Imagine you are working for a multinational company, headquartered in the U.S. You were recently hired to lead a newly-established virtual product development division. Your job is to develop new inno

Journal of Policy Modeling 37 (2015) 65–91 Available online at www.sciencedirect.com ScienceDirect Why some countries are slow in acquiring new technologies? A model of trade-led diffusion and absorption Gouranga Gopal Das ∗ Department of Economics, Hanyang University, Erica Campus, Kyunggi-Do 426-791, South Korea Received 12 July 2014; received in revised form 4 November 2014; accepted 10 January 2015 Available online 21 January 2015 Abstract Drawing on the stylized facts and evidences, in a computable general equilibrium (CGE) model, this paper examines the impact on TFP of North–South, North–North trade-related triangular R&D spillovers.

By constructing different technology appropriation parameters based on embodied and disembodied R&D, absorption and learning effects, it shows: (i) North–South R&D ows have a positive impact on TFP; (ii) human capital-induced skill facilitates North–South R&D ows; (iii) socio-institutional and technology adoption parameters do play roles for knowledge ows, its capture, and transmission. Such technology diffusion and assimilation counters the adverse impact of North–South geographical distance on productivity dynamics.

© 2015 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved. JEL classification: J24; O33; O43; F43; D58 Keywords: Technology diffusion; Trade; Syndromes; Capture; R&D “In all the cases of sustained, high growth, the [developing] economies have rapidly absorbed knowhow, technology, and, more generally, knowledge from the rest of the world. With the usual disclaimer, I bene ted from helpful discussions with Ron Davies, University College, Dublin; Bruce Blonigen, University of Oregon. Comments from conference participants at the City University of Hong Kong and Izmir University, Turkey are extremely useful. ∗Tel.: +82 31 400 5628; fax: +82 31 400 5591.

E-mail address: [email protected] http://dx.doi.org/10.1016/j.jpolmod.2015.01.0010161-8938/© 2015 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved. 66 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 These economies did not have to originate much of this knowledge, but they did have to assimilate it at a tremendous pace. That we know. What we do not know—at least not as well as we would like—is precisely how they did it, and how policy makers can hurry the process along. This is an obvious priority for research. [Economies] can learn faster than they can invent. Knowledge acquired from the global economy is thus the fundamental basis of economic catch-up and sustained growth.”—Michael Spence, Spence Commission on Growth and Development, World Bank (2008) , p. 41. 1. Introduction and motivation Drawing on the idiosyncratic regional growth experiences and historical trajectories of devel- opment episodes, Spence (2011) , and World Bank (2008) highlighted, inter alia, the importance of pursuing context- and case-speci c policies—as emphasized in the quote that opened this article—related to trade and industry development, education, knowledge creation, technological improvement, inequality, etc.; in particular, it emphasized the preponderant role of engagement in the global economy as well as extending the knowledge frontier via investment in human capital, institutional quality, R&D, and the educative effect of trade to achieve Millennium Development Goals. In particular, Goals 2, 8A and 8F emphasize the roles of not only global partnership for technological access to reduce disparities in ICT diffusion and usage, but also on the importance of good governance and institutions (United Nations, October 2010 ). In a constantly evolving world of ideas, science, technology, and innovation policy aims at dispensing with the growth inertia by sparking mankind’s innate ‘lack of patience with status quo, or sitzfleisch’ (Griliches, 2000 ). Government policy plays role for orchestration of knowledge, entrepreneurial know-how, cross-fertilization of ideas to enrich the knowledge-capital, and institutional innovation. The important visionary role that the state could play as enabling transformer of an economy through innovation-oriented policy trajectory is widely evidenced in the literature. Mazzucato (2013) has emphasized the role played by state—like Prometheus, not Leviathan—in undertaking invest- ments in innovation via policy instruments to create an innovation ecosystems, where symbiotic coexistence of public–private investments, and other supportive policies like education policy, industrial policy work concurrently. Szirmai, Gebreeyesus, Guadagno, and Verspagen (2013) has discussed in the context of Africa that wide range of policies related to trade, sectoral innovation, population, employment, and labor market are important for productive employment. Bengoa and Sanchez-Robles (2005) and Bluedorn, Duttagupta, Guajardo, and Mwase (2014) have discussed the importance of favorable economic policies conducive for structural reforms, institutions, and domestic factors causing take-off and sustained growth of the less-developed countries’ (LDCs).

In particular, proper region-speci c investment climate via policy instruments in uencing infra- structure, governance features, scarcity of tertiary human capital, weak institutional capacity, and other pertinent socio-institutional factors in the LDCs matter. 1 Why, despite the State-Prometheus having a visionary role to undertake strategic policy instru- ments, the things are not taking shape? In this paper, we ask this policy-guided research question grounded upon theoretical rationale. We enunciate the role of host of factors underlying seizure of the potential knowledge-diffusion. Lack of education, skill mismatch, inappropriate structural change, educational bottlenecks, inadequate innovation, and lack of attention to SMEs with growth 1World Bank’s ‘Doing Business Project’, benchmarking different indicators for regulatory regime for 130 countries, offered Investment Climate Surveys conducted at the rm level for 26,000 rms in 53 developing nations. G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 67 potentials, all balk the growth spurt for the LDCs. Interestingly, Jones and Romer (2010) offered stylized facts based on the interactions between increased market integration (via trade) and four state variables, namely: ideas, human capital, population and institutions. LDCs’ growth and development in the long-term depends on their capabilities, for effectively identifying, procuring, assimilating it (absorptive capacity, AC) and applying the state-of-the-art (Cohen & Levinthal, 1989; Das, 2002, 2010 ), and structural congruence, socio-institutional features like governance, corruption, domestic circumstances. World Bank (2010) has stressed the importance of innovation lying in the core of economic progress of the developing nations and Southern engines. Recently, Asian Development Bank Report (2014a, 2014b) has constructed the Creative Productivity Index (CPI), by ranking developing countries like Indonesia, China, India, Bangladesh, Cambodia, according to creative inputs and outputs such as skill, education, R&D, infrastructure, etc., and urged for ‘strong, coordinated government policies’ for tertiary education, skill-enhancement, ICT investment, innovation and favorable institution so as to develop a knowledge-based economy.

In LDCs, policymakers have focused on these factors in their policy agenda for inclusive growth and development, and policy instruments for accumulation of human and physical capital, innovation, and institution-building are prioritized. In the current paper’s taxonomy, lack of these makes deleterious syndromes to occur and inhibit the growth process. Policy shocks inducing those factors could be a source of endogenous growth, and this paper considers role of tech- nology shocks and trade in intermediates (Salvatore, 2007 ). This paper elicits a mechanism of North–North–South triangular knowledge diffusion for reducing the technological gap via syn- ergistic spillover capture, and how all these country-speci c factors create a conducive policy climate for nourishing and nurturing the innovation potential. For unraveling the interlinkages, we employ multi-sectoral, multi-regional computable general equilibrium (CGE) Global Trade Analysis Project’s (GTAP) model tailored to formalize a stylized mechanism. This paper adds value to the literature by: (i) offering an analysis of interplay between trade, technology dissemi- nation and highlighting the importance of enabling factors for such absorption factors affecting its assimilation; (ii) making a contribution to the mechanism underlying adoption of trade-mediated technology diffusion; and (iii) thus, expounding theoretical underpinning for growth-oriented public policy. As the paper unfolds, Sections 2 and 3 outlines the literature, and considers the data, and Section 4 offers the model. Simulation design and principal ndings are discussed in Section 5. Section 6 summarizes with policy discussion. 2. North–south trade-led technology flows and adoption factors: a bird’s eye (re-) view The important issues of creation of new technology, its diffusion and actual adoption have been discussed on both theoretical and empirical planes. Conscious efforts by policy-makers around the developing world for embracing globalization, and investment in human capital, innovation, and macroeconomic policy has led to the emergence of dynamic emerging economies like Brazil, China, India, Korea, and others, making a New Economic Order conspicuous by 2020. The recon- guration of the world economy via such transformation of the dynamic economies (contrary to the next tier laggards)—as discussed by Duval and de la Maisonneuve (2010) , Jorgenson and Vu (2011, 2013) , OECD (2010a, 2010b, 2014) , Klein and Salvatore (2013) —are attributed gener- ally to: global integration, ICT and non-ICT capital, total factor productivity (TFP), technology transfer, human and non-human capital accumulation, and institutional qualities. However, we observe differences in technology assimilation across countries.

Growth-spillover bene ts the trade-partners so long as different countries engage in trade. Spence (2011, p. 5) has talked highly about this aspect of dissemination and its bene ts to the 68 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 developing world, by mentioning ‘two parallel revolutions’: industrial revolution in advanced world and inclusive revolution, spreading to the south for its quantum leap. For current evidences of such trade-mediated technology diffusion in the developing world, plethora of empirical sup- ports are available—see, Global Economic Prospects (World Bank, 2008 ), Hoekman and Javorcik (2006) , World Bank (1998, 2010) . Despite debates about relative ef cacies of different channels, it is well-established in the literature that trade (especially, imports and to some extent, exports) is signi cant channel of technology diffusion (Eaton & Kortum, 1996; Keller, 2004 ).2Consid- ering a panel of 101 developing and developed countries, Arora and Vamvakidis (2004, 2005) have explored the importance of growth spillover via trade. Additionally, with intermediate trade representing 56% of trade in goods and 73% of trade in services in OECD and emerging engines being well-integrated into production networks, this diffusion improves productive ef ciency of the recipients via intermediate goods and services (OECD, 2010b ). The interest in exploring the North–South trade-mediated ‘indirect’ spillover has drawn attention of researchers, which culmi- nated into papers analyzing the relative merits of indirect technology ows via trade-embodiment, FDI and ‘direct’ disembodied ows via technological proximity such as internet technology, telephone, or effective teledensity (Lee, 2005; Tang & Koveos, 2008 ). Schiff and Wang (2006) considered direct and indirect North–South R&D spillovers between 15 OECD and 24 develop- ing nations across 16 manufacturing industries. We consider domestic R&D as well as foreign sourced R&D in DCs as well as in LDCs, along with heterogeneities in assimilation rates.

Additionally, even with global integration localized nature of technology makes domestic investments in enabling factors such as indigenous innovative capabilities, absorptive capacity, sound macroeconomic policies, human capital and creating better institutions are necessary for effective dissemination (Caselli & Coleman, 2006; Cosar, 2011; Keller, 2004; Lall, 2001; López- Puyeo & Mancebón, 2010; Lucas, 2009a, 2009b; Onyeiwu, 2011; Schiff & Wang, 2006 , to name a few). Quantifying absorptive capacity (AC) by human capital, in the context of Sub-Saharan African (SSA) manufacturing rms Foster-McGregor, Isaksson, and Kaulich (2014) has shown that AC is instrumental for importer rms to have positive productivity spillover. Clark, High lle, de Oliveira, and Rehman (2011) have surveyed FDI-related technological spillovers and growth impact; however, they emphasized the importance of trade channel and the role of AC, mentioning the lacunae in the literature in incorporating these aspects operationally. Onyeiuw (March 2011) has shown that for 31 Sub-Saharan African countries apart from political stability and good governance, absorptive capability to utilize new technology matter much for development. Role of adoption for using a speci c set of ICT skills for teacher trainers is discussed in the context of Cambodia by Richardson (2011) . World Bank (p. 2, 2010) reports that: “Innovation depends signi cantly on overall conditions in the economy, governance, education, and infrastructure. Such framework conditions are particularly problematic in developing countries, but experience shows not only those proactive innovation policies are possible and effective but also that they help create an environment for broader reforms. “Spence (2011, p. 109) has mentioned about ‘key internal ingredients’ for sustained high-growth and the recipes involve factors such as infrastructure, good institutions, governance, education, and facilitating structural change. The negative impacts of corruption, instability, ‘internal snags’ like ‘bad’ rule of law, and bad governance are highlighted ( Collier, 2009 ). In what follows, we offer some pertinent stylized data. 2Keller (2004: p. 769) has mentioned that: “[A] number of surveys have recently concluded that there is no evidence for substantial FDI spillover.” Amiti and Konings (2007) studied the positive effects on productivity of a tariff-cut in trade in intermediates for Indonesian manufacturing. G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 69 3. Database and stylized facts: a macro lens view 3.1. Sectoral and regional dimensions Quoting Lucas (2009a, p. 1): “a study of economic growth in the world as a whole must be a study of the diffusion of the industrial revolution across economies, a study of the cross- country ows of production-related knowledge from the successful economies to the unsuccessful ones.” Literatures abound with information and communication technology’s (ICT) role as a general-purpose-technology (GPT) for facilitating growth via impacts on nanotechnology and biotechnology (OECD, 2006a ). Following Qiang (2007) and Grace, Kenny, and Qiang (2004) , ‘informatization’ is a complex process for achieving ‘critical development goals’ via ‘ICT-driven economic and social transformation’, and investment in social and economic infrastructure.

In the technology diffusion literature, the dynamism and persistence of technological compet- itiveness across industrial clusters are quite common. In this paper, we divide the entire global economy into composite regions—broadly into two groups, the North (N) and the South (S), according to their status of development. Based on variation of diffusion and adoption of new technologies across geographical regions, there has been considerable evidence on differences in technology transfer between the North and the South. OECD nations account for largest of total world R&D and within them, 7 largest (G7) account for major share (Coe, Helpman, & Hoffmaister, 2008 ). UNESCO (2009) has shown that the number of researchers in developing countries has increased by 45% as compared to 9% in the DCs; however, in America, Europe and Oceania researchers per million inhabitants were still far higher than the world average. Consid- ering R&D intensity (i.e., R&D expenditure as a percentage of the GDP), the Americas accounted for 37.6% of World R&D expenditure (mainly attributed to R&D spending in USA and Canada) followed by Europe and Asia. Rapid globalization of science and technological invention has been accompanied not only by concentration of such activities in only OECD regions, but non-OECD economies also have exhibited fastest growth and sizable contribution to global R&D. 3How- ever, Asia’s increasing R&D intensity is largely dominated by China’s contribution—registering increase from 1.1% in 2002 to 1.5% in 2007, thus, accounting for 39% of R&D expenditure and 53% of researchers in the LDCs; but in case of India, it is about 0.8%. In Sub-Saharan Africa (SSA), the intensity is much less, about 0.3%, whereas South Africa invested almost 1% of GDP for R&D. On the world as a whole, R&D expenditure has increased (1.7% of world GDP). Thus, we see that G7 countries have the signi cant share of R&D and still developing countries need to make signi cant strides in their innovative skills. G7 countries’ ‘technological readiness’ index score is far higher than those in Asia and Africa (Global Competitiveness Report 2009–2010).

For our model, we categorize the whole range of 57 product categories in the GTAP database into 7 broad R&D-intensive technology clusters—namely, ICT, Transport Equipment, Mate- rials, Consumption goods, Fabrication and Services (Das, 2002 ). We follow OECD (2003a, 2005) classi cation of manufacturing activities according to technological intensity. Based on Hatzichronoglou (1997) , using ISIC Rev.3 OECD (2003a, 2005) methodology considers both ‘technology-producer’ and ‘technology-user’ aspects based on three technological intensity indi- cators, namely, R&D expenditures as proportion to value-added, production and R&D plus technology embodied in capital goods and intermediates as proportion of production, to determine 3According to OECD (2006a) , compared to 7% in 1995 China, Israel, Russia, and South Africa contribute combined 17% of R&D expenditure of OECD nations in 2004. 70 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 Table 1 Technology clusters and industries. Technology clusters Industries Information and Communications TechnologyComputers and related equipment, Telecommunication and Semiconductor Equipment, Electrical Machinery, Audio and Video Equipment, Instruments Transport Technology Shipbuilding, Aircraft, Motor Vehicles, Other Transportation Consumer goods Technology Food, Beverages and Tobacco, Textiles, Apparel and Footwear Materials Technology Agriculture, Construction, Mining, Paper and Printing, Wood Fabrication Technology Fabricated Metal Products, Other non-electrical machinery, Others Source: OECD (1997), Science, Technology and Industry – Scoreboard of Indicators. pp. 40–41. Ta k e n from Das (2008). ‘technological criteria’. By adopting a narrow de nition (ISIC Rev 3.) and based on idea of embod- ied technology ows estimated from input-output tables, market service activities like ‘Finance and Insurance (Divn 65–67)’, ‘Business activities (71–74)’, ‘Post and telecommunications (Divn 64)’ are considered knowledge-intensive. Table 1 presents the taxonomy of grouping industries into technology clusters (OECD, 2004 ).4IT cluster belongs to the hi-tech cluster. Consumer goods and Fabrication are in the medium-low and low technology categories, respectively. We consider technology clusters—‘industries sharing a number of common characteristics’ (OECD, 2000 ). Table 2 presents the regional concordance and geographical matching of regions/nations. OECD Outlook (2004) identi es ve broad categories of ICT goods. ICT services, based on industry-based de nition and ISIC, Rev 3., are grouped together in the ‘services’ cluster com- prising mainly telecommunications, IT-enabled and related services facilitate trade. 3.2. Stylized facts The GTAP database (Version 7) divides the world economy into 113 regions, 57 sectors and 2 labor classes. Based on SITC, Revision 3, and Commodity Product Classi cation and Harmonised System (HS, Rev 2.), WPIIS, OECD (2003a) has developed a classi cation of ICT-goods sep- arately from ICT-services. 5Tables 3 and 4 present annual growth rates of global and regional trade in the clusters over 1965–2004. As per OECD (2003a) , high-technology industries like electronic equipment and computers represent about 25% of total OECD trade and registered highest growth rates in manufacturing trade. Together with Medium high-technology (transporta- tion cluster, chemicals, machinery and equipment), the share is 65% of manufactures trade. 6This is attributed to rise in investment in knowledge (i.e., R&D expenditures, software investment, human resources via higher education) especially in ICT sector accounting for 5.2% of GDP in the OECD economies in 2002. However, this has also been accompanied by closer integration of OECD and non-OECD countries leading to increase in internationalization of R&D activities.

A cursory look con rms that although the some regions have higher chances of enrichment via trade, the lower bilateral technology achievement indexes, relative to G7 composite region, could reduce the chance; on the contrary, the emerging southern engines of growth with higher bilat- eral technology appropriation factor and inventive capabilities have better scope of enrichment in contrast with the regions with lower R&D (for example, SSA, MENA, other South Asia, South 4OECD (1997) , Science, Technology and Industry Scoreboard of Indicators, pp. 40–41. 5Working Party on Indicators for the Information Society (WPIIS, OECD, 2003a ). 6OECD (2003b) , p. 147. G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 71 Table 2 Sectoral and regional aggregations adopted for the simulation. Regions and elements Sectors and descriptions 1 G7 G7-North Developed 1 ConsumerGood Sugar cane, sugar beet, Plant-based bers, Cattle, sheep, goats, horses 2 OtherEU EU minus G7 4 members Animal products nec, Raw milk, Wool, silk-worm cocoons 3 Brazil Brazil Fishing, Meat: cattle, sheep, goats, horse, Meat products nec 4 Russia Russian Federation Dairy products, Processed rice, Sugar, Food products nec 5 India India. Beverages and tobacco products, Textiles, Wearing apparel 6 China China. Leather products, Wood products, Paper products, publishing 7 Hkg Tw n HongKong Taiwan 2 AgBioTech Paddy rice, Wheat, Cereal grains nec, Vegetables, fruit, nuts 8 SouthKorea South Korea Oil seeds, Crops nec, Forestry, Vegetable oils and fats 9 SouthEAsia Developing Asia 3 ElectronicIT Electronic equipment, Machinery and equipment nec 10 RSA Rest of South Asia Manufactures nec 11 ECA Europe and Central Asia 4 Nano Matrls Coal, Oil, Gas, Minerals nec, Petroleum, coal products 12 SouthAfrica South Africa Chemical, rubber, plastic prods 13 LAC LatinAmerica&Caribbean 5 TransportTec Motor vehicles and parts, Transport equipment nec 14 Mexico Mexico 6 Metal MedTec Mineral products nec, Ferrous metals 15 OthrOECD OECD minus G7 minus EU Metals nec, Metal products 16 MENA MiddleEastNorthAfrica 7 Svces Electricity, Gas manufacture, distribution 17 SSA Sub-Saharan Africa Water, Construction, Trade, Transport nec 18 ROW All other regions Sea transport, Air transport, Communication Financial services nec, Insurance, Business services nec Recreation and other services, PubAdmin/Defence/Health/Educat Dwellings Source: This is based on 18 × 7 Aggmap.Txt le based on author’s aggregation of augmented GTAP Version 7 database. Table 3 Average annual growth rates for global trade in technology clusters, 1965–2004. Technology clusters Average annual growth rates (%) Information and communication technology 12 Consumer goods 9.1 Biotechnology Cluster 6.1 Nanotechnology Cluster 10.4 Transport Equipment 11.2 Fabrication 9.1 Source: Calculated from the time-series trade data for the aggregated GTAP Database. 72 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 Table 4 Average annual growth rates for global trade in technology clusters, 1992–2006. Technology clusters Average annual growth (%) in trade from Source region G7 Other EU Other OECD ICT 5.3 6.9 4.9 Consumer goods 3.8 3.8 4.0 Biotechnology Cluster 2.4 4.1 4.9 Nanotechnology Cluster 8.3 10.2 8.9 Transport Equipment 6.4 6.8 7.7 Fabrication 5.8 5.9 7.2 Source: Author’s calculations as above. East Asia). 7Why? That is explained in the model in Section 4. Global integration has led to technology ows—indirectly via embodied traded intermediate inputs and/or, directly via disem- bodied through IT-enabled services. Thus, effectiveness of trans-border technology diffusion are contingent on several factors that, amongst a tall order, we identify as: recipient’s (any region ‘r’ specific) own domestic R&D (R&D d), foreign-trade induced R&D ows via imported intermedi- ates (R&D f) connectivity with advanced world via superior network (disembodied, R&D disemb ), human development (HDI, a broad multidimensional concept incorporating literacy, health, edu- cation), human-capital induced skill-intensity proxying learning capacity (AC), latest technology (TA), socio-institutional parameter (SIP) encompassing Governance (GP), transparency (T), and competitiveness indicator (C). Apart from that, the role of geographical proximity and contiguity in facilitating trade ows between regions is important. 8Krugman (1991, 1995) has emphasized the role of geography and locational choices in formation of trade blocs (termed ‘natural’ blocs). Frankel (1997) , Diao, Rattso, and Stokke (2005) have focused on the role of adjacency in promot- ing trade between neighboring countries. In Asia and Latin America’s context, the importance of physical (hard) and institutional (soft) infrastructure (institutional bottlenecks like legal, admin- istrative, regulatory, rent-seeking behavior, human capital and customs clearance) along with complementary structural reforms and investments in productive capacity for trading linkages are emphasized. As geographical clustering bene ts easy transmission of technology—either via trade-embodiment, disembodied channel and/or, cultural proximity—between the trade partners, we also consider the role of bilateral distance (D ns) between North (‘n’, G7 here) and other client regions (‘s’). In a single framework we consider the roles of these parameters and relative perfor- mance of DCs vis-à-vis LDCs in harnessing technology vehicled via domestic R&D as well as foreign sourced R&D. 4. Technology spillover and effective assimilation 4.1. Core model The pattern and magnitude of the trans-border ows can be discerned by constellation of the conducive parameters that enable superseding the ‘barriers to riches’ (Parente and Prescott, 7Emerging Southern Giants are new economic center of gravity and hence, recently scope of South-South, South-North cooperation in a declining North-South trend is often explored (see OECD, 2010a; Santos-Paulino and Wan, 2010 ). In this paper, for parsimony we do not consider such scheme. 8Proximity is proxied by physical distance between two members or margins of trade. G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 73 1994 ). Acquisition of state-of-the-art technologies from the industrialized nations (North) to the developing economies (South) is a dominant mode of fostering productivity growth via knowledge propagation (Coe et al., 2008; Eaton & Kortum, 1996; Lucas, 2009a, 2009b; Spence, 2011 ). As Tang and Koveos (2008) has shown that both trade-embodied and dis- embodied information technology-induced knowledge spillovers have larger impact than FDI effect, I consider conjoint impacts of both channels of diffusion. In fact, Tang and Koveos (2008) and Bitzer and Kerekes (2008) have shown that ICT cluster and trade, as compared to FDI, has larger impact on technology spillover from G7 to other destination nations. Vita (2013) , Jorgenson and Vu (2013) , López-Puyeo and Mancebón (2010) , and Bayraktar-Sa glam and Yetkiner (2014) has mentioned the necessity of policies for innovation, capital accumu- lation and spillover assimilation for labor-productivity growth in ICT industry as a leading sector.

Contrasting performances across countries in seizure of the potential bene ts of trans-boundary technology spillovers has been ascribed, inter alia, to their heterogeneous assimilation and idiosyncratic features related to enabling ‘systems of learning and diffusion’, like governance, learning, indigenous technological and absorptive capability (AC), geographical distance, tech- nological and structural congruence. On the contrary, lack of such factors or bottlenecks will aggravate the syndromes, impede absorption and effective utilization of ideas, and make devel- opment disorderly. All these are encapsulated into a learning effect and adoption parameter (LeAP).

Regarding the source region (generically, North), we consider G7 as the major progenitor of technological change. Typically, destination regions comprise the amalgam of heteroge- neous nations, viz., other ‘Norths’ (like other EU and other OECD nations), dynamic adopters or emerging economies, as well as the relatively laggards encapsulated into generic ‘South’.

We reserve ‘r’ for all regions (say, REG) where r ∈ {n, k, u} where ‘n’ is the source (unique), ‘k’ and ‘u’ are other northern and southern recipients of technological change respec- tively. Thus, s ∈ [k, u] represents generic ‘destinations’. One pertinent point to note is that we will use ‘r’ to represent all regions and then, as per the requirements of speci cation we will use the moving subscripts/superscripts viz., ‘n’, ‘s’, ‘k’ and ‘u’ (s ⊂ r and n / = k, u / = k).

Investment in skill helps unlocking the potential of technological capability and technology absorption, diffusion of ideas and innovation-entrepreneurship (Caselli & Coleman, 2006; Cosar, 2011; Goldberg, Branstetter, Goddard, & Kuriakose, 2008; Herrerias & Orts, 2013; Jones, 2008; Kosempel, 2007 ). Destinations’ growth depends on the extent of technology propagation as well as on skill intensity-induced absorptive capacity (AC). AC rindex is region ‘r’ speci c 0 ≤ AC r≤ 1. Following Sen (2004) , this ‘capability’ translates into important functioning of accessing technological improvement, and converting into well-defined action via productiv- ity.

As successful adoption depends on a combination of factors such as, educational attainment, intensity of R&D activity, knowledge creation, we need to consider a broader aggregative multi- dimensional ‘Innovation Capability Index’ (0 < ICC r< 1) as developed by the World Investment Report (2005). Not only that, technology availability to a region (0 < TA r< 7) is also crucial for technology acquisition, derived from scores of such measure from Global Competitiveness Report (2009–2010). Kosempel (2007) has shown that elasticity of human capital acquisition in response to technology, a measure of learning propensity, determines TFP. Thus, productivity of skill 74 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 depends on availability of learning scope or technology (TA r) R&D, ICC and AC, determining regional ‘Technology Absorption Parameter (TAP r)’.9Thus, we write: TA P r= AC r· TA r· ICC r (1) Bi-lateral, with respect to TA P of G7 (n = North) is de ned as: TA P ns = min 1,TA P s TA P n , where r = n, s = k, u(2) Here, also TA P ns∈ [0,1] with zero implying further away from the technological absorption capacity of the relatively advanced nation/s and unity implying closer to innovating partners. Keller (2004) mentions, ‘speeding up the diffusion of technology [will] depend on the dis- tribution of human capital and R&D across countries (pp. 774–775).’ As 96% of world’s R&D expenditure ows take place in the developed North, abstracting from Southern R&D does not necessarily undermine our primary emphasis. I here consider domestic R&D expenditure data (as % of GDP) as ow variable measure. For own R&D (R&D d r), we combine data—domestic R&D expenditure as percentage of GDP (GERD) from UNESCO (2008) and Human Development Report (2008, 2009) for the base year of Gtap data (i.e., 2004)—to match the single region and derive for composite regions a simple average of their component-wise gures. We take GTAP database’s regional imports, exports and bi-lateral intermediate import shares in total imports in the destination to get foreign R&D ows (R&D f r) according to the formula: R&D f s= R&D d n· TFP n· Ψns, where r = n and s = k, u (3) where n = G7 and ‘s’ is destination. Ψns= bilateral intermediate import shares in value-added.

As there are substantial inter- and intra-regional trade ows among the composite North and their partners, we adopt the same formula to derive intra-regional ows. Thus, total R&D for a particular region ‘r’ is: R&D r= R&D d r+ R&D f r ∀r = n, k, u (4) Therefore, bi-lateral, with respect to R&D of G7 (n = North), is de ned as: R&D ns = min 1,R&D s R&D n , here r = n (5) Here, also R&D ns∈ [0,1] with zero implying further away from the invention frontier.

A measure of foreign R&D ow that transmits across borders directly (disembodied, DISEMB r) via effective teledensity—international telephone traf c, broadband penetration, cellular sub- scription encapsulated in telecommunications infrastructure penetration per 100 inhabitants in a region—is also computed to measure bilateral technological proximity. Ours coverage of disembodied ows is derived as: between source ‘n’ and destination ‘s’ : DISEMB ns = R&D d n· dn· ds (6a) between norths : ‘n’ and ‘k’ : DISEMB n= k / = n R&D d n· dn· dk (6b) 9Wang (2007) has measured the elasticity of AC wrt Human capital as quite high for South—3.3 whereas elasticity of TFP wrt Human capital is 1.1 for almost all the regions. Das (2002) explores relationships between AC and TFP. G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 75 Here, dr(∀r) is a composite measure for the development of teledensity such as telephone and internet subscribers penetration obtained from Telecommunications Database (World Telecom- munications ICT Indicator, International Telecommunications Union, ITCU 2008).

Therefore, bi-lateral, with respect to R&D of G7 (n = North) is de ned as: DISEMB ns = min 1,DISEMB ns DISEMB n , (where r = n) (7) Thus, from earlier discussion and speci cation we infer that foreign R&D gives scope for cross- border learning, and subsequent innovations build on cumulative R&D experience. As regards the bilateral trade intensity, following Bergstrand (1985) & Linnemann (1966) , we calculate the bilateral distance between ‘n’ and ‘s’ (D ns) using a simple speci cation as below: Dns = exp − d ns dmax ns (8) where dns= distance between ‘n’ and ‘s’ and dmax ns is the largest absolute distance between all pairs of regions. This formulation scales the differences in binary distances on the unit interval so that the function takes the value ‘unity’ if two countries are nearest, whereas declines exponentially to ‘zero’ if they are farthest. Thus, the lower values indicate farther distant whereas higher values imply geographical proximity.

The cross-country heterogeneity in the perceived effect of technology transfer depends, inter alia, on factors encapsulated in Socio-Institutional Parameter (SIP r), such as rule of law, regulatory constraints, violence, corruption, transparency, labor strikes, governance and social capital. These are important policy-syndrome variables affecting growth and causing reversal of fortunes in LDCs in Asia, Latin America or Africa (Lee and Kim: World Development 2010, Lucas 2009).

However, acceptance of ‘foreign technology’ depends also on social capital, social cohesion and cultural af nity based on network and trust (Dasgupta, 2009 ). We construct such measure by the UN’s human development index (HDI) embracing multi-faceted nature of social acceptance via factors in uencing human capital as well as income characteristics. This is speci ed via bi-lateral proximity parameter HD (0 ≤ HD rs≤ 1): HD rs= min 1,HD s HD r (9) Domestic invention and foreign-sourced technological spillovers depend, inter alia, on domes- tic institutional setting based on political dimensions and good governance, institutional factors like legal side protecting intellectual property rights (IPRs), habits and even languages. Dasgupta (p. 3, 2009) argues that “that a natural place to look for the worth of social capital in macro- economic statistics is “total factor productivity” (TFP). TFP is an amalgam of technology and institutions.” We incorporate these socio-institutional factors via a binary parameter of the index of governance (Kaufmann, Kraay, & Mastruzzi, 2009 ), (GP rs, −1 ≤ GP rs≤ 1) as comparative measure of institutional quality indicator between partners as: GP rs= min 1,GP s GP r (10) If destination ‘s’ has higher GP sthan that of source ‘r’ i.e., GP s> GP r, then it is conducive governance structure for ‘s’. Otherwise, if the client region lags in institutional quality behind the advanced source [i.e., GP s< GP r], then it poses hindrance in ‘s’ even with high TA P. Regarding 76 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 corruption or transparency, we take Transparency International’s (2008/9) Global Corruption Barometer data on Corruption perception Index (0 < Tr< 10). Also, a composite indicator of national competitiveness (1 < Cr< 7) which encapsulates different aspects of a nation’s technolog- ical readiness in terms of socio-economic variables is taken from Global Competitiveness Report (World Economic forum 2008/09). Hence, the socio-institutional parameter (SIP r) is de ned as: SIP r= Tr· Cr· GP r (11) As before, we convert it into a binary parameter [0 < SIP ns< 1] w.r.t. n = G7 as below: SIP ns = min 1,SIP s SIP n (12) Having described the arrays of ‘barriers to adoption and diffusion’ parameters, we specify a binary Learning effects and Distance Function (LeAD ns) between ‘n’ and ‘s’ as: LeAD ns = R&D βnsns · HD φnsns · Disemb ns· Dns ∀s = k, u and n = G7(unique) (13) where Φs= elasticity of TFP/R&D with human capital or education is taken from prior study ( Wang, 2007 , World Bank) and βs= elasticity of TFP with foreign R&D. We set Φs= 1 (Wang, 2007 estimates it 0.99–1.06) and βs= 0.12 (for advanced economies based on CHH 1995) and β s= 0.3 for all emerging and dynamic adopters (Wang, 2007 ), and βs= 0.04 for least developed ones.

For laggards the values will be lower than the dynamic adopters or advanced northern recipients (i.e., LeAD nk> LeAD nu). Now, we de ne the structural congruence index (0 < SCI ns< 1) and Technological congruence Index (0 < TCI ns< 1) parameters between ‘n’ and ‘s’ as: SCI ns = LeAD ns· SIP ns (14) and TCI ns = LeAD ns· TA P ns (15) For the advanced North (here G7 composite), the major source of knowledge or current vintage technology, we de ne a function representing its invention capability. We call it ‘Indigenous, D isembodied and Embodied R&D, and Schooling parameter (IDEAS r) where IDEAS r= R&D r· HDI r· DISEMB r This is, in fact, related to LeAD rvariables. This is, in fact, related and isomorphic to LeAD rvariables. Next we discuss the crucial parameters for technology capture and assimilation—L earning-enabled Absorption Parameter (LeAP)—for North (LeAP n) and north vis-à-vis recipients ‘s’ (LeAP ns) as follows: LeAP n= AC n· ICC n· TA n· GP n· Cn· Tn· IDEAS n (16) and LeAP ns = AC s· ICC ns· TA ns· GP ns· Cns· Tns· LeAD ns, ∀s = k, u = TA P ns· LeAD ns· SIP ns (17) It is different from (13) as the dimensions of Syndrome-variables enter into (13) to determine the extent of capture from accessible technology spectrum, failing which emerges the development G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 77 and growth disorder. Following an exogenous Hicks-neutral technological improvement in one sector of a region (i.e., the ICT sector in the North), all other sectors in the source and destinations experience trade-induced endogenous TFP improvement. We de ne embodiment index [Eijrs ] as the ow of imported intermediate in sector ‘i’ in source region ‘r’ and exported to rms in sector ‘j’ in recipient ‘s’, [Firjs ], per unit of composite intermediate input (domestic as well as composite imported inputs) of ‘i’ used by ‘j’ in ‘s’, [M ijs]. Thus, Eirjs =F ijrs Mij s (18) Spillover coef cient for ‘j’ in destination ‘s’ (γijrs ) is: γijrs (E ijrs , θs) = E1−θ s ijrs (s ∈ {k, u}) (19) γ s(0) = 0, γs(1) = 1, γ s= (1 − θ) · E−θsrs > 0, γ s= −θ s(1 − θs) E1+θ s rs < 0 where θ = LeAP ∀ r ∈ {n, s} and s ∈ {k, u}. The realized productivity level from the potential ows of ‘current technology’ depends on LeAP ns∈ [0, 1], LeAP ns= 1 implies full appropriation. For the destination region ‘s’, θsand Ersjointly determine the value of the ‘Spillover Coefficient’ γs(Ers, θ s), where primes indicate the rst ( ) and the second ( ) derivatives with respect to Ers. Thus, if F irjs indicates usage in region ‘s’ by ‘j’ of imported intermediate ‘i’ from ‘r’, we assume that the share of imported input ‘i’ from source ‘r’ in receiving region ‘s’ holds for all industries ‘j’ in ‘s’ using imported input ‘i’: Firjs Fij s =F irs Fis (20) where Fisis the aggregate imports of tradeable commodity ‘i’ in region ‘s’ from all source regions.

Source ‘r’ reaps technological spillover via inputs embodying technology so that: Eij n =D ij n Mjn, i / = j(21) where Dijn is the domestic tradeable ‘i’ used by j of ‘r = n’. Mjrrepresents domestic production of ‘j’ in ‘r = n’. Given constellation of parameters, higher IDEAS nand LeAP ninduce knowledge- spillover via: γij n(E ij n, LeAP n) = E1−LeAP n ij n (22) θnhas one-to-one correspondence with LeAP n, 0 ≤ LeAP n≤ 1. TFP transmission equation for ‘s’ is: ajs= E1−LeAP ns ijns ain (23) where ain– an exogenous TFP improvement in sector ‘i’ (IT-cluster) of ‘r’ (r = n) – induces endoge- nous TFP changes ajs, translating into induced-innovation in user-clusters. Such a mechanism is invoked via:

afij s = E1−LeAP ns ijns afij n (24) where afijn is ith (unique i ) input-augmenting technical change in ‘j’ sector in ‘n’. 78 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 4.2. Construction of parameters Tables 5 and 6 show the Technology appropriation parameters for each region as well as the binary values vis-à-vis G7 (North = n). Look at the pattern in the Table/s: compare BRIC, although TA is almost similar in the range of 4.1–5.5, but as AC is highest in Brazil and ICC ris higher its TA P parameter value is higher than other 3 in the group. Thus, although ‘available tech’ is comparable, differences in Innovative Capability and Absorptive Capacity make the difference.

It shows the Socio-institutional parameters as described in Section 3. R&D-expenditures data are from published sources (for example, ANBERD or STAN or OECD’s Main Science and Technology Indicators). However, we use UNESCO (2008) and UN’s Human Development Report (2008) (see Tables 5 and 6). Table 7 gives the measures of structural and technological congruence between the north (G7 here) and other destinations, showing that relatively advanced nations and dynamic adopters (Hong Kong, South Korea, BRIC) have higher TCI and SCI values compared to the followers or laggards, for example, MENA, RSA or SSA.

Comparing TCI nsand TAP ns, we see that they are not equal also the ranking is preserved more or less. This is because magnitude of LeAD ns values—encapsulating learning, human develop- ment, research efforts and spillover—impact on the technology appropriability and magnitude of those factors for technology assimilation. In case of structural symmetry, we see that G7 is more congruent to EU and other OECD, as compared to, for example, SSA, India, China, RSA, LAC or MENA. Typically, rapidly industrializing economies, like South Korea and Hong Kong- Taiwan, have higher bilateral SIP values (and hence, SCI) with G7 because of higher indices for the constituent parameters like GP, transparency, and lower Corruption perception index.

Thus, the constituents of the learning and distance function (via Eq. (13) ) determine the extent of congruence—either institutional or technological—between north and the southern recipients.

From Table 7, we observe that usually countries/regions with better enabling factors have higher bi-lateral LeAP and LeAD values, for example, dynamic Asian economies like S. Korea, other OECD and EU vis-à-vis G7. Comparing India among BRIC and LAC, the values for the embodied and disembodied spillover are lower. With low values for HDI, she registers low magnitude for ‘Ideas’ parameter. But, if we compare India with developing South East Asia and Latin America, disembodied spillover is lower in India. HDI values are lower, too. Thus, even with higher embodied spillover LeAD and IDEAS for India are lower than these emerging economies.

It con rms our conjecture that disembodied spillover and better governance and institutions are crucial factor for ‘IDEAS’ to develop. 4.3. Illustrative simulation design: productivity experiment The model is a special tailor-made version of a CGE trade model-GTAP (Hertel, 1997 ). For capturing direct and indirect intersectoral effects based on well-de ned production and demand structure, the CGE model scores over the simplistic input-output speci cation and the Social Accounting Matrix (SAM) based models, and enable us to account for behavioral responses of each representative economic agent in response to relative price changes following policy changes.

Because of our enhancement of theory via technology spillover mechanism, an augmented version of the comparative static multi-regional, multi-sectoral model is solved using General Equilibrium Modeling Package (GEMPACK) (Harrison & Pearson, 1996 ). We just consider 5% Total factor productivity shock in the north, G7. From the current literature on TFP, we see that sources of TFP growth are mainly governed by ICT growth in most sectors (see EUKLEMS database, release March 2008 and Groningen Growth and Development Center (GGDC 2006), and OECD G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 79 Table 5 Region-wise Technology Absorption and Socio-institutional Parameter derived from Eqs.

(1), (2), (11) and (12) in the text. GTAP code GTAP regions Tr, CPI r Cr, GCI r GP r SIP ns TAr AC r ICC r TA P r TA P ns G7 G7-North Developed 7.228571 5.184509 1.31564 1 6.128571 0.724223 0.870286 3.862724 1 OtherEU EU minus G7 4 members 7.745455 4.979621 1.528518 1 6.054545 0.705468 0.853545 3.64574 0.94 Brazil Brazil 3.5 4.227963 0.75 0.22 5.3 0.561541 0.529 1.574394 0.41 Russia Russian Federation 2.1 4.152973 0.28 0.05 4.1 0.466861 0.79 1.512162 0.39 India India 3.4 4.303131 0.59 0.18 5.5 0.314229 0.29 0.501195 0.13 China China 3.6 4.736538 0.38 0.13 4.3 0.279342 0.358 0.430019 0.11 Hkg Tw n HongKong Taiwan 6.9 5.211573 1.128491 0.82 6 0.720604 0.714 3.087069 0.8 SouthKorea South Korea 5.6 5.003964 1.42 0.8 5.9 0.447592 0.839 2.215624 0.57 SouthEAsia Developing Asia 3.27 4.168391 0.49 0.14 4.75 0.353306 0.3612 0.606166 0.16 RSA Rest of South Asia 2.7125 3.645467 0.18 0.04 3.75 0.346921 0.17375 0.226041 0.06 ECA Europe and Central Asia 4.004667 4.068955 0.86 0.28 4.513542 0.426924 0.60328 1.162483 0.3 SouthAfrica South Africa 4.9 4.340151 1.16 0.5 5.5 0.564539 0.55 1.707731 0.44 LAC LatinAmerica&Caribbean 3.640686 3.857779 0.62 0.18 4.138125 0.430442 0.373625 0.665509 0.17 Mexico Mexico 3.6 4.189043 0.76 0.23 4.6 0.617532 0.47 1.335103 0.35 OthrOECD OECD minus G7 minus EU 8.7 5.13939 1.732302 1 6.38 0.698383 0.896 3.992294 1 MENA MiddleEastNorthAfrica 3.356944 4.177408 0.47 0.13 4.995833 0.481806 0.3715 0.89421 0.23 SSA Sub-Saharan Africa 2.984375 3.48093 0.28 0.06 4.207639 0.276345 0.146813 0.170708 0.04 Source:

Author’s calculation by matching with GTAP Regions based on external Datasources; AC (from GTAP), ICC (World Investment Report 2005), TA (Global Competitiveness Report 2008/2009), T (transparency international), GFI (Global competitiveness Report 2009), GP (World Bank 2009).

See text. 80 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 Table 6 R&D ows comprising own R&D and Trade-mediated ows (via Eqs.

(3)–(5) ) with Intra-Regional ows in 3 Composite Norths n1, n2, and n3. GTAP regions n1 = G7 n2 = Other EU n3 = Other OECD Total foreign R&D spilloverOwn R&D Total R&D f+ R&D d Own Domestic Bi-lateral R&D ns North–South Bi-lateral R&D ns TRDFn1s TRDFn2s TRDFn3s Aggr R&D f R&D d Total R&D r 1 G7 0.818344 0.248468 0.040196 1.107009 2.234765 3.341774 1 1 2 OtherEU 1.013433 0.341471 0.035673 1.390577 1.923093 3.31367 0.86 1 3 Brazil 0.917653 0.182791 0.030486 1.130929 0.968866 2.099795 0.43 0.63 4 Russia 0.815808 0.244339 0.020218 1.080364 1.067194 2.147559 0.477542 0.64 5 India 0.499535 0.149897 0.08031 0.729742 0.80374 1.533481 0.359653 0.46 6 China 0.762083 0.076567 0.026729 0.865379 1.44 2.305379 0.644363 0.69 7 Hkg Tw n 0.907377 0.069333 0.039655 1.016365 0.694976 1.711341 0.310984 0.51 8 SouthKorea 0.925408 0.060453 0.046127 1.031989 2.6 3.631989 1 1 9 SouthEAsia 0.73246 0.099003 0.033905 0.865368 0.394665 1.260033 0.176603 0.38 10 RSA 0.528673 0.106296 0.036234 0.671203 0.366886 1.038089 0.164172 0.31 11 ECA 0.831126 0.251604 0.02322 1.10595 0.576384 1.682335 0.257917 0.5 12 SouthAfrica 0.882286 0.166564 0.035765 1.084616 0.919225 2.003841 0.41133 0.6 13 LAC 0.87102 0.141309 0.018794 1.031123 0.270354 1.301478 0.120977 0.39 14 Mexico 1.52725 0.077711 0.011637 1.616598 0.504215 2.120814 0.225623 0.63 15 OthrOECD 1.01026 0.300632 0.042889 1.353781 1.873612 3.227393 0.838393 0.97 16 MENA 0.865824 0.200827 0.037445 1.104097 0.541462 1.645558 0.24229 0.49 17 SSA 0.692301 0.218337 0.019576 0.930214 0.243271 1.173484 0.108857 0.35 Source:

Author’s calculations based on R&D data sources, UNESCO (2009) and UN’s HDR 2008/2009, and Trade shares derived from GTAP Version 7 database. G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 81 Table 7 Values of binary Structural and Technological Congruence Indices, LeAP, LeAD, Disembodied R&D and the elasticity values based on Eqs.

(5)–(7), (9), (13)–(16) . GTAP regionsHDI r HDI rs IDEAS r = RD r· HD r· DISEMB r LEAD rs TCI rs = LEAD rs· TA P rs SCI rs = LEAD rs· SIP rs TCI r = IDEAS r· TA P r SCI r = IDEAS r· SIP r LeAP ns DISEMB nsDns Φrs βrs G7-North 0.948429 1 0.035614 1 1 1 1 0.12 OtherEU 0.943545 0.99 0.038248 0.617547 0.580494 0.617547 0.139441 2.254848 0.580494 1 0.623784 1 0.12 Brazil 0.8 0.84 0.027024 0.162527 0.066636 0.035756 0.042546 0.006573 0.01466 0.526977 0.421747 1 0.3 Russia 0.802 0.84 0.030121 0.249959 0.097484 0.012498 0.045548 0.073555 0.004874 0.572885 0.593837 1 0.3 India 0.62 0.65 0.011277 0.093653 0.012175 0.016858 0.005643 0.097184 0.002191 0.388527 0.468125 1 0.3 China 0.777 0.82 0.026815 0.184067 0.020247 0.023929 0.011531 0.173747 0.002632 0.490363 0.511669 1 0.3 Hkg Tw n 0.937 0.98 0.029296 0.229298 0.183439 0.188025 0.090438 1.188829 0.15042 0.598465 0.478481 1 0.3 SouthKorea 0.921 0.97 0.058989 0.295867 0.168644 0.236694 0.130697 1.130771 0.134915 0.577664 0.52802 1 0.3 SouthEAsia 0.7232 0.76 0.01258 0.107272 0.017163 0.015018 0.007626 0.084021 0.002403 0.452218 0.417243 1 0.3 RSA 0.59525 0.62 0.007731 0.118231 0.007094 0.004729 0.001748 0.013761 0.000284 0.409845 0.487601 1 0.04 ECA 0.82825 0.87 0.023019 0.225121 0.067536 0.063034 0.026759 0.045981 0.01891 0.541157 0.588685 1 0.3 SouthAfrica 0.674 0.71 0.020776 0.112919 0.049684 0.05646 0.03548 0.183865 0.024842 0.503916 0.367879 1 0.3 LAC 0.777063 0.8 0.015994 0.149407 0.025399 0.026893 0.010644 0.139272 0.004572 0.518048 0.47818 1 0.3 Mexico 0.829 0.87 0.027162 0.220763 0.077267 0.050775 0.036263 0.007975 0.017771 0.506068 0.575964 1 0.3 OthrOECD 0.9582 1 0.03819 0.376426 0.376426 0.376426 0.152465 2.958018 0.376426 1 0.377804 1 0.12 MENA 0.7415 0.78 0.018966 0.208292 0.047907 0.027078 0.01696 0.125007 0.006228 0.509182 0.539631 1 0.04 SSA 0.497625 0.52 0.00677 0.087011 0.00348 0.005221 0.001156 0.019691 0.000209 0.379742 0.459537 1 0.04 Source:

Author’s computations based on data constructed as described in the text and in other tables. 82 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 Productivity Database 2008). For simulation, however, we consider GGDC (2006)’s Productivity monitor to get estimated TFP growth in ICT (Table 12, GGDC 2006). 5. Selective simulation results 5.1. Macroeconomic impacts Macroeconomic impacts show that following the TFP-led endogenous spillover via trade- embodiment in all traded sectors (both domestically and abroad), all the regions register higher TFP improvement—although magnitudes differ across regions—owing to differential embodied knowledge-spillover via intermediates (Table 8). We see the largest bene ciary is the other two developed regions, namely, other EU and OECD. They have higher embodiment and spillover coef cients (see Table 9). LDCs like MENA, Mexico, SSA have highest region-wide trade- embodiment index (Eirs). Nevertheless, higher capture-parameter (LeAP nk> LeAP nu) magni es E irs to higher spillover-coef cient, 0.52 and 0.42 by more than 100% compared to Low and middle income economies (LMIE) (varying on an average between 0.01% in case of SSA, MENA, Mexico, India to 25% in case of S Korea, Hong Kong Taiwan).

Despite having low LeAP(θ nu), with post-simulation technological bene ts higher, γirs and Eirs result into higher TFP, exports and GDP growth in LMIE. In G7, principal bene ciary of tech- nological change, highest value of LeAP(θ n) ampli es spillover γirre ected in highest GDP and TFP-growth (rows 1, 2 and 3, Table 8). North, thus, reaps the maximum productivity growth by sourcing relatively high proportion of its own ‘technological improvement-bearing’ input. TFP- growth acts as an export supply shifter for each generic commodity so that output and global trade increases (rows 4 and 5, Table 8).

Regional differences are explained in terms of differences in the economy-wide embodiment index and spillover coef cient (Table 9). As conjectured, G7, EU, other OECD and dynamic adopters of foreign-improvement in technology, like Korea, Taiwan, China, Hkg have registered much higher regional index of technical change as opposed to the relatively laggard, viz., Rest of South Asia, Africa and most of the Latin American countries experiencing modest region- wide TFP performance. This is attributed not only to lower embodiment index, but also to lower capture-parameter thanks to higher R&D-enabled ows, as well as the better institutional set up facilitating higher learning effect and assimilation by these emerging economies.

However, for MENA and SSA, although being outliers in case of having low LeAP values or SIP, Technology Appropriation Parameters, they show better performance in terms of almost all macroeconomic variables reported in Tables 8 and 9. This is due to higher values of trade-led embodiment via traded intermediates. USA, EU, Canada and Japan in G7-being more structurally congruent and hence, along with trade-embodiment-reaped most of the bene ts of own and foreign induced spillover because of highest values of socio-economic parameters, viz., GP, HDI, R&D ows and ICC (see the values in Tables 5–7 ). On the contrary, for India, SSA, MENA, Rest of Africa, and South Africa, these values are considerably low. In fact, for GP values showing poor governance quality, the capture parameter drives the magnitude of structural-congruence (SCI rs) to low (see Tables 5 and 7). Table 9 shows that BRIC has lower SCI rsvalues than those in S. Korea, HKG Twn, and other North; lower SIP ns and LeAD ns caused lower capture values (LeAP ns) so that they were unable to tap the potential spillover in foreign-technology. As conjectured, regions with higher AC s, HD rs, R&D rs, DISEMB rs, GP rsand ICC rshad higher absorption and adoption of technology vehicle via trade, for example, the case of South Korea, Other OECD, Hkg Twn, ECA, etc. In case of G7, the more pronounced TFP-enhancement is attributed to the fact that most G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 83 Table 8 Simulated macroeconomic impact of 5% TFP shock in G7 (North) in ICT sectors and its repercussions in destinations for some selected macroeconomic variables. Variables (% change)G7 OtherEU Brazil Russia India China Hkg Tw n SKorea SE Asia ECA SAfrica LAC Mexico OthrOECD MENA SSA 1 Region-wide Te c Change5.79 4.11 0.53 0.88 0.52 0.59 2.41 1.13 1.40 1.66 1.13 1.28 1.45 3.05 1.92 2.08 2 Regional Income (real)5.00 3.11 0.34 0.87 0.42 0.66 2.20 1.03 1.48 1.45 0.99 1.20 1.48 2.56 1.85 2.02 3 Real GDP 4.44 2.79 0.36 0.79 0.51 0.64 2.08 1.07 1.29 1.31 0.89 1.10 1.33 2.30 1.61 1.76 4 Regional exports (volume)1.71 3.22 5.55 1.96 5.64 3.33 2.33 3.63 2.16 2.76 3.12 3.61 3.42 3.59 2.55 2.85 5 Regional imports (volume)4.11 2.38 −1.30 0.21 −0.82 0.24 1.19 0.75 0.82 0.70 −0.10 0.42 1.38 1.64 0.64 1.05 6 Regional export Price Index−3.31 −3.44 −3.90 −3.26 −3.89 −3.51 −3.51 −3.74 −3.41 −3.43 −3.48 −3.48 −3.42 −3.49 −3.33 −3.34 7 Regional import Price Index−3.41 −3.37 −3.39 −3.39 −3.37 −3.43 −3.40 −3.35 −3.41 −3.37 −3.42 −3.44 −3.38 −3.39 −3.41 −3.41 8 Terms-of-Trade 0.11 −0.07 −0.53 0.14 −0.54 −0.09 −0.11 −0.40 −0.01 −0.06 −0.06 −0.04 −0.05 −0.10 0.08 0.07 9 Welfare (in US $ million)1,154,495 99,822 1839 4719 2483 9938 9397 6163 10,072 15,709 2342 9283 8938 30,764 18,471 4510 10 Change in trade Balance ($Usmln)−99,537 11,322 5414 1310 6892 17,044 1960 5980 5844 12,650 2510 6812 3614 6491 7531 1746 11 Per capita real income (HHLD)5.00 3.11 0.34 0.87 0.42 0.66 2.20 1.03 1.46 1.45 0.99 1.20 1.48 2.56 1.85 2.01 12 Regional investment demand5.12 3.16 0.33 0.90 0.57 0.69 2.22 1.06 1.55 1.54 0.99 1.22 1.86 2.57 1.91 2.10 Source:

Author’s simulation results.

Except welfare and change in trade balance, gures present post-simulation %-deviation from base case scenario. 84 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 Table 9 Values of economy-wide spillover, embodiment indexes and constituents of learning and effective capture parameters. G7 OtherEU Brazil Russia India China Hkg Tw n SKorea SEAsia ECA SAfrica LAC Mexico OthrOECD MENA SSA Aggregate Embodiment 0.092662 0.241771 0.076912 0.138281 0.093122 0.108811 0.326736 0.146455 0.244025 0.249281 0.173661 0.220713 0.24376 0.268228 0.294053 0.323586 Aggregate Spillover 0.803137 0.515655 0.079615 0.139549 0.093516 0.109389 0.381974 0.184688 0.244789 0.255421 0.180978 0.222078 0.248918 0.421236 0.29607 0.323653 LeAD ns 1 0.617547 0.162527 0.249959 0.093653 0.184067 0.229298 0.295867 0.107272 0.225121 0.112919 0.149407 0.220763 0.376426 0.208292 0.087011 LeAP ns= θ 0.95559 0.580494 0.01466 0.004874 0.002191 0.002632 0.15042 0.134915 0.002403 0.01891 0.024842 0.004572 0.017771 0.376426 0.006228 0.000209 SIP ns 1 1 0.22 0.05 0.18 0.13 0.82 0.8 0.14 0.28 0.5 0.18 0.23 1 0.13 0.06 TCI rs 0.580494 0.066636 0.097484 0.012175 0.020247 0.183439 0.168644 0.017163 0.067536 0.049684 0.025399 0.077267 0.376426 0.039797 0.002649 AC r 0.724223 0.705468 0.561541 0.466861 0.314229 0.279342 0.720604 0.447592 0.353306 0.426924 0.564539 0.430442 0.617532 0.698383 0.481806 0.276345 Source:

Author’s calculation based on databases and adapted for GTAP databases regional aggregation as documented in the text. G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 85 of the spillover is domestically sourced with indirect spillover with the trading partners, namely, EU, Japan, High-income Asia and Canada (a NAFTA member). Table 8 (3rd item) shows that, region by region, the overall technical change translates into approximately similar percentage increase in real GDP at factor cost. As the shock is factor-neutral, with xity of regional supplies of all the components of value-added, the percentage deviation in real GDP at factor cost is almost equal to the respective region-wide TFP changes.

Real income increases in all the scenarios in all regions, caused by relatively pronounced TFP-gains (row 2, Table 8). More predominant effect occurs in USA, Canada, LAEEX, EU, Brazil and high-income Asia as they experience higher doses of trade-induced spillovers. For Canada and Mexico, since they belong to NAFTA with USA as the hub, the induced spillover is more dominant. However, in case of South Korea, the trade-induced spillover is high due to higher trade in ICT, biotech intermediates and higher AC and LeAP rs, SCI rs. After the technology shocks, there are changes in price relativities across regions which induce changes in regional TOT (row 8, Table 8), perturbing the pattern of inter-regional competition. This indicates that due to technological bene ts there is substantial cost reductions leading to decline in export price indexes in all the regions—the extent of fall depending on the magnitude of technology transmission and its actual capture by the sectors in destinations. From Tables 8 and 10, we observe that aggregate regional export price indexes fall in almost all the regions with more fall observed in the major bene ciaries of such spillover and its higher capture such as Canada, Argentina, Japan, China, Brazil, India, and EU, compared to SSA, RoAFR, MENA and RoSAsia, bene ting from cost-reduction technological inventions. This resulted in increase in volume of regional exports (trade as a whole)—see rows 4 and 5 in Table 8. In case of Brazil, India, other Asian countries considered, and Africa, regional imports shrank a bit due to increase in own production via TFP-gains, acting as an export-supply shifter for each generic commodity. Not surprisingly, all the countries have a positive change in trade balance, with degrees differing owing to differences in realization of trade-induced bene ts. Regional investment demand, following the technological bene ts augmented with lowest impact in India and China—this is surprising, but given the base period data (2004), it could be attributed to the post-simulation lesser TFP growth bene ts accrual—caused by lower sets of host of parameter values. The aggregate spillover index gives us an average overall magnitude of technology appropriated by all user sectors in G7 as well as host regions from the ICT sector via intermediate inputs. From Table 9, it is evident that the aggregate spillover index in G7 is highest and domestic spillover is higher than the LDCs.

The capture-parameter (θn) in G7 is higher than θsin all the destinations, so that it reaps the maximum spillover (γir) compared to most of the LDCs and other EU, and OECD. For EU, Japan and BRIC, although the values of Eirs are of the same order of magnitude, the aggregate spillover coef cient (γirs) for EU and OECD is of much higher magnitude than in most of the LDCs in South America and South Asia and the composite African regions. This is because the higher value of the capture parameter (θr) magni es the value of the embodiment index and hence enables them to record a much higher rate of TFP improvement. From Tables 8 and 9, it is evident that in conformity with our theory the relatively laggard and less congruous regions like Argentina, Brazil, South Asia, Mexico and Africa register moderate growth effects. Note that the ordering of the spillover coef cient in Table 9 matches the ordering of the real GDP results in Table 8. Regarding Welfare effects, in all the simulations it leads to welfare-augmentation with much higher welfare improvement in case of concomitant productivity improvement of ICT. It is true in most of the regions exception being South Asia, Africa and Thailand capturing less trade-induced bene ts due to lower capture (Table 8, Row 9). 86 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 Table 10 Simulated impact on output supply and supply prices of selected sectors under the designated scenario along with the embodied spillover coef cients, TFP and sectors’ total exports. G7 OtherEU Brazil Russia India China Hkg Tw n SKorea SEAsia SAfrica LAC Mexico OthrOECD MENA SSA Output supply ElectronicIT 1.684141 1.564169 1.515986 −0.06163 0.914797 1.363866 0.660896 3.082189 1.464043 −0.70159 0.750626 1.719393 2.792393 1.13524 1.07523 Nano Matrls 3.486675 3.827113 2.294131 1.293233 1.626885 1.362416 2.092595 1.837482 1.840176 2.31871 2.912059 1.925469 3.197725 1.94574 2.669626 TransportTec 4.08713 2.127445 0.868632 −0.17769 −0.96964 −0.31345 1.074714 0.806255 −0.45266 −1.05208 −0.53035 1.440262 0.83883 0.026598 0.984384 Svces 4.507883 2.846243 −0.13396 0.576299 0.256751 −0.00123 2.231367 0.577624 1.038605 0.737413 0.678662 1.322104 2.154094 1.530571 1.408554 CGDS 4.165318 0.199055 −5.53456 −2.4364 −3.6582 −2.41853 −2.20215 −3.30653 −3.91163 −5.67351 −3.58932 −1.87392 −1.14356 −2.79036 −2.56045 Supply prices ps AgBioTech−3.46632−3.50879 −3.61378 −3.31781 −3.21403 −2.99942 −3.36392 −2.85037 −3.41872 −3.5853 −3.51019 −3.47396 −3.4925 −3.51044 −3.56526 ElectronicIT −3.38304 −3.47757 −4.34954 −3.50561 −4.11489 −3.62464 −3.47068 −3.89258 −3.54901 −3.46891 −3.63969 −3.50858 −3.70436 −3.55024 −3.42163 Nano Matrls −3.16046 −3.38083 −3.7189 −3.16953 −3.63847 −3.43692 −3.31602 −3.38402 −3.27029 −3.41994 −3.49844 −3.47162 −3.30856 −3.19095 −3.3089 TransportTec−3.75058 −3.58604 −3.96537 −3.59393 −4.19491 −3.57907 −3.4815 −3.69962 −3.41774 −3.44339 −3.52202 −3.67894 −3.52518 −3.68159 −3.51473 Svces −3.09876 −3.42432 −4.10873 −3.36418 −4.24096 −3.58843 −3.61997 −3.8152 −3.40628 −3.47987 −3.33115 −2.73285 −3.52303 −3.66381 −3.27985 CGDS −3.24948 −3.4566 −4.05061 −3.40775 −4.10306 −3.57985 −3.56683 −3.78003 −3.4578 −3.49012 −3.42108 −3.36055 −3.52785 −3.60615 −3.38989 Commodity-wise regional exports qxw ElectronicIT 0.314352 1.417106 8.093122 0.149507 6.742657 2.98725 0.724954 4.358434 1.658072 0.305252 2.508863 2.736252 3.459097 1.626718 1.181503 Nano Matrls 2.101226 4.264537 5.842737 1.895737 5.000992 4.255849 2.012942 2.800885 2.286862 4.045146 4.795204 5.313208 3.677576 2.120432 2.949326 TransportTec 3.179523 2.315035 3.54658 −0.51263 4.721197 1.838908 1.065863 2.003881 −0.57672 0.658994 0.214333 4.160388 0.843385 0.981344 1.226547 Metal MedTec 1.925837 2.837139 6.54308 1.870188 5.442841 2.861921 2.299726 2.719613 1.959651 2.747458 1.260477 −0.44785 2.761592 1.932792 1.183893 Svces 2.104148 3.455882 6.058083 3.066243 6.488312 4.012369 3.902191 4.149473 3.303499 3.776216 3.069859 1.319321 3.790215 4.471451 2.984968 Sectoral TFP (%) changes avaall AgBioTech 4.127485 2.875599 0.329549 0.671444 0.000666 0.182098 1.594373 0.236366 1.468324 1.17698 1.389678 1.457524 2.199038 1.490045 1.819754 ElectronicIT 5 3.071155 1.350795 1.025996 1.208258 0.996439 1.865391 1.745403 1.721374 0.878606 1.720621 2.250799 2.84877 1.6564 1.957118 Nano Matrls 4.11206 2.957302 0.334545 0.73672 0.323689 0.611827 1.67186 0.464999 1.225114 1.120991 1.722745 2.108583 2.290256 1.422197 2.113146 TransportTec 4.460036 2.896465 0.353648 1.374397 1.020911 0.554007 1.687535 0.820315 1.05413 0.449754 1.007364 1.56859 2.137932 1.849614 1.975464 Metal MedTec 4.296549 2.705159 0.343764 0.749156 0.251708 0.295329 1.921561 0.68097 1.420476 1.010909 0.544202 0.291073 2.295599 1.520918 2.022779 Svces 4.19724 2.827511 0.416733 0.725229 0.768205 0.65467 2.143546 1.047371 1.294663 0.894721 0.936262 0.776027 2.285038 1.811285 1.624791 Sectoral spillover coefficients SPLCOEFFT AgBioTech 0.733227 0.581898 0.068558 0.137671 0.000139 0.037548 0.325582 0.048799 0.30032 0.240755 0.284831 0.297628 0.447308 0.304668 0.370944 ElectronicIT 0.922728 0.620871 0.278202 0.209972 0.248436 0.204484 0.380441 0.356643 0.351606 0.179986 0.352006 0.457731 0.577265 0.338383 0.398692 Nano Matrls 0.802921 0.59819 0.069593 0.151003 0.067142 0.125839 0.341281 0.095857 0.250898 0.229365 0.352437 0.429125 0.465615 0.2909 0.430166 TransportTec 0.859164 0.58606 0.073554 0.280754 0.210304 0.113985 0.344455 0.168696 0.216078 0.09233 0.206916 0.320123 0.435033 0.377467 0.402395 Metal MedTec 0.828311 0.547869 0.071505 0.153542 0.052249 0.060856 0.391794 0.140173 0.290607 0.206954 0.112074 0.0598 0.466687 0.310929 0.411942 Svces 0.912772 0.572303 0.086621 0.148657 0.158644 0.134617 0.436601 0.215054 0.265044 0.183273 0.192389 0.159029 0.464569 0.36972 0.331502 CGDS 0.937794 0.609238 0.277299 0.312476 0.116009 0.185398 0.435001 0.296036 0.335883 0.275088 0.373101 0.158793 0.541632 0.439147 0.487766 Source:

Author’s simulations.

Except spillover coef cients, the induced productivity effects are post-shock % changes from base-case. G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 87 5.2. Sector-wise effects It is evident that spillover indexes depend on the source and user sector-speci c trade- embodiment index via the Equations (Section 3). The spillover coef cients for ICT, heavy manufacturing and services are higher in G7 and other DCs than those in the less developed host for these industries (Table 10). These technological improvements translate into productivity gains leading to increased supply in all regions except for MENA, some LAC and SSA. As dis- cussed before, this is attributed to lower technological spillover, lower capture and lack of R&D causing less-pronounced TFP-effects (Tables 7 and 10).

However, in all scenarios productivity gains and capture has larger impact on supply prices to go down—the impact differentials being caused by disparate TFP-augmentation and its assimilation.

The changes in price relativities coupled with the Armington (1969) speci cation of commodity substitution triggers inter-regional competition. For the global economy as a whole, in all the scenarios there has been an increase in the quantity index of world trade by almost 2.6%. In effect, supply prices for all the commodities fall in USA and other developed economies whereas for those experiencing lesser transmitted spillover bene ts the gains are limited. Concomitantly, output response is strongly positive in all the regions except very few negligible negative impacts due to general equilibrium competition among regions and sectors. Not surprisingly, India, Rest of South Asia, Vietnam, SSA, ROAFR, MENA, being laggard in technology capture and lower transmission, register modest increase in supply. As expected, we see that this has been gov- erned by the magnitude of the embodiment and spillover coef cients and uneven distribution of productivity enhancements across sectors. A glance at Table 10 reveals that the impact of the tech- nological improvement is not as uniform across sectors and other regions although the direction of change matches expectation, largely governed by the magnitude of the sector-wise spillover coef cients—in uencing differentials in commodity-wise regional export performances. This can be ascribed-one-to-one correspondence—to the differentials in the bi-lateral sectoral embod- iment indexes [Eirjs ]. Largest bene ciaries among the developing economies are South Korea, Hkg Twn, South Africa, Mexico with BRIC being relatively unsuccessful due to lower spillover.

As expected, cost-reducing spillovers cause rise in regional exports (Tables 8 and 10). 6. Conclusion and policy insights Empirical growth literature—based on Solovian and endogenous Lucas-Romer theoretical paradigms—has embarked upon policy-recommendations for catching-up and convergence of economies with per capita income differences, and brings into light the role of economic policy mix for successful transition to growth. Complementarities between economic policy and structural features such as institutions, nancial development, infrastructure, governance as well as R&D- policy, education are crucial for growth process (Calderón & Fuentes, 2012 ). Hence, the public policies to stimulate structural change toward investments in R&D, human capital, infrastructure, as well as trans-border technology diffusions are necessary for growth and avoiding middle- income trap (MIT). According to the New Growth Theory framework (Barro & Sala-I-Martin, 1995; Lucas, 1998 ), the key requirements of technology-driven development are not simply global integration and new knowledge as economic development requires education, combinations of technical skills, and a whole series of institutions, networks, and capabilities that enable the effective use of existing knowledge, all of which must be part of, or even precede, any serious effort to create new knowledge (Das, 2002, 2010; World Bank, 2010 ). 88 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 Along this maintained hypothesis, in this paper, we construct an empirical general equilib- rium model (CGE) to highlight the role of skill, institutions, innovation policy for assimilating transferred technology, and avoiding middle-income trap via catch-up. The paper designs a model where a change in policy facilitating trade, technology, different propensities and heterogeneous assimilation to adopt frontier technology cause relative convergence and diver- gence of regions to the world technology frontier. Based on evidences of North–South and South–South trade patterns, the results well accord with our a priori expectations: (i) in general, North–South and South–South technology diffusion embodied in traded goods have a positive impact on TFP; (ii) conducive economic policies make the emerging economies superior in the league in acquisition of spillover bene ts, while the policy bottlenecks roadblock further development of the poor economies in Africa or Asia; (iii) trade and associated enabling factors are crucial for enrichment of recipients. The result supports the conjecture that learning and effective assimilation—attributed to governance, R&D, institu- tions, technological capability, human capital, skill—are important for harnessing North–South transmission of knowledge. The model shows that policies targeting investments in R&D, human and non-human capital, structural diversi cation are important for closing the dif- ferences in income gap, but for economies lacking resources for such allocation policy makers should also enable technology dissemination and assimilation for conditional conver- gence.

Thus, as expected, a proper combination of local technology, socio-institutional structure and human capital-induced skill intensity, and indigenous inventive capability (R&D) is sine-qua-non for effective absorption of current vintage technology. For jumpstarting take-off, the LDCs need appropriate economic policies focusing on these factors that can initiate the take-off into sustained growth, without reversal of fortunes (Bluedorn et al., 2014 ). It depends not only on trade potential, but also importantly on other factors like human capital, research capability or inventive capacity, and institutions. These depend also on nurturing domestic usability and absorption capacity to harness foreign-improvement in technology, which is couched in terms of socio-structural fea- tures. According to OECD (2006b) , ICT has an important role for facilitating interconnectedness and convergence of diverse technological applications via spin-off effects, skills and compe- tencies requisite for appropriating the bene ts from ICT-use. Several evidences—for example, in case of South Korea, Taiwan, Singapore, Japan and lately, the BRIC—exist supporting the role of integrated economic policy in promoting these candidate factors for trade-led technol- ogy assimilation and thus, avoiding MIT (Lee, 2013; Ojha, Pradhan, & Ghosh, 2013; Vivarelli, 2014 ).

The policy lesson is that boosting long-term investments in the spheres of human capital, skill formation, technological infrastructure, better governance, institutions, science and technol- ogy policy for innovation capture, and global trade and nancial linkages matter immensely for catching-up (Salvatore, 2004, 2007 ). The CGE result elicits the importance for catalyz- ing TFP-improvement via R&D, education, enhancing better governance and development of logistics infrastructure for making transition to a syndrome-free, policy regime. A compre- hensive package of policy response beyond trade policy is required. Innovation policy goes beyond the science and technology policy per se, as with global trade, these domestic fac- tors could work aplomb for output and employment growth, and improving welfare. Thus, the model provides a conceptual framework for public support policies in the evolution of international competitiveness, technological innovativeness, and inclusiveness for economic development. G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 89 References Amiti, M., & Konings, J. (2007). Trade liberalization, intermediate inputs, and productivity: Evidence from Indonesia.

American Economic Review, 97(3), 1611–1638. Armington, P. A. (1969). A Theory of Demand for Products Distinguished by Place of Production. IMF Staff Papers, 16, 159–178. Arora, V. , & Vamvakidis, A. (2004). The impact of U.S. economic growth on the rest of the world: How much does it matter? Journal of Economic Integration, 19(March), 1–18. Arora, V. , & Vamvakidis, A. (2005). How much do trading partners matter for economic growth? IMF Staff Papers, 52(April), 24–40. Asian Development Bank. (2014, August). Innovative Asia: Advancing the knowledge-based economy – The next policy agenda. Manila, Philippines. Asian Development Bank. (2014, August). Creative Productivity Index: Analysing creativity and innovation in Asia. Barro, & Sala-I-Martin. (1995). Economic Growth. McGraw-Hill. Bayraktar-Sa glam, B., & Yetkiner, H. (2014). A Romerian contribution to the empirics of economic growth. Journal of Policy Modeling, 36, 257–272. Bengoa, M., & Sanchez-Robles, B. (2005). Policy shocks as a source of endogenous growth. Journal of Policy Modeling, 27, 249–261. Bergstrand, J. H. (1985). The Gravity Equation in International Trade: Some Microeocnomic Foundations and Empirical Evidence. Review of Economics and Statistics, 67(3), 474–481. Bitzer, J., & Kerekes, M. (2008). Does foreign direct investment transfer technology across borders? New evidence.

Economics Letters, 100(3), 355–358. Bluedorn, J., Duttagupta, R., Guajardo, J., & Mwase, N. (2014). What underlies the recent growth comeback in developing economies? Journal of Policy Modeling, 36, 717–744. Calderón, C., & Rodrigo Fuentes, J. (2012). Removing the constraints for growth: Some guidelines. Journal of Policy Modeling, 34, 948–970. Caselli, F. , & Coleman, W. , II. (2006). The World Technology Frontier. American Economic Review, 96, 499–522. Clark, D. P. , High lle, J., de Oliveira, J., & Rehman, S. S. (2011). FDI, technology spillovers, growth, and income inequality: A selective survey. Global Economy Journal, 11(2) http://dx.doi.org/10.2202/1524-5861.1773 . Article 1 Coe, D. T. , Helpman, E., & Hoffmaister, A. (2008). International R&D Spillovers and Institutions. IMF Working Paper WP/08/104. Cohen, W. M., & Levinthal, D. A. (1989). Innovation and learning: The two faces of R&D. Economic Journal, 99(September), 569–596. Collier, P. (2009). Wa r s , guns, and votes: Democracy in dangerous places. Harper Collins. Cosar, A. K. (2011). Human capital, technology adoption and development. The B.E. Journal of Macroeconomics, 11(1 (Contributions)). Article 5. Das, G. G. (2002). Trade, technology and human capital: Stylized facts and quantitative evidence. World Economy, 25(2), 257–281. Das, G. G. (2010). How to reap the induced technological bonus? A mechanism and illustrative implementation. Modern Economy (Scientific Research), 1(2), 80–88. Dasgupta, P. (2009). A matter of trust: Social capital and economic development. In Annual bank conference on develop- ment economics (ABCDE) June 2009, Seoul, (pp. 1–47). Diao, X., Rattso, J., & Stokke, H. E. (2005). International spillovers, productivity growth and openness in Thailand: An intertemporal general equilibrium analysis. Journal of Development Economics, 76(2), 429–450. Duval, R., & de la Maisonneuve, C. (2010). Long-run growth scenarios for the world economy. Journal of Policy Modeling, 32, 64–80. Eaton, J., & Kortum, S. (1996). Trade in ideas: Patenting and productivity in the OECD. Journal of International Economics, 40, 251–278. Foster-McGregor, N., Isaksson, A., & Kaulich, F. (2014). Importing, exporting and performance in sub-Saharan African manufacturing rms. Review of World Economics, 150(2), 309–336. Frankel, J. A. (1997). Regional Trading Blocs in the World Economic System. Washington, DC: Institute for International Economics. Goldberg, I., Branstetter, L., Goddard, J. G., & Kuriakose, S. (2008, June 20). Globalization and Technology Absorption in Europe and Central Asia the Role of Trade, FDI, and Cross-border Knowledge Flows. World Bank Working Paper No. 150. 90 G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 Grace, J., Kenny, C., & Qiang, C. Z.-W. (2004). Information and communication technologies and broad-based develop- ment: A partial review of the evidence. Washington, DC: The World Bank. Griliches, Z. (2000). R&D, education and productivity: A retrospective. Cambridge, MA: Harvard University Press. Harrison, W. J., & Pearson, K. R. (1996). Computing solutions for large general equilibrium models using GEMPACK.

Computational Economics, 9(2), 83–127. Hatzichronoglou, T. (1997). Revision of the High Technology Sector and Product Classification. OECD Science, Tech- nology and Industry Working Papers, 1997/02. OECD Publishing. http://dx.doi.org/10.1787/134337307632 Herrerias, M. J., & Orts, V. (2013). Capital goods imports and long-run growth: Is the Chinese experience relevant to developing countries? Journal of Policy Modeling, 35, 781–797. Hertel, T. W. (Ed.). (1997). Global trade analysis: Modeling and applications. Cambridge, MA: Cambridge University Press. Hoekman, B., & Smarzynska Javorcik, B. (2006). Global integration and technology transfer. Palgrave Macmillan/World Bank. Jones, B. F. (2008, June). The Knowledge Trap: Human Capital and Development Reconsidered. NBER Working Paper No. W14138. Jones, C. I., & Romer, P. M. (2010). The new Kaldor facts: Ideas, institutions, population, and human capital. American Economic Journal: Macroeconomics, 2(1), 224–245. Jorgenson, D. W. , & Vu , K. M. (2013). The emergence of the new economic order: Growth in the G7 and the G20. Journal of Policy Modeling, 35, 389–399. Jorgenson, D. W. , & Vu , K. M. (2011). The rise of developing Asia and the new economic order. Journal of Policy Modeling, 33, 698–716. Kaufmann, D., Kraay, A., & Mastruzzi, M. (2009, June). Governance Matters VIII: Aggregate and Individual Governance Indicators for 1996–2008. World Bank Policy Research Working Paper WPS 4978. Keller, W. (2004). International technology diffusion. Journal of Economic Literature, XLII, 752–782. Klein, L., & Salvatore, D. (2013). Shift in the world economic center of gravity from G7 to G20. Journal of Policy Modeling, 35, 416–424. Kosempel, S. (2007). Interaction between knowledge and technology: A contribution to the theory of development.

Canadian Journal of Economics, 40(4), 1237–1260. Krugman, P. (1991). Increasing returns and economic geography. Journal of Political Economy, 99(3), 483–499. Krugman, P. (1995). Development. In Geography and Economic Theory. Cambridge, Mass.: MIT Press. Lee, G. (2005). Direct versus indirect international R&D spillovers. Information Economics and Policy, 17, 334–348. Lee, K. (2013). Schumpeterian analysis of economic catch-up: Knowledge, path-creation, and the middle-income trap.

UK: Cambridge University Press. Lall, S. (2001). New technology, competitiveness, and poverty reduction. In Asia Pacific forum on poverty: Reforming policies and institutions for poverty reduction. Asian Development Bank. Linnemann, H. (1966). An econometric study of international trade flows. Amsterdam: North-Holland. López-Puyeo, C., & Mancebón, M.-J. (2010). Innovation, accumulation and assimilation: Three sources of productivity growth in ICT industries. Journal of Policy Modeling, 32, 268–285. Lucas, R. E., Jr. (1998). On the mechanics of economic development. Journal of Monetary Economics, 22, 3–42. Lucas, R. E., Jr. (2009a). Trade and the diffusion of the Industrial Revolution. American Economic Journal: Macroeco- nomics, 1(1), 1–25. Lucas, R. E., Jr. (2009b). Ideas and growth. Economica, 76(1), 1–19. Mazzucato, M. (2013). The entrepreneurial state—Debunking public vs. private sector myths. Anthem Press. OECD. (1997). Science, Technology and Industry Scoreboard of Indicators. Paris. OECD. (2000). Science, Technology and Industry Scoreboard of Indicators. Paris. OECD. (2003a). OECD Science, Technology and Industry Scoreboard. Paris. OECD. (2003b). OECD Electronic Commerce Business Impact Project: Summary of the methodology for assessing the dynamics and impacts of electronic commerce. Paris. OECD. (2004). OECD Information Technology Outlook 2004. Paris. OECD. (2005). OECD Science, Technology and Industry Scoreboard – Towards a Knowledge-based Economy. Paris. OECD. (2006a). OECD Science, Technology and Industry Outlook – Highlights. Paris. OECD. (2006b). OECD Information Technology Outlook 2006 – Highlights. Paris. OECD. (2010a). Perspectives on Global Development 2010: Shifting Wealth. Paris. OECD. (2010b, May). How imports improve productivity and competitiveness. Paris. OECD (2014, July). Policy challenges for the next 50 years. OECD Economic Policy Paper. No. 9. Paris. G.G. Das / Journal of Policy Modeling 37 (2015) 65–91 91 Ojha, V. P. , Pradhan, B. K., & Ghosh, J. (2013). Growth, inequality and innovation: A CGE analysis of India. Journal of Policy Modeling, 35, 909–927. Onyeiwu, S. (2011, March). Does lack of Innovation and Absorptive Capacity Retard Economic Growth in Africa?

UNU-WIDER Working Paper No. 2011/19. ISSN 1798-7237. Parente, L. S., & Prescott, E. C. (1994). Barriers to technology adoption and development. Journal of Political Economy, 102, 298–321. Qiang, C. Z.-W. (2007). China’s information revolution: Managing the economic and social transformation. Washington, DC: The World Bank. Richardson, J. W. (2011). Technology adoption in Cambodia: Measuring factors impacting adoption rates. Journal of International Development, 23, 697–710. Santos-Paulino, A. U., & Wan, G. (2010). Southern engines of global growth. Oxford University Press. Salvatore, D. (2004). Growth and poverty in a globalizing world. Journal of Policy Modeling, 26, 543–551. Salvatore, D. (2007). Growth, international inequalities, and poverty in a globalizing world. Journal of Policy Modeling, 29, 635–641. Schiff, M., & Wang, Y. (2006). North–South and South–South trade-related technology diffusion: An industry-level analysis of direct and indirect effects. Sen, A. (2004). Development as freedom. Oxford University Press. Spence, M. (2011). The next convergence: The future of economic growth in a multispeed world. Farrar, Strauss and Giroux. Szirmai, A., Gebreeyesus, M., Guadagno, F. , & Verspagen, B. (2013). Promoting productive employment in Sub-Saharan Africa: A review of the literature. UNU-MERIT Working Paper ## 2013-062. Tang, L., & Koveos, P. E. (2008). Embodied and disembodied R&D spillovers to developed and developing countries.

International Business Review, http://dx.doi.org/10.1016/j.busrev.2008.03.002 UNESCO. (2009, September). A global perspective on research and development. UNESCO Institute for Statistics. UN Fact Sheet, October 2009, No. 2. United Nations. (2010, October). Millennium Development Goal 8, The Global Partnership for Development at a Critical Juncture. MDG Gap Task Fo r c e Report 2010. Vita, G. D. (2013). The TRIPS agreement and technological innovation. Journal of Policy Modeling, 35, 964–977. Vivarelli, M. (2014, April). Structural change and Innovation as Exit strategies from the Middle Income Tra p. IZA DP No. 8148. Wang, Y. (2007). Trade, human capital, and technology spillovers: An industry-level analysis. Review of International Economics, 15(2), 269–283. World Bank. (1998). Knowledge for development. World Development Report 1998/9. New York: Oxford University Press. World Bank. (2008). Global economic prospects: Technology diffusion in the developing world. Washington, DC. World Bank. (2010). Innovation policy: A guide for developing countries. Washington, DC: The World Bank.