27-0.05 0.83-0.027 0.34-0.086 2.2* 1.807 1.29 Oslo Agreement (93-96) 0.04 0.27 0.063-1.7 0.052 1.064 0.04 2.02* 0.009 0.07 Intifadha 1 (1987-90) 0.044 0.24 0.014 0.28 0.046 0.72 0.031 1.09 0.004 0.07 Iraq-Iran war (84-87) -0.172 1.028-0.004 0.06-0.27 0.34-0.042 1.12-0.045 0.72 Kuwait War (91-92) 0.33 1.52 0.026 0.52 0.058 0.904 0.046 1.76 0.016 0.23 Dummy var. 2001 0.085 0.437 Dummy var. 2002-0.15 0.618 C 6.7 4.14-2.07 4.01-1.15 1.963-0.52 1.87-0.46 0.99 F-Statistic 4.84 (10,11) 4.83 (9,12) 2.88 (9,12) 5.15 (9,12) 1.6 (9,12) AR(1) 0.69 1.54-0.46 1.39-0.15 0.49-0.63 1.76 0.23 0.76 Observations 22 81-2002 22 81-2002 22 81-2002 22 81-2002 22 81-2002 * EU4=UK+Italy+France+Germany, EU2=France+Germany Source: Direction of Trade Statistics (IMF) Trade Share is defined as US/(US+X) where US= US exports to the target country or region. Table 3: Saudi Arabia 16

Dependent Variable ---> 1) 2) 3) Egypt's imports from the U.S. (Log) t-stat U.S. share of exports to Egypt EU4)* (Log) t-stat U.S. share of exports to Egypt EU2)* (Log) t-stat National currency/$ Exchange Rate (Log) -0.41 3.71* -0.34 3.73* euro/$ exchange rate (Log) -0.57 0.98-0.017 0.105 0.04 0.31 Country GDP$(Log) 0.4 2.85* 0.5 3.5* 0.43 3.67* 9-11 (Dummy Var.) 0.139 0.65 0.22 3.29* 0.16 2.99* Lebanon War (1982-84) 0.087 0.34-0.4 4.5* -0.39 4.66* Oslo Agreement (93-96) -0.18 1.15 0.3 4.14* 0.25 4.15* Intifadha 1 (1987-90) -0.52 3.11* -0.43 4.12* -0.4 4.71* Iraq-Iran war (84-87) -0.618 4.51* -0.62 4.76* -0.53 4.94* Kuwait War (91-92) -0.38 1.84 0.4 3.03* 0.34 3.16* Dummy Var. 2001 Dummy Var. 2002 C 5.93 10.18-2.71 5.35-2.2 5.23 F-Statistic 3.78 (9,12) 4.5 (10,11) 5.11 (10,11) AR(1) -0.36 1.4-0.67 2.66-0.7 3.03 Observations 22 81-2002 22 81-2002 22 (81-2002) * EU4=UK+Italy+France+Germany, EU2=France+Germany Source: Direction of Trade Statistics (IMF) Trade Share is defined as US/(US+X) where US= US exports to the target country or region. Table 4: Egypt 17

Regression Analysis Based on Monthly Data (12-month moving average) 1) 2) 3) 4) Dependent Variable ---> U.S. share of exports to Egypt EU4)* (Log) t-stat U.S. share of exports to ME (US vs. EU4)* (Log) t-stat U.S. share of exports to GCC EU4) (Log) t-stat U.S. share of exports to Saudi Arabia EU4) (Log) euro/$ exchange rate (Log) 12-month moving average -0.25 1.25-0.028 0.28 0.02 0.13-0.23 1.69 t-stat 9-11 (Dummy Var.) 0.025 1.41-0.017 1.81-0.029 1.86-0.04 2.82 Oslo Agreement (93-96) 0.011 0.8 0.004 0.61-0.002 0.21-0.004 0.36 Kuwait War (91-92) 0.0004 0.0211-0.007-0.7-0.018 1.149-0.02 1.131 C -0.93 15.27-1.13 42.34-1.05 9.01-0.85 30.98 F-Statistic 649 (5, 138) 1224 (5, 138) 703 (5, 138) 396 (5, 138) AR(1) 0.96 57.78 0.96 72.2 0.98 55.83 0.94 31.99 Observations 144 Jan91-Dec02 144 Jan91-Dec02 144 Jan91-Dec02 144 Jan91-Dec02 * EU4=UK+Italy+France+Germany, EU2=France+Germany GCC= Gulf Cooperation Council: Saudi Arabia, Oman, UAE, Bahrain, Qatar, Kuwait. Source: Direction of Trade Statistics (IMF) Trade Share is defined as US/(US+X) where US= US exports to the target country or region. Table 5: Regression Analysis Based in Monthly Data (12-month moving averages) 18

6) Summary and Conclusion In this study I used time series regression analysis to see if variations of U.S. market share in the merchandise imports of Middle Eastern countries were correlated with major political and diplomatic events in the region. The main argument was that since U.S. had played an important role in many of the regional conflicts, particularly the Arab-Israeli dispute, the Arab countries demand for U.S. products must have been sensitive to the U.S. position during these conflicts. I further hypothesized that since major European countries were more sympathetic to the Palestinian cause in comparison to the United States, we must observe a gradual substitution of European products for U.S. products in Middle Eastern markets. (This is known as Europe s sympathy dividend.) The regression results based on annual trade data suggest that during 1980-2002 period the U.S. export share in Middle East market as a whole, was mostly insensitive to major political events of the region. The only exception is the 2001-2002 interval, which is associated with the September 11 attacks and the second Palestinian uprising. Both of these events had a negative impact on U.S. Arab relations. During these two years U.S. has lost market share to Europeans in most Middle Eastern countries with the exception of Egypt. During the Oslo peace negotiations the U.S. gained a modest amount of market share in Middle East at the expense of France but not Germany. In the Saudi market, which is important to both the U.S. and Europe, the 1982-84 Lebanon war led to a market gain for France at the expense of U.S. while the Oslo peace negotiations and the first Gulf war had the opposite effect. However, the period that led to the largest shift in Saudi import demand from American to European suppliers, was the 2001-2002 period, which witnessed a sharp increase in anti- American sentiments in that country. While opinion poles in Egypt showed similar types of sentiments, our regression models pointed to an increase in U.S. market share of Egyptian imports during 2001-2002. This could be due to the conditions attached to U.S. economic aid to Egypt which require the money to be spent on American products. Overall, our analysis detected some mild evidence that Middle East imports from U.S. and Western European countries were sensitive to these countries foreign policies towards the region. The study, however, had a limited scope and can be expanded in several directions in the future. First, a larger number of Middle Eastern countries can be brought into the analysis. Currently, the country regressions are limited to Egypt and Saudi Arabia. The model can be estimated for other major Middle Eastern economies as well. Second, we can go beyond aggregate imports and repeat the analysis for specific commodities that are offered for export by both U.S. and European countries such as manufactured goods, agricultural products and medicine. We might observe different political sensitivities of import demand for different product groups. Third, this paper does not cover the impact of bilateral and multilateral trade agreements between Middle Eastern countries and their European and American trade partners. Both U.S. and Europe are in various stages of trade negotiations with various subregions of Middle East. As these negotiations result in new trade arrangements, their impact must also be taken into account. Further analysis of this topic is important because the Middle East has a substantial potential for economic growth and as the region grows its share of international trade will increase. Neither U.S. nor Europe can ignore the Middle East market in the future and if more detailed analysis of this topic shows that Middle East importers are sensitive to an exporter s foreign policy then the United States and the European Union will both have to take this factor into account. 19

Footnotes 1) The Direct Arab boycott of American products began soon after the start of second Palestinian uprising in September 2000. Within a matter of months National Boycott Committees were set up is UAE, Egypt and many other Arab countries. The NGOs, religious leaders, political activists and even high school students joined the grass-roots campaign against American products. Well known brand American brand names such as McDonalds, Berger King and Coca Cola were targeted for boycott. The boycott organizers provided Arab consumers with list of American products and recommended European or domestic substitutes. As a result of boycotts UAE retailers reported a 50% decline (Reuters, July 25, 2003). According to U.S. officials the volume of trade between U.S. and the Arab world declined by more than 205 between September 2001 and September 2002. (World Tribune, Sep. 16, 2003). Despite significant domestic pressure no Arab government other than Syria endorsed the boycott of U.S. goods. However, Saudi government expressed its frustration with the U.S. policies by delaying a major oil and gas development agreement with a consortium that included several major American oil companies. The Saudies also expressed an interest in Eurofighter and other European weapon systems in response to the growing accusations in the U.S. media about their support for terrorist organizations. (Gulf News, Dec. 6, 2002). 2) One of the major differences between U.S. and European policies is on the role of Yasser Arafat. While U.S. supports Israel s accusation that Arafat was implicitly responsible for the suicide attacks against Israeli civilians, European countries insisted that Arafat was a positive and stabilizing force in the occupied territories. For more details see EU challenges U.S. over Israel, Guardian, 2/1/2002. 3) Several recent studies (Aristotelous (2001), Barkolar et.al. (2002), Sercu and Appal (2003) ) have suggested that in addition to exchange rate itself, the volatility of exchange rate also affects the volume of trade among nations. However, since the countries that I have included in my sample; Egypt and GCC nations have maintained highly stable or fixed exchange rates against U.S. dollar I decided not to take this consideration into account. 20

Reference Aristotelous, Kyriacos, Exchange Rate Volatility, Exchange rate Regime, and Trade Volume: Evidence from UK-US Export Functions (1889-1990), Economic Letters, Volume 72, Issue 1, July 2001, pp87-94. Barkoulas, John T., C.F.Baum amd M. Caglayan, Exchange Rate Effects on the Volume and Variability of Trade Flows, Journal of International Money and Finance, Vol. 21, Issue 4, August 2002, pp481-496. Lim,Kang-Taeg and Kim, Jae-Young, Economic and Political Changes and Import Demand Behavior of North Korea, Journal of Economic Development, Vol.27, Number 1, June 2002 Sercu, Piet and Raman Uppal, Exchange Rate Volatility and International Trade: A General-Equilibrium Analysis, European Economic Review, Volume 47, Issue 3, June 2003, pp429-441. 21

22