Clash of civilizations and the impact of cultural differences on trade

Author’s Accepted Manuscript Clash of Civilizations and the Impact of Cultural Differences on Trade Gunes Gokmen

www.elsevier.com

PII: DOI: Reference:

S0304-3878(16)30115-8 http://dx.doi.org/10.1016/j.jdeveco.2016.12.008 DEVEC2118

To appear in: Journal of Development Economics Received date: 23 December 2015 Revised date: 22 December 2016 Accepted date: 24 December 2016 Cite this article as: Gunes Gokmen, Clash of Civilizations and the Impact of Cultural Differences on Trade, Journal of Development Economics, http://dx.doi.org/10.1016/j.jdeveco.2016.12.008 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Clash of Civilizations and the Impact of Cultural Di¤erences on Trade Gunes Gokmen December 2016

Abstract This paper studies the Clash of Civilizations hypothesis from an economic perspective. Using data on bilateral trade and measures of culture, we evaluate how the impact of cultural di¤erences on trade evolves over time during and after the Cold War. Evidence suggests that the negative in‡uence of cultural di¤erences on trade is more prominent in the post-Cold War era than during the Cold War. For instance, ethnic di¤erences reduce trade by 24% during the Cold War, whereas this reduction is 52% in the post-Cold War period. We also suggest a channel for the di¤erential impact of cultural di¤erences over time. By studying the evolution of the e¤ects of cultural di¤erence and cold-war blocs on trade, we provide evidence consistent with the hypothesis that cold-war blocs have trumped cultural di¤erences during the Cold War. Thus, cultural determinants of trade replace cold-war blocs as a major impediment to international trade only after the end of the Cold War. Keywords: Cold War, culture, economic clash, trade. JEL Classi…cation: F1, Z10. New Economic School. [email protected]. I am indebted to the editor Nathan Nunn, three anonymous referees, Alberto Alesina, Maristella Botticini, Klaus Desmet, Marc Dincecco, Simeon Djankov, Paolo Epifani, Alessandra Fogli, Selim Gulesci, Andreas Madestam, Kiminori Matsuyama, Matthias Messner, Annaig Morin, Tommaso Nannicini, Michele Pelizzari, Barbara Petrongolo, Giovanni Pica, Jean-Philippe Platteau, Massimiliano Onorato, Guido Tabellini, Silvana Tenreyro, Pierre-Louis Vezina, Romain Wacziarg, Shlomo Weber, and numerous seminar and conference participants for their help and support. I acknowledge the support of the Ministry of Education and Science of the Russian Federation, grant No. 14.U04.31.0002 administered through the NES CSDSI. The usual disclaimers apply.

1

1

Introduction

Cultural di¤erences play an important role in economic exchange.1 In this context, cultural barriers to trade are associated both with transaction costs, due to informal barriers, trust, business networks, informational costs and misunderstandings related to non-verbal communication, and with dissimilarity of preferences and tastes of culturally distant consumers. We add to this line of research by studying how the impact of cultural di¤erences on trade evolves over time, and how this impact interacts with the Cold War. In particular, this paper analyzes how the e¤ect of cultural di¤erences on trade changes during and after the Cold War. This relates to the Clash of Civilizations hypothesis by Huntington (1993a). This hypothesis puts forward that in the post-Cold War period the dominating source of discord will be cultural, and dissimilarity in culture will lead to clashes over a range of issues including trade. While Huntington argues for an increase in both violent and non-violent competition among cultural groups, the Clash of Civilizations hypothesis has so far received attention from a military con‡ict angle only,2 and it has not been empirically tested from an economic perspective. This is the aim of the present paper. To that end, using civilizations, religion, ethnicity and language as proxies of culture, we evaluate whether the negative e¤ect of cultural di¤erences on trade ampli…ed in the post-Cold War era.3 Employing bilateral imports data over 1962-2012, we provide evidence that the negative in‡uence of cultural di¤erences on trade is larger in the post-Cold War era than during the Cold War. For instance, ethnic di¤erences reduce trade by 24% during the Cold War, whereas this reduction is 52% in the post-Cold War period. Additionally, we quantify the tari¤ equivalent costs of cultural di¤erences for standard levels of elasticities of substitution in the literature. For example, with an elasticity of substitution of eight, the tari¤ equivalent cost of cultural di¤erences varies between 1.3% and 7.4% during the Cold War, while this additional cost is between 9.4% and 19.4% in the post-Cold War era. Furthermore, we explore the mechanism for the di¤erential impact of cultural di¤erence 1

See Felbermayr and Toubal (2010), Guiso et al. (2009), Melitz (2008), Melitz and Toubal (2014), and Rauch and Trindade (2002). 2 See Chiozza (2002), Henderson and Tucker (2001), Russett et al. (2000). 3 Throughout the paper, post-Cold War refers to post-1991.

2

over time. Huntington (1993a) argues that dissimilarity in culture gives rise to di¤erences in how we perceive and carry out a multitude of issues, including economic exchange. Businessmen make deals with people they can understand and trust; states surrender sovereignty to international organizations composed of like-minded states they understand and trust. Thus, the roots of economic cooperation are in cultural commonality. However, such tendencies, he claims, were held in check by the Cold War. Cold War institutions repressed these more fundamental channels of culture, and arti…cially promoted trade among countries of the same Cold War bloc. Therefore, only by the end of the Cold War, cultural cleavages resurface to increasingly prevail over ideological ones.4 To better understand this mechanism, we assign each country to a cold-war bloc and create an indicator of di¤erent blocs. Then, we track the evolution of the e¤ects of cultural di¤erences and di¤erent blocs on trade over time. The strong negative e¤ect of di¤erent blocs on trade over the Cold War disappears by the end of the Cold War, and instead, cultural di¤erence gains signi…cance as a trade barrier. We also show that the di¤erential impact of cultural di¤erence in the post-Cold War era is largely driven by former same-bloc countries. The evidence we provide is consistent with the hypothesis that cold-war blocs have trumped cultural di¤erences during the Cold War. Therefore, long-term cultural determinants of trade gain more signi…cance and replace cold-war blocs as a major impediment to international trade only after the end of the Cold War. Our main …ndings are robust to alternative speci…cations. We estimate a gravity model of international trade accounting for time-varying multilateral resistance terms5 as well as country-pair …xed e¤ects. We employ a set of cultural-di¤erence measures that allow us to capture di¤erent aspects of culture. Unlike other existing studies (Felbermayr and Toubal, 2010; Giuliano et al., 2006; Guiso et al., 2009; Rauch and Trindade, 2002),6 our data set 4 This can be interpreted in more economic terms as follows. Although trade frictions tend to be larger for countries with di¤erent cultures, during the Cold War there were incentives to trade within the same ideological bloc ignoring the trade frictions associated with di¤erent cultures. However, after the end of the Cold War, such political mechanisms disappear, and cultural barriers to trade converge to their market equilibrium. 5 Omission of which leads to biased estimates. See Anderson and Van Wincoop (2003), and Baldwin and Taglioni (2007). 6 Melitz and Toubal (2014) is an exception. They collect linguistic data on 195 countries and show that the in‡uence of language on trade works not only through cultural components but also through ease of

3

contains most of the countries of the world. We control for an array of measures of geographic barriers as well as historical and policy-related determinants of trade. Results are also robust to taking into account time-varying e¤ect of distance, genetic distance as an alternative measure of culture, political proximity, communication channel, zero trade ‡ows, and a rich set of geographic controls. This study contributes to the literature on the Clash of Civilizations thesis by adding an economic perspective. This strand of the literature has focused on the militarized disputes aspect of the thesis and its implications for economic interaction among cultural groups remained unexamined.7 To our knowledge, we are the …rst to study this thesis from an economic perspective. This paper also adds to the literature on trade and culture by bringing in the dynamics and showing the evolution of the e¤ects of culture. In this strand of the literature, for instance, Felbermayr and Toubal (2010) establish a correlation between culture and trade using scores from the Eurovision Song Contest as a proxy for cultural proximity. Guiso et al. (2009) show that bilateral trust between pairs of European countries leads to higher trade between them. However, the dynamic aspect of the in‡uence of culture is absent in these analyses. This could be important, for instance, to explain the recent regionalization phenomenon. Moreover, we know that trade a¤ects con‡ict involvement (Martin et al., 2008), and our results suggest that, in the post-Cold War era, cultural di¤erences might have an additional indirect e¤ect on the probability of con‡ict by reducing bilateral trade. Another strand of related literature looks at trade in the context of the Cold War. Importantly, Berger et al. (2013) show that during the Cold War imports from the US increased as a result of stronger political in‡uence arising from CIA interventions. Alternatively, Djankov and Freund (2002) study trade between Russian regions and former Soviet republics, and …nd that there is an increasing bias toward domestic trade after the disintegration of the Soviet Union. The paper proceeds as follows. Section 2 lays out the methodology and describes the data. communication. 7 For a discussion on the militarized con‡ict aspect of the thesis, see Chiozza (2002), Henderson and Tucker (2001), Russett et al. (2000).

4

Section 3 provides results. Section 4 proposes a potential channel. Section 5 concludes.

2

Econometric Speci…cation and Data

2.1

Econometric Speci…cation

We estimate a standard gravity equation (Anderson and Van Wincoop, 2003). The theoretical gravity equation is:

Mij = G

Yi Yj 1

; where G

ij

1 Y

1 i Pj

(1

)

(1)

Mij is the nominal value of imports from country i to country j; Yi and Yj are country i’s and country j’s economic sizes, respectively; income;

ij

is bilateral trade costs; Y is world nominal

> 1 is the elasticity of substitution between goods;

i

and Pj can be thought of as

price indices.8 ij

re‡ects all trade costs, natural and man-made, between countries i and j: In addition

to transportation costs, these trade costs might re‡ect legal costs, regulatory and institutional costs, and all the remaining costs that form bilateral trade barriers. This is where we see our measures of cultural di¤erence come into play as a cultural barrier to trade. Cultural variables re‡ect, among other things, business norms, customs, beliefs, trust and information costs that might act as a source of informational cost and uncertainty, and thus, impede trade relations between countries. Log-linearization of equation (1) gives the empirical gravity equation:

log Mij =

log Y + log Yi Yj + (1

) log

ij

+(

1) log

i Pj

(2)

Moreover, Feenstra (2002) shows that an estimation strategy with exporting and importing 8

Notice that the G term bears the price indices of the two countries. Although i and Pj could be interpreted as price indices in the model, they cannot be interpreted as price levels in general. These unobservable variables should be better thought of as nonpecuniary trade costs a country has with all its trading partners. Hence, i and Pj represent average trade barriers of country i and country j, respectively, which are referred to as "multilateral resistance" terms. Omission of multilateral resistance terms leads to biased estimates (Anderson and Van Wincoop, 2003). See Head and Mayer (2013) for more details and a discussion of the state-of-the-art.

5

country …xed e¤ects produces consistent estimates, whereas Baldwin and Taglioni (2007) show that with panels importing and exporting country …xed e¤ects should be time-varying. Thus, our empirical speci…cation is a log-linearized version of equation (1) together with time-varying importing and exporting country …xed e¤ects, and country-pair …xed e¤ects where applicable:9

log Mijt = a + Cij +

P

k kijt

+ Rit + Rjt +

ijt

(3)

k

where Cij is our variable of interest, which is a binary variable that captures cultural di¤erences across country pairs;

kijt

represents k control variables; Rit is time-varying ex-

porting country …xed e¤ects; Rjt is time-varying importing country …xed e¤ects; and

ijt

is

the unaccounted-for error term.

2.2

Data

Measure of Trade. Trade data between 1962-2012 are from the UNComtrade Trade Data Set. Measures of Culture. As a …rst measure of culture 179 countries are classi…ed as members of various civilizations. These civilizations are Western, Sinic, Islamic, Hindu, Orthodox, Latin American, African, Buddhist and "Lone" States. The classi…cation and the construction of civilization membership is based on Huntington (1998). Accordingly, each country is assigned to a civilization.10 Then, to indicate civilizational dissimilarity within a pair we construct an indicator variable, "Di¤erent Civilizations," denoting whether a pair of countries belong to di¤erent civilizations. Out of 15931 country-pairs, 2875 pairs are formed of countries belonging to the same civilization and 13056 pairs to di¤erent civilizations. In addition, Tanja Ellingsen’s "Ethnic Witches’Brew Data Set" provides us with data on religious, linguistic and ethnic fragmentation within countries between 1945-2012.11 Ellingsen (2000) collected data on the dominant linguistic, religious, and ethnic groups of countries.12 9

Guimarães and Portugal (2010) provide an algorithm to run estimations with high dimensional …xed e¤ects. See Table 1A in the Appendix for details. 11 The original data by Tanja Ellingsen runs from 1945 to 1994. We use the version of the data by Gartzke and Gleditsch (2006) and extend it further. 12 She has obtained information from three reference books: Handbook of the Nations, Britannica Book of 10

6

What is important for our purposes in this data set is the information on the name of the largest linguistic, religious, and ethnic groups. We recode these variables so that they take value one when two countries have di¤erent majority religion, "Di¤erent Religion", or different majority ethnicity, "Di¤erent Ethnicity", or di¤erent majority language, "Di¤erent Language," zero otherwise.13 Other Determinants of Trade. Geographic barriers proxy transportation as well as information costs. Correspondingly, we control for contiguity14 and geodesic distance15 between the major cities of the countries. To control for historical, political and institutional links we include indicator variables for whether a pair of countries ever had a colonial relationship; had a common colonizer after 1945; and whether the two countries have been part of the same polity.16 In addition, same legal origins in a pair of countries might reduce information costs related to legal and regulatory systems, and sharing the same legal origins might enhance trust between interacting parties (Guiso et al., 2009). Hence, we include an indicator variable for two countries with the same legal origins.17 We also take into account policy related dyadic variables, such as free trade area (FTA), GATT/WTO membership, common currency and generalized system of preferences (GSP).18;19;20 the Year and Demographic Yearbook. 13 Table A.10 in the Online Appendix presents simple unconditional correlations between imports and measures of cultural di¤erence over the entire time period. All measures of cultural dissimilarity are negatively associated with trade. Moreover, there is a strong correlation between the di¤erent civilizations indicator, and di¤erent religion and di¤erent language, as well as between di¤erent language and di¤erent ethnicity. 14 Contiguity data come from Correlates of War Project, Direct Contiguity Data, 1816-2006, Version 3.1 (Stinnett et al., 2002). 15 See Head and Mayer (2002) for details. 16 These data come from CEPII. The data are available at http://www.cepii.fr/anglaisgraph/bdd/distances.htm. 17 Legal origin indicators (common law, French civil law, German civil law, Scandinavian law, and Socialist law) are from La Porta et al. (1999). 18 Data are from Martin, Mayer and Thoenig (2008). Available at http://econ.sciences-po.fr/node/131. 19 As noted by Anderson and van Wincoop (2004), regional trade agreements may not be exogenous, and therefore, FTA included contemporaneously may su¤er from reverse causality. A reasoning for this is that countries might have agreed on a trade agreement since they already have been trading lots for many reasons that are not observed by the econometrician. Consequently, we tried lagging FTA variable to overcome reverse causality up to four-period lags. The results concerning our variables of interest carry over. 20 Summary statistics are in the Appendix Table 2A.

7

3

Results

In Table 1, we start with con…rming the general negative e¤ect of cultural di¤erences on trade established in the literature. Over the entire sample, accounting for a full set of geographical, historical, political, and economic barriers to trade as well as time-varying multilateral resistance terms, the e¤ect of di¤erent civilizations indicator is both economically and statistically signi…cant. If two countries in a pair belong to di¤erent civilizations their import ‡ows are 32.7% lower (column (1)).21 Furthermore, columns (2); (3); and (4) show that having di¤erent majority religion or di¤erent majority ethnicities or di¤erent majority languages negatively a¤ect trade relations. For instance, when the two countries have di¤erent dominant ethnicities they tend to have, on average, about 42% lower import ‡ows than the two countries with the same dominant ethnicity, while this e¤ect is 59% for di¤erent languages. In column (5), we include all of the measures of cultural dissimilarity together. We observe that the e¤ect of di¤erent civilizations is diminished to a great extent, and hence, most of the variation in cultural dissimilarity can be explained by religion, ethnicity and language. All the estimates of the control variables have the expected signs.22

3.1

Main Results

In this subsection, we present results on how the e¤ect of cultural di¤erences on trade evolves during and after the Cold War. This relates to the Clash of Civilizations hypothesis by Huntington (1993a). He argues that in the post-Cold War period the dominating source of clashes will be cultural, and dissimilarity in culture will engender clashes over a range of issues including economic interaction. While Huntington argues for an increase in both violent and non-violent competition among civilizations, the Clash of Civilizations hypothesis has so far been tested from a military con‡ict angle only,23 and the economic clash24 aspect has never been put to rigorous econometric testing. Therefore, using civilizations, religion, ethnicity and language as proxies of culture, we test whether there has been an ampli…cation in the 21

[exp( 0:397) 1] 100 ' 32:7% See, for example, Blomberg and Hess (2006), and Glick and Taylor (2005). 23 Chiozza (2002), Henderson and Tucker (2001), Russett et al. (2000). 24 In reference to Huntington’s thesis, we use the term "economic clash" just to refer to lower trade. 22

8

negative e¤ect of cultural di¤erences on trade in the post-Cold War era. Tables 2 and 3 present the main results of the paper. In Table 2, each cell of a row reports the coe¢ cient of a cultural variable of interest from a separate regression in the two respective time periods. Evidence suggests that there is a surge in economic clashes of di¤erent cultures in the post-Cold War era. The negative e¤ect of belonging to di¤erent cultural groups on bilateral trade is much larger in the post-Cold War era than in the Cold War era. This …nding is not subject to the de…nition of culture, and all four measures of cultural dissimilarity have a stronger impact in the post-Cold War era. For instance, when the two trading partners do not share the same dominant ethnicity, their imports are reduced, on average, by 24% during the Cold War, whereas this reduction is 52% in the post-Cold War period. Alternatively, in the post-Cold War period, two countries with distinct religious majorities tend to have 40% lower imports than those sharing the same religion, whereas this negative e¤ect is 23% during the Cold War.25 These results show that the end of the Cold War brought about more con‡ictual economic relations among countries of heterogeneous cultural backgrounds. Chow tests con…rm that coe¢ cients in the two periods are statistically di¤erent. In Table 3, we carry out a similar analysis with a di¤erence-in-di¤erence strategy instead of splitting the sample. We interact di¤erent culture variables with a post-Cold War indicator and run regressions on the entire sample. Columns (1) to (4) include controls, timevarying importer and exporter …xed e¤ects, whereas columns (5) to (8) additionally control for country-pair …xed e¤ects. Previous results and interpretations remain the same. The negative e¤ect of cultural di¤erence on trade is larger in the post-Cold War than in the Cold War period.26 For example, the di¤erential e¤ect of religion in column (6) is -10%, whereas that of ethnicity in column (7) is -37%. It is also important to note that when we run the same regressions excluding those countries that gained their independence only after the end 25 These …ndings are not in‡ated due to the time-invariant nature of our variables of interest. On the contrary, they are closer to the lower bound estimates. When we collapse the data to a cross-section by taking the mean imports as dependent variable, the results are qualitatively the same, and in some cases the coe¢ cients on cultural di¤erence variables are even bigger. This is because when we run the regressions in a panel setup we control for many dyadic time-varying determinants of trade as well as time-varying importer and exporter …xed e¤ects. 26 For columns (1) to (4), we test whether the coe¢ cients of cultural di¤erence are the same during and after the Cold War. This test is rejected for every measure of culture. P-values are reported in Table 3.

9

of the Cold War, regressions yield the same conclusions. Therefore, our …ndings are not an artefact of the newly independent countries. In Table 4, we ask what if cultural di¤erence was a tari¤ and calculate the tari¤ equivalent costs of cultural dissimilarity for di¤erent elasticities in the two time periods.27 suggests that the regression coe¢ cients correspond to the estimates of [(1 (

Theory

) ln ], and

1) is the tari¤ equivalent of the cultural barriers to trade. In line with the literature, we

calculate the tari¤ equivalent of cultural trade barriers for elasticities of

= 5;

= 8;

= 10

(Anderson and van Wincoop, 2004). We observe in Table 4 that the minimum tari¤ equivalent cost of cultural dissimilarity is 1% during the Cold War, whereas this lower bound estimate is about 5.8% in the post-Cold War era. On the other hand, the maximum tari¤ equivalent cost of culture during the Cold War is about 13%, while this upper bound estimate is about 36% in the post-Cold War era. For example, for an elasticity of 5, the tari¤ equivalent cost of di¤erent ethnicities is 7% during the Cold War, whereas it is about 20% in the post-Cold War. To put these numbers in perspective, Anderson and van Wincoop (2003), for instance, calculate the tari¤ equivalent cost of national borders as 48% (for

= 5). In our case, in the

post-Cold War period, di¤erent languages account for more than half of the estimate of the national border barrier, while di¤erent religions and di¤erent ethnicities roughly equal one forth and forty percent of the estimate of the national border barrier, respectively.

3.2

Robustness

In this subsection, we discuss the robustness of our main …nding.28 Distance. Disdier and Head (2008) and Carrère et al. (2009) show that the elasticity of trade to distance has increased over time. Therefore, we probe that what cultural di¤erence captures over time is not just a re‡ection of distance e¤ect. When we take into account the di¤erential e¤ect of distance over time in the post-Cold War period, previous …ndings on cultural di¤erence variables carry over (see Table A.1 in the Online Appendix).29 27

See, for instance, Blomberg and Hess (2006) and Rose and van Wincoop (2001) for examples on the tari¤ equivalent costs of trade barriers. 28 Please refer to the Online Appendix for extended discussions and the results on the robustness analysis. 29 We also included an interaction of cultural di¤erence and trend in order to show that the di¤erential e¤ect

10

Genetic Distance.

Genetic distance as a proxy for cultural distance has recently re-

ceived widespread attention from researchers (Giuliano, Spilimbergo and Tonon, 2006; Guiso, Sapienza and Zingales, 2009; Spolaore and Wacziarg, 2009a, 2009b). Moreover, Desmet et al. (2006) provide empirical support for choosing genetic distance as a proxy for cultural di¤erences measured by the World Values Survey. To that end, we would like to understand the sensitivity of our measures of culture against genetic distance variable.30 Previous …ndings on the e¤ect of cultural di¤erences on trade in the post-Cold War hold when we control for genetic distance as a proxy for culture. If genetic distance captures an element of culture, our measures of culture explain some additional constituent of it that is not explained by genetic distance. Moreover, genetic distance itself as a proxy for cultural distance has a signi…cantly negative e¤ect on trade in the post-Cold War era, a …nding that supports our main conclusion (see Table A.2 in the Online Appendix).31 Political Proximity. Political factors might be an in‡uential determinant of trade ‡ows between countries. For instance, political proximity, measured as correlations of votes at the United Nations General Assembly, positively impacts bilateral trade (Umana Dajud, 2013). At the same time, the liberal peace argument suggests that democratic countries trade more (Bliss and Russett, 1998; Yu, 2010). Relatedly, Umana Dajud (2013) shows that countries that are closer on the democracy/autocracy axis trade more. Furthermore, Long (2003) and Morrow et al. (1998) provide evidence that countries in mutual security alliances have greater levels of trade. On these grounds, we examine how political proximity in‡uences the e¤ect of cultural dissimilarity, and thus, bring United Nations voting correlations,32 regime of cultural di¤erence during and after the Cold War is not just a time trend. Previous results carry over. 30 Genetic distance data we use are from Spolaore and Wacziarg (2009a) as the genetic distance information on populations is mapped onto countries. 31 We have also run the same speci…cation replacing current genetic distance with genetic distance in 1500. The conclusion is the same. Taking genetic distance in 1500 as a long run component of cultural distance, the negative e¤ect is much stronger in the post-Cold War period compared to the Cold War period. 32 UN voting correlations data are from Erik Gartzke, who created The A¢ nity of Nations Index based on the United Nations General Assembly roll-call data. This index takes values between -1 and 1 for the correlation of votes between countries at the United Nations General Assembly over the period 1948-2006. Available at http://dss.ucsd.edu/~egartzke/htmlpages/data.html and at http://www9.georgetown.edu/faculty/ev42/UNVoting.htm.

11

di¤erences,33 and security alliances34 into the analysis. Previous …ndings on the e¤ect of cultural dissimilarity on trade remain the same (see Table A.3 in the Online Appendix). Communication Channel. Melitz and Toubal (2014) argue that the in‡uence of language on trade works not only through cultural components but also through ease of communication. Therefore, we test whether what we capture is the e¤ect of communication rather than culture. We include in the regression common spoken language variable of Melitz and Toubal (2014) together with our measures of cultural di¤erence.35 The di¤erential impact of culture in the post-Cold War is not driven by communication, and the previous …ndings hold (see Table A.4 in the Online Appendix). Zero Trade Flows. The question of how to deal with zero trade ‡ows is an on-going debate in the literature.36 Thus, we test the robustness of our results to the inclusion of zero trade ‡ows. We apply alternative methods: adding one to trade ‡ows, applying inverse hyperbolic sine transformation to trade ‡ows,37 and the Heckman two-step selection method.38 By and large, our results are robust to the inclusion of zero trade ‡ows and are not driven by the omission of zero trade ‡ows from the estimation analysis (see Tables A.5 through A.7 in the Online Appendix).39 33

As in Umana Dajud (2013), we generate a variable labelled "Regime Di¤erence," which equals the absolute value of the di¤erence between two countries’democracy scores from the Polity IV project. 34 We employ an indicator variable for whether a pair of countries are in an alliance. Alliances data are Version 3.03 from Correlates of War Project (Gibler, 2009). 35 Results are identical when we also include common o¢ cial language. 36 See Head and Mayer (2013) for a discussion of the state of the art. 37 Inverse hyperbolic sine transformation is an easy-to-apply method that is de…ned for any real number and formally de…ned as: sinh 1 (x) = log(x + (x2 + 1)1=2 ) (see Burbidge et al., 1988). Burbidge et al. (1988) shows that inverse hyperbolic sine transformation is a viable alternative to log transformation when the dependent variable can take on zero values. We apply this transformation to import ‡ows so that the log function is de…ned for the zero values of the dependent variable as well. Moreover, this way we refrain from adding the same ad hoc constant to each observation of import ‡ows. Instead, each value to be added to the dependent variable changes and is determined by the dependent variable itself. For an example of inverse hyperbolic sine transformation in gravity models, see Kristjánsdóttir (2012). 38 Following Martínez-Zarzoso (2013), we use the number of islands in the pair as an exclusion restriction with the assumption that islands a¤ect the probability of trading but not how much you trade conditional on trading. 39 In addition to the robustness checks that are presented in this subsection, we also run robustness tests (not shown) for con‡ict involvement, role of minorities and di¤erences in GDP per capita of countries. Previous results on the e¤ect of cultural di¤erences on trade remain the same.

12

4

A Potential Channel

In this section, we explore a potential channel for the time-varying e¤ect of culture on trade. A possible mechanism for the di¤erential impact of cultural dissimilarity during and after the Cold War could be the role ideology and Cold War blocs play in these two time periods. One can argue that although cultural di¤erences have always been present, they were subdued during the Cold War. With the end of the Cold War, however, cultural di¤erences resurface and become more salient. In more economic terms, although trade frictions tend to be larger for countries with di¤erent cultures, during the Cold War there were incentives to trade within the same ideological bloc ignoring the trade frictions associated with di¤erent cultures. However, after the end of the Cold War, such political mechanisms disappear, and cultural barriers to trade converge to their market equilibrium. Thus, to assess how the e¤ect of Cold War blocs, in conjunction with culture, on bilateral trade evolves, we construct a di¤erent blocs indicator. First, based on Huntington (1998), every country is assigned to either the …rst world or the second world or the third world as they were in the heights of the Cold War. The …rst world is composed of the United States and its allies; the second world is composed of the Soviet Union and its allies; and the third world is composed of unaligned countries.40 Then, we create an indicator, "Di¤erent Blocs," that takes one if the two countries belong to two di¤erent superpower camps. In other words, this variable equals one if one country in the pair belongs to the …rst world and the other one belongs to the second world, zero otherwise. We present the …rst set of results in Table 5. Columns (1) to (4) suggest that countries that were in di¤erent blocs during the Cold War had lower import ‡ows than those of the same bloc. Also, the large e¤ect of di¤erent blocs during the Cold War dwarfs that of cultural dissimilarity, and the impact of blocs is at least four times greater than the impact of any measure of culture. For example, during the Cold War, the negative e¤ect of Di¤erent Blocs is four times greater than that of Di¤erent Religion (Table 5, column (2)), and this corresponds to -52% and -16% reductions in trade, respectively. 40

See Table 3A in the Appendix for details.

13

A logical interpretation from Table 5 is that the impact of ideological di¤erences were so great during the Cold War that cultural dissimilarities were trumped and had a less important role in trade relations. However, in the post-Cold War period, country pairs that were formerly in di¤erent blocs start trading and making up for their low levels of prior trade.41 In the remaining columns (5) to (8), we also control for country-pair …xed e¤ects. The di¤erential impacts of cultural di¤erence and di¤erent blocs over time are not driven by country-pair speci…c unobserved heterogeneity. Furthermore, we track the evolution of the impact of cultural dissimilarity and di¤erent blocs on trade over time from 1962 to 2012. Figure 1 plots the coe¢ cients of di¤erent religion and di¤erent blocs together with the 95% con…dence interval from a regression of log imports on di¤erent religion-year interactions, di¤erent blocs-year interactions, control variables, and time-varying country …xed e¤ects.42 The results are striking. Being part of di¤erent blocs impedes trade relations during the Cold War, and this e¤ect is both economically and statistically signi…cant. Toward the …nal years of the Cold War, however, we observe a decreasing trend (in absolute values) in the negative e¤ect of di¤erent blocs, and after the demise of the Cold War it is not signi…cant anymore. On the other hand, throughout most of the Cold War, the e¤ect of di¤erent religious backgrounds on bilateral trade lingers around zero. However, towards the end of the Cold War the impact of di¤erent religion variable exhibits a jump and doubles. This jump is in the year 1984. In 1983 the coe¢ cient on di¤erent religion is about -0.16, whereas in 1984 it is -0.34, signi…cant in both cases. This evidence in the data overlaps with the historical signs of the end of the Cold War accumulating in the mid 80s. For example, Mikhail Gorbachev assumed power in the Soviet Union in 1985, and immediately after, liberal-minded Gorbachev started implementing reforms. Consequently, both economic (Perestroika) and political (Glasnost) 41

In a similar vein, Berger et al. (2013), for example, study political in‡uence on trade, and show that, during the Cold War, when the CIA intervened in another country successfully, imports from the US increased dramatically. 42 For space considerations, we present results with Di¤erent Religion, although we carry out the same exercise for di¤erent civilizations, di¤erent ethnicity and di¤erent language variables. The results are similar and available upon request.

14

liberalization packages were put into e¤ect. At the same time, the relations with the leaders of the U.S. and the U.K. at the time -Ronald Reagan and Margaret Thatcher, respectivelyimproved considerably.43 All of these developments signalled the de facto end of the Cold War. Thus, long-term cultural determinants of trade start gaining more signi…cance in the mid 80s toward the end of the Cold War, and replace di¤erent blocs as a major impediment to international trade. After the end of the Cold War, there is a stable negative relation between cultural di¤erence and trade. Next, we perform a triple interaction exercise. Table 6 shows regressions of trade on the triple interaction of cultural di¤erence, di¤erent blocs, and post-Cold War. If in the Cold War era cultural di¤erences were curbed by ideology, then this implies that inter-cultural dyads that belonged to the same Cold War bloc must have set aside their cultural di¤erences. Therefore, the di¤erential impact of culture on trade in the post-Cold War should be more accentuated for former same-bloc pairs. The evidence in Table 6 is consistent with this hypothesis. The di¤erential post-Cold War e¤ect of cultural di¤erence is, by and large, more negative for former same-bloc countries. This is further evidence that Cold War blocs might have played a role in the di¤erential impact of cultural dissimilarity on trade after the Cold War.44 Lastly, we construct a synthetic measure of latent cultural di¤erence. One might think that our measures of cultural di¤erence are correlated, and hence, should not be included in the same regression. On the other hand, if they are included in separate regressions, certain elements of culture might be ignored. To that end, we create a variable, "Cultural Di¤erence," as a weighted average of our four measures of cultural di¤erence with the loadings of the …rst principal component as weights.45;46 43

For example, the Reykjavík Summit between Ronald Reagan and Mikhail Gorbachev led to the eventual Intermediate-Range Nuclear Forces Treaty between the U.S. and the Soviet Union in 1987. Another example, Margaret Thatcher addressed Mikhail Gorbachev as a man she can do business with. 44 To give an anecdotal example, average total trade to GDP ratio between Hungary and Mongolia, who were in the same Cold War bloc but obviously belonged to di¤erent cultural backgrounds, is 0.087 before 1991, while it is 0.02 after 1991. 45 Cultural Di¤ erence= 0:5416 N (Di¤ erent Civilizations)+0:4351 N (Di¤ erent Religion)+0:4642 N (Di¤ erent Ethnicity ) + 0:5495 N (Di¤ erent Language) where N ( ) is a function of standard normalization. For example, United States-United Kingdom, Guatemala-Bolivia, and Sierra Leone-Tanzania pairs have the lowest cultural di¤erence, while Kenya-Japan, Bulgaria-Libya, and Israel-Sri Lanka pairs have the highest cultural di¤erence. 46 The correlation coe¢ cients of the cultural di¤erence variable with: di¤erent civilizations=0.76, di¤erent

15

The results with Cultural Di¤erence, Di¤erent Blocs, and post-Cold War interactions are in Table 7. First column shows the negative impact of cultural di¤erence on trade over the entire sample, while the second column shows that this negative e¤ect is stronger in the post-Cold War period. In column (3), we contrast as before the evolution of cultural di¤erence with that of di¤erent blocs and previous interpretations carry over. Importantly, the triple interaction speci…cation in column (4) suggests that the di¤erential impact of cultural di¤erence in the post-Cold War is signi…cant for former same-bloc countries, while this e¤ect is insigni…cant for former di¤erent-blocs countries (with a composite coe¢ cient of -0.069 and p-value of 0.28). This suggests that inter-cultural dyads that belonged to the same Cold War bloc must have suppressed cultural barriers to trade, and as a consequence, the di¤erential impact of culture on trade in the post-Cold War is more accentuated for former same-bloc pairs. Thus, the evidence is consistent with the hypothesis that Cold War blocs trumped cultural dissimilarity and that is why culture plays a more signi…cant role in trade relations only after the end of the Cold War. However, one should be cautious with the interpretations of these results. Although we provide suggestive evidence, we do not directly test whether ideology suppressed cultural di¤erences during the Cold War.

5

Conclusion

The main novelty of this study is to test Huntington’s the Clash of Civilizations hypothesis from an economic perspective. We study the dynamics of the e¤ect of cultural di¤erences on trade and show that the negative in‡uence of cultural di¤erences on trade has increased over time. More speci…cally, cultural di¤erences are a larger barrier to international trade in the post-Cold War period than in the Cold War period. For instance, when two countries do not share the same dominant ethnicity, their imports are 24% lower during the Cold War, whereas this reduction is 52% in the post-Cold War period. Furthermore, we suggest an explanation for the di¤erential impact of cultural dissimilarity over time. By studying the evolution of the e¤ects of cultural dissimilarity and Cold War blocs, religions=0.61, di¤erent ethnicities=0.65, and di¤erent languages=0.78.

16

we show that cold-war ideological blocs might have trumped cultural di¤erences during the Cold War. Therefore, long-term cultural determinants of trade gain more signi…cance only by the end of the Cold War and replace ideological di¤erences as a major impediment to international trade. Unstable trade relations is a source of concern for policy makers. This paper highlights a threat to the world trade system as found in cultural di¤erences. If this is an emergent phenomenon, then we might observe a shift in the behavior of the mass of individual economic actors via considerations of cultural and ideological identity. Such a destabilizing phenomenon at a global scale needs better understanding. A natural line of further investigation would be to look in more detail at the causes underneath the evolution of the impact of cultural dissimilarity. More disaggregated data, for example, could shed some more light on the main drivers of this phenomenon. Also, the role of recent migration patterns and minorities is worth investigating.

17

-1.5

Parameter estimate -1 -.5 0

.5

Evolution of the Effects of Different Religion and Different Blocs on Trade over Time

1960

1970

1980

1990

2000

2010

year Different Blocs

Different Religion

Figure 1: Parameter Estimates and 95% Con…dence Bands of Di¤erent Religion and Di¤erent Blocs Variables Throughout Years. The values are from the following regression speci…cation. Regressand: log Imports. Other Regressors: ln Distance, Contiguity, Colonial Link, Same Country, Common Colonizer, Same Legal Origin, FTA, Both in WTO, Common Currency, GSP and time-varying importing country and exporting country …xed e¤ects.

18

Table 1: Impact of Culture on Bilateral Trade (1) Di¤erent Civilizations -0.397 (0.038) Di¤erent Religion

(2)

(3)

-0.404 (0.034)

Di¤erent Ethnicity

-0.543 (0.088)

Di¤erent Language Additional Controls Importer-Year E¤ects Exporter-Year E¤ects N R2

(4)

YES YES YES 343714 0.722

YES YES YES 343714 0.722

YES YES YES 343714 0.721

-0.906 (0.070) YES YES YES 343714 0.723

(5) -0.111 (0.045) -0.259 (0.039) -0.169 (0.088) -0.689 (0.077) YES YES YES 343714 0.724

Regressand: log Imports. Regressors included but with unrecorded coe¢ cients: ln Distance, Contiguity, Colonial Link, Same Country, Common Colonizer, Same Legal Origin, FTA, Both in WTO, Common Currency, GSP, and time-varying importing and exporting country …xed e¤ects. Robust standard errors (clustered at the countrypair level) are in parentheses. p < 0:10,

p < 0:05,

p < 0:01

19

Table 2: Impact of Culture on Trade: Cold War vs. post-Cold War Comparisons

(1) Cold War -0.094 (0.048)

(2) post-Cold War -0.633 (0.044)

(3) Chow P-value 0.000

Di¤erent Religion

-0.268 (0.046)

-0.513 (0.040)

0.000

Di¤erent Ethnicity

-0.277 (0.107)

-0.749 (0.093)

0.000

Di¤erent Language

-0.500 (0.085)

-1.246 (0.076)

0.000

YES YES YES

YES YES YES

Di¤erent Civilizations

Additional Controls Importer-Year E¤ects Exporter-Year E¤ects

Each cell of a row reports the coe¢ cient of a cultural variable of interest from a separate regression in the two respective time periods. Regressand: log Imports. Regressors included but with unrecorded coe¢ cients: ln Distance, Contiguity, Colonial Link, Same Country, Common Colonizer, Same Legal Origin, FTA, Both in WTO, Common Currency, GSP and time-varying importing and exporting country …xed effects. Robust standard errors (clustered at the country-pair level) are in parentheses. Number of observations: Cold War=149434; post-Cold War=194280. p < 0:10,

p < 0:05,

p < 0:01

20

Table 3: Impact of Culture in the post-Cold War Di¤erent Civilizations Di¤erent Civilizations Post-Cold War

(1) 0.009 (0.045) -0.712 (0.040)

Di¤erent Religion Di¤erent Religion Post-Cold War

(2)

(4)

(5)

(6)

(7)

-0.176 (0.045) -0.395 (0.048)

-0.104 (0.042) -0.096 (0.105) -0.800 (0.092)

Di¤erent Ethnicity Post-Cold War Di¤erent Language Di¤erent Language Post-Cold War YES YES YES NO 343714 0.723 0.009

-0.353 (0.082) -1.008 (0.070) YES YES YES YES YES YES NO NO 343714 343714 0.722 0.724 0.000 0.000

-0.464 (0.079)

YES YES YES YES 343714 0.868

YES YES YES YES 343714 0.867

-0.518 (0.063) YES YES YES YES YES YES YES YES 343714 343714 0.867 0.867

Regressand: log Imports. Regressors included but with unrecorded coe¢ cients: ln Distance, Contiguity, Colonial Link, Same Country, Common Colonizer, Same Legal Origin, FTA, Both in WTO, Common Currency, GSP, and time-varying importing and exporting country …xed e¤ects. Columns (5)-(8) include also country-pair …xed e¤ects. Robust standard errors (clustered at the country-pair level) are in parentheses. p < 0:10,

p < 0:05,

(8)

-0.432 (0.036)

Di¤erent Ethnicity

Additional Controls YES Importer-Year E¤ects YES Exporter-Year E¤ects YES Country-Pair E¤ects NO N 343714 R2 0.724 P value Culture CW=Culture post-CW 0.000

(3)

p < 0:01

21

Table 4: Tari¤ Equivalent Costs of Cultural Barriers to Trade (in percentages)

Di¤erent Di¤erent Di¤erent Di¤erent

Civilizations Religion Ethnicity Language

Cold War (2) (3) (1) =5 =8 =10 2.37 1.35 1.04 6.92 3.90 3.02 7.17 4.03 3.12 13.31 7.40 5.71

post-Cold War (4) (5) (6) =5 =8 =10 17.14 9.46 7.28 13.68 7.60 5.86 20.59 11.29 8.67 36.54 19.48 14.84

The results in this table are based on the estimates from Table 2.

22

Table 5: Culture, Cold War Blocs, and Trade (1) (2) -0.009 (0.045) Di¤erent Civilizations Post-Cold War -0.690 (0.040) Di¤erent Religion -0.183 (0.045) Di¤erent Religion Post-Cold War -0.385 (0.048) Di¤erent Ethnicity

(3)

(4)

(5)

(6)

(7)

(8)

Di¤erent Civilizations

Di¤erent Ethnicity Post-Cold War

-0.408 (0.036)

-0.097 (0.042) -0.114 (0.105) -0.778 (0.092)

-0.373 (0.082) Di¤erent Language Post-Cold War -0.991 (0.070) Di¤erent Blocs -0.710 -0.740 -0.742 -0.768 (0.094) (0.094) (0.093) (0.093) Di¤erent Blocs Post-Cold War 0.643 0.706 0.682 0.685 (0.102) (0.104) (0.102) (0.102) Additional Controls YES YES YES YES Importer-Year E¤ects YES YES YES YES Exporter-Year E¤ects YES YES YES YES Country-Pair E¤ects NO NO NO NO N 343714 343714 343714 343714 R2 0.724 0.723 0.722 0.725 P value Culture=Blocs in CW 0.000 0.000 0.000 0.001 P value Culture=Blocs in post-CW 0.000 0.000 0.000 0.000

-0.440 (0.079)

Di¤erent Language

-0.498 (0.062)

0.998 1.073 1.059 (0.109) (0.110) (0.110) YES YES YES YES YES YES YES YES YES YES YES YES 343714 343714 343714 0.868 0.868 0.868 0.000

0.000

0.000

1.051 (0.110) YES YES YES YES 343714 0.868 0.000

Regressand: log Imports. Regressors included but with unrecorded coe¢ cients: ln Distance, Contiguity, Colonial Link, Same Country, Common Colonizer, Same Legal Origin, FTA, Both in WTO, Common Currency, GSP and time-varying importing and exporting country …xed e¤ects. Columns (5)-(8) include also country-pair …xed e¤ects. Robust standard errors (clustered at the country-pair level) are in parentheses. p < 0:10,

p < 0:05,

p < 0:01

23

Table 6: Cultural Di¤erence, Di¤erent Blocs, and post-Cold War Triple Interaction (1)

(2)

(3)

(4)

Di¤erent Civilizations Post-Cold War

-0.403 (0.036) Di¤erent Civilizations Di¤erent Blocs Post-Cold War -0.151 (0.185) Di¤erent Religion Post-Cold War -0.092 (0.042) Di¤erent Religion Di¤erent Blocs Post-Cold War -0.246 (0.193) Di¤erent Ethnicity Post-Cold War -0.456 (0.080) Di¤erent Ethnicity Di¤erent Blocs Post-Cold War 1.138 (0.578) Di¤erent Language Post-Cold War -0.502 (0.062) Di¤erent Language Di¤erent Blocs Post-Cold War 1.338 (0.251) Di¤erent Blocs Post-Cold War 1.097 1.235 -0.056 -0.278 (0.152) (0.160) (0.582) (0.260) Additional Controls YES YES YES YES Importer-Year E¤ects YES YES YES YES Exporter-Year E¤ects YES YES YES YES Country-Pair E¤ects YES YES YES YES N 343714 343714 343714 343714 R2 0.868 0.868 0.868 0.868

Regressand: log Imports. Regressors included but with unrecorded coe¢ cients: FTA, Both in WTO, Common Currency, GSP, time-varying importing and exporting country …xed e¤ects, and country-pair …xed e¤ects. Robust standard errors (clustered at the country-pair level) are in parentheses. p < 0:10,

p < 0:05,

p < 0:01

24

Table 7: Principal Component Analysis of Cultural Di¤erence (1) -0.184 (0.012)

Cultural Di¤erence

(2)

(3)

(4)

Cultural Di¤erence Post-Cold War

-0.119 -0.113 -0.112 (0.011) (0.011) (0.011) Di¤erent Blocs Post-Cold War 1.018 1.022 (0.110) (0.109) Cultural Di¤erence Di¤erent Blocs Post-Cold War -0.048 (0.086) Additional Controls YES YES YES YES Importer-Year E¤ects YES YES YES YES Exporter-Year E¤ects YES YES YES YES Country-Pair E¤ects NO YES YES YES N 343714 343714 343714 343714 R2 0.724 0.868 0.868 0.868 Regressand: log Imports. Regressors included but with unrecorded coe¢ cients: ln Distance, Contiguity, Colonial Link, Same Country, Common Colonizer, Same Legal Origin, FTA, Both in WTO, Common Currency, GSP, and time-varying importing and exporting country …xed e¤ects. Columns (2)-(4) include also country-pair …xed e¤ects. Robust standard errors (clustered at the country-pair level) are in parentheses. p < 0:10,

p < 0:05,

p < 0:01

25

Appendix A TABLE 1A. Civilization Membership Country Civilization Andorra, Australia, Austria, Barbados, Belgium, Canada, CroaWestern tia, Czech Rep., Denmark, Dominica, Estonia, Finland, France, French Guiana, Germany, Greenland, Grenada, Hungary, Iceland, Ireland, Israel, Italy, Jamaica, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Monaco, Netherlands, New Zealand, Norway, Papua New Guinea, Philippines, Poland, Portugal, San Marino, Slovakia, Slovenia, Solomon Islands, Spain, Sweden, Switzerland, Trinidad and Tobago, United Kingdom, United States, Vanuatu. Sinic

China, Hong Kong, North Korea, South Korea, Taiwan, Vietnam.

Islamic

Afghanistan, Albania, Algeria, Azerbaijan, Bahrain, Bangladesh, Bosnia and Herzegovina, Brunei, Burkina Faso, Chad, Djibouti, Egypt, Eritrea, Gambia, Guinea, Guinea-Bissau, Indonesia, Iran, Iraq, Jordan, Kyrgyzstan, Kuwait, Lebanon, Libya, Malaysia, Mali, Mauritania, Morocco, Niger, Oman, Pakistan, Qatar, Saudi Arabia, Senegal, Somalia, Sudan, Syria, Tajikistan, Tunisia, Turkey, Turkmenistan, United Arab Emirates, Uzbekistan, Yemen.

Hindu

Guyana, India, Nepal.

Orthodox

Armenia, Belarus, Bulgaria, Cyprus, Georgia, Greece, Kazakhstan, Macedonia, Moldova, Romania, Russia, Serbia, Ukraine.

Latin American

Antigua and Barbuda, Argentina, Bahamas, Belize, Bolivia, Brazil, Chile, Colombia, Costa Rica, Cuba, Dominican Rep., Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Puerto Rico, Saint Lucia, St.Vincent & Grenadines, Uruguay, Venezuela.

African

Angola, Benin, Botswana, Burundi, Cameroon, Cape Verde, Central African Republic, Comoros, Congo, Congo Dem.

Rep.

(Zaire), Equatorial Guinea, Gabon, Ghana, Ivory Coast, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Nigeria, Rwanda, Sao Tome and Principe, Sierra Leone, South Africa, Suriname, Swaziland, Tanzania, Togo, Uganda, Zambia, Zimbabwe. Buddhist

Bhutan, Cambodia, Lao People’s Dem. Rep., Mongolia, Myanmar, Singapore, Sri Lanka, Thailand.

"Lone" States Ethiopia, Haiti, Japan. Source: Author’s own construction based on Huntington (1998).

26

TABLE 2A: Summary Statistics Mean

Std.

Min

Max

Observations

Log Imports

8.55

3.43

-4.60

19.91

343714

Di¤erent Civilizations

0.78

0.41

0

1

343714

Di¤erent Religion

0.54

0.49

0

1

343714

Di¤erent Ethnicity

0.96

0.19

0

1

343714

Di¤erent Language

0.93

0.25

0

1

343714

Cultural Di¤erence

-0.01

1.43

-5.90

0.91

343714

Di¤erent Blocs

0.04

0.20

0

1

343714

Log Distance

8.62

0.83

4.65

9.89

343714

Contiguity

0.06

0.23

0

1

343714

Colonial Link

0.03

0.17

0

1

343714

Same Country

0.01

0.12

0

1

343714

Common Colonizer

0.07

0.26

0

1

343714

Same Legal Origin

0.36

0.48

0

1

343714

FTA

0.04

0.20

0

1

343714

Both in WTO

0.66

0.47

0

1

343714

Common Currency

0.01

0.13

0

1

343714

GSP

0.15

0.35

0

1

343714

27

TABLE 3A. Blocs of Countries Bloc Country 1st World Andorra, Australia, Belgium, Canada, Denmark, France, Germany, Greece, Iceland, Israel, Italy, Japan, Luxembourg, Malta, Monaco, Netherlands, New Zealand, Norway, Philippines, Portugal, San Marino, South Korea, Spain, Taiwan, Thailand, Turkey, United Kingdom, United States. 2nd World

Albania, Armenia, Azerbaijan, Belarus, Bulgaria, China, Cuba, Czech Rep., Estonia, Georgia, Hungary, Kazakhstan, Kyrgyzstan, Lao People’s Dem. Rep., Latvia, Lithuania, Moldova, Mongolia, North Korea, Poland, Romania, Russia, Slovakia, Turkmenistan, Ukraine, Uzbekistan, Vietnam.

3rd World

Afghanistan, Algeria, Angola, Antigua and Barbuda, Argentina, Austria, Bahamas, Bahrain, Bangladesh, Barbados, Belize, Benin, Bhutan, Bolivia, Bosnia and Herzegovina, Botswana, Brazil, Brunei, Burkina Faso, Burundi, Cambodia, Cameroon, Cape Verde, Central African Republic, Chad, Chile, Colombia, Comoros, Congo, Costa Rica, Croatia, Cyprus, Congo Dem. Rep. (Zaire), Djibouti, Dominica, Dominican Rep., Ecuador, Egypt, El Salvador, Equatorial Guinea, Eritrea, Ethiopia, Fiji, Finland, Gabon, Gambia, Ghana, Grenada, Guatemala, Guinea, GuineaBissau, Guyana, Haiti, Honduras, India, Indonesia, Iran, Iraq, Ireland, Ivory Coast, Jamaica, Jordan, Kenya, Kuwait, Lebanon, Lesotho, Liberia, Libya, Liechtenstein, Macedonia, Madagascar, Malawi, Malaysia, Mali, Mauritania, Mauritius, Mexico, Morocco, Mozambique, Myanmar, Namibia, Nepal, Nicaragua, Niger, Nigeria, Oman, Pakistan, Panama, Papua New Guinea, Paraguay, Peru, Puerto Rico, Qatar, Rwanda, Saint Lucia, St.Vincent and Grenadines, Sao Tome and Principe, Saudi Arabia, Senegal, Serbia, Sierra Leone, Singapore, Slovenia, Solomon Islands, Somalia, South Africa, Sri Lanka, Sudan, Suriname, Swaziland, Sweden, Switzerland, Syria, Tajikistan, Tanzania, Togo, Trinidad and Tobago, Tunisia, Uganda, United Arab Emirates, Uruguay, Vanuatu,

Venezuela, Yemen, Zambia, Zimbabwe. Source: Author’s own construction based on Huntington (1998).

28

References [1] Anderson, J. E. and van Wincoop, E. 2003. "Gravity with Gravitas: A Solution to the Border Puzzle." The American Economic Review, 93 (1), 170–192. [2] Anderson, J. E. and van Wincoop, E. 2004. "Trade Costs." Journal of Economic Literature, Vol. XLII, 691-751. [3] Baldwin, R. and Taglioni, D. 2007. "Trade E¤ects of the Euro: A Comparison of Estimators." Journal of Economic Integration, 22 (4), 780-818. [4] Berger, D., Easterly, W., Nunn, N. and Satyanath, S. 2013. "Commercial Imperialism? Political In‡uence and Trade During the Cold War." The American Economic Review, Vol. 103(2), 863-896. [5] Bliss, H. and Russett, B. 1998. "Democratic Trading Partners: The Liberal Connection, 1962-1989." The Journal of Politics, 60(4), 1126-1147. [6] Blomberg, S. B. and Hess, G. D. 2006. "How Much Does Violence Tax Trade?" The Review of Economics and Statistics, 88(4), 599-612. [7] Burbidge, J. B., Magee L. and Robb, A. L. 1988. "Alternative Transformations to Handle Extreme Values of the Dependent Variable." Journal of the American Statistical Association, 83, 401, 123-127. [8] Carrère, de Melo and Wilson. 2009. “The Distance E¤ect and the Regionalization of the Trade of Developing Countries.” CEPR DP 7458. [9] Chiozza, G. 2002. "Is There a Clash of Civilizations? Evidence from Patterns of International Con‡ict Involvement, 1946–97." Journal of Peace Research 39: 711–34. [10] Desmet, K., Le Breton, M., Ortuno-Ortin, I. and Weber, S. 2006. "Nation Formation and Genetic Diversity." CORE Discussion Papers 2006095, Université Catholique de Louvain, Center for Operations Research and Econometrics (CORE). [11] Djankov, S. and Freund, C. 2002. "Trade Flows in the Former Soviet Union, 1987 to 1996." Journal of Comparative Economics, Vol. 30(1), 76-90. [12] Disdier, A. C. and Head, K. 2008.“The Puzzling Persistence of the Distance E¤ect on Bilateral Trade.” The Review of Economics and Statistics, 90 (1), 37-48. [13] Ellingsen, T. 2000. "Colorful Community or Ethnic Witches’Brew? Multiethnicity and Domestic Con‡ict during and after the Cold War." Journal of Con‡ict Resolution, 44(2), 228-249. [14] Feenstra, R. C. 2002. "Border E¤ects and the Gravity Equation: Consistent Methods for Estimation." Scottish Journal of Political Economy, Vol. 49, No.5, 491-506. [15] Felbermayr, G. J. and Toubal, F. 2010. "Cultural Proximity and Trade." European Economic Review, Vol. 54, 279-293.

29

[16] Gartzke, E. 2010. "The A¢ nity of Nations: Similarity of State Voting Positions in the UNGA." http://dss.ucsd.edu/~egartzke/htmlpages/data.html. [17] Gartzke, E. and Gleditsch, K. S. 2006. "Identity and Con‡ict: Ties that Bind and Differences that Divide." European Journal of International Relations, 12(1): 53-87. [18] Gibler, D. M. 2009. "International Military Alliances from 1648 to 2008." Washington: Congressional Quarterly Press. [19] Giuliano, P., Spilimbergo, A. and Tonon, G. 2006. "Genetic, Cultural and Geographical Distances." IZA Discussion Paper No. 2229. [20] Glick, R. and Taylor, A. 2005. "Collateral Damage: Trade Disruption and the Economic Impact of War." The Review of Economics and Statistics, 92(1), 102-127. [21] Guimarães, P. and Portugal, P. 2010. "A Simple Feasible Alternative Procedure to Estimate Models with High-Dimensional Fixed E¤ects." Stata Journal, 10(4), 628-649. [22] Guiso, L., Sapienza, P. and Zingales, L. 2009. "Cultural Biases in Economic Exchange." The Quarterly Journal of Economics, 124 (3): 1095-1131. [23] Head, K. and Mayer, T. 2002. "Illusory Border E¤ects: Distance Mismeasurement In‡ates Estimates of Home Bias in Trade." CEPII, Working Paper No 2002-01. [24] Head, K. and Mayer, T. 2013. "Gravity equations: Toolkit, cookbook, workhorse." Handbook of international economics, 4: 131-195. [25] Helpman, E., Melitz, M. and Rubinstein, Y. 2008. "Estimating Trade Flows: Trading Partners and Trading Volumes." The Quarterly Journal of Economics, 123, 441–87. [26] Henderson, E., A. and Tucker, R., M. 2001. "Clear and Present Strangers: The Clash of Civilizations and International Con‡ict." International Studies Quarterly 45: 317–38. [27] Huntington, S. P. 1993a. "The Clash of Civilizations?" Foreign A¤airs 72(3): 22-49. [28] Huntington, S. P. 1993b. "If Not Civilizations, What? Paradigms of the Post-Cold War World." Foreign A¤airs 72(5): 186–94. [29] Huntington, S. P. 1998. "The Clash of Civilizations and the Remaking of the World Order." Simon & Schuster Ltd. West Garden Place Kendal Street London W2 2AQ. [30] Kristjánsdóttir, H. 2012. "Exports From a Remote Developed Region: Analysed by an Inverse Hyperbolic Sine Transformation of the Gravity Model." The World Economy, 35, 7, 953-966. [31] La Porta, R., Lopez-de-Silanes, F., Shleifer, A. and Vishny, R. 1999. "The Quality of Government." Journal of Law, Economics and Organization 15(1), 222-279. [32] Long, A. 2003. "Defense Pacts and International Trade." Journal of Peace Research, 40, 5, 537-552.

30

[33] Martin, P., Mayer, T. and Thoenig, M. 2008. "Make Trade Not War?" Review of Economic Studies, 75, 865-900. [34] Martínez-Zarzoso, I. 2013. "The Log of Gravity Revisited." Applied Economics, 45, 311– 327. [35] Melitz, J. 2008. "Language and Foreign Trade." European Economic Review, 52, 667-699. [36] Melitz, J. and Toubal, F. 2014. "Native language, spoken language, translation and trade." Journal of International Economics 93 (2): 351-363. [37] Morrow, J. D., Siverson, R. M. and Tabares, T. E. 1998. "The Political Determinants of International Trade: The Major Powers, 1907-90." The American Political Science Review, 92(3), 649-661. [38] Rauch, J. and Trindade, V. 2002. "Ethnic Chinese Networks in International Trade." Review of Economics and Statistics, 84, 116–130. [39] Rose, A. K. and van Wincoop, E. 2001. "National Money as a Barrier to International Trade: The Real Case for Currency Union." The American Economic Review, Vol. 91, No. 2, 386-390. [40] Russett, B., M., Oneal, J., R. and Cox, M. 2000. "Clash of Civilizations, or Realism and Liberalism D‘ej´a Vu?" Journal of Peace Research, 37(5): 583–608. [41] Spolaore, E. and Wacziarg, R. 2009a. "The Di¤usion of Development." The Quarterly Journal of Economics, 124 (2): 469-529. [42] Spolaore, E. and Wacziarg, R. 2009b. "War and Relatedness." NBER Working Paper Series, Working Paper 15095. [43] Stinnett, D. M., Tir, J., Schafer, P., Diehl, P. F. and Gochman, C. 2002. "The Correlates of War Project Direct Contiguity Data, Version 3." Con‡ict Management and Peace Science 19 (2):58-66. [44] Umana Dajud, C. 2013. "Political Proximity and International Trade." Economics and Politics, Vol 25(3), 283-312. [45] Yu, M. 2010. "Trade, Democracy and the Gravity Equation." Journal of Development Economics, 91(2), 289-300. New Economic School, Moscow

31