Information and Communication Technology (Evidence from Turkey)

INFORMATION AND COMMUNICATION TECHNOLOGY (ICT) AND INTERNATIONAL TRADE: EVIDENCE FROM TURKEY Abstract This study analyzes the impacts of information and communication technology (ICT) on international trade between Turkey and its trading partners. Based on an extended panel gravity model, the effects of four ICT indices on Turkish bilateral exports and imports are examined with static and dynamic panel data models for the period 2000-2014. The sample includes 35 countries that import Turkish goods and 34 countries that export goods to Turkey. The results indicate that ICT has positive and significant impacts on both Turkish import and export volumes. Additionally, ICT has a larger effect on imports than on exports. Among ICT indices, ICT access has the largest effect on exports while ICT skills have the strongest impact on imports. In contrast, ICT use has the least impact on both Turkish exports and imports. These results are robust to alternative specifications and estimation methods. Based on these results, some policy implications can be derived. For instance, Turkey may develop strategic trading partnerships with countries that have high levels of ICT endowments, in order to increase its overall trade. 1. INTRODUCTION Information and communications technology (ICT) in general, and the Internet, in particular, has contributed significantly to the globalization of the world economy by allowing the crossborder flow of ideas, knowledge, expertise, and innovations (Choi, 2010). In particular, the post2000 period has witnessed remarkable changes such as enhanced economic activity and accelerated productivity growth as a result of the revolution in ICT. For instance, the significant increases in world trade in services have coincided with the unprecedented advances in ICT (Liu and Nath, 2016). The technological progress in telecommunications and associated declines in communication costs have frequently been cited as the leading causes of growing world trade in the last quarter of the 20th century (Fink et al. 2005). Over time, concerning the relationship between ICT and trade, the advancements in ICT sector has brightened the discussions in the popular press of the “death of distance” (see Caroline and Weinhold, 2000; Demirkan et al., 2009). Thanks to the advancements in ICT, physical distance as an impediment to trade has less importance. The proximity requirement for face-toface interaction between business partners is anymore not a necessary condition because innovations in ICT such as telephone, email, and virtual conferences have become substitutes for face-to-face interactions (Dettmer, 2014). The developments in ICT sector have eroded the impediments to mobility and inaccessibility to information, and thus entrepreneurs and small businesses have now better access to international markets (Vemuri and Siddiqi, 2009). In particular, poor and developing countries have considerable geographic distances, and cultural and political barriers to their trading partners (Demirkan et al., 2009). In this respect, ICT could play a significant role in increasing trade ties by compensating for the lack of strong historical trade linkages (Freund and Weinhold, 2000). Thus, countries that invest more in their ICT 1 infrastructures can overcome these barriers and increase their bilateral trade volumes with their trading partners. Several plausible mechanisms through which ICT may affect the flows of international trade could be suggested. These mechanisms essentially decrease transaction costs associated with trade. These include the costs of search for information, bargaining, policy and enforcement, and coordination costs. These are the real costs of trade which are the important determinants of a country's ability to participate fully in the world economy (Limao and Venables, 2001). By improving information flows and lowering transaction costs, ICT makes markets more competitive and efficient (Ahmad et al., 201; Jungmittag and Welfens, 2009; Park and Koo, 2005). Thus, it has trade-creation (trade-enhancing) effects. ICT reduces transportation costs significantly and thereby provides a channel of exchange that overcomes historical trading bottlenecks for many services (Freund and Weinhold, 2002). Moreover, ICT, through organized exchanges with several buyers and sellers, and through powerful search engines providing for sellers and buyers to find each other at a low cost, has the potential to lower market-specific fixed entry cost (Freund and Weinhold, 2004; Lin, 2015). By reducing the fixed entry costs, such as the costs of search cost, advertising, and establishing a distribution network into a market, the use of ICT stimulates exports (Liu and Nath, 2013). Therefore, it creates a large global market for specific traded goods. Regarding communication costs, telecommunication creates an avenue to maintain fast and efficient communication with trade partners to sustain business competitiveness (Bankole et al. 2015). Sellers and buyers can easily access to international markets with reduced international communication costs. Besides, cheaper and faster communication may boost market transactions and also enlarge the radius of international trade (Jungmittag and Welfens, 2009). In case of 2 information costs, they have turned to be invaluable because the Internet has provided a new medium to transmit information related to commercial transactions (Vemuri and Siddiqi, 2009). The delays in acquiring and transmitting information are reduced and planning is now more efficient and accurate due to advances in ICT (Liu and Nath, 2013, 2016). In this respect, ICT is a low-cost channel for information gathering, processing, and dissemination. It also leads to welfare enhancement by reducing information asymmetries because all members of any given exchange share the same information (Ahmad et al., 2011; Freund and Weinhold, 2000). In this study, we aim to analyze the impact of ICT on Turkey’s bilateral exports and imports for the period 2000-2014 based on an augmented panel data gravity model. For that aim, we select the top 35 importing and 34 exporting partners of Turkey. As proxies for ICT, we compute and use four ICT indices. The contributions of the study to the literature are as follows. This is the first panel study for Turkey that analyzes the issue in a gravity model framework. It also sets an example and provides policy implications for developing countries at the same development level as Turkey. Second, there are only a few studies (Liu and Nath, 2016; Mattes et al., 2012) that consider all aspects of ICT (by computing and using ICT indices) while examining its impact on trade. The rest of the paper is organized as follows. Section 2 describes Turkey’s international trade structure in the 2000s. Section 3 briefly reviews the related empirical literature while Section 4 explains the model and data. The results from the empirical analysis are presented in Section 5. Finally, Section 6 concludes this study with a summary of the research findings and some important policy implications. 3 2. LITERATURE REVIEW With the seminal studies of Freund and Weinhold (2002, 2004), many studies have analyzed the impact of ICT on international trade. Freund and Weinhold (2002) estimate both a crosssection model and a panel data gravity model for 1995 and 1999 years to analyze the impact of the Internet on the 14 service industries’ trade between U.S. and 31 countries. The results indicate that the effect of the Internet on services trade growth is positive and significant. In another study, Freund and Weinhold (2004) examine the impact of the Internet on bilateral merchandise trade among 56 countries from 1995 to 1999 by first using a theoretical model and then employing both a cross-section and a panel data gravity model. They find that the Internet stimulates trade. Over time, the number of cross-section studies has increased. For instance, using the framework of cross-sectional data analysis, Clarke and Wallsten (2006) analyzed the impact of the Internet usage on the export of a group of high-income and developing countries for 2001. Their results indicate that the Internet has a positive and significant impact on the export from developing countries. In another study, Clarke (2008) assesses whether Internet access affects the export performance using pooled cross-section data from enterprises in low and middle-income economies in Eastern Europe and Central Asia for 1999, 2002, and 2005 and he finds a strong positive impact of the Internet on export. Demirkan et al. (2009) examine the impact of the Internet use on bilateral trade flows among 175 countries for 2005 in the framework of a crosssectional gravity model. They find that greater bilateral trade flows occur between countries with higher Internet use. Using a cross-sectional gravity model, Kurihara and Fukushima (2013) show that the Internet promotes international trade both in 34 developed and 24 developing Asian countries for 2005 and 2010. Yushkova (2014) examines the role of ICT deployment by the 4 business sector in international trade between 40 countries for 2011 based on a cross-sectional gravity model. The results indicate that ICT implementation by business sector has a positive impact on the export flows between the countries. The recent studies in the literature have used panel data models. For instance, Vemuri and Siddiqi (2009), using a panel data gravity model, analyze the impact of the Internet and ICT infrastructure on bilateral trade among 64 countries for the years from 1985 to 2005 and obtain a positive and a significant impact of ICT infrastructure on international trade. Using a panel gravity model, Ahmad et al. (2011) utilize different indicators as proxies for ICT infrastructure and examine the impact of ICT on bilateral trade between Malaysia and its 36 trading partners from 1980 to 2008. The results favor the positive and significant impacts of ICT on bilateral trade. Mattes et al. (2012) analyze the impact of ICT development index on trade within the European Union (EU) and between the EU and its main trading partners using a panel gravity model for the period 1995 to 2007. The results indicate that ICT has a significant impact on inter- and extra-European trade. A similar ICT development index is also used by Liu and Nath (2016), applying a dynamic panel data model to examine the impacts of ICT on exports and imports of 10 service categories for 49 countries from 2000 to 2013. The results suggest that ICT development affects only 25% of all services trade positively and significantly. In another study, Liu and Nath (2013) estimate the effect of ICT on both exports and imports in 40 emerging markets from 1995 to 2010. Their results show that Internet subscriptions and Internet hosts have significant positive effects on both exports and imports. There are some panel studies (see Choi, 2010; Lin, 2015; Timmis, 2012; Thiemann et al., 2012) that focus only on the impact of the Internet on trade. The panel study by Lin (2015) estimates the impact of the Internet users on 200 countries’ bilateral trade from 1990 to 2006 in a 5 panel gravity model. The Internet use has a positive and statistically significant effect on international trade, and its effect on export is stronger than that on import. Choi (2010) also utilizes the Internet users per 100 people to investigate the impact of ICT on service trade for 151 countries from 1990 to 2006 and finds that an increase in the Internet users promotes total service trade, as well as service export and service import. Timmis (2012) employs a panel gravity model to assess the role of the Internet adoption on trade within 34 OECD countries over the period 1990-2010. However, he finds that the Internet has a less clear-cut impact on international trade. Based on a panel gravity model, Thiemann et al. (2012) also conclude that none of the ICT variables in exporting and importing countries has a high positive impact on vegetable and fruit trade between major exporting and importing countries for the period 1995 to 2009. In the related literature, there are also studies that use different methodologies (see Bankole et al., 2015; Dettmer, 2014). For instance, Dettmer (2014) applies a Tobit model to show that the effect of time zones on bilateral business services trade among 27 OECD countries depends on the access to ICT over the period of 1999 to 2006. They find that the effect of time zones on service trade is positive when ICT infrastructure is taken into consideration. In a recent study, Bankole et al. (2015) use structural equation models to investigate how ICT infrastructure impacts trade among 27 African countries between 1998 and 2007. The findings indicate that telecommunication infrastructure has a positive and statistically significant impact on intraAfrican trade. Furthermore, a few studies (see Fink et al., 2005; Jungmittag and Welfens, 2009; Tang, 2006;) examine the impact of ICT on international trade in an indirect way by searching for the effect of communication costs on bilateral trade. Fink et al. (2005) examine the impact of 6 communication costs (per-minute two-way calling price) on bilateral trade between 107 countries for 1999 and find that international variations in communication costs have a significant and negative influence on trade patterns. By doing so, they estimate the impact of the telecommunication sector on bilateral trade. Tang (2006) applies a panel gravity model including bilateral US import data from 103 countries from 1975 to 2000 to estimate how the decline in communication costs affects the pattern of trade in differentiated goods. The results suggest that development of IT has significantly increased US imports of differentiated goods. By using bilateral telephone minutes, Jungmittag and Welfens (2009) analyze the impact of the liberalization of telecommunication on international trade and show that international telecommunication has a positive and significant impact on trade for 12 OECD countries between 1995 and 1997. Some studies in the literature analyze the impact of institutions and infrastructure on international trade (see Francois and Manchin, 2013; Limao and Venables, 2001; Nardas and Piermartini, 2004; Portugal-Perez and Wilson, 2012). They use ICT infrastructure as one of the aspects of infrastructure quality. For instance, Francois and Manchin (2013), in a panel gravity model including both developing and developed countries, find that export depends on institutional quality as well as communications infrastructure. In a similar vein, by using a panel gravity model, Portugal-Perez and Wilson (2012) estimate the impact of indicators of “soft” and “hard” infrastructure on the export performance of 101 developing countries over the period 2004–2007. They find that ICT as a proxy for the hard dimension of infrastructure seems to be necessary for export for richer countries. Nardas and Piermartini (2004) analyze the role of the quality of infrastructure on a country's trade by estimating a gravity model for 138 countries for 2000. Their results show that access to telecommunication plays a crucial role in determining 7 bilateral trade in the automotive sector. Limao and Venables (2001) use main telephone lines per person as one of the items of their infrastructure index for 103 countries for 1990 and find that if a country improves its infrastructure, its trade cost will fall, and trade volume will rise. 3. TURKEY’S INTERNATIONAL TRADE STRUCTURE IN THE 2000S Turkish economic structure was in disarray in the first half of the 2000s. Due to the earthquake in 1999 and the domestic financial crises in 2000-01, the end of the 20th century and the early 2000s were drastically destructive for Turkey (Babacan, 2010). However, through the enforcement of the “Transition to the Strong Economy Program” in May 2001, many structural reforms such as market liberalization, privatization, banking system strengthening, fiscal discipline, tight monetary policy, inflation targeting, and a floating exchange rate regime were undertaken to alleviate the impacts of these severe crises and to recover the Turkish economy. Thereafter, a robust banking sector and a strong fiscal position allowed the Turkish Government to provide fiscal stimulus to restore economic growth. Furthermore, Turkey witnessed a new era through the victory of AK Party after 2002. Since then, there have been new alignments in both Turkey’s foreign policy and its trade orientation. Re-defining Turkey’s position as central rather than peripheral or acting as a bridge brought Turkey forward to a new era (Babacan, 2010). Since 2002, Turkey’s imports and exports of goods and services have followed upward trends except for the world financial crisis period (2008-09) (see Figure 1). 8 2.5E+11 2E+11 1.5E+11 1E+11 5E+10 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Figure 1: Imports and Exports of Goods and Services, constant 2010 US$ exports imports Source: World Bank Development Indicators, 2016. With respect to the shares of trading partner countries in Turkey’s trade, the top 20 importers and 20 exporters of Turkey account for 68.5% and 72.9% of Turkey’s exports and imports, respectively. Among them, Germany with a value of 9.3 percent and China with a value of 12 percent rank first in Turkey’s exports and imports, respectively (see Figures 2 and 3). Intermediate goods (processed materials incidental to industry) and capital goods (except transportations vehicles) are the largest import items into Turkey whereas intermediate products (processed materials incidental to industry) and consumption products (durable, semi-durable and non-durable consumption goods) are the most important export items from Turkey. 9 Germany England Iraq Italy USA France Sweden Spain UAE Iran Russia Saudi Arabia Netherlands Egypt Romain Israel Belgium China Polond Azarbaijan Figure 2: Top 20 Countries in Exports, 2015 1.7 1.9 2.0 1.8 1.6 1.3 9.3 2.2 2.4 2.2 7.3 2.5 5.9 2.5 4.8 3.3 3.9 4.1 3.3 4.4 Source: TURKSTAT, 2016 China Germany Russia USA Italy France South Korea Iran India Spain England Ukraine Belgium Japan Polond Netherlands Romain Sweden Bulgaria Czech Republic Figure 3: Top 20 Countries in Imports, 2015 1.3 1.4 1.4 1.5 1.1 1.2 1.1 1.5 12 1.7 2.7 2.7 2.7 2.9 3.4 3.7 10.3 9.8 5.1 5.4 Source: TURKSTAT, 2016 As regards country groups, EU countries have the largest share in both Turkish exports and imports. For instance, EU alone accounts for 44.5% and 38% of the Turkish exports and imports in 2015, respectively (see Figs. 4 and 5). 10 EU-28 Figure 4: Exports by country group and year, share in total (%) in 2015 Europe (except EU) Free Trade Zones 0.4 North Africa Other Africa 0.1 7.2 North America Central and South America and Caribbean Near and Middle East 21.6 44.5 Other Asia Australia and New Zeland Other countries and regions 1.5 4.9 5.9 2.7 9.8 1.3 Source: TURKSTAT, 2016 Figure 5: Imports by country groups and year, share in total (%) in 2015 EU-28 0.3 Europe (except EU) 4.7 Free Trade Zones North Africa 25.7 38.0 Other Africa North America Central and South America and Caribbean Near and Middle East Other Asia Australia and New Zeland 6.6 2.3 5.8 1.0 13.6 0.6 1.5 Other countries and regions Source: TURKSTAT, 2016 11 However, these shares were even larger before the last global financial crisis. For instance, EU made up 56.3% of Turkish exports and 42.5% of Turkish imports in 2006. Turkey has changed its trade policy substantially due to the recession in Europe after the global financial crisis of 2008. It has started a search for alternative markets. As Civan et al. (2013) state, Turkey has improved its relations with neighboring countries and economies in Africa even though its relationship with the West is still significant. In particular, the respective shares of Asian, Near and Middle East countries, members of Black Sea Economic Cooperation, and Organization of the Islamic Conference have increased substantially in the post-crisis periods. The share of Near and Middle East countries in Turkey’s exports rose from 13.2 % in 2006 to 21.6 % in 2015. The other Asian countries’ share in Turkey’s imports rose from 18.4 % in 2006 to 25.7 % in 2015. 4. MODEL AND DATA 4.1 Model This study utilizes an augmented version of the gravity model as a standard analytical tool for the prediction of bilateral trade flows. In the applied international trade literature, gravity model is accepted as the workhorse and as much as 75% of the variability in the levels of international trade could be explained by the gravity model (Frankel 1997, Shepherd, 2012). The gravity model is used in the analysis of trade-related issues, e.g. to determine the trade potential of a country or to evaluate the effect of specific policy variables on trade (Mattes et al., 2012). As such, it has been widely applied to identify the trade flow impacts of institutions such as customs unions, exchange rate mechanisms, ethnic ties, linguistic identity and international borders (Anderson and Wincoop, 2003). Tinbergen (1962) was the first to propose a gravity equation for bilateral trade as an empirical specification, of course taking clue from Newton’s universal law of gravitation. The 12 extraordinary stability of the gravity equation prompted the search for a theoretical explanation for the gravity model (WTO, 2012) and Anderson (1979) laid out its theoretical foundation. In the universal law of gravitation, the attractive force between two objects is proportional to the product of their masses and inversely proportional to the square of the distance between the objects (Vemuri and Siddiqi, 2009). Over time, scholars in international trade, development economics, and political economy have modified the Newton’s law to estimate the impacts of distance, GDP, and population on people, commodities trade, and the exchange of ideas. (Demirkan et al. 2009). In the gravity model of international trade, bilateral trade flows between countries are positively related to the market sizes (economic masses) of the exporting and importing economies and are negatively linked to the distance between these countries. The orginal gravity equation of international trade is specified as follows: Yij = G ( M i M j) Dij (1) In Eq. (1), Yij is the volume of trade between countries i and j ; M i and M j are the sizes of economies (economic masses); Dij is the physical distance between countries and captures trade costs; and G is a constant. GDP is used as a proxy for the size of the economy and expected to have positive coefficient. In this regard, the larger the size of the economy, the higher will be the volume of trade between country pairs (Ahmad et al. 2011; Biswas and Kendy 2016). The distance between countries is expected to have a negative impact on bilateral trade because of lower transportation costs. On its functional form, a log-linear specification of gravity model fits the data fairly well (Frankel and Romer 1999). The original version of the gravity model in Tinbergen’s (1962) study is defined in a log-log form so that the parameters are the elasticity of 13 the trade flows with respect to the explanatory variables. The logarithmic version of Eq. (1) is specified as follows: ln Yij = log G + α 1 ln(M i ) + α 2 ln(M j ) − α 3 ln( Dij ) + ε ij (2) Later on, Eq. (2) has been extended to include additional variables that can potentially explain the remaining variability in international trade. These variables are included to capture trade costs such as other transportation costs and information costs (see Nordas and Piermartini, 2004). In general, empirical studies use bilateral distance as a proxy for transportation costs. However, island, landlocked, and common border dummies are also used to reflect shipping costs which are higher for landlocked and island countries and lower for neighboring countries. As stated by Limao and Venables (2001), neighboring countries have more integrated transport networks that reduce the number of transshipments, and they are more likely to have transit and customs agreements that reduce transit times and translate into lower shipping and insurance costs. However, island countries are unable to use rail and road transportation in their trade relationships while landlocked countries cannot use ocean freight for the first leg of transportation to the final destination (Nordas and Piermartini, 2004). Additionally, binary dummy variables for a common language and colonial history indicate cultural proximity between countries and capture information costs that go down with sharing a common language or some colonial linkages (WTO, 2012). Based on the explanations above, we specify our model in Eq. (3) as an extended version of the original gravity equation.1 1 The variables in our model are similar to those in the studies by Biswas and Kennedy (2016), Choi (2010), Freund and Weinhold (2002, 2004). However, we could not differentiate the impacts of GDP, population and ICT variables for Turkey and its partners separately because Turkey’s data are same across all trading partners. Instead of doing this, we searched for the impacts of GDP mass, population mass and ICT mass on bilateral trade between Turkey and its partner countries. 2 Variables and their weights used to compute IDI index are reported in Table (1) in Appendix. 14 lnYijt = α 0 + α1 ln(GDPit xGDPjt ) + α 2 ln( Popit xPop jt ) + α 3 ln( ICTit xICT jt ) + α 4 ln Dis tan ceij + α 5 Borderij +α 6 Languageij + α 7 Colonyij + α 8 RTAijt + α 9 Landlocked j + α10 Island j + γ t + ε ijt (3) where i and j indicate Turkey and its trading partners respectively, while t is the period. Yijt denotes the volume of export from/import to Turkey to/from country j in period t . ln(GDP it xGDPjt ) is the GDP mass that measures the real GDP of the Turkey and country j . It is expected to have positive impacts on export and import. ln( POP it xPOPjt ) is the population mass which is used as a proxy for country size and measures the populations of Turkey and country j . The impact of population on export is not clear a priori (see Liu and Nath, 2013). Because a growing population may raise domestic production and export by increasing labor supply; however, at the same time, by creating domestic demand, it may also reduce export. Similarly, its impact on import is also not certain. On one hand, increasing domestic demand may rise import demand. On the other hand, country may decide to produce domestically instead of importing, resulting decreases in import volume. ln( ICTit xICT jt ) refers to mass ICT and is expected to have positive effects on export and import. Dis tan ceij is the weighted distance between Turkey and country j that is computed by using population weights of those countries. Borderij is a binary dummy variable that is unity if Turkey and country j share a common border; Language ij is a binary dummy variable that is unity if a language is spoken by at least nine percent of the population in each of the two countries. Colony ij is a binary dummy variable that takes the vale 1 if Turkey and j have been any colonial relationship; RTAij is a binary dummy variable that is unity if there is a regional trade agreement (RTA) between Turkey and country j in some periods of time; Landlocked j and Island j are binary dummy variables taking the value of unity 15 if country j is a landlocked and an island country, respectively. Additionally, γ t is a time effect and ε ijt is an error term that is assumed to be normally distributed with zero mean. There are two separate models for exports and imports in this study. The export model refers to Turkey’s bilateral export to its trading partners, while the import model is used for the Turkey’s bilateral import from its exporters. Time effects are included into models to capture the effects of global events that may have impacts on trade. Among binary dummy variables, distance, landlocked and island variables are expected to have negative coefficients while language, border, RTA, and colony variables to have positive coefficients. In the sensitivity analyses, we utilize additional control variables. For instance, the difference between GDP per capita of Turkey and its trading partners as a proxy for similarity in lifestyles is included. By doing this, we aim to test the trade theories of Linder (1961) and Heckscher (1919)-Ohlin (H-O, 1933). Linder theory asserts that bilateral trade will only occur between two nations which have the same levels of economic development, whereas H-O theory suggests that bilateral trade happens between two countries that have different levels of economic development. Therefore, the negative sign of the coefficient for the difference in GDP per capita variable provides support for the Linder theory whereas positive sign of this variable favors the H-O theory. Further, real exchange rate (national currency unit of trading partners per Turkish Lira after adjusting for inflations) is included in the models. A higher value of real exchange rate means that Turkish Lira (TL) appreciates while a lower value indicates the deprecation of TL. The coefficient of real exchange rate is expected to be positive in the import model as foreign goods are relatively cheaper than domestic goods and purchasing power of TL is high. In contrast, the real exchange rate is expected to have a negative coefficient in the export model 16 because the appreciation of TL indicates that the domestic goods become relatively more expensive than foreign goods. Tariff rate (a proxy for trade cost) is also used in the sensitivity analyses. The sign of its coefficient is expected to be negative both for import and export models. Due to lower tariff rates, the costs of imported goods go down, and import becomes cheaper. Thereby, as stated by Nath and Liu (2013), a country may increase imports of capital goods that enhance its ability to export. Further, railway is used as a proxy for physical infrastructure quality and higher quality of physical infrastructure is expected to boost trade between countries. Thus, the estimated coefficient for railway is supposed to have a positive sign. Finally, to measure political openness, the index of voice and accountability from Kaufmann et al. (2003) - which is a measure of political and civil rights in the country - is added to the models. The level of political openness of a country signals the democracy level in that country. Democratization will lead to more liberal trade policies and improve institutional quality, which in turn foster product quality and the reputation of a country’s exports (see Krenz, 2016). 4.2 Data Bilateral merchandise import and export data are obtained from the Direction of Trade Statistics (DOTS) database of International Monetary Funds (IMF) and are measured in current US dollars. We use export and import unit value indices (2005=100) from IMF’s International Financial Statistics (IFS) to convert trade data into 2005 constant USD. GDP (constant 2005 US dollars) and population data are obtained from the World Development Indicators (WDI) database of the World Bank. Our binary dummy variables are from CEPII GeoDist Dataset and Gravity Dataset. However, CEPII Gravity Dataset contains data for RTA only for the period 17 1948-2006. Therefore, this information has been supplemented by data from the World Trade Organization (WTO)’s RTA database for the years 2007-2014. ICT variables are obtained from the International Telecommunication Union-World Telecommunication ICT Indicator database (ITU 2015). We constructed four ICT indices: ICT development index (IDI) and its three sub-indices: ICT Access, ICT Use, and ICT Skills. Note that ITU has been publishing data on IDI index, a composite index that combines 11 indicators into one benchmark measure for countries since 2009. Among the sub-indices, ICT access index captures ICT readiness and includes indicators that provide an indication of the available ICT infrastructure and individuals’ access to basic ICTs (fixed- telephone subscriptions per 100 inhabitants, mobile-cellular telephone subscriptions per 100 inhabitants, international internet bandwidth per internet user, percentage of households with a computer and percentage of households with internet access). ICT use index measures ICT intensity and consists of three intensity and usage indicators (individuals using the Internet, fixed broadband subscriptions, and mobile-broadband subscriptions). Finally, ICT skills index reflects capabilities or skills that are necessary for ICTs, and it includes three proxy indicators (adult literacy, gross secondary enrolment, and gross tertiary enrolment). Since ITU’s data on the IDI and the three sub-indices do not cover the period before 2009, we follow ITU’s methodology to construct the composite index and its three sub-indices for the entire sample period we consider here. Linear interpolation is used to fill the missing values of the component variables. However, we have to exclude some variables used by ITU while constructing these indices, primarily due to a lack data for the entire sample of countries and the sample period. For instance, we do not include the percentage of 18 households with a computer in the ICT access index; active mobile-broadband subscriptions per 100 inhabitants in the ICT use index; and adult literacy rate in the ICT skills index.2 For sensitivity analyses, exchange rate data are obtained from IMF’s International Financial Statistics (IFS) database. However, we convert exchange rate data to national currency per TL. Also, consumer price indices of Turkey and the partner countries were taken from IFS to calculate the real values of nominal exchange rate. Weighted average tariff rates data (in percentages) are extracted from the World Bank’s World Integrated Trade Solution (WITS system) database. Railways, goods transported (million ton-km) data are available from the WDI database. The voice and accountability index data are from the Kaufmann et al (2003). World Governance Indicators are obtained from the World Bank’s website. We scale up this index by two so that we can take logarithmic values of this indicator as its value ranges between -1.86 to 1.81 for the countries in our sample. We include all these control variables to verify that ICT indices are not picking up the effects of these variables. The choice of the sample period from 2000 to 2014 is dictated by the availability of ICT data. We select the most traded 40 exporting and 40 importing countries of Turkey based on the data availability. However, we exclude the United Arab Emirates (UAE), Brazil, Vietnam, Kazakhstan, Canada and South Africa from the estimation of the import models; and Turkmenistan, Iraq, Libya, UAE, and Kazakhstan from the export models primarily due to unavailability of ICT data. Therefore, we include only 35 countries that account for 73.7% of Turkey’s exports in the export models and only 34 countries that account for 82.7% of Turkey’s imports in the import models. The list of the countries included in our sample is presented in Appendix Table 2. 2 Variables and their weights used to compute IDI index are reported in Table (1) in Appendix. 19