WPS8450 Policy Research Working Paper 8450 Services Development and Comparative Advantage in Manufacturing Xuepeng Liu Aaditya Mattoo Zhi Wang Shang-Jin Wei Development Economics Development Research Group May 2018 Policy Research Working Paper 8450 Abstract Most manufacturing activities use inputs from the finan- of revealed comparative advantage based on domestic value cial and business services sectors. But these services sectors added in gross exports. The paper shows that the devel- also compete for resources with manufacturing activities, opment of financial and business services enhances the provoking concerns about de-industrialization—inancial revealed comparative advantage of manufacturing sectors services in industrial countries like the United States and that use these services intensively but not that of other man- the United Kingdom, and business services in develop- ufacturing sectors. It also finds that a country can partially ing countries like India and the Philippines. This paper overcome the handicap of an underdeveloped domes- examines the implications of services development for the tic services sector by relying more on imported services export performance of manufacturing sectors. It develops inputs. Thus, lower services trade barriers in developing a methodology to quantify the indirect role of services in countries can help to promote their manufacturing exports. international trade in goods and constructs new measures This paper is a product of the Development Research Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/research. The authors may be contacted at xuepengliu@gmail.com, amattoo@worldbank.org, zhi.wangusa@gmail.com, and shangjin.wei@columbia. edu. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Services Development and Comparative Advantage in Manufacturing * Xuepeng Liu Kennesaw State University Aaditya Mattoo World Bank Zhi Wang University of International Business & Economics Shang-Jin Wei Columbia University [JEL Code]: F1 Keywords: Services, trade, value added, comparative advantage * We thank the participants at the Conference on “National Competitiveness, Scalability of International Value Chains and Location of Production” at the Peterson Institute of International Economics, the CEP-IMF-World Bank-WTO Workshop on “Trade Policy, Inclusiveness and the Rise of the Service Economy,” the Association of International Business (AIB) at Georgia Institute of Technology, Kansas State University, and the Conference of China Development Studies at Shanghai Jiaotong University, for comments and suggestions. Research for this paper has been supported in part by the Multidonor Trust Fund for Trade and Development, and by the Strategic Research Program. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors and not necessarily those of the institutions to which they belong. I. Introduction On the face of it, services play a relatively small role in international trade. Conventional trade statistics show that services trade accounts for only one-fifth of cross-border trade (Loungani et al., 2017). However, a significant part of goods trade includes trade in embodied services. In the United States, for example, more than a quarter of intermediate inputs purchased by manufacturers were from the services sector (USITC, 2013). For certain manufacturing sectors, such as computers and electronic products, this percentage — a measure of “services intensity” — is as high as 47.6 percent. Drawing on the Trade in Value Added Database (TiVA, 1995-2011), Miroudot and Cadestin (2017) show that services inputs account for about 37 percent of the value of manufacturing exports in the sample of countries covered. The development of the domestic services sector, as well as access to imported services inputs, can, therefore, be expected to influence comparative advantage in manufacturing trade. This paper seeks to understand this indirect role of services development drawing upon new data and new techniques. The impact of services development is interesting because it is not straightforward. Since services are used as inputs in the production of manufactured goods, services development can help to increase manufacturing production. But since services and manufacturing compete for resources, the development of the former can be at the expense of the latter. For example, it is evident that the development of the services sector has drawn resources away from manufacturing not just in industrial countries like the United States and the United Kingdom, but also in developing countries like India.1 We focus on two key services sectors: financial services and business services. Both have emerged as skill-intensive, dynamic, internationally traded services. These two services sectors are often regarded as the pillars of modern economies, and their value added shares in GDP have a strong positive correlation with countries’ income levels (see Figure 1). These are also the two sectors which represent the tension we discussed above in the sharpest form. On the one hand, manufacturing performance is critically dependent on the domestic availability of these services. On the other, these are the sectors that often provoke “de-industrialization” concerns – financial services in industrial countries like the United States and United Kingdom, and business services in developing countries like India and the Philippines.                                                              1 See, for example, Kochar et al. (2006). 2      Well-functioning financial sectors are critical in mobilizing resources, stimulating investment, and at the same time helping firms (and households) better manage their risks. As shown in Appendix 3C, the business services sector covers a variety of critical of services activities, ranging from software consulting and data processing to management consultancy, engineering and R&D services. Intensive use of these modern services can help manufacturing firms increase productivity, reduce the cost of doing business, expand their choices within a longer geographic distance, differentiate their products from those of their competitors,2 strengthen their after-sale customer services, etc.3 USITC (2013) shows that business services accounted for nearly half of all services purchased by manufacturing sectors in the U.S. in 2008. Our first hypothesis is that, while the overall effect of services development on the performance of manufacturing sectors can be ambiguous, its effect is more likely to be positive for manufacturing sectors that use services inputs more intensively. Furthermore, we distinguish embodied domestic services inputs from embodied foreign services inputs. When domestic firms have access to foreign services, they may at least partially bypass their own inefficient services provision by relying more on imported services inputs. As our second hypothesis, we expect to see a more positive effect of access to foreign services inputs on manufacturing export performance in countries with lower levels of domestic services development. We develop a methodology to quantify the indirect role of services in international trade in goods using a method developed by Koopman, Wang, and Wei (2014) and Wang, Wei, and Zhu (2013) that generalizes the vertical specialization measures proposed by Hummels, Ishii and Yi (2001). We use revealed comparative advantage (RCA) to measure the competitiveness of manufacturing sectors. Following Koopman, Wang, and Wei (2014) and Wang, Wei, and Zhu (2013), we improve on the traditional Balassa (1965) RCA and construct new measures of RCA based on domestic value added in gross exports by taking into account both domestic production sharing and international production sharing. In our econometric analysis of the impact of services development on RCA of manufacturing sectors, the key explanatory variable is the interaction between a measure of the development of                                                              2 To differentiate a product from others, firms need to invest more in R&D, quality-upgrading, and advertisement. The groups of manufacturing sectors with high embodied financial and business services as listed in Appendix 4 indeed produce more differentiated products than those sectors with low services input intensity. In addition, combining pure manufacturing and after-sale services is also a way to differentiate itself from competitors.   3 See the next section for discussion in the literature on how producer services may affect firms’ productivity. 3      financial (or business) services and the financial (or business) services-intensity of each manufacturing sector. We find that domestic services development has a mixed effect on manufacturing export RCA: in manufacturing sectors with low embodied services, services development reduces manufacturing export RCA; however, in sectors with a high degree of embodied services, services development increases manufacturing RCA. Figure 2 provides a visual illustration of this relationship in the case of financial services. We see a negative association between manufacturing RCA and a measure of financial development for a sector with low embodied financial services, but a positive association for a sector with high embodied financial services. In the second hypothesis, we consider the role of services imports in helping overcome the limitations of domestic services markets. We begin by showing that a country’s access to foreign services markets measured by the share of foreign embodied services is negatively correlated to countries’ services trade barriers. Using the World Bank Services Trade Restriction Indexes (STRI) (Borchert, Gootiiz and Mattoo, 2012), Figure 3 shows a negative relationship between financial services trade barriers and the share of embodied foreign financial services among embodied domestic and foreign financial services for the textile sector of 40 countries in 2000, using data from the World Input-Output Database (WIOD).4 A similar pattern holds for other manufacturing sectors and years. We then find that in countries with lower levels of services development, manufacturing exports benefit more from access to foreign services inputs. Our result suggests that lower services trade barriers may help developing countries to bypass their inefficient domestic services provision and promote their manufacturing exports through inter-sectoral linkages. The rest of the paper is organized as follows. We review the relevant literature in Section II. In Section III, we present our hypotheses and carry out the empirical analysis. We conclude in Section IV. II. Literature review This paper is related to at least two strands in the literature: one is on the estimation of services embodied in traded goods; the other is on the role of services in economic development.                                                              4 See Dietzenbacher et al. (2014) and Timmer et al. (2015) for more information on the construction of the WIOD. 4      Research on services embodied in traded goods, based on the Leontief inverse, can be traced back to Grubel (1988) who examined Canadian exports in 1973 and 1983. He found that, over that decade, Canadian embodied services exports had increased substantially to the point where Canada enjoyed a surplus in embodied services trade but had a deficit in direct trade in services. Urata and Kiyota (2003) examined the embodied services in total gross trade for several major services categories of five Asian economies – China, Malaysia, the Philippines, Singapore, and Thailand − in 1990. They found that embodied services accounted for a large share of total services trade for each country. Francois and Woerz (2008) examined the role of services as inputs in manufacturing sectors. They found a significant and strong positive effect of increased business services openness (i.e. greater levels of imports) on some industries, supporting the notion that offshoring of business services may promote the competitiveness of the most skill and technology intensive industries in the OECD countries. Recently, Francois et al. (2013) demonstrated that the ratio of value added exports to gross exports is significantly higher than one in services sectors, suggesting an important role of services sectors in downstream sectors through forward inter-industrial linkages. Their studies cover many countries and provide some interesting insights. These early studies used single national input-output tables, rather than an international input-output table as in this paper, so they could not break down the inputs according to their origins or consider the mismeasurement in services inputs due to two-way trade in intermediate products. In addition, they can only consider how much a service sector’s value-added is embodied in manufacturing exports regardless whether parts of the exported value-added return back to the exporting country or not. In the current paper, we make use of the newly constructed international input-output tables by the WIOD team to measure more precisely the embodied services and indirect trade through other sectors. With the multi-country input-output table and the information about the origins of inputs, we can study embodied domestic and foreign services and their interaction with domestic services development. Stehrer, Foster, and Vries (2012) and Timmer et al. (2013) also use a similar method and the WIOD data to estimate the shares of services, income and jobs in a country that are directly and indirectly related to the production of manufacturing goods, but their work is primarily descriptive without connecting embodied services to the performance of manufacturing sectors. 5      On the role of services in economic development, Hoekman and Mattoo (2008) review the literature, focusing on channels through which openness to trade in services may increase the productivity of a firm, an industry and an economy as a whole. The existing studies show that access to low-cost and high-quality producer services can promote economic growth. Based on an industry-level analysis of the U.S., Amit and Wei (2009a) find that services offshoring by high- income countries tends to raise their manufacturing sectors’ productivity. While services offshoring has both positive and negative effects on domestic employment, Amiti and Wei (2009b) show that, at least for the case of the United States, it tends to enhance domestic employment on average. Arnold, Javorcik, and Mattoo (2011), using firm-level data from the Czech Republic for the period 1998-2003, find a positive effect of services sector reforms on the productivity of domestic firms in downstream manufacturing. The manufacturing-services linkage is measured using information on the degree to which manufacturing firms rely on intermediate inputs from services industries. Arnold et al. (2012) use a similar methodology to show that services reforms had significant and positive effects on the productivity of manufacturing firms in India. Fernandes and Paunov (2012), using the annual manufacturing survey of Chilean firms, find a positive effect of substantial FDI inflows in producer services sectors on the total factor productivity (TFP) of Chilean manufacturing firms. Their findings also suggest that services FDI fosters innovation activities in manufacturing and offers opportunities for laggard firms to catch up with industry leaders. Debaere et al. (2013) find that greater availability of services increases manufacturing firms’ foreign sourcing of materials, which may in turn enhance manufacturing productivity. Using Swedish firm-level data, Lodefalk (2014) shows that in-house and outsourced services help to increase export intensity measured by the share of merchandise exports in total sales. Finally, a recent paper by Bamieh et al. (2017) shows that more intensive use of producer services appears to be positively associated with resilience to greater import competition. In this paper, we study particularly the roles of financial services and business services in manufacturing production. On financial services, Rajan and Zingales (1998) and a number of follow-up studies find that industries that are particularly dependent on financing grow relatively faster in countries with more developed financial markets.5 Our approach differs from Rajan and                                                              5 Based on a non-parametric estimation, however, Shen (2013) shows that the effect is even stronger for financially underdeveloped countries than financially developed countries due to diminishing returns. 6      Zingales (1998) in two major ways. First, we consider modern business services sectors in addition to financial services. For most countries in our sample, business services as a share of GDP are generally on par with or greater than financial services. Second, even for financial services, we measure the intensity of its use in manufacturing sectors differently from Rajan and Zingales in order to maintain consistency with our measure of business services intensity. In particular, their measure of financial dependence is about the intrinsic needs for externally raised funds relative to total funding needs for long-term investment. In an input-output context, the financial services sector only provides financial services in value added terms, rather than the amount of external finance raised. Financial services may facilitate an investment deal, but it is different from investment. Therefore, their measures and ours reflect two different concepts, and the scatter plot in Figure 4 shows a weak correlation between the two measures.6 Some more recent papers also examine the role of finance in the economy. Ju and Wei (2011) show in a general equilibrium model that, for economies with low-quality institutions, finance is a key driver of the real economy and a source of comparative advantage. Buera et al. (2011) demonstrate in a model that sectors with more financing needs are disproportionately vulnerable to financial frictions. A growing recent literature on credit constraints demonstrates that access to external finance helps to increase firms’ export performance (see, Amiti and Weinstein, 2011, among others). Business services cover a wide range of activities as listed in Appendix 3C. There are many case studies on how a certain type of business services promotes economic performance at the firm, state or national level (see USITC, 2013). However, comprehensive empirical analyses covering most of the major economies at a detailed industry level are rare, probably owing to the lack of detailed services data. In the existing literature, the estimation of embodied services and the recent empirical analyses on their linkage to manufacturing export performance are somewhat disconnected. The former estimates the embodied services but does not examine empirically how services input                                                              6  We compare our embodied financial services measures with the external financial dependence measures used by Rajan and Zingales (1998) for the U.S. and find a very weak correlation between them, using a concordance between ISIC Rev. 1 and the WIOD sectors (constructed by authors). The simple correlation coefficient is actually negative at -0.32 or -0.36, depending on whether we consider only embodied domestic financial services inputs or embodied domestic and foreign financial services inputs. A note of caution is that our sample period (1995-2007) differs from theirs (1970s and 1980s). Although we tried narrowing the gap as much as we can by picking their measure for year 1980 and ours for 1995, the weak correlation can be partially due to the different time coverages.   7      intensity affects the performance of downstream sectors. The latter, on the other hand, uses some proxies of inter-sectoral linkage or the direct inputs in gross output to examine the effects of services reforms on downstream manufacturing sectors without quantifying precisely services input intensity. The current paper connects the two literatures: we measure precisely services input intensity as the ratio of embodied services to manufacturing value-added, considering both direct and indirect input usages; then, we directly quantify the effect of services development on the export performance of manufacturing sectors. In addition, we also consider the interaction between embodied domestic services and embodied foreign services and how they affect manufacturing export performance, depending on countries' domestic services development and services input intensity. Finally, the second hypothesis in this paper considers how access to foreign services markets may help developing countries to bypass their possibly inefficient domestic services provision. By distinguishing domestic from foreign services input, we implicitly assume that they are incomplete substitutes.7 Such a bypass effect is also discussed in a theoretical model by Ju and Wei (2010), which derives the conditions under which financial globalization can serve as a substitute for reforms of the domestic financial system. This is also broadly consistent with the theory of comparative advantage – countries with underdeveloped services sectors benefit from imported services, but our paper shows that these benefits may go beyond services sectors through inter- sectoral linkages. III. Empirical analysis In this section, we test empirically the following two hypotheses. Hypothesis 1: the effect of domestic services development on manufacturing export competitiveness is larger (more positive) for manufacturing sectors that use services as inputs more intensively.                                                              7 The magnitude of the Armington elasticity of substitution between domestic and foreign varieties depends on several factors such as the time windows (long run vs. short run) and the level of product disaggregation. In general, estimates of the elasticity are usually quite small at the macroeconomic level. This is why, for example, Obstfeld and Rogoff (2007) found that rebalancing the U.S. current account would require a 30 percent depreciation of the U.S. dollar. Even at the sector level, the suggested Armington elasticity in Global Trade Analysis Project (GTAP Commodity Model) is less than two for most of the services categories, generally lower than those of manufacturing sectors (Hertel, 1997). The U.S. International Trade Commission (e.g., USITC-128 Sector Model) uses similar estimates for financial and business services sectors (Donnelly et al., 2004). 8      Hypothesis 2: the effect of embodied foreign services inputs on manufacturing export competitiveness is more positive in countries with lower levels of domestic services development, especially for manufacturing sectors with high services input intensity. Although the above two hypotheses seem to be straightforward, the theoretical predictions are actually not certain, as discussed in the introduction section. The development in services can draw resources away from manufacturing sectors and can also enhance the productivity of manufacturing when more productive services are used as inputs. Whether the net effect is positive or negative becomes an empirical question. As for the second hypothesis, the effects of foreign services on domestic manufacturing sectors can also be manifold and conflicting among them. The net effect depends on many factors, such as the development level of domestic services sectors. We expect to see a more beneficial role of imported services inputs in countries with less efficient services sectors. In the following, we will lay out our empirical strategy, explain the measures of the key variables, describe the data, and discuss the regression results. III.1 Empirical strategy In our empirical analysis, we use RCA to measure the export competitiveness of individual manufacturing sectors. We will explain later in this paper how we modify the conventional definition of RCA after stating our specification. To test Hypothesis 1, we estimate the effect of services development (D) on manufacturing export performance (RCA), and analyze how this effect depends on service input intensity as measured by the ratio of embodied domestic services in total final demand to manufacturing value- added (or simply SII; see a later subsection for more details). Our baseline regression specification is: (1) ∗ where subscripts i, s, and t refer to country, manufacturing sector and year respectively; SII may measure a benchmark country’s or each country’s own services input intensity, being averaged over time or time-varying; Z is a vector for other control variables; , , are the country, 9      manufacturing sector and year fixed effects; and is an error term. As a robustness check, we also use time-varying country and sector fixed effects (i.e., Country*Year and Sector*Year).8 Hypothesis 1 suggests a positive . can be negative because a more developed services sector (a higher D) in a country could imply a higher services export RCA which in turn could lead to a lower manufacturing export RCA. Our second hypothesis suggests that the effect of D on RCA depends not only on SII, but also on the access to foreign services markets. To capture the relative importance of foreign services inputs compared to domestic services inputs, we measure access to foreign services markets by the share of embodied foreign services in total embodied (domestic and foreign) services in a manufacturing sector of a country (forsh). 9 To ease the interpretation of the results, we run regressions using the subsample for only the manufacturing sectors with high services input intensity because services development and services inputs are less relevant when a sector uses little services as inputs. We also run the same regressions for all of the other sectors with low SII to show how the results differ. The specification of the regressions is similar to equation (1), except that we replace SII with forsh as follows: (2) ∗ According to Hypothesis 2, coefficient is expected to be positive, while should be negative. III.2 Measures of RCA                                                              8 We do not use Country*Sector fixed effects for two reasons. First, the positions of countries in terms of RCA and key explanatory variables are quite stable during our sample period and there is limited variation in these variables’ over time. For example, the variations of RCA within Country*Sector are less than a quarter of the variations between Country*Sector. Second, interpolation is often used to fill the data between benchmark years for the WIOD, so the within variations for a sector of a country may not be very informative (Timmer 2012). Nevertheless, we include in our regressions several variables that vary across countries and sectors to control for the heterogeneity at Country*Sector level. We also tried including other similar variables such as Sector*GDP/capita interaction term but they are always insignificant, and we choose to exclude them from our preferred specification. 9 For instance, a country with low embodied foreign services does not necessarily mean that this country is not open to foreign markets, especially when it also uses limited domestic services inputs. The low foreign services input intensity of this country is probably just because the technology it adopts requires little services inputs. Therefore, the share of embodied foreign services can capture better a country’s openness or access to foreign services markets. 10      The conventional definition of the RCA measure was first proposed by Balassa (1965). Export RCA of country j’s sector k is defined as the share of exports (X) of sector k in j’s total exports relative to the world average share of the same sector k in world exports as follows: ∑ ∑ / ∑ ∑ , where country i, j = 1, 2, ..G; sector k=1, 2, ..K where G is the total number of countries in the world. The RCA measure has been used extensively in the literature to measure the competitiveness of a country in a particular sector. When the RCA exceeds one, the country is deemed to have a revealed comparative advantage in that sector; when it is below one, the country is deemed to have a revealed comparative disadvantage in that sector. Koopman, Wang, and Wei (2014) and Wang, Wei, and Zhu (2013) point out that the traditional RCA ignores both domestic production sharing and international production sharing. First, it ignores the fact that a country-sector’s value added may be exported indirectly via the country’s exports in other sectors. Second, it ignores the fact that a country-sector’s gross exports partly reflect foreign content. A conceptually correct measure of comparative advantage needs to exclude foreign-originated value added and pure double counted terms in gross exports, and to include indirect exports of a sector’s value added through other sectors of the exporting country. When a country uses imported intermediate goods intensively to produce for its exports, Koopman, Wang and Wei (2014) show that RCA based on gross exports can be misleading. The problem of double counting of certain value added components in the official trade statistics suggests that the traditional computation of RCA could be noisy. The gross export decomposition method suggested by Koopman, Wang and Wei (2014) provides a way to remove the distortion of double counting by focusing on domestic value-added in exports. Following Wang, Wei, and Zhu (2013), we calculate RCA based on domestic valued added (DVA) in gross exports, rather than gross exports, for country i in sector k as follows (i = 1, 2, …, G; k = 1, 2, …, N).  DVAi    G DVAki  RCA   N k i i   N i 1 G  k   DVAk     DVAki   k 1   k 1 i 1  The above new RCA measure is the share of a country-sector’s forward linkage-based measure of domestic value added in exports in the country’s total domestic value added in exports relative to that sector’s total forward linkage-based domestic value added in exports from all countries as a share of global value added in exports. The domestic value added (DVA) in gross 11      exports in the above formula is the sum of value added exports (VAX) and returned domestic value added consumed at home (RDV). Because it describes the characteristics of a country’s production or total domestic factor content in output, it does not depend on where the output is absorbed. By comparison, VAX are produced at home but ultimately absorbed abroad. For those applications in which a production-based RCA is the right measure as in this paper, we should use DVA in exports rather than VAX to compute RCA. In addition, RCA based on gross exports (the dependent variable) can cause an endogeneity problem because the embodied services (an explanatory variable) are part of gross manufacturing exports. In our paper, manufacturing RCA is based on the value added by the factors employed in manufacturing sectors, not including the embodied services in gross exports which are contributed by the factors employed in services sectors, so our approach is free from the above-mentioned endogeneity problem. Intuitively, we focus on how services help factors employed in manufacturing sectors to create value by improving their productivity, reducing costs, or both.10 III.3: Measurement of embodied services and services input intensity (SII) We compute embodied services in manufacturing sectors using a method developed by Koopman, Wang and Wei (2014) and Wang, Wei, and Zhu (2013) that generalizes the vertical specialization measures proposed by Hummels, Ishii and Yi (2001). Assume a world with G countries, in which each country produces goods in N tradable sectors. Goods and services produced in each sector can be consumed directly or used as intermediate inputs, and each country exports both intermediate and final goods to other countries. All gross outputs (X) produced by a country must be used as intermediate goods/services or as final goods/services (F), i.e., Xi   j ( Aij X j  Fij ), G (3) i, j = 1,2, ... G where Xi is the N×1 gross output vector of country i, Fij is the N×1 vector for final goods and services produced in country i and consumed in country j, and Aij is the N×N input-output coefficient matrix, giving intermediate use in j of goods and services produced in i.                                                              10 Miroudot and Cadestin (2017) provide a detailed discussion on how services help manufacturing sectors to create values by facilitating exchange among users and by solving problems and bringing tailored solutions. See also Heuser and Mattoo (2017) for a review of services in global value chains. 12      The G-country, N-sector production and trade system can be written as an inter-country input-output (ICIO) model in block matrix notation as follows.  X1   A11 A12  A1G   X1   F11  F12    F1G  X  A A22  A2G       2    21   X 2    F21  F22    F2G                (4)         X G   AG1 AG 2  AGG   X G   FG1  FG 2    FGG  After rearranging, we have 1   G F1 j   X 1   I  A11  A12   A1G   j 1   B11 B12  B1G   F1   X   A   A2G   G F  B  B2 G      j 1 2 j    21  2 21 I  A22  B22   F2  (5)                               X G    AG1  AG 2  I  AGG    G FGj   BG1 BG 2  BGG   FG   j 1  where I is an NxN identity matrix, and Bij denotes the N×N block Leontief inverse matrix, which is the total requirement matrix that gives the amount of gross outputs in producing country i required for a one-unit increase in final demand in destination country j. Let Vi be the N×1 direct value-added coefficient vector. Each element of Vi gives the ratio of direct domestic value-added to gross output (exports) for country i at the sector level. This is equal to one minus the intermediate input share from all countries (including domestically produced intermediates):  G (6) Vi  (u  j 1 Ajiu) where u is an Nx1 unit vector of 1. Putting all Vi in the diagonal and denoting it with a hat-  symbol ( V i ), we can define a GN×GN matrix of direct domestic value-added coefficients for all countries as,  0  0 V 1   ˆ   0 V2  0 (7) V          0 0  VG  Putting final demand in the diagonals, we can define another GN×GN matrix of all countries’ final demand as 13      Fˆ 0  0 1   ˆ  0  0 F    F (8) 2        0 0 ˆ   F  G Then the decomposition of value-added in final demand can be conducted by the following equation:    B1G   F1 0   V1 0  0   B11 B12   0    B2G   ˆ  0 V  0   B21 B22   0 F2  0 VB F         2                 BGG   0 0   0 0  VG   BG1 BG 2   FG  (9)         V1 B11 F1 V 1 B12 F 2  V1 B1G FG         V2 B2G FG    V2 B21 F1 V2 B22 F2                G G1 1 G G 2 2 V B F V B F  VG BGG FG   where VˆB F is a GN×GN square matrix that gives the estimates of sector and country sources of   value-added in a country's total final demand. Each block matrix Vi Bij F jj is an N×N square matrix, with each element representing the value-added from a source sector of a source country directly or indirectly used by an absorbing sector in a destination country's total final demand (both domestic and foreign). Because we assume that the same technology is used in the production meeting a country’s domestic demand and foreign demand (exports), we use total final demand, which is the sum of domestic final demand and final export demand, to calculate embodied services ratios. Based on equation (9), we create the following measure of domestic services input intensity in manufacturing sectors in country j: (10) / , j = 1, 2, …, G where , an element in equation (10), refers to country j’s domestic services (superscript s) values embodied in country j’s total final demand in the manufacturing sector (superscript m); is the total value-added created by the factors employed in the manufacturing sector of the absorbing country j (or the manufacturing GDP in country j). SII defined in formula (10) is a scalar 14      if s and m refer to a specific services and manufacturing sector respectively, in a country j in a given year. The numerator on the right-hand side of formula (10) refers to the value added contributed directly and indirectly by the factors employed in a services sector, while the denominator measures the value added contributed by factors employed in a manufacturing sector. Therefore, the denominator is not a part of the numerator and the SII measure is not bounded by one, although it is always less than one in the data. It would be bounded by one if we used the gross manufacturing output in the denominator, and the SII of one services sector would likely be negatively correlated to the SII of other goods or services sectors, and so omitted variable bias can be a problem if we do not include all other sectors in our analysis. The strategy we adopt to measure SII as in formula (10) can help us to avoid this problem and keep our specification simple. It is tempting to use a country’s own services input intensity (SII) directly in the regression. But there are a number of issues with such a strategy. SII of a country with underdeveloped services sectors (e.g., financial repression) may not be able to capture the required services input intensity along the manufacturing production possibility frontier. Hence, instead of using countries’ own services input intensities, we use U.S. services input intensity for all the countries under the assumption that the U.S. is among the countries with the least financial and business services transaction costs and frictions. If inter-sectoral linkage is considered as a feature of the production technology, it should be the same across countries in the absence of services under-development. Adopting a similar strategy, Rajan and Zingales (1998) measure industries’ dependence on external funds using only U.S. data for all countries covered by their analysis. Figure 5 shows a scatter plot of the domestic financial services input intensity in manufacturing against the business services input intensity for each of the WIOD countries in 2005. As we expect, U.S. embodied services ratios are among the highest for both financial and business services. Another problem of using countries’ own services input intensity is a potential endogeneity issue because a country’s embodied services and services development can also be affected by its own manufacturing performance. For example, a country like India with comparative disadvantage in manufacturing may choose to specialize in services, which in turn will promote services development and reduce embodied services due to the weakness of the manufacturing sectors. When we use only U.S. embodied services, the feedback or reverse causality to the U.S. embodied services from other countries’ manufacturing export RCA will be less a concern. In addition, we 15      will drop U.S. observations from our regressions to further alleviate the endogeneity problem. Finally, as another justification for using U.S. measures, the U.S. is arguably one of the countries with the most reliable data. We will either use the time-varying U.S. services input intensities or take their averages over years. An advantage of the former measure is that it retains the time variations, while the later measure can smooth temporal fluctuations and hence is less sensitive to outliers. The variations in the U.S. services input intensities over the years are small for most of the WIOD sectors and some of the input-output data in the WIOD are filled in based on interpolation, so we will take the averaged measure as the benchmark and use the time-varying measure only as a robustness check. When average U.S. SIIi is used, this variable will drop out of regressions with sector or time- varying sector fixed effects. When time-varying sector fixed effects are used, SIIit will also be dropped. An important caveat is that even a measure based on U.S. data is still a proxy intending to capture the potential linkage between services and manufacturing sectors. A noisy measure, however, should create a bias against finding a significant effect of services intensity on manufacturing RCA. Should we be able to find a better measure, the effect is likely to be even stronger. In our empirical analysis, we also use the share of foreign embodied services in the total embodied domestic and foreign services as follows (for country j): (11) ∑, /∑ The denominator in equation (11) sums over all source countries i=1, 2, …, G, including j itself, while the numerator leaves out country j’s own (domestic) embodied services. III.4 Measures of domestic services development (D) Our main services development measure (D) is defined as the ratio of services value-added to GDP. Figure 1 shows a clear positive relationship between income level and the shares of financial and business services in GDP. It seems reasonable to use their shares in GDP to measure the level of development of these sectors. Alternatively, we use the average value added per worker to measure domestic services development. It is calculated as total value added divided by total number of employees for 16      financial or business services based on the data from the WIOD and its Socio Economic Accounts. It is commonly used as a measure for labor productivity in services sectors, which should be closely linked to the levels of services development. We also use other measures of services development to check the robustness of our results when data are available. Following the tradition in the literature, as in Rajan and Zingales (1998), we adopt two alternative measures for financial services development using the data from the World Bank Global Financial Development Database (GFDD). GFDD is an extensive data set of financial system characteristics for 203 economies from 1960 to 2010. The first measure is the bank private credit to GDP ratio, which is defined as the share of financial resources provided to the private sector by domestic banks in a country’s GDP, originally from the International Financial Statistics of the IMF.11 The second measure is the share of bank private credit and stock market capitalization in GDP. Stock market capitalization refers to the total value of all listed shares in a stock market based on Standard & Poor's Global Stock Markets Factbook and supplemental S&P data. III.5 Data and Some Stylized Facts on Embodied Services The primary data source for this study is the WIOD (2013 Version) which covers 35 industries for 40 countries over 1995-2007, so our data structure is a panel at the country-sector level over 13 years (see Appendixes 2-3C for lists of WIOD countries and sectors).12 The original 2103 version of the WIOD data covers years 1995-2009, but we drop the data for 2008-2009 to avoid potential complication resulting from the 2008 global financial crisis. We consider all the manufacturing sectors (WIOD sectors 3-16), and focus on two types of modern services in this paper: financial intermediation services (WIOD sector 28) and other business services sector (WIOD sector 30). To illustrate the importance of embodied services and to motivate our empirical analysis, we first show in Appendix 1A some data on the gross exports (X) and value added exports (VAX) of                                                              11  Domestic money banks comprise commercial banks and other financial institutions that accept transferable deposits, such as demand deposits.  12 China and Romania are not covered by the regressions due to missing wage or employment data. Because the U.S. is used as the benchmark country to define services input intensity, it is also dropped from most of the regressions to alleviate potential endogeneity problem as explained in Section III.2.   17      financial services for some WIOD countries over 1995-2007. We further separate VAX into direct value-added exports (dVAX) and indirect value-added exports through all other sectors (indVXP).13 The last row reports the world total for all the WIOD economies. Overall, VAX of services are 53 percent higher than the gross exports, and indirect VAX are 88 percent higher than the direct VAX. Among the 40 WIOD countries/regions excluding ROW, only three of them (Ireland, Luxembourg, and U.K.) have direct VAX higher than indirect VAX. The BRICs (Brazil, the Russian Federation, India, and China), Japan, the Republic of Korea, Lithuania, Turkey, and Taiwan, China, have much higher indirect VAX than direct VAX (especially China, Russia, and Turkey). Financial services in these countries may have reached an intermediate level of development at which they can compete in the domestic market but not yet internationally. It could also be that restrictions on cross-border imports in these countries oblige goods producers to use domestically produced services. For instance, if domestic firms in China have no easy access to foreign financial services due to high trade barriers, they will have to use domestic financial services (e.g., loans from state-owned banks). Appendix 1B for business services, analogous to Appendix 1A, offers a similar pattern. Some emerging economics (e.g., Mexico, Russia, and especially Turkey) and Japan have much higher indirect business services VAX than direct VAX.14 Most of the high-income countries such as the U.S. and the U.K. export large magnitudes of business services both directly and indirectly. By comparison, developing or emerging economies export significantly less business services, with the exception of India. India has developed an internationally competitive business services industry which has large direct VAX but its indirect VAX are small due to the relatively weak manufacturing sectors. Because financial intermediation services and other business services sectors cover many different types of services, as listed in Appendices 3B and 3C, measuring their level of development is not straightforward. In this paper, we measure financial or business services development as the share of financial or business services value-added in a country’s total value- added in all sectors (or GDP). The logic is simple: services sectors, especially modern ones like                                                              13 tvaexp can be bigger than gexp because it includes not only direct exports of a service sector, but also the indirect value added exports of services through other sectors. 14 Japan is well-known for its competitive manufacturing but relatively inefficient services sectors. See, for example, a report at https://www.economist.com/node/3219857. As a result, Japan exports business services mainly indirectly through manufacturing sectors. 18      financial and business services, usually account for larger shares in total value added in countries with more developed services sectors. Keeping in mind that this may not always be the case for other industries such as agricultural and manufacturing industries as suggested by the literature on structural change (see, e.g., Kongsamut, Rebelo and Xie (2001), among others). We control for the levels of countries’ overall development using the GDP per capita data from the Penn World Tables. Other determinants of manufacturing RCA considered in this paper include the following: productivity measured by total factor productivity (TFP), scale economy measured by a manufacturing sector’s employment in logarithms, factor endowment variables including capital-labor ratio (K/L) and skill ratio (SKratio, defined as the share of the wage payment to high skill workers in total wage payment), relative wages in manufacturing sectors defined as a country’s average wage per worker over world average wage per worker. These variables vary across countries and sectors. The data for these variables are obtained from or estimated based on the WIOD Social-Economic Account database (SEA). The total factor productivity (TFP) growth rate for each WIOD manufacturing sector is estimated using the dual approach as in Hsieh (2002). It is calculated as a weighted average of the growth rates of labor prices (w) and capital prices (r), weighted by the share of payment to labor (L) and capital (K). For this method to be valid, no assumptions are needed for the relations of factor prices to social marginal products or about the production function form as long as the total factor payments add up to total output (i.e., Y = r*K + w*L). Finally, we also include a measure for GVC participation. Wang et al. (2017) propose a framework to decompose total production activities to different types, depending on whether they are for pure domestic demand, traditional international trade, simple GVC activities, and complex GVC activities. Then they construct indices of GVC participation to measure the degree of a sectors’ GVC participation – a concept similar to the vertical specialization (VS1) as in Hummels, Ishii, and Yu (2001) but with a few important improvements. We include a measure of forward industrial linkage-based GVC participation to estimate how a country/sector’s engagement in GVC activities strengthens its overall export performance. Table 1 provides the descriptive statistics of these variables and their definitions. III.6 Empirical results 19      In Table 2, we estimate the specification in equation (1). The dependent variable is manufacturing export RCA calculated based on DVA in gross exports. The U.S. domestic services input intensity is averaged over 1995-2007 and treated as time-invariant. The financial (business) services development measure is defined as the ratio of U.S. financial (business) services value- added to U.S. GDP. Because the embodied services measures are based on U.S. data, we drop the observations for the U.S. from the regressions to alleviate the potential endogeneity problem. In the first three columns, we consider financial services (f), business services (b), and the combined financial and business services (fb) respectively. Country fixed effects, year dummies, and manufacturing sector dummies are all included in the first three regressions. Standard errors are always robust to heteroscedasticity and are also clustered by country*sector to address the potential serial correlation in the error terms for a particular country-sector across years. The coefficient of services development is negative and significant in the first regression for financial services, but not significant for business services in the second regression. The coefficient of the key interaction term is always positive and highly significant. The results imply that financial services development reduces manufacturing RCA when embodied financial services are sufficiently low. This is not surprising given the definition of RCA: services development tends to increase a country’s services export RCA and in turn should lead to lower manufacturing export RCA when manufacturing sectors do not benefit much from services development due to low services input intensity. When embodied services are sufficiently high, services development can actually increase manufacturing RCA. These results provide strong support for our first hypothesis. We can calculate easily the cutoff value of SII. Taking regression (1) as an example, the cutoff SIIf is about 0.046 as compared to its average value (0.035) reported in Table 1. The last three columns of Table 2 are analogous to the first three regressions except that we include time-varying country and time-varying sector fixed effects. As a result, services development measures and log(GDP/capita) are dropped from the regressions. The three interaction terms remain positive and highly significant, with similar magnitude as in the first three regressions. The control variables in Table 2 have the expected signs. Manufacturing productivity (TFP), the measure of scale economy (log(emp)), capital–labor ratio (K/L), and GVC participation 20      increase manufacturing RCA.15 Other variables, including log(GDP/capita), relative wage, and the skill ratio, do not have significant effects. In Table 3A, Table 3B and Table 4, we perform various robustness checks. In Table 3A, we use an alternative measure of services development defined as average value added per worker in financial or business services. Our previous results continue to hold well. The interaction term D*SII is always positive and significant at the 1% or 5% level. Their magnitude is much smaller because the average values of the new services development measures are much bigger as shown in Table 1. In Table 3B, we replace the services development measures in Table 2 by another two alternative measures for financial services as discussed in Section III.4. Because such a measure is not available for the corresponding WIOD business services sector, we perform this robustness check only for financial services. Our previous findings hold very well with or without time- varying fixed effects. The estimated cutoff SII (about 0.04) is similar to what we got from Table 2. In Table 4, we examine the sensitivity of our results to alternative measures of services input intensity. The time-varying country and sector fixed effects are used in all of the regressions, so both SII and D variables are dropped. We consider here financial and business services together. In the first regression, we replace the average U.S. SII with time-varying U.S. SII; our main findings remain unchanged, with a slightly smaller coefficient of the interaction term than the corresponding one reported in column (3) of Table 2. Although the services input intensity of the U.S. is arguably the best choice to capture the role of financial and business services in manufacturing sectors, it is still useful to check the robustness of the results when countries’ own SII measures are used. Regression (2) in Table 4 is analogous to those in the first column, except that we replace U.S. SII with each country’s own SII (time-varying). We no longer drop the U.S. observations from this regression. The interaction term remains positive and significant at the 1% level, but the magnitude of the coefficient is much smaller than the one reported in the first column, probably because a country’s own SII may not capture well the potential role of services in manufacturing sectors if services sectors are under-developed as we would expect. In the last                                                              15 Wang et al. (2017) construct indices for shallow, deep and overall GVC participation. We use only the overall measure in our regressions. The results are robust to other measures. 21      column, we use the average SII of the U.K., another developed country with competitive services sectors. The results are similar to those when the U.S. data are used: the magnitude of the D*SII’s coefficient is similar to what is reported in Table 2 (30.3 vs. 27.72).16 Next, we test for the second hypothesis, which states that countries may bypass their own inefficient domestic services sectors by relying on imported foreign services. As defined in equation (11), the share of embodied foreign services in total embodied services (forsh) is used to measure the degree of a country’s access to foreign services markets. Because our story is relevant only to the sectors that use a significant amount of services as inputs, we consider only the first seven manufacturing sectors with high services input intensity as listed in Appendix 4, 17 and expect to see a stronger bypass effect than from sectors with lower services input intensity. We examine how the interaction between foreign services and domestic services development affects manufacturing export RCA based on specification (2) and report the results in Table 5A. As in the previous table, in the first three columns, we include separate country, year, and sector fixed effects; the first two regressions consider financial and business services respectively; and the third regression combines the two types of services. The coefficients of D*forsh are always negative and significant at the 1% or 5% level. This shows that the benefit of foreign services inputs on manufacturing export RCA decreases with the level of domestic services development, suggesting that foreign and domestic services inputs are at least partially substitutable. Together with a positive coefficient of forsh, this also implies that the access to foreign services can help a country to bypass under-developed domestic services provision. In the last three columns of Table 5A, we include time-varying country and sector fixed effects. As a result, we cannot estimate the coefficient of D any longer. The absolute value of the estimated coefficients of the interaction term is even larger, although it is statistically less significant for financial services. If we include some additional sectors with medium levels of services input intensity, the above results are still robust, although a bit weaker as expected. For instance, including also sectors 9 and 7 does not lead to a dramatic change in the results, except that the interaction term turns                                                              16 We use the export RCA as our preferred measure for the dependent variable. This measure has the advantage of being comparable across sectors and countries. Nevertheless, we also try an alternative measure for the dependent variable – domestic value added in manufacturing exports in logarithms. Our previous findings are robust to this alternative measure. 17 The sector rankings are identical if we consider only financial or only business services, or if we consider both embodied domestic and foreign services. 22      slightly less significant in the regression for financial services. We also run similar regressions as in Table 5A for the other seven manufacturing sectors with low financial and business services input intensity as listed in Appendix 4. For these sectors, services development and access to foreign services markets should matter less. The results are reported in Table 5B. As expected, forsh and its interaction with D are mostly insignificant at the 10% level. Although the interaction term is significant at the 10% level for business services in column (2), the magnitude of the coefficient in absolute value is smaller than the corresponding coefficient in Table 5A. These results provide further support to the second hypothesis.18 IV. Concluding remarks In this paper, we examine how the development of domestic services sectors may affect the export performance of downstream manufacturing sectors, taking into account the services input intensities of manufacturing sectors. We focus on two types of modern services, i.e., financial services and business services, whose shares in an economy normally increase with the level of a country’s development. We show that the indirect exports of services are surprisingly high for a number of countries, especially developing or emerging economies, even though most of these countries’ direct exports of services are relatively small. We also find that the manufacturing sectors that use these services intensively as inputs benefit more from domestic services development. These findings suggest that policy makers should take into account the linkages among sectors, not look at them in isolation on a single sector basis as can happen with the “silo” approach to trade negotiations (Hoekman and Jackson, 2013). Industrial countries have been strong in exporting services, both directly and indirectly. For example, according to Appendix 1B, the U.S. is not only the largest direct exporter of business services in the world, but also the largest indirect exporter of business services (actually twice as large), suggesting an important role of business services in U.S. manufacturing activities. However, developing and emerging economies have significantly lagged, with the only the exception of India in direct exports of business services. Services development in these countries not only strengthens                                                              18 Hypothesis 2 suggests a triple interaction between D, SII and forsh. With all their combinations as regressors in the regressions, it would be significantly more difficult to interpret the coefficients and partial effects. Therefore, we chose to run the regressions for subsamples using only one double interaction term. 23      their services sectors but also promotes manufacturing and other goods producing sectors. Countries like China that may be concerned with the sustainability of their manufacturing export success may consider building stronger services sectors as a way to upgrade their manufacturing sectors to an even higher level. According to Appendix 1B, China’s business services exports in value added terms, relative to its exports in gross terms, are less impressive compared to the corresponding figure for financial services shown in Appendix 1A. Both its direct and indirect business services exports are only 8-9 percent of the corresponding numbers of the U.S. Drawing from the firm-level data in ORBIS, Miroudot and Cadestin (2017) show that China is the only country in their sample which has a majority of the manufacturing firms (77 percent in 2013) selling only goods, with little bundling of goods and services such as manufacturing and distribution services, as seen with Apple iPhones/iPads and Apple Stores. To strengthen the manufacturing sector, countries need to focus not just on manufacturing production, but also on services upgrading including but not limited to R&D, marketing, advertising, inventory management, quality control, production scheduling, after-sale technical supports, and follow-up customer services. With significant improvement in transportation and communication technologies and increasing services outsourcing activities, some developing countries such as India have developed competitive services sectors. For example, Indian services RCA calculated based on DVA in gross exports are either greater than one (financial services) or close to one (business services), much bigger than the corresponding numbers for other developing economies such as China. For countries like India, our paper suggests that the manufacturing sectors that use these services intensively tend to have a comparative advantage. However, different from most of the other WIOD countries, Indian gross exports of business services are actually larger than its total value added exports, suggesting relatively little embodied business services in other sectors, as shown by the direct and indirect ratio of Indian business services exports in Appendix 1B. There is plenty of room left for India and similar countries to take advantage of their competitive services sectors during their industrialization process. This illustrates the importance for policy makers and entrepreneurs to understand the implications of inter-sectoral linkages. We also provide evidence for a bypass effect, that is, countries may bypass their inefficient domestic services sectors by relying more on imported services inputs. This suggests that nations 24      with under-developed services may take advantage of globalization in services. Countries that hesitate to liberalize their services sectors in hopes of protecting their inefficient domestic services sectors may hurt the competitiveness of their manufacturing sectors. Although this paper focuses only on the services-manufacturing linkages, many other important research questions could also be studied using a similar methodology. With the inter- country input-output tables, we have complete information on how countries and sectors are inter- linked to each other. We expect to see more and more studies along this line of research. References Ali-Yrkkö, Jyrki, Petri Rouvinen, Timo Seppälä, and Pekka Ylä-Anttila, 2011. “Who Captures Value in Global Supply Chains? Case Nokia N95 Smartphone”, Journal of Industry, Competition and Trade 11(3): 263-278. Amiti, Mary, and Shang-Jin Wei, 2005, “Fear of outsourcing: Is it justified?” Economic Policy 20 (April): 308–48. Amiti, Mary, and Shang-Jin Wei, 2009a, "Service Offshoring and Productivity: Evidence from the United States," The World Economy, Blackwell Publishing, vol. 32(2), pages 203-220, 02. Amiti, Mary, and Shang-Jin Wei, 2009b, "Does Service Offshoring Lead to Job Losses? Evidence from the United States,” Chapter in NBER book: International Trade in Services and Intangibles in the Era of Globalization, edited by Marshall Reinsdorf and Matthew J. Slaughter (p. 227 - 243). Amiti, Mary, and David E. Weinstein, 2011. "Exports and Financial Shocks," The Quarterly Journal of Economics 126(4): 1841-1877. Arnold, Jens M., Beata Javorcik, Molly Lipscomb, Aaditya Mattoo, 2012. “Services Reform and Manufacturing Performance: Evidence from India,” World Bank Policy Research Working Paper 4048. Arnold, Jens M., Beata Javorcik, Aaditya Mattoo, 2011. “Does Services Liberalization Benefit Manufacturing Firms? Evidence from the Czech Republic,” Journal of International Economics 85(1): 136-46 Balassa, Bela. 1965. “Trade Liberalization and ‘Revealed’ Comparative Advantage.” Manchester School of Economic and Social Studies 33: 99-123. 25      Bamieh, Omar, Matteo Fiorini, Bernard M. Hoekman, Adam Jakubik, 2017. “Services Input Intensity and US Manufacturing Employment. Responses to the China Shock.” EUI Working Paper RSCAS 2017/39 Borchert, Ingo, Batshur Gootiiz, and Aaditya Mattoo, 2012. “Guide to the Services Trade Restrictions Database”, World Bank Policy Research Working Paper 6108. Buera, Francisco J., Joseph P. Kaboski, and Yongseok Shin. 2011. "Finance and Development: A Tale of Two Sectors." American Economic Review 101(5): 1964-2002. Debaere, Peter, Holger Görg, and Horst Raff, 2013. “Greasing the wheels of international commerce: how services facilitate firms’ international sourcing.” Canadian Journal of Economics 46(1):78–102.   Dietzenbacher, Eric, Bart Los, Robert Stehrer, Marcel Timmer, and Gaatzen de Vries, 2013. "The Construction of World Input-Output Tables in the WIOD Project", Economic Systems Research 25: 71-98. Donnelly, William A., Kyle Johnson, Marinos E. Tsigas, and David Ingersoll, 2004. “Revised Armington Elasticities of Substitution for the USITC Model and the Concordance for Constructing a Consistent Set for the GTAP Model,” USITC Office of Economics Research Note No. 2004-01-A. Fernandes, Ana M., and Caroline Paunov, 2012. “Foreign Direct Investment in Services and Manufacturing Productivity: Evidence for Chile,” Journal of Development Economics 97(2): 305-21. Francois, Joseph, Mirian Manchin, and Patrick Tomberger, 2013. “Services Linkages and the Value Added Content of Trade,” World Bank Policy Research Working Paper No. 6432. Francois, Joseph, and Julia Woerz, 2008. “Producer Services, Manufacturing Linkages, and Trade,” Journal of Industry, Competition and Trade 8: 199-229. Grubel, Herbert G., 1988. “Direct and Embodied Trade in Services,” in C. H. Lee and N. Seiji (eds), 1988, Trade and Investment in Services in the Asia-Pacific Region, Westview Press, Boulder Co, pp. 53-76. Hertel, Thomas W. (ed.), 1997. Global Trade Analysis: Modeling and Applications, Cambridge, Cambridge University Press.   Heuser, Cecilia, and Aaditya Mattoo, 2017. “Services Trade and Global Value Chains,” Chapter 6 in David Dollar and Zhi Wang (eds.), Measuring and Analyzing the Impact of GVCs on Economic Development, World Bank, Washington, D.C. 26      Hoekman, Bernard, and Selina Jackson, 2013. “Shifting Focus in Trade Agreements – From Market Access to Value-Chain Barriers.” World Bank Blogs. Available online at http://blogs.worldbank.org/trade/shifting-focus-in-trade-agreements-from-market-access-to- value-chain-barriers Hoekman, Bernard, and Aaditya Mattoo, 2008. “Services Trade and Growth,” World Bank Policy Research Working Paper 4461. Hsieh, Chang-Tai, 2002. "What Explains the Industrial Revolution in East Asia? Evidence From the Factor Markets." American Economic Review 92(3): 502-526. Hummels, David, Jun Ishii, and Kei-Mu Yi, 2001. “The Nature and Growth of Vertical Specialization in World Trade.” Journal of International Economics 54:75–96. Jensen, J. Bradford, 2011. Global Trade in Services: Fear, Facts, and Offshoring, Washington, DC, Peterson Institute for International Economics Press. Ju, Jiandong, and Shang-Jin Wei, 2010. “Domestic Institutions and the Bypass Effect of Financial Globalization.” American Economic Journal: Economic Policy 2(4): 173-204. Ju, Jiandong, and Shang-Jin Wei, 2011. “When is Quality of Financial Institutions a Source of Comparative Advantage?” Journal of International Economics 84 (2): 178-187. Kochhar, Kalpana, Utsav Kumar, Raghuram Rajan, Arvind Subramanian, and Ioannis Tokatlidis, 2006. “India's Pattern of Development: What Happened, What Follows?” Journal of Monetary Economics 53(5): 981–1019. Kongsamut, Piyabha, Sergio Rebelo, and Danyang Xie, 2001. "Beyond Balanced Growth," Review of Economic Studies 68(4): 869-82.   Koopman, Robert, Zhi Wang, and Shang-Jin Wei, 2012. "Estimating domestic content in exports when processing trade is pervasive." Journal of Development Economics 99(1): 178-189. Koopman, Robert, Zhi Wang, and Shang-Jin Wei, 2014. “Tracing Value-added and Double Counting in Gross Exports.” American Economic Review 104(2):459-494. Also available as NBER Working Paper No. 18579. Lodefalk, Magnus, 2014. “The Role of Services for Manufacturing Firm Exports.” Review of World Economics 150(1): 59-82. Loungani, Prakash, Saurabh Mishra, Chris Papageorgiou, and Ke Wang, 2017. "World Trade in Services: Evidence from A New Dataset." IMF Working Papers 17/77. Miroudot, Sébastien, and Charles Cadestin, 2017. “Services In Global Value Chains: From Inputs to Value-Creating Activities.” OECD Trade Policy Papers, No. 197. 27      Obstfeld, Maurice, and Kenneth Rogoff, 2007. “The Unsustainable US Current Account Position Revisited,” in Richard Clarida (ed.), G7 Current Account Imbalances: Sustainability and Adjustment, University of Chicago Press. Rajan, Raghuram G., and Luigi Zingales, 1998. “Financial Dependence and Growth.” The American Economic Review 88(3): 559-586. Shen, Leilei, 2013. “Financial Dependence and Growth: Diminishing Returns to Improvement in Financial Development.” Economics Letters 120(2): 215-19. Stehrer, Robert, Neil Foster, and Gaaitzen J. de Vries, 2012. “Value Added and Factors in Trade: A Comprehensive Approach,” WIOD Working Paper No. 7. Timmer, Marcel P., 2012. “The World Input–Output Database (WIOD): Contents, Sources, and Methods.” Version 0.9. Timmer, Marcel P., Bart Los, Robert Stehrer, and Gaaitzen J. de Vries. 2013. "Fragmentation, Incomes and Jobs: An analysis of European competitiveness," Economic Policy 28(76): 613–61. Timmer, Marcel P., Erik Dietzenbacher, Bart Los, Robert Stehrer, and Gaaitzen J. de Vries, 2015. "An Illustrated User Guide to the World Input–Output Database: the Case of Global Automotive Production", Review of International Economics 23: 575–605   Urata, Shujiro, and Kozo Kiyota, 2003. “Services Trade in East Asia,” in T. Ito and A. O. Krueger (eds.), 2003, Trade in Services in the Asia-Pacific Region, NBER-East Asia Seminar on Economics, Vol. 11, University of Chicago Press, Chicago and London. U.S. International Trade Commission (USITC), 2013. “The Economic Effects of Significant U.S. Import Restraints.” Investigation #332-325, Publication 4440. Wang, Zhi, Shang-Jin Wei, and Kunfu Zhu, 2013. “Quantifying International Production Sharing at the Bilateral and Sector Level.” NBER Working Paper No. 19677.   Wang, Zhi, Shang-Jin Wei, Xinding Yu, and Kunfu Zhu, 2017. “Measures of Participation in Global Value Chains and Global Business Cycles.” NBER Working Paper No. 23222.         28      Figure 1: Scatter plots of the shares of services value added in GDP against income level, for financial and business services, year 2005 .15 .4 LUX .25 GBR FRA LUX USA BEL DEU NLD .2 IRL .3 Financial & Business Serives Value Added / GDP SWE Business Services Value Added / GDP Financial Serives Value Added / GDP .1 LUX AUS ITA .15 SVN AUT DNK GBR HUN USA IRL EST CZE FIN MLT JPN .2 NLD ESP PRT BEL FRA POL DEU BRA MEX SVK AUS CAN TWN KOR .1 SWE IRL LVA RUS CYP AUT JPN DNK ITA USA .05 BRA PRT HUNSVN TWN KOR CAN GBR CYP MLT AUS LVAEST ESP NLD LTU POL SVKCZE BRA CYP MEX KOR CAN IND TUR FIN ROM .1 PRT JPN CHN IND RUS LVA BEL GRC AUT .05 GRC IND BGR DNK BGR HUN GRCFRA DEU ITA MLT SWE ESP BGR CHN TURLTU SVK SVN POL IDN MEX TWN ROM RUS IDN EST CHN TUR CZE FIN IDN ROM LTU 0 0 0 8 9 10 11 8 9 10 11 8 9 10 11 log(GDP/capita) log(GDP/capita) log(GDP/capita) Notes: The vertical axis represents the share of services value added in GDP for financial & business services combined, financial services only, and business services only, respectively. The horizontal axis represents log(GDP/capita). Only the data for year 2005 are used. Data sources: WIOD and PWT. 29      Figure 2: RCA against Bank Private Credit to GDP (Df1), 2005 Basic & Fabricated Metals (low SII) Food, Beverages & Tobacco (high SII) 2 3 IRL DNK SVK SVN BRA NLD JPN CZE BRA 1.5 KOR ITA IDN CYP 2 TUR AUT FIN DEU LTU BEL CHN ESP AUS RUS ROM SWE POL RCA RCA POL BGR FRA 1 BEL AUS EST MEX CAN LVA FRAMLT ESP PRT IND USA TUR CZE AUT HUN PRT CHN ROMSVK GBR 1 LUX GBR CAN BGR ITA NLD DEU DNK LVA GRC HUN GRC .5 MEX SVN LUX EST SWE FIN IND USA IDN CYP LTU KOR JPN MLT IRL RUS 0 0 0 .5 1 1.5 0 .5 1 1.5 (Bank Private Credit)/GDP (Bank Private Credit)/GDP Notes: The vertical axis represents the RCA based on DVA in gross exports for Basic & Fabricated Metals sector and Food, Beverages & Tobacco sector, respectively. The horizontal axis represents a measure of financial development – the ratio of bank private credit to GDP. The data for bank private credit are from the World Bank Global Financial Development Database (GFDD). Only the data for year 2005 are used in this diagram. 30      Figure 3: Correlation between services trade barriers and the share of embodied foreign services for textiles sector in year 2000 1 Share of embodied foreign financial services .8 IRL RUS HUN .6 LTU GRC DNK CZE BEL MEX CAN DEU SWE POL NLD IDN .4 BGR AUT FIN ESP PRT ROM AUS KOR CHN GBR FRA ITA .2 BRA TUR USA IND JPN 0 0 10 20 30 40 50 Financial services trade restriction index Notes: The vertical axis represents the share of embodied foreign financial services among the embodied foreign and domestic financial services for year 2000. The horizontal axis represents financial services trade restriction index, using the data from the World Bank Services Trade Restriction Indexes (STRI). See Appendix 2 for the country names corresponding to these country codes. 31      Figure 4: Scatter plot of our financial services input intensity against Rajan and Zingales (1998) external financial dependence measure for the U.S. Electrical & Optical .2 Transport Equipment Coke & Refined Petroleum Rubber and Plastics External Financial Dependence 0 Chemicals Basic Metals Machinery, Nec WoodPulp & Paper Manufacturing, Nec Mineral -.2 Food, Beverages & T -.4 Textiles Leather & Fo -.6 0 .02 .04 .06 .08 Financial Services Input Intensity Notes: Our financial services input intensity measure is based on the WIOD data in 1995, while Rajan and Zingales (1998) external financial dependence measure is for year 1980. 32      Figure 5: Scatter plot of average SIIf against SIIb for all manufacturing sectors, 2005 IND .04 PRT FRA TWN Financial Services Input Intensity GRC USA BGR CYP ITA BRAJPN .03 MEX GBR HUN .02 TUR DNK MLT ESP KOR LVA AUS NLD BEL DEU AUT IRL CHN POL SVN CAN SWE ROM LTU EST .01 ZOW CZE IDN SVK FIN RUS LUX 0 0 .05 .1 .15 Business Services Input Intensity Note: The services input intensity measures are based on the WIOD data in 2005. 33      Table 1: Descriptive statistics of key variables (for WIOD countries over 1995-2007) Variables Variable Description Obs Mean S.D. Min Max RCA RCA based on manufacturing domestic value added (DVA) in gross exports 6,734 1.112 1.037 0.000 12.507 SIIf (U.S.) ratio of U.S. embodied domestic financial (f) services to U.S. manuf. total value added (totva) 6,734 0.036 0.024 0.004 0.100 SIIb (U.S.) ratio of U.S. embodied domestic business (b) services to U.S. manufacturing totva 6,734 0.109 0.074 0.011 0.309 SIIfb (U.S.) ratio of U.S. embodied domestic financial & business (fb) services to U.S. manuf. totva 6,734 0.145 0.098 0.014 0.406 SIIf (U.S. avg.) average U.S. SIIf over 1995-2007 6,734 0.035 0.023 0.005 0.080 SIIb (U.S. avg.) average U.S. SIIb over 1995-2007 6,734 0.107 0.068 0.013 0.215 SIIfb (U.S. avg.) average U.S. SIIfb over 1995-2007 6,734 0.142 0.089 0.018 0.294 SIIf country’s own SIIf, time-varying 6,704 0.026 0.031 0.000 0.530 SIIb country’s own SIIb, time-varying 6,704 0.044 0.044 0.000 0.352 SIIfb country’s own SIIfb, time-varying 6,704 0.070 0.062 0.000 0.559 forshf share of embodied foreign financial services in total embodied domestic & foreign F services 6,674 0.416 0.192 0.036 0.999 forshb share of embodied foreign business services in total embodied domestic & foreign B services 6,674 0.496 0.207 0.076 1.000 forshfb share of embodied foreign F&B services in total embodied domestic & foreign F&B services 6,674 0.450 0.188 0.066 0.999 Df (VA/GDP) share of financial services totva among a country’s GDP 6,734 0.056 0.035 0.012 0.289 Db (VA/GDP) share of business services totva among a country’s GDP 6,734 0.069 0.031 0.011 0.154 Dfb (VA/GDP) share of financial & business services totva among a country’s GDP (= Df + Db) 6,734 0.125 0.050 0.036 0.391 Df (VA/worker) services value added per worker for financial services sector (in thousand USD) 6,720 4.053 0.807 1.607 5.882 Db (VA/worker) services value added per worker for business services sector (in thousand USD) 6,720 3.407 0.828 0.871 5.091 Dfb (VA/worker) services value added per worker for financial & business services sectors (in thousand USD) 6,720 3.649 0.776 1.244 5.352 Df1 Alternative measure 1 for Df = ratio of bank private credit to GDP 6,300 0.693 0.457 0.072 2.010 Df2 Alternative measure 2 for Df = (Bank Private Credit + Stock market capitalization)/GDP 6,202 1.216 0.767 0.086 4.072 log(GDP/capita) log(GDP/capita) 6,734 9.788 0.714 7.357 11.363 TFP manufacture TFP estimated by the dual approach 5,686 -0.001 0.446 -4.282 4.656 log(emp) log(manufacture employment) 6,704 4.070 2.026 -5.717 9.749 SKratio Skill ratio = payment to high skill workers / payment to all workers (in manufacturing sectors) 6,734 0.200 0.091 0.028 0.631 K/L capital/labor ratio in manufacture 6,697 0.932 1.917 -0.876 62.323 rwage Relative wage = average wage per worker in a country / world average wage per worker 6,704 0.964 4.671 0.000 75.502 GVA Participation GVC participation index based on forward linkage 6,704 0.304 0.196 0.000 3.870 Notes: This table is based on all of the available data for 14 manufacturing sectors of 37 WIOD countries over 1995-2007 (# of observations can be up to 14*37*13=6734). Three WIOD countries are not covered: U.S. is dropped as the benchmark country to define services input intensity; China and Romania are also dropped due to lack of wage and employment data. The statistics for the samples used in regressions are very similar. 34      Table 2: The effects of services development on manufacturing export RCA, using U.S. services input intensity averaged over 1995-2007 (1) (2) (3) (4) (5) (6) Df -5.102*** (1.733) Df*SIIf 111.109*** 111.829*** (35.993) (37.762) Db 0.857 (2.066) Db*SIIb 38.719*** 39.306*** (12.895) (13.937) Dfb -3.240** (1.405) Dfb*SIIfb 27.217*** 27.721*** (6.695) (7.182) log(GDP/capita) 0.153 0.008 0.084 (0.168) (0.160) (0.168) TFP 0.069*** 0.075*** 0.072*** 0.064** 0.066** 0.065** (0.022) (0.021) (0.022) (0.029) (0.028) (0.028) log(emp) 0.938*** 0.935*** 0.952*** 0.958*** 0.952*** 0.970*** (0.091) (0.092) (0.092) (0.098) (0.099) (0.099) SKratio 0.588 0.521 0.542 0.802 0.833 0.796 (0.621) (0.636) (0.638) (0.801) (0.826) (0.822) K/L 0.069*** 0.066*** 0.067*** 0.071*** 0.068*** 0.069*** (0.025) (0.025) (0.025) (0.026) (0.026) (0.026) rwage -0.001 -0.001 -0.001 -0.000 -0.001 0.001 (0.007) (0.007) (0.007) (0.009) (0.008) (0.009) GVC Participation 2.463*** 2.452*** 2.470*** 2.577*** 2.558*** 2.584*** (0.448) (0.445) (0.444) (0.500) (0.497) (0.495) Country FEs Yes Yes Yes Sector FEs Yes Yes Yes Year FEs Yes Yes Yes Country*Year FEs Yes Yes Yes Sector*Year FEs Yes Yes Yes Observations 5,686 5,686 5,686 5,686 5,686 5,686 R-squared 0.575 0.575 0.581 0.593 0.593 0.599 Notes: The dependent variable is manufacturing export RCA. Df (Db) refers to the share of financial (business) services value added in GDP. Dfb equals the sum of Df and Db. Sf is the ratio of the U.S. embodied domestic financial services to U.S.’ manufacturing value added, averaged over 1995-2007. All WIOD manufacturing sectors 3-16 are covered (not grouped together). Robust standard errors in parentheses, clustered by country*sector. *** p<0.01, ** p<0.05, * p<0.1. 35      Table 3A: Robustness check (1), alternative measure of services development (services value added per worker) (1) (2) (3) (4) (5) (6) Df 0.050 (0.076) Df*SIIf 3.667** 3.992** (1.690) (1.948) Db 0.106 (0.068) Db*SIIb 1.237** 1.247** (0.574) (0.621) Dfb 0.132* (0.079) Dfb*SIIfb 1.125** 1.160** (0.457) (0.504) log(GDP/capita) -0.221 -0.320** -0.428** (0.168) (0.163) (0.170) TFP 0.069*** 0.062*** 0.063*** 0.063** 0.062** 0.062** (0.022) (0.021) (0.021) (0.029) (0.029) (0.029) log(emp) 0.937*** 0.934*** 0.939*** 0.953*** 0.950*** 0.955*** (0.093) (0.092) (0.093) (0.100) (0.099) (0.100) SKratio 0.586 0.512 0.514 0.788 0.708 0.689 (0.629) (0.630) (0.629) (0.823) (0.822) (0.823) K/L 0.069*** 0.068*** 0.069*** 0.070*** 0.069*** 0.070*** (0.026) (0.025) (0.025) (0.026) (0.025) (0.026) rwage -0.002 -0.000 -0.000 -0.001 -0.001 -0.001 (0.007) (0.007) (0.007) (0.008) (0.008) (0.008) GVC Participation 2.428*** 2.433*** 2.425*** 2.530*** 2.512*** 2.506*** (0.445) (0.443) (0.442) (0.495) (0.490) (0.490) Country FEs Yes Yes Yes Sector FEs Yes Yes Yes Year FEs Yes Yes Yes Country*Year FEs Yes Yes Yes Sector*Year FEs Yes Yes Yes Observations 5,674 5,674 5,674 5,674 5,674 5,674 R-squared 0.573 0.574 0.576 0.591 0.591 0.592 Notes: The dependent variable is manufacturing export RCA. Df (Db) refers to the share of financial (business) services value added per worker. Dfb is the services value added per worker for both financial and business services sectors combined. SIIf (SIIb) is the ratio of the U.S. embodied domestic financial (business) services to U.S.’ manufacturing value added, averaged over 1995-2007. SIIfb is the combined measure for both financial and business services. All WIOD manufacturing sectors 3-16 are covered (not grouped together). Robust standard errors in parentheses, clustered by country*sector. *** p<0.01, ** p<0.05, * p<0.1. 36      Table 3B: Robustness check (2), alternative measures of financial development (Df1 & Df2) (1) (2) (3) (4) Df1 -0.292** (0.120) Df1*SIIf 7.724*** 7.668** (2.841) (3.065) Df2 -0.206*** (0.072) Df2*SIIf 4.962*** 5.121*** (1.662) (1.863) log(GDP/capita) 0.175 0.097 (0.173) (0.153) TFP 0.068*** 0.061*** 0.068** 0.059* (0.022) (0.023) (0.030) (0.030) log(emp) 0.935*** 0.937*** 0.954*** 0.954*** (0.094) (0.093) (0.101) (0.100) SKratio 0.629 0.662 0.866 0.845 (0.636) (0.635) (0.823) (0.811) K/L 0.066*** 0.065*** 0.068*** 0.066*** (0.024) (0.023) (0.025) (0.024) rwage -0.002 -0.002 -0.001 -0.001 (0.007) (0.007) (0.009) (0.009) GVC Participation 2.549*** 2.536*** 2.668*** 2.658*** (0.473) (0.466) (0.528) (0.518) Country FEs Yes Yes Sector FEs Yes Yes Year FEs Yes Yes Country*Year FEs Yes Yes Sector*Year FEs Yes Yes Observations 5,758 5,678 5,758 5,678 R-squared 0.509 0.511 0.530 0.532 Notes: The dependent variable is manufacturing export RCA. Df1 refers to the ratio of bank credits to private sectors to GDP. Df2 is the ratio of bank credits to private sectors and stock market capitalization to GDP. SIIf is the ratio of the U.S. embodied domestic financial services in U.S.’ manufacturing value added, averaged over 1995-2007. All WIOD manufacturing sectors 3-16 are covered (not grouped together). Robust standard errors in parentheses, clustered by country*sector. *** p<0.01, ** p<0.05, * p<0.1. 37      Table 4: Robustness check (3), using alternative services input intensity measures for financial & business services combined (fb) (1) (2) (3) Use time-varying U.S. Use countries’ own time- Use U.K.’s average services input intensity varying services intensity Services input intensity Dfb*SIIfb 23.941*** 11.965*** 30.300** (6.546) (4.147) (14.332) TFP 0.064** 0.067** 0.067** (0.028) (0.029) (0.029) log(emp) 0.964*** 0.944*** 0.969*** (0.099) (0.095) (0.099) SKratio 0.800 0.730 0.746 (0.818) (0.682) (0.687) K/L 0.070*** 0.075*** 0.070*** (0.026) (0.027) (0.025) rwage 0.000 -0.001 0.001 (0.009) (0.008) (0.008) GVC Participation 2.575*** 2.572*** 2.556*** (0.496) (0.499) (0.498) Country*Year FEs Yes Yes Yes Sector*Year FEs Yes Yes Yes Observations 5,686 5,855 5,714 R-squared 0.597 0.591 0.593 Notes: The dependent variable is manufacturing export RCA. Dfb refers to the share of financial & business services value added in GDP. In regression (1), SIIfb is the ratio of the U.S. embodied domestic financial & business services to U.S.’ manufacturing value added (not averaged over years). In regression (2), SIIfb measures countries’ own services input intensity (not averaged over years). In regression (3), SIIfb is the ratio of the U.K. embodied domestic financial & business services to U.K.’ manufacturing value added, averaged over 1995-2007. All WIOD manufacturing sectors 3-16 are covered (not grouped together). U.S. observations are dropped from regression (1) and the U.K. observations are dropped from regression (3). Robust standard errors in parentheses, clustered by country*sector. *** p<0.01, ** p<0.05, * p<0.1. 38      Table 5A: Effects of embodied foreign services on manufacturing export RCA, for MORE services intensive manufacturing sectors (1) (2) (3) (4) (5) (6) Df 2.445 (3.766) forshf 0.246 0.989 (0.499) (1.295) Df*forshf -8.962** -32.749 (4.376) (22.928) Db 16.921*** (5.080) forshb 1.070 2.801** (0.660) (1.113) Db*forshb -20.873*** -45.683*** (8.025) (15.308) Dfb 6.959** (3.374) forshfb 0.991 3.265** (0.820) (1.548) Dfb*forshfb -12.261*** -35.116*** (4.499) (12.198) log(GDP/capita) -0.183 -0.182 -0.168 (0.204) (0.220) (0.217) TFP 0.071** 0.078** 0.073** 0.051 0.057 0.050 (0.034) (0.032) (0.034) (0.050) (0.050) (0.051) log(emp) 0.880*** 0.888*** 0.889*** 0.894*** 0.907*** 0.918*** (0.080) (0.081) (0.080) (0.089) (0.089) (0.090) SKratio 0.320 0.179 0.256 0.317 0.283 0.315 (0.713) (0.689) (0.716) (0.935) (0.910) (0.924) K/L 0.145** 0.150** 0.151** 0.178** 0.189** 0.182** (0.072) (0.072) (0.072) (0.090) (0.088) (0.088) rwage -0.018* -0.016* -0.017* -0.021 -0.017 -0.018 (0.010) (0.010) (0.010) (0.014) (0.013) (0.013) GVC Participation 0.913*** 0.940*** 0.928*** 1.033*** 1.004*** 1.024*** (0.289) (0.293) (0.290) (0.335) (0.338) (0.337) Country FEs Yes Yes Yes Sector FEs Yes Yes Yes Year FEs Yes Yes Yes Country*Year FEs Yes Yes Yes Sector*Year FEs Yes Yes Yes Observations 2,901 2,901 2,901 2,901 2,901 2,901 R-squared 0.596 0.598 0.598 0.622 0.626 0.626 Notes: Dependent variable is manufacturing export RCA. We keep only the last seven sectors with high financial and business service input intensity, as listed in Appendix 4. Robust standard errors in parentheses, clustered by country*sector. *** p<0.01, ** p<0.05, * p<0.1. 39      Table 5B: Effects of embodied foreign services on manufacturing export RCA, for LESS services intensive manufacturing sectors (1) (2) (3) (4) (5) (6) Df 0.359 (4.761) forshf -0.233 -1.231 (0.560) (1.549) Df*forshf -2.248 15.603 (9.165) (25.722) Db 9.509** (4.404) forshb -0.525 -0.390 (0.589) (0.984) Db*forshb -13.422* -17.385 (7.452) (14.001) Dfb 4.016 (3.735) forshfb -0.120 0.334 (0.786) (1.844) Dfb*forshfb -6.610 -10.217 (5.555) (15.852) log(GDP/capita) 0.371 0.284 0.325 (0.273) (0.250) (0.260) TFP 0.058** 0.056** 0.057** 0.067* 0.057 0.064* (0.028) (0.028) (0.028) (0.038) (0.035) (0.036) log(emp) 0.959*** 0.931*** 0.940*** 0.966*** 0.944*** 0.963*** (0.155) (0.162) (0.158) (0.169) (0.183) (0.177) SKratio 0.196 0.226 0.187 0.147 0.152 0.153 (0.653) (0.666) (0.667) (0.831) (0.871) (0.862) K/L 0.068*** 0.073*** 0.071*** 0.070*** 0.076*** 0.073*** (0.023) (0.025) (0.024) (0.022) (0.025) (0.024) rwage 0.018** 0.019** 0.018** 0.022* 0.022* 0.021* (0.009) (0.009) (0.009) (0.012) (0.012) (0.012) GVC Participation 3.972*** 4.098*** 4.056*** 4.254*** 4.373*** 4.306*** (0.706) (0.696) (0.708) (0.847) (0.838) (0.844) Country FEs Yes Yes Yes Sector FEs Yes Yes Yes Year FEs Yes Yes Yes Country*Year FEs Yes Yes Yes Sector*Year FEs Yes Yes Yes Observations 2,931 2,931 2,931 2,931 2,931 2,931 R-squared 0.637 0.644 0.640 0.665 0.669 0.666 Notes: Dependent variable is manufacturing export RCA. We keep only the first seven sectors with high financial and business service input intensity, as listed in Appendix 4. Robust standard errors in parentheses, clustered by country*sector. *** p<0.01, ** p<0.05, * p<0.1. 40      Appendix 1A: Total direct & indirect value added export (VAX) of financial service, 1995-2007 X Ratio1 = Ratio2 = Country (gross export) VAX VAX/X dVAX indVAX indVAX/dVAX AUS 210 579 2.76 133 447 3.74 AUT 564 686 1.22 334 352 1.00 BEL 620 896 1.45 327 569 1.65 BGR 20 47 2.37 13 34 2.69 BRA 72 457 6.30 47 410 8.68 CAN 487 984 2.02 275 709 2.59 CHN 53 1774 33.62 34 1740 42.94 CYP 4 12 2.72 3 8 2.79 CZE 45 104 2.32 20 84 4.77 DEU 1151 2644 2.30 529 2115 3.56 DNK 122 282 2.30 76 206 2.83 ESP 438 847 1.93 284 563 1.82 EST 8 15 1.91 5 10 1.87 FIN 33 122 3.76 21 101 3.56 FRA 790 1925 2.44 405 1520 3.96 GBR 5050 4339 0.86 2591 1748 0.54 GRC 39 87 2.27 28 59 2.21 HUN 52 118 2.28 29 89 3.74 IDN 139 271 1.95 103 168 1.95 IND 101 683 6.77 78 605 7.46 IRL 1597 1171 0.73 831 341 0.38 ITA 552 1604 2.90 331 1273 3.57 JPN 789 3603 4.57 522 3081 6.25 KOR 199 976 4.91 119 857 6.58 LTU 2 12 6.45 1 11 9.35 LUX 2910 849 0.29 640 209 0.32 LVA 8 16 1.96 5 10 2.09 MEX 181 512 2.83 119 393 3.70 MLT 7 11 1.49 4 7 1.97 NLD 820 1244 1.52 442 803 1.58 POL 114 219 1.92 70 148 2.55 PRT 93 235 2.52 65 170 2.76 ROM 24 67 2.82 17 50 3.07 RUS 5 152 29.25 4 149 49.65 SVK 20 37 1.89 12 25 2.01 SVN 9 38 4.35 6 33 4.66 SWE 372 532 1.43 247 285 1.08 TUR 7 251 34.48 5 247 45.46 TWN 95 1197 12.55 75 1122 13.78 USA 10116 11897 1.18 5624 6273 1.07 ROW 2382 4798 2.01 1598 3199 1.93 TOT 30300 46293 1.53 16070 30223 1.88 Notes: The export values in this table are for financial services sector (WIOD sector 28) in 100 million U.S. dollars at 2005 constant price, using the U.S. GDP deflator from the Federal Reserve Economic Data (website address: http://research.stlouisfed.org/fred2). X is total gross exports. VAX is total value added exports. dVAX is direct value added exports. indVAX is indirect value added exports through other sectors. 41      Appendix 1B: Total direct & indirect value added export (VAX) of business service, 1995-2007 X Ratio1 = Ratio2 = Country (gross export) VAX VAX/X dVAX indVAX indVAX/dVAX AUS 512 1128 2.20 250 879 3.60 AUT 1012 1184 1.17 579 604 1.11 BEL 1964 2657 1.35 964 1693 1.76 BGR 11 26 2.41 8 18 2.23 BRA 303 665 2.19 193 472 2.38 CAN 1322 2380 1.80 812 1568 1.93 CHN 1765 1921 1.09 757 1164 1.33 CYP 12 21 1.78 8 13 1.66 CZE 274 352 1.28 121 231 1.97 DEU 3821 12761 3.34 2596 10166 3.74 DNK 429 680 1.59 236 444 1.80 ESP 1576 1957 1.24 924 1033 1.08 EST 27 38 1.41 15 24 1.62 FIN 474 654 1.38 278 376 1.25 FRA 3138 8361 2.66 1801 6560 3.69 GBR 6748 9602 1.42 4635 4966 1.06 GRC 99 130 1.31 53 77 1.39 HUN 287 388 1.35 163 225 1.38 IDN 29 48 1.63 17 32 1.31 IND 1588 1279 0.81 1084 195 0.19 IRL 1563 1321 0.85 862 458 0.51 ITA 1713 4178 2.44 965 3212 3.29 JPN 1011 4660 4.61 608 4052 7.06 KOR 591 1624 2.75 374 1249 3.27 LTU 16 25 1.55 10 15 1.87 LUX 231 240 1.04 121 118 0.90 LVA 14 25 1.85 7 18 2.67 MEX 88 793 9.06 63 730 12.17 MLT 24 24 1.02 15 9 0.67 NLD 3266 3954 1.21 1896 2058 1.06 POL 258 511 1.98 148 363 2.28 PRT 175 299 1.70 92 206 2.16 ROM 106 119 1.12 52 66 1.23 RUS 72 524 7.27 45 480 12.01 SVK 85 121 1.42 43 78 1.86 SVN 50 99 2.00 27 72 2.68 SWE 1444 1903 1.32 811 1092 1.27 TUR 1.34 146 109.57 1 146 148.14 TWN 423 516 1.22 216 300 1.61 USA 9517 20777 2.18 6230 14547 2.20 ROW 9776 8234 0.84 5458 2776 0.51 TOTAL 55815 96323 1.73 33535 62788 1.87 Notes: The export values in this table are for business service sector (WIOD sector 30) in 100 million U.S. dollars at 2005 constant price, using the U.S. GDP deflator from the Federal Reserve Economic Data (website address: http://research.stlouisfed.org/fred2). X is total gross exports. VAX is total value added exports. dVAX is direct value added exports. indVAX is indirect value added exports through other sectors. 42      Appendix 2: Countries & Codes in WIOD Code Country Code Country Code Country Code Country AUS Australia DNK Denmark IRL Ireland POL Poland AUT Austria ESP Spain ITA Italy PRT Portugal BEL Belgium EST Estonia JPN Japan ROM Romania Russian BGR Bulgaria FIN Finland KOR Korea, Rep. RUS Federation Slovak BRA Brazil FRA France LTU Lithuania SVK Republic United CAN Canada GBR Kingdom LUX Luxembourg SVN Slovenia CHN China GRC Greece LVA Latvia SWE Sweden CYP Cyprus HUN Hungary MEX Mexico TUR Turkey Taiwan, CZE Czech Rep. IDN Indonesia MLT Malta TWN China DEU Germany IND India NLD Netherlands USA United States Appendix 3A: Manufacture (sec 3-16) and service sectors (sec 28 & 30) covered by this paper sec descriptions c3 Food, Beverages and Tobacco c4 Textiles and Textile Products c5 Leather, Leather and Footwear c6 Wood and Products of Wood and Cork c7 Pulp, Paper, Paper , Printing and Publishing c8 Coke, Refined Petroleum and Nuclear Fuel c9 Chemicals and Chemical Products c10 Rubber and Plastics c11 Other Non-Metallic Mineral c12 Basic Metals and Fabricated Metal c13 Machinery, Nec c14 Electrical and Optical Equipment c15 Transport Equipment c16 Manufacturing, Nec; Recycling c28 Financial Intermediation (see Appendix 3B for its coverage) c30 Renting of M&Eq and Other Business Activities (see Appendix 3C for its coverage) 43      Appendix 3B: Detailed ISIC sectors inside financial services (WIOD sector 28) 6511 Central banking 6519 Other monetary intermediation 6591 Financial leasing 6592 Other credit granting 6599 Other financial intermediation n.e.c. 6601 Life insurance 6602 Pension funding 6603 Non life insurance 6711 Administration of financial markets 6712 Security dealing activities 6719 Activities auxiliary to financial intermediation n.e.c. 6720 Activities auxiliary to insurance and pension funding Appendix 3C: Detailed ISIC sectors inside business services (WIOD sector 30) 7111 Renting of land transport equipment 7112 Renting of water transport equipment 7113 Renting of air transport equipment 7121 Renting of agricultural machinery and equipment 7122 Renting of construction and civil engineering machinery and equipment 7123 Renting of office machinery and equipment (including computers) 7129 Renting of other machinery and equipment n.e.c. 7130 Renting of personal and household goods n.e.c. 7210 Hardware consultancy 7220 Software consultancy and supply 7230 Data processing 7240 Data base activities 7250 Maintenance and repair of office, accounting and computing machinery 7290 Other computer related activities 7310 Research and experimental development on natural sciences and engineering (NSE) 7320 Research and experimental development on social sciences and humanities (SSH) 7411 Legal activities 7412 Accounting, book-keeping and auditing activities; tax consultancy 7413 Market research and public opinion polling 7414 Business and management consultancy activities 7421 Architectural and engineering activities and related technical consultancy 7422 Technical testing and analysis 7430 Advertising 7491 Labour recruitment and provision of personnel 7492 Investigation and security activities 44      7493 Building-cleaning activities 7494 Photographic activities 7495 Packaging activities 7499 Other business activities n.e.c. Appendix 4: Manufacturing sector classification based on service input intensity Average ratio of embodied WIOD Sector domestic financial & business sector Description services to total manufacturing High SII Sectors value added for the U.S. over 1995-2007 5 Leather, Leather and Footwear .315 0 3 Food, Beverages and Tobacco .291 0 15 Transport Equipment .236 0 13 Machinery, Nec .196 0 16 Manufacturing, Nec; Recycling .189 0 4 Textiles and Textile Products .171 0 14 Electrical and Optical Equipment .168 0 9 Chemicals and Chemical Products .128 1 7 Pulp, Paper, Paper, Printing and Publishing .122 1 8 Coke, Refined Petroleum and Nuclear Fuel .091 1 6 Wood and Products of Wood and Cork .045 1 10 Rubber and Plastics .043 1 11 Other Non-Metallic Mineral .020 1 12 Basic Metals and Fabricated Metal .019 1 45