Liberalisation and Technological Accumulation Strategy in Industries from Developing Economies: An Analysis of Indian Capital Goods Manufacturing Firms

Abstract

Technological accumulation is a complex process of correct mix of indigenous and/or imported technology, mainly for the firms belonging to developing economies after opening their economies. To understand these issues, the case of Indian capital goods industry is taken for the years 1994–1995 to 2015–2016. It was found that only 39% of the firms belonging to Indian capital goods sector are technologically active, that is, resorting to either embodied or disembodied technology acquisition. Multinomial logit model is estimated to find the impact of various variables in determining the strategy of technology accumulation. The factors such as age, size, technology spillovers and outward foreign direct investment were found to have a positive impact on the firms to resort to ‘indigenous R&D only’ as a strategy of technology accumulation. While ‘disembodied technology import only’ is influenced by factors like embodied and disembodied technology imports. However, strategy ‘both indigenous R&D and disembodied technology import’ is found to be influenced by foreign equity participation, mergers and acquisitions.

Keywords

Strategies of technology accumulation ‘make or/and buy’ strategy Indian Capital Goods sector outward foreign direct investment embodied and disembodied technology imports

I. Introduction

Technological accumulation is widely acknowledged (Nelson & Winter, 1982; Rosenberg, 1982) as a determining factor of long-run growth and sustainability in integrated world economies with fast changing technological paradigms. Firms across countries are competing continuously to increase their share in regional or global value chains. The increasing participation in value chains reflects the underlying upgrading process that could be product upgrading, process upgrading, functional upgrading or inter-chain upgrading (Pietrobelli & Rabellotti, 2011) which is a complex process of inter-active association of numerous factors and actors (Jurowetzki et al., 2018). ‘Learning’ is at the centre that evolves the ‘system of innovation’ (Freeman, 1987; Lundvall, 2016) overtime accompanying increase in productivity, employment opportunities (Edquit, 2011) competitiveness and innovative capabilities (Lee, 2013). The core of evolutionary transformation is continuous accumulation of ‘technological capabilities.’ With endogenous growth models (Lucas, 1988; Romer, 1986) in open economies (Grossman & Helpman, 1991), the concerns for understanding the nature and factors determining innovation gained inter-disciplinary perspective with the work of Freeman (1987); Nelson and Winter (1982) and Lundvall (1992, 2016). The developed countries in global North usually have more resources for building technological capabilities as compared to their South counterparts. The industrial revolution 4.0 is also bringing unknowing challenges for growth and employment opportunities, mainly in firms belonging to the global South. Moreover, the re-emergance of ‘protectionist’ policies (Akyuz, 2009) to ‘bring-manufacturing back’ (Aiginger & Rodrik, 2020) is also changing the production system. After WWII, many developing countries too adopted protectionist policies as a strategy of ‘self-reliance’ that started fading away in the mid of 1980s and early 1990s when the Washington Consensus took the centre stage with models depicting international technology spillovers (Keller, 2002) and productivity increase (Pack, 1988), which also showed contrary outcomes (Stiglitz, 2002). Henceforth, it remain inconclusive as how technologies get accumulated in firms, mainly in developing countries. Broadly, Classical approach revolves around deciding whether the firm ‘make’ or ‘buy’ innovation (Veugelers, 1997). The firms can also combine the two internal and external sources of innovation as ‘make and buy’ strategy (Cohen & Levinthal, 1989). Here, ‘making’ innovation means investing in Research and Development (R&D) while ‘buying’ connotes the direct technology purchase from other firms through licensing and royalty payments. Although, apart from the above direct means of technology accumulation, there are numerous indirect means and channels of technology flows among different firms including the purchase of Capital Goods such as machineries, components and parts embodying technologies,¹ movement of employees, access to published materials, attending conferences and fairs, interacting with different actors among others. It is assumed that in the initial stages, the laggard firms resort more to ‘buying’ technologies’ that they assimilate by ‘investing in indigenous R&D.’ Only those firms that remain technologically active can transform in later stages into ‘making’ technologies with the wealth of technological capabilities accumulated over the years. To examine these issues and to contribute in the literature of ‘make or/and buy’ hypothesis, the case of Indian Capital Goods industry and its sub-sectors is taken for analysis. It is hypothesised that firms in technology accumulation process largely relied on combination of both in-house R&D and technology imports during post-liberalisation period. Indian Capital Goods industry remained ‘protected’ for about more than three decades (Ahluwalia, 1991; Kambampati, 1996) before adopting the Structural Adjustment Program in 1991 opening the doors for greater trade and movement of resources including capital. To encourage indigenous production, the Government of India has given a call for ‘Make in India’ in 2014 with refinements as ‘AatmaNirbhar Bharat Abhiyan’ program in 2020. Therefore, the main objective of the article is to examine the technology accumulation strategy adopted by the industries belonging to the Global South during the contemporary globalised era by taking the case of Indian Capital Goods industries. The rest of the article is structured as follows: Besides the introductory Section I, Section II presents the theoretical framework along with the discussion of the existent literature on the subject. Section III discusses the methodology used for empirical analysis. Section IV analyses the empirical findings and finally Section V concludes the article.

II. Theoretical Framework: ‘Make or/and Buy’ Hypothesis

Technological Divide

Technology is an intangible source of production. It remained ‘inside the black box’ (Rosenberg, 1982) unless endogenous growth models (Lucas, 1988; Romer, 1986), evolutionary approach (Nelson & Winter, 1982) and innovation systems framework (Freeman, 1987) started evolving. The focus was to explore the nature and determinants of technology, technology spillovers and relationship of technology with various socio-economic-political factors. Neo-classical school of thought (Grossman & Helpman, 1991; Keller, 2002) argued about the ‘public good’ character of technology that can flow through numerous direct and indirect means including trade of goods embodying technology, movement of employees and technicians, learning by exporting, participating in conferences and fairs, etc. The ‘tacitness’ of technology also determines the mechanism of its flow. The more tacit the technology is, closer the interaction between developers and users of technology is needed for technological inflows, even after its direct purchase. With technological diffusion in due course of time, it started assuming public good characteristics leading to its spillover more easily. ‘Innovation systems’ approach delve deep to bring forth the complexity of the innovation process as composed of inter-relationship between various actors, organisations and institutional dynamics (Lundvall, 1992).

Technological Accumulation

Technology accumulation is important for the firms to compete and survive in the dynamic market structure and fragmented production chains. Therefore, the strategies of technology accumulation vary in different firms. Both incumbents and laggards are the reflection of the technology capabilities accumulated over the years. This includes the multiplicity of factors, actors and their close interactive relationship (Lundvall, 1992). If new entrants and laggards aims to catch-up with the incumbent firms, they may resort to various technological accumulation strategies (Lee, 2013). However, the choice of the appropriate technology accumulation strategy depends upon a complex mix of cost, quality and availability of technology along with short-term and long-term growth plans. The firms make an array of choice for accumulating technologies composed of indigenous investment in R&D and/or purchase of technologies from abroad (either directly or indirectly). The firms need to be technologically active to absorb, assimilate and contribute in indegneous technology accumulation process (Basant, 1997a; Cohen & Levinthal, 1989).

Make and/or Buy Technology

The external sources of technology may involve transaction cost (Veugelers, 1997 citing the work of Williamson, 1981) that divert the resources away from the internal investment in developing and absorbing technologies. The substitution of internal and external means of technologies is contradicted with the ‘absorptive capacity’ argument (Cohen & Levinthal, 1989). Veugelers (1997) examined closely the relationship between internal and external sources of technology in case of Flemish manufacturing firms and found that external technologies stimulate internal R&D expenditure. Henceforth, these results confirm the ‘absorptive capacity’ hypothesis of complementary relation between internal and external sources of technology. Veugelers and Cassiman (1999) also examined the innovation-specific characteristics of Belgian company data to understand the nature of choice for appropriate innovation strategies of different firms. They found that small firms largely restrict to ‘make or buy’ strategy, while the large firms largely combine both internal and external sources of technology.

In Indian context, Katrak (1985) explored engineering firm’s behaviour and found that relationship between technology purchase and R&D expenditure existed to a limited extent probably due to licensing system and barriers imposed by technology supplier firms. Basant (1993) also found that investment in indigenous R&D build absorptive capabilities in firms to get advantage of technological imports. Basant (1997b) analyzed the technical strategies of non-electrical machinery industries and did not find complement relationship between foreign technology purchase and investment in in-house R&D. Siddharthan and Pandit (1998) also found that technology imports and R&D investment encouraged the capacity creation of firms in chemical, pharmaceutical and machinery industries during post-liberalisation period. Narayanan (2004) found that liberalisation policies enhanced competition for Indian automobile sector that induced local firms to invest in R&D and import advanced technologies. It is thus evident that literature exists that examined the strategies of technology accumulation for different firms and industries belonging to different manufacturing industries for both developed and developing countries. But, it is important to re-examine the technology accumulation strategy adopted by the firms in the recent past to capture the recent dynamic scenario, mainly for the firms belonging to the industries of developing countries. In this context, the aim of the article is to examine the relationship between indigenous R&D investment and technology purchase in light of the various factors such as mergers and acquisitions, outward foreign direct investment (OFDI), foreign equity participation, concentration and technology spillover. The case of Indian Capital Goods industry is taken for the post-reform period from 1994–1995 to 2015–2016 with a focus upon the impact of Asian crisis of 2008.

Data

For empirical analysis of Indian Capital Goods sector, four sub-sectors namely electronics, electrical machinery, non-electrical machinery and transport equipment industries were considered. The selection of firms was based on industry groups which are given under three-digit classification of National Industrial Classification (NIC) 2008. The undeflated series of variables has been extracted for industries of Capital Goods sector from disaggregated database of CMIE Prowess at five digit level (categorized at three digit level). The unbalanced panel data set consisting of 2,292 firms was taken for the period of 1994–1995 to 2015–2016, that was cleaned for further analytical exploration by first dropping the firms with zero ‘sales’ or missing values in entire period, followed with dropping those firms which had less than 5 years data for ‘sales.’ The final sample consists of 1,273 firms.

III. Empirical Analysis: Methodology

Multinomial logistic regression model (Baltagi, 2005) is used to estimate the model. The model regards the strategies of technology acquisition (π_ij) adopted by a firm i for selection of particular technology strategy j as follows:

π_{i j} = β^{'}_{j} ​ . X_{i} + ε_{i j}

(1)

where X_i denotes industrial characteristics in terms of qualitative and quantitative vector of non-stochastic component of firm and ε is the stochastic component. The probability P_it of ith firm for making technological choice can be 1, 2, 3 and 4, respectively, for four different choices such as S₀, S₁, S₂ and S₃. Let V_i be the random variable to indicate the technology choices of firms. To predict Pb_ij probability, the study assumes logit (U) linear explanation of given variables,

U = β^{'}_{j} ​ . X_{i}

(2)

The maximum likelihood principle required the parameters values that would also maximize the function of likelihood, given as follows:

L o g U = \sum_{i} \sum_{j} V_{i j} \log P_{i j}

(3)

However, the probabilities of Equation (2) for model are given as follows:

P_{i j} = P (V_{j} = j) = e^{β^{'}_{j} X_{i}} / \sum_{k = 0}^{m} e^{β^{'}_{k} X_{i}}

(4)

for i = 1, 2, 3, …, n and j = 1, 2, …, m. The J log odd-ratios can be computed with,

L o g (P_{i j} / P_{i j}) = x_{i} (β_{j} - β_{k}) = x_{i} β_{j} i f k = 0

(5)

In the model, an odds ratio (P_j/P_k) does not depend on other choices that follow independence of disturbance. Except the constant term, coefficients show the change of predicted logit. The odds ratio helps to interpret the explanatory variables more intuitively, especially for the interpretation of dummy variables (Mukharjee et al., 1998).

The following model is the expansion of Equation (2):

\begin{matrix} U_{i t} = α + β_{1} A g e_{i j t} + β_{2} E \underset{─}{​} i m p_{i j t} + β_{3} D \underset{─}{​} i m p_{i j t} + β_{4} O \underset{─}{​} i m p_{i j t} + \\ β_{5} S i z e_{i j t} + β_{6} H H I_{i j t} + β_{7} M & A + β_{8} F \underset{─}{​} e q u_{i j t} + \\ β_{9} T e c h \underset{─}{​} s p i l_{i j t} + β_{10} O F D I_{i j t} + β_{11} \exp ​ o r t_{i j t} + \\ β_{12} \sec \underset{─}{​} 1_{i j t} + β_{13} \sec \underset{─}{​} 2_{i j t} + β_{14} \sec \underset{─}{​} 3_{i j t} u_{i j t} \end{matrix}

(6)

where U is the logit, i = 1, 2, …, n are indexes of firms, j = 1, 2, …, N are indexes of industry, t = 1, 2, …, T represents time and u_ijt represents the error term.

E_imp is the embodied technology imports intensity, D_imp is the import of technical know-how, royalty intensity, O_imp is the raw material imports intensity, HHI is the Herfindahl-Hirschman index, M&A is the mergers and acquisitions, F_equ is the foreign equity participation, Tech_spil is the foreign technology spillovers, OFDI is the outward foreign direct investment, sec_1 is the electronics sub-sector, sec_2 is the electrical machinery sub-sector, sec_3 is the non-electrical machinery sub-sector and sec_4 is the transport equipment sub-sector is reference state. The rationale of the dependent and independent variables chosen and the construction of the chosen variables are discussed below.

Dependent Variables

Following Basant (1997b) and Chaurgudi (2008), the following empirical framework is used to analyze discrete technological choices of the firm for Indian Capital Goods sector. The firm has the following four explicit choices:

S₀ = if firm selected to indigenous R&D investment only;

S₁ = if firm selected to disembodied technology import only;

S₂ = if firm selected to invest in both S₀ and S₁; and

S₃ = if firm selected neither S₀ nor S₁;

Firm i can opt for either of the four strategies in the time period t. To elaborate, firm i can choose to invest in ‘R&D only’ by opting for strategy S₀. Likewise, it can choose to import disembodied technology, that is, option S₁. Similarly, if in case, the investment in R&D along with import of disembodied technology is done by the firm, it is considered that it has made a choice for option S₂. Again if same ‘firm i’ do not choose to invest in indigenous R&D nor purchased technical know-how in year t, it is considered to have chosen the option S₃. The first three strategies S₀, S₁ and S₂ represent that the firms are technologically active, whereas the last option S₃ represents technologically passive firms. It is important to examine that what factors lead different firms to make different technological choices.

Variable Construction (Independent variables)

Disembodied Technology Import Intensity (D_imp)

Direct purchase of technologies through royalty payments is an important strategy of buying complex technologies, that are hard to spillover (Parameswaran, 2002; Rijesh, 2015). Either the imported technology can substitute the indigenous R&D investment or it can complement it. Disembodied technology intensity is calculated as a ratio of ‘royalty and technical know-how’ to the ‘sales’ of a firm. The expenditure on disembodied technology imports of each firm was deflated with the United States (US) R&D deflator, collected from the Analytical Business Enterprise Research and Development (ANBERD) database.

Embodied Technology Imports Intensity (E_imp)

The imported Capital Goods embodied with advanced technologies also works as an important agent of technology spillovers and expected to boost the domestic production process (Herrerias & Orts, 2011). The firms that want to compete and ‘catch-up’ with those in the frontiers resort to embodied technologies as the means of ‘know-what’ (Lundvall & Johnson, 1994) that they assimilate and accumulate further with indigenous R&D investments. The firms in the process of catching-up may resort more to external technologies, unlike the ones want to ‘leap-frog’ the incumbents. We have estimated the ‘embodied technology imports intensity’ as the ratio of ‘import of capital goods’ to ‘sales.’

Raw Material Import Intensity (O_imp)

Following the literature (Satpathy et al., 2017; Sharma, 2010), ‘raw material import intensity’ is obtained as the ratio of ‘import of raw materials’ to ‘sales’ of firm. The raw material import intensity is an important factor to determine the appropriate strategy of the technology accumulation by the firm. The import of raw materials may require indigenous R&D investments to assimilate or these can lead to imports of technologies for better outcomes. Although a mix of both strategies are never ruled out.

Exports Intensity (Exports)

Learning through exports is an important component of technology accumulation. India has emerged as one of the important exporters of machinery including heavy and light engineering equipment (Sharma, 2010). Various studies (Mahambare & Balasubramanyam, 2005; Mondal & Pant, 2018) showed that exporting firms are technologically active as they improve export performances. The export intensity of firms is captured as the ratio of ‘exports’ of related industries divided by ‘sales.’

Age of the Firms (AGE)

‘Age’ is an important factor determining the choice for optimal technology strategy by the firm. Technological accumulation is a time-consuming process evolving over time with development of indigenous distinct tacit knowledge and managerial abilities (Bhaduri & Ray, 2004). The older firms are expected to have greater ability to understand market complexities to foresee the future technological dynamics. Therefore, these firms may adopt different strategies of technology accumulation depending upon their existing capabilities and objectives. ‘Age’ of the firm is measured by subtracting the ‘year of incorporation’ from the available year (at the time of doing this research), that is, 2016.

Size of the Firm (Size)

The size of a firm indicates the quantum of resources and opportunities available to a firm for expansion (Penrose, 1959). The firms with more resources are expected to be ‘technologically active.’ ‘Size’ of a firm is obtained as logarithm of ‘sales’ at constant prices 2011–2012 using Wholesale Price Index (WPI) deflators.

Market Structure (HHI)

The integrated economies are expected to change the degree of ‘concentration’ of the market. More concentrated markets may either encourage or discourage firms to invest in indigenous R&D, if technology inflows in the latter case. Following Malik (2015), ‘Herfindahl-Hirschman index’ (HHI) was calculated for concentration as the sum of square of market shares in real value of ‘sales’ for each firm in industry at 2011–2012 prices.

Mergers and Acquisitions

Mergers and Acquisitions (M&A) may impact the technological capabilities of the firm in numerous ways. It may invest in indigenous R&D to assimilate the technologies of acquired/merged firms; or they can buy the technologies to reap the benefits of the technologies of acquired/merged firms; or the combination of the two. Following Chaurgudi (2008), the value 1 is assigned if firm has ‘M&A’ deal and 0 if not.

Foreign Equity Participation (F_equ)

The firms with foreign equity are expected to have resources to invest in building technological capabilities either through embodied or disembodied means of technology. It is also highlighted in the literature (Cohen & Levinthal, 1989) that to assimilate foreign technologies indigenous investment in R&D is necessary. ‘Ownership group’ from CMIE Prowess has been used to find the foreign equity firms with assigning the dummy 1 to a ‘foreign firms’ and 0 otherwise.

Foreign Technology Spillovers (Tech_spil)

Foreign technology spillover is an important determinant that helps the firm in deciding the strategy of technology accumulation. Following Kathuria (2002), foreign technology spillovers are measured as ratio of ‘sales’ of foreign firms to ‘sales’ of whole industry.

Outward Foreign Direct Investment (OFDI)

OFDI opens new avenues of ‘learning’ from host countries that also impact the choice of means of technology accumulation (Kumar, 2001; Pradhan & Singh, 2009). Dummy variable is included that takes the value 1 if firm has ‘OFDI’ and 0 otherwise.

IV. Technology Strategies of Indian Capital Goods Sector: Empirical Findings

The firms mainly make a choice between ‘Make and/or Buy’ strategy of technology accumulation. Whereas, ‘make or buy’ depicts substituting possibilities, the option of ‘make and buy’ can combine different arrays of technology accumulation options. These choices are the reflection of cost-benefit comparison of numerous factors. Table 1 highlights the quantum and percentage of ‘technologically active firms’ belonging to Capital Goods industries in India from 1994–1995 to 2015–2016, with a comparison of pre-Asian crisis period (1994–1995 to 2007–2008) with post-Asian crisis period (2008–2009 to 2015–2016). It is noteworthy that Asian crisis erupted in 2008 that impacted the production and trade in the region.

Table 1.
Technologically Active Firms in Capital Goods Sector (in %)

Industries 1994–1995 to 2007–2008 2008–2009 to 2015–2016 1994–1995 to 2015–2016

Electronics 38.08 (1,454) 34.22 (1,030) 36.51 (2,484)

Electrical machinery 34.99 (2,642) 28.95 (2,185) 32.33 (4,827)

Non-electrical machinery 43.89 (4,002) 39.92 (3,059) 42.21 (7,061)

Transport equipment 40.88 (3,245) 42.32 (2,734) 41.53 (5,979)

Capital goods 40.29 (11,343) 37.51 (9,008) 39.08 (20,351)

Industries	1994–1995 to 2007–2008	2008–2009 to 2015–2016	1994–1995 to 2015–2016
Electronics	38.08 (1,454)	34.22 (1,030)	36.51 (2,484)
Electrical machinery	34.99 (2,642)	28.95 (2,185)	32.33 (4,827)
Non-electrical machinery	43.89 (4,002)	39.92 (3,059)	42.21 (7,061)
Transport equipment	40.88 (3,245)	42.32 (2,734)	41.53 (5,979)
Capital goods	40.29 (11,343)	37.51 (9,008)	39.08 (20,351)

Source: Authors’ own calculations based on CMIE Prowess database.

Note: Total numbers of observation are given in parenthesis.

For the period from 1994–1995 to 2015–2016, 39.08% of firms belonging to Capital Goods sector were technologically active, with about one-third firms are technically active during the post-Asian crisis period. The firms in transport equipment sub-sector have the higher proportion of technologically active firms with 40.88% in pre-Asian crisis period that increased further to 42.32% during the post-Asian crisis period. Furthermore, Table 2 shows that technology acquisition strategies of different subsectors within Indian Capital Goods industry change over the period from 1994–1995 to 2015–2016.

For the aggregated Capital Goods sector, the share of firms opted for S₀ ‘indigenous R&D investment only’ S₁ of ‘disembodied technology imports only’ increased whereas the share of firms opting for S₂, that is, ‘both indigenous R&D and disembodied technology import’ declined from pre-Asian crisis period to post-Asian crisis period (Table 2). Overall, only 18.42% of the firms belonging to the Indian Capital Goods sector choose to invest in S₂, that is, ‘both indigenous R&D and disembodied technology imports’ as a strategy of technology acquisition.

Furthermore, Table 2 also shows the strategy of technology acquisition by different sub-sectors of Indian Capital Goods sector. It was found that more than 50.46% of electronics and electrical machinery industries rely on option S₀ of ‘indigenous R&D investment only’ for overall period from 1994–1995 to 2015–2016. The share of electrical machinery opting for S₁, that is, ‘disembodied technology import only’ shows a marginal increase from 34.59% to 35.88%, depicting their increasing dependence upon imported technologies,. Furthermore, it was found that only in case of non-electrical machinery industries, 42.70% opted for S₀ in 1994–1995 to 2007–2008 that declined to 41.88% in 2008–2009 to 2015–2016. Overall, the firms belonging to Indian Capital Goods sector opted for strategy S₀, that is, ‘indigenous R&D investment only’ as a dominant means of technology accumulation followed by strategy S₁ of ‘disembodied technology import only.’ To examine the choice of the Indian Capital Goods firms for optimal technology acquisition strategy, it would be important to know that with which magnitude does the firms opted for option S₀ over option S₁.

Table 2.

Technology Accumulation Strategies Adopted by Technologically active firms (in %)

Industrial Sectors	Strategies	1994–1995 to 2007–2008	2008–2009 to 2015–2016	1994–1995 to 2015–2016
Electronics	S₀	48.34	54.17	50.46
	S₁	35.14	32.62	34.23
	S₂	16.52	13.21	15.31
Electrical machinery	S₀	47.45	50.49	48.56
	S₁	34.59	35.88	35.06
	S₂	17.96	13.63	16.38
Non-electrical machinery	S₀	42.70	41.88	42.39
	S₁	37.56	40.43	38.77
	S₂	19.74	17.69	18.84
Transport equipment	S₀	40.11	41.04	40.53
	S₁	39.39	40.71	39.53
	S₂	20.50	18.25	19.94
Capital Goods sector	S₀	43.39	44.06	43.66
	S₁	37.32	38.82	37.92
	S₂	19.29	17.12	18.42

Source: Authors’ own calculations based on CMIE Prowess database.

Based upon the relative preference of the firms belonging to different sub-sectors of Indian Capital Goods sector, Table 3 shows that the proportion of firms preferring to invest in ‘indigenous R&D only’ as compared to ‘disembodied technical imports only’ is higher. These observations also strengthened the results presented in Table 2. It shows that majority of firms preferred indigenous investment in R&D over the disembodied technology imports only. Considering the trend of investment in R&D at the national level in India, it was found that around 28% of R&D expenditure was done by public institutions (Department of Science and Technology, 1999), although R&D expenses as a percentage of GNP declined from 0.98% to 0.71% from period 1987–1988 to 1995–1996 (DST, 1999).² As per report³ of ‘Ministry of Heavy Industry and Public Enterprise’ 2012, R&D remains constant at 0.9% for Capital Goods which is also a cause for emergence of technological gap between India and technologically advanced countries.

Table 3.

Relative Preferences of Firms: Make or Buy Technology (in %)

Sub-sectors	Spending Preference	1994–1995 to 2007–2008	2008–2009 to 2015–2016	1994–1995 to 2015–2016
Electronics	R&D > DISEMB	62.85	69.30	65.03
Electronics	DISEMB > R&D	37.15	38.80	34.97
Electrical machinery	R&D > DISEMB	63.83	63.32	63.19
Electrical machinery	DISEMB > R&D	36.17	37.68	36.81
Non-electrical machinery	R&D > DISEMB	55.96	51.04	53.89
Non-electrical machinery	DISEMB > R&D	44.04	48.96	46.11
Transport equipment	R&D > DISEMB	50.89	51.64	51.25
Transport equipment	DISEMB > R&D	49.11	48.36	48.75
Capital goods sector	R&D > DISEMB	57.04	55.37	56.32
Capital goods sector	DISEMB > R&D	42.96	44.63	43.68

Source: Authors’ own calculations based on CMIE Prowess database.

Note: R&D and DISEMB indicate R&D expenses only and disembodied technology import only.

Multinomial Regression Analysis: Results and Discussion

With dummies for four different technological strategies taken as the dependent variable, Table 4 presents the results of multinomial regression analysis for Capital Goods industries for the period from 1994–1995 to 2015–2016. ‘Age’ has a significant and positive impact with better log odds for firms’ decision to choose S₀ and S₂ strategies. This implies that firms older in age prefer to invest in S₀ and S₂. On the other hand, older firms have negative and significant impact for ‘disembodied technology imports only’ (S₁) showing less probability for choosing option S₁. However, increase in ‘embodied technology import intensity’ (E_imp) was found to have significant impact for options S₁ and S₂, implying that the imports that embodied advanced technologies results in direct purchase of technology to assimilate the advanced technologies properly. But the impact of ‘disembodied technology import intensity’ (D_imp) have positive and significant impact on firms’ choices to opt ‘both indigenous R&D and disembodied technology imports’ (S₂).

Table 4.

Technologically Active Firms in Capital Goods Sector: Estimation of Multinomial Logit Model

Regressors	S₀		S₁		S₂
Regressors	Coefficients	Odds	Coefficients	Odds	Coefficients	Odds
(1)	(2)	(3)	(4)	(5)	(6)	(7)
Intercept	−2.293* (0.147)	0.06	−1.722* (0.167)	0.06	−2.570* (0.150)	0.03
Age	0.026* (0.001)	1.03	−0.013* (0.001)	0.99	0.016* (0.001)	1.01
E_imp	0.001 (0.000)	1.00	0.094* (0.000)	1.00	0.135* (0.188)	0.99
D_imp	−16.007 (0.957)	0.00	0.852* (0.025)	2.34	0.251* (0.019)	1.91
O_imp	0.000 (0.000)	1.00	0.056* (0.015)	0.64	0.112* (0.126)	0.99
Size	0.448* (0.048)	1.00	0.000 (0.000)	1.00	0.025* (0.005)	1.00
HHI	−0.001*** (0.001)	1.00	0.008** (0.001)	1.00	−0.005* (0.000)	1.00
M&A	0.448* (0.048)	1.62	−0.151* (0.057)	0.86	0.747* (0.048)	2.13
F_equ	0.199* (0.081)	1.22	0.429* (0.075)	1.54	0.852* (0.066)	1.87
Tech_spil	0.028** (0.013)	1.05	−0.076* (0.022)	0.93	−0.016* (0.007)	0.98
OFDI	0.919* (0.070)	2.73	−0.349* (0.101)	0.71	0.787* (0.074)	2.34
Export	−0.008* (0.001)	0.99	−0.001** (0.000)	1.00	−0.037* (0.004)	0.95
Sec_1	0.347* (0.066)	1.45	−0.445* (0.083)	0.64	−0.534* (0.087)	0.65
Sec_2	−0.119** (0.056)	0.85	−0.320* (0.065)	0.73	−0.655* (0.066)	0.07
Sec_3	0.028 (0.052)	1.08	0.048 (0.054)	1.05	−0.323* (0.055)	0.86
Log-likelihood	−7832.32
LR x²	3295.22
No. of observations	20,351

Source: Authors’ own calculations based on CMIE Prowess database.

Note:

1. The standard errors are given in parenthesis for coefficients.

2. p-Values for test are *, ** and *** that indicates level of statistical significance at 1%, 5% and 10% level, respectively.

3. D_int, Disembodied technology import intensity; E_int, embodied technology import intensity; F_eq, foreign equity participation; HHI, market structure; M&A, mergers and acquisitions; OFDI, outward foreign direct investment; O_imp, raw material import intensity; Tech_spil, foreign technology spillover; Sec_1, electronics; Sec_2, electrical machinery and Sec_3, non-electrical machinery.

4. The strategies S₀, S₁ and S₂ represent strategies to invest in ‘indigenous R&D only,’ ‘disembodied technology imports only’ and ‘both indigenous R&D investment and disembodied technology imports’ respectively.

The positive and significant impact of ‘size’ on firm’s decision to opt for S₀ and S₂ strategies shows that large-sized firms may take more advantages of economies of scale through ‘indigenous R&D only’ and strategy of ‘both indigenous R&D and disembodied technology import.’ The ‘raw material import intensity’ (O_imp) of firms positively effects the choice of the firms to opt for strategies S₁ and S₂. Further increase of concentration in ‘market structure’ (HHI) was found to have positive and significant impact on choices of firms to opt for investment in S₀ and S₂. As concerning the impact of ‘mergers and acquisitions’ (M&A) on making the choice for technology strategy, it was found that positive and significant coefficient for S₂ strategy indicated with better odds values followed by S₀ strategy. The coefficient of ‘foreign equity participation’ (F_equ) with better log odds have a positive impact on option S₂. The ‘foreign technology spillovers’ (Tech_spill) have negative and significant impact on strategies S₁ and S₂ signifying a decrease in the probability of direct technology purchase and ‘both indigenous R&D and disembodied technology import.’ Rather, ‘foreign technology spillovers’ (Tech_spill) were found to have a positive impact on choosing the strategy S₀. It was also found that ‘outward foreign direct investment’ (OFDI) have a significant and positive impact on choosing option S₀ and negative impact on option S₁ indicating that the firms engaged in OFDI are investing more in ‘indigenous R&D only’ to be competitive at the international level.

The electronics sub-sector (Sec_1) and non-electrical machinery sub-sector (Sec_3) have better log odds for opting ‘indigenous R&D only’ (S₀) as compared with electrical machinery sub-sector (Sec_2) whose chances to opt S₀ reduced by 15%.⁴ The electronics industries (Sec_1) have 45% higher chances to opt S₀. The firms belongs to non-electrical machinery industry (Sec_3) have 5% higher chances to opt S₁ that are fairly better as compared to electronics (Sec_1) and electrical machinery industries (Sec_2). Overall, the findings render the advantage from S₀ strategy for all sub-sectors.

Tables 5–8 present the results of multinomial regression analysis for different sub-sectors of Indian Capital Goods industries.

Table 5 shows the impact of different factors that lead firms of electronics sub-sector to choose among the three strategic choices for technology acquisition. It was found that positive and significant impact of ‘age’ and ‘mergers and acquisitions’ (M&A) lead the firms to choose strategy S₀ but not S₁ and S₂. The OFDI with better log odds lead to choose S₀ followed by S₂ which are positive and significant coefficients with better log odds.

Table 5.

Technologically Active Firms in Electronics Industry: Estimation of Multinomial Logit Model

Regressors	S₀		S₁		S₂
Regressors	Coefficients	Odds	Coefficients	Odds	Coefficients	Odds
(1)	(2)	(3)	(4)	(5)	(6)	(7)
Intercept	−2.515* (0.498)	0.08	−2.106* (0.440)	0.12	−1.902* (0.481)	0.15
Age	0.029* (0.005)	1.03	0.007 (0.006)	1.01	−0.044* (0.003)	1.00
E_imp	0.000 (0.001)	1.00	−0.001 (0.001)	1.00	0.000 (0.001)	1.00
D_imp	−6.980* (1.328)	0.00	0.435* (0.054)	1.55	0.501* (0.058)	1.65
O_imp	0.054* (0.002)	1.54	0.023* (0.04)	1.44	0.076 (0.09)	1.07
Size	0.001*** (0.000)	1.00	0.001* (0.000)	1.00	−0.004** (0.000)	1.00
HHI	0.000 (0.001)	1.00	0.002 (0.001)	1.00	−0.001 (0.001)	1.00
M&A	0.794* (0.133)	2.21	−0.449** (0.214)	0.64	−0.486** (0.233)	0.62
F_equ	−0.336 (0.329)	0.71	0.587 (0.684)	1.62	1.914* (0.281)	6.78
Tech_spil	−0.005* (0.031)	0.99	−1.066** (0.556)	0.34	0.031*** (0.019)	1.03
OFDI	1.474* (0.164)	4.37	−0.620** (0.280)	0.53	1.051* (0.233)	2.86
Export	−0.005* (0.031)	0.99	0.089* (0.038)	1.00	−0.022* (0.006)	0.98
Log-likelihood	−1062.19
LR x²	375.98
No. of observations	2,484

Source: Authors’ own calculations based on CMIE Prowess database.

Note:

1. The standard errors are given in parenthesis for coefficients.

2. p-Values for test are *, ** and *** that indicates level of statistical significance at 1%, 5% and 10% level, respectively.

The ‘disembodied technology import intensity’ (D_imp) has a positive and significant impact on the firms’ decision to invest in ‘both indigenous R&D and disembodied technology imports’ (S₂) with 65% possibility as compared to ‘disembodied technology import only’ (S₁) which has positive and significant coefficient with 55% possibility. However, a positive and significant coefficient impact of ‘size’ and ‘raw material import intensity’ (O_imp) is found for firms’ choices to opt for strategy S₀ with better log odds as compared to S₁ decision of firms. The firms have high probability for ‘foreign equity participation’ (F_equ) to choose S₂ as technology decisions of firms with better log odds. The ‘foreign technology spillovers’ (Tech_spill) has influenced positively and significantly to S₂ technology choice of firms. The increase in ‘export intensity’ of electronics sub-sector effects positively the firms to opt S₁ for technology acquisition. The ‘embodied technology import intensity’ (E_imp) has insignificant impact and weak odds to adopt any of strategy.

The maximum likelihood results for multinomial logit model for electrical machinery industry are shown in Table 6. It was found that strategy S₂ emerged dominant with high probability with positive and significant impact of coefficients such as age, size, market structure (HHI), disembodied technology import intensity (D_imp), raw materials import intensity (O_imp), foreign equity participation (F_equ) and exports intensity. The ‘mergers and acquisitions’ (M&A) is found profitable with 76% higher chances to gain advantages with making a choice of opting for option S₂ with better odds followed by S₀ with 23% chances. However, increase of ‘foreign technology spillovers’ (Tech_spil) have positive and significant influence on choices of firms to invest in S₀. For S₀ and S₂ technology strategies, ‘outward foreign direct investment’ (OFDI) of firms is found to have better odds for technologically active firms.

Table 6.

Technologically Active Firms in Electrical Machinery Industry: Estimation of Multinomial Logit Model

Regressors	S₀		S₁		S₂
Regressors	Coefficients	Odds	Coefficients	Odds	Coefficients	Odds
(1)	(2)	(3)	(4)	(5)	(6)	(7)
Intercept	−2.421* (0.285)	1.03	−2.924* (0.400)	0.05	−3.926* (0.393)	0.04
Age	0.031* (0.002)	1.00	0.001 (0.003)	1.00	0.021* (0.003)	1.02
E_imp	0.000 (0.000)	0.00	0.000 (0.001)	1.00	0.000 (0.001)	1.00
D_imp	−5.329* (0.747)	1.00	1.124* (0.082)	1.00	1.123* (0.084)	1.62
O_imp	0.000 (0.000)	1.00	0.010* (0.053)	2.55	1.023* (0.134)	2.16
Size	0.015 (0.000)	1.00	−0.003** (0.000)	1.00	0.003* (0.000)	1.00
HHI	−0.002 (0.002)	1.54	−0.001 (0.002)	1.00	0.014* (0.005)	1.01
M&A	0.433* (0.098)	1.23	0.050 (0.131)	1.05	0.649* (0.127)	1.76
F_equ	0.205 (0.167)	1.03	−0.165 (0.201)	0.85	0.396** (0.201)	2.34
Tech_spil	0.029** (0.013)	1.50	0.001 (0.014)	1.00	−0.039* (0.012)	0.96
OFDI	0.407** (0.175)	1.00	0.139 (0.226)	1.15	1.058* (0.208)	2.82
Export	−0.005* (0.002)	0.09	0.000 (0.000)	1.00	0.089* (0.038)	1.00
Log-likelihood	−1378.75
LR x²	338
No. of observations	4,827

Source: Authors’ own calculations based on CMIE Prowess database.

Note:

1. The standard errors are given in parenthesis for coefficients.

2. p-Values for test are * and ** that indicates level of statistical significance at 1% and 5% level, respectively.

Table 7 presents the estimation results for the non-electrical machinery sub-sector of the Indian Capital Goods industry. The impact of ‘age’ and ‘size’ of non-electrical machinery manufacturing firms resulted in equal probabilities of opting for S₀ and S₂ with similar log odds ratios. The ‘foreign equity participation’ (F_equ) and OFDI were found to have a significant and positive impact on the firms to opt for strategy S₂ with better log odds followed by S₁ strategy. However, ‘merger and acquisitions’ (M&A) deals have higher probability for opting for option S₂ followed by S₀ for technology acquisition. The variables such as ‘embodied technology import intensity’ (E_imp), ‘disembodied technology import intensity’ (D_imp) and ‘raw material imports intensity’ (O_imp) have significant impact on firms’ decision to opt for S₁ with better odds followed by strategy S₂. The presence of foreign firms in ‘foreign technology spillovers’ (Tech_spil) and ‘outward foreign direct investment’ (OFDI) of firms have observed high probability to opt for S₀ with better log odds.

Table 7.

Technologically Active Firms in Non-electrical Machinery Industry: Estimation of Multinomial Logit Model

Regressors	S₀		S₁		S₂
Regressors	Coefficients	Odds	Coefficients	Odds	Coefficients	Odds
(1)	(2)	(3)	(4)	(5)	(6)	(7)
Intercept	−2.111* (0.233)	0.12	−1.559* (0.243)	0.21	−3.302* (0.238)	0.03
Age	0.024* (0.002)	1.02	−0.013* (0.002)	0.99	0.020* (0.002)	1.02
E_imp	0.000 (0.001)	1.00	0.104* (0.031)	1.00	0.208* (0.034)	0.99
D_imp	−49.821* (5.137)	0.00	0.809* (0.040)	2.24	0.246* (0.029)	1.28
O_imp	0.434 (0.010)	1.23	0.055* (0.013)	1.76	0.065* (0.013)	1.43
Size	0.001* (0.000)	1.00	0.000 (0.000)	1.00	0.001* (0.000)	1.00
HHI	−0.007 (0.031)	0.99	0.008 (0.017)	1.01	0.016 (0.017)	1.01
M&A	0.367* (0.087)	1.44	−0.016 (0.091)	0.98	0.713* (0.080)	2.05
F_equ	0.130 (0.126)	1.15	0.609* (0.101)	1.84	0.664* (0.101)	1.96
Tech_spil	0.161*** (0.093)	1.16	−0.160* (0.055)	0.85	0.001 (0.040)	1.00
OFDI	0.616* (0.138)	1.84	−0.233 (0.164)	0.79	0.597* (0.126)	1.82
Export	−0.019* (0.003)	0.98	−0.003** (0.001)	1.00	−0.031* (0.006)	0.96
Log-likelihood	−2509.62
LR x²	1690.23
No. of observations	7,061

Source: Authors’ own calculations based on CMIE Prowess database.

Note:

1. The standard errors are given in parenthesis for coefficients.

2. p-Values for test are *, ** and *** that indicates level of statistical significance at 1%, 5% and 10% level, respectively.

In Table 8, results of maximum likelihood for multinomial regression model for transport equipment industries show that the impact of disembodied technology imports intensity (D_imp) is positive and significant for firms to opt for strategy S₂. The coefficient of age, foreign technology spillovers (Tech_spil) and OFDI of firms have positive and significant impact for the firm to opt for strategy S₀ as compared to other technology choices like S₁ and S₂. It is noticeable that a coefficient of ‘embodied technology import intensity’ (E_imp) has positive and significant impact with better odds for options S₁ and S₂. However, for the ‘size’ of firms with similar positive and significant coefficients (0.001) along with similar value of odds (1.00) showed equal possibilities to choose between the options S₀ and S₁.

Table 8.

Technologically Active Firms in Transport Equipment Industry: Estimation of Multinomial Logit Model

Regressors	S₀		S₁		S₂
Regressors	Coefficients	Odds	Coefficients	Odds	Coefficients	Odds
(1)	(2)	(3)	(4)	(5)	(6)	(7)
Intercept	−2.374* (0.292)	0.08	−1.494* (0.358)	0.22	−2.471* (0.284)	0.08
Age	0.024* (0.002)	1.02	−0.026* (0.003)	0.97	0.041* (0.005)	1.01
E_imp	−0.001 (0.001)	1.00	0.232* (0.042)	1.00	0.221* (0.032)	1.00
D_imp	−11.865* (17.318)	0.00	1.091* (0.051)	2.98	0.115* (0.017)	1.12
O_imp	−2.304* (0.585)	0.15	0.241* (0.054)	1.00	0.010* (0.000)	1.15
Size	0.001* (0.000)	1.00	−0.001* (0.000)	1.00	0.002* (0.000)	1.00
HHI	−0.001 (0.003)	1.00	0.006* (0.002)	1.01	−0.001 (0.001)	1.00
M&A	0.400* (0.095)	1.51	−0.319* (0.102)	0.73	1.054* (0.081)	2.87
F_equ	−0.941** (0.439)	0.34	0.493* (0.171)	1.64	1.142* (0.154)	3.13
Tech_spil	1.853* (0.523)	7.13	0.007 (0.068)	1.01	−0.169* (0.063)	0.84
OFDI	0.998* (0.143)	2.25	−1.193* (0.254)	0.30	0.616* (0.128)	1.85
Export	−0.006* (0.002)	0.99	−0.001 (0.001)	1.00	−0.149* (0.019)	0.86
Log-likelihood	−2060.79
LR x²	1235.52
No. of observations	5,979

Source: Authors’ own calculations based on CMIE Prowess database.

Note:

1. The standard errors are given in parenthesis for coefficients.

2. p-Values for test are * and ** that indicates level of statistical significance at 1% and 5% level, respectively.

As regarding the technological choices, it is evident from the analysis that electronics industries showed higher probabilities to opt for S₀ followed by electrical machinery industry, non-electrical machinery industry and transport equipment industry. With increasingly integrated economies and fast advancing technologies, it is important for the firms to increase their technological capabilities over time. Models of economic growth have largely emphasized the role of technological capabilities for long-run growth and sustainability. Therefore, the firms should strategize the process of technological capability building with the correct mix of internal and external sources of technology. In this context, the article examined the case of Indian Capital Goods industry. It was found that there is a need to increase the number of technologically active firms within the Indian Capital Goods industry. The strategies to invest in indigenous R&D and investment in disembodied technology imports both are related with high returns that improves technological capabilities and encourages the firms for innovation. This is an important implication with the call for ‘AatmaNirbhar Bharat Abhiyan’ policy of Government of India that aims to increase the contribution of manufacturing sector in global value chains.

V. Conclusion

With the objective of examining the strategy of technology accumulation by industries belonging to the Global South during the liberalised period, the case of Indian Capital Goods industry was examined. The idea is to contribute to the debate of ‘make and/or buy’ strategy as adopted by industries at different levels of development. India opened her economy in 1991 after remaining largely protected for about four decades since Independence in 1947. The analysis showed that a little over one-third of firms are technologically active in Indian Capital Goods sector, with 43% of the firms preferring for ‘investment in indigenous R&D’ as the strategy of technology accumulation during the period from 1994–1995 to 2015–2016. However, ‘disembodied technology import only’ was found to be the second major dominant strategy and very few firms opting for ‘both indigenous R&D and disembodied technology import’ at the same reported year. Hence, it can be inferred that firms either prefer ‘indigenous R&D only’ or ‘disembodied technology import only.’ Furthermore, multinomial logit regression model was estimated to examine the impact of various factors determining different strategies of technology accumulation. The results show that factors such as technology spillovers from foreign firms, raw material import intensity, concentration, age, size and OFDI have positive and significant impact on firms to opt for strategy (S₀), that is, investment in ‘indigenous R&D only.’ With endogenous growth models emphasizing the importance of indigenous R&D investment that helps not only in breakthrough innovations but also helps in assimilating the embodying and disembodying knowledge. Henceforth, given the importance of indigenous R&D, it is important to focus on the factors that determine the choice of the firms for opting the strategy of investment in ‘indigenous R&D only.’ On the other hand, technology purchase or strategy option, S₁, that is, ‘disembodied technology import only’ is also important, moreso to adopt the frontier technologies. The results show that ‘disembodied technology import intensity’ of Indian Capital Goods industry is merely 0.4% over the period of analysis. However, the increase of foreign equity participation, mergers and acquisitions of firms and export intensity has improved probability for strategy S₂, that is, opting for ‘both indigenous R&D and disembodied technology import.’ Overall, firms belonging to Indian Capital Goods industry opt for different strategies of technology accumulation depending upon numerous factors. The nature and complexity of technologies also determine the technology choices of firms.

Footnotes

Acknowledgements

The authors are highly thankful to the editors and two anonymous reviewers of the article that helps immensely in improving the article. Usual disclaimer applies.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Notes

References

Ahluwalia

I. J.

(1991). Productivity and growth in Indian manufacturing. Oxford University Press.

Aiginger

, & Rodrik

(2020). Rebirth of industrial policy and an agenda for the twenty-first century. Journal of Industry, Competition and Trade, 20, 189–207.

Akyuz

(2009). Industrial tariffs, international trade, and development. In Cimoli

, Dosi

, & Stiglitz

J. E.

(Eds.), Industrial policy and development: The political economy of capability accumulation (pp. 144–174). Oxford University Press.

Baltagi

B. H.

(2005). Econometric analysis of panel data (3rd ed.). Wiley.

Basant

(1993). R&D, foreign technology purchase and technology spillovers in Indian industry: Some explorations (Working Paper No. 8). United Nations University Institute of New Technologies.

Basant

(1997a). Analysing technology strategy: Some issues. Economic and Political Weekly, 32(48), M111–M120.

Basant

(1997b). Technology strategies of large enterprises in Indian industry: Some explorations. World Development, 25(10), 1683–1700.

Bhaduri

, & Ray

A. S.

(2004). Exporting through technological capability: Econometric evidence from India’s pharmaceutical and electrical/electronic firms. Oxford Development Studies, 32, 87–100.

Chaurgudi

(2008). Technology strategies under globalisation: A study of India’s manufacturing sector (Unpublished dissertation). Centre for Development Studies.

10.

Cohen

, & Levinthal

(1989). Innovation and learning: Two faces of R&D. Economic Journal, 99(397), 569–596.

11.

Department of Science and Technology. (1999). Research and development statistics, 1996–97. Department of Science and Technology.

12.

Edquit

(2011). Innovation and employment: Process versus product innovation. Edward Elgar Publishing.

13.

Freeman

(1987). Technology policy and economic performance: Lessons from Japan. Pinter.

14.

Grossman

G. M.

, & Helpman

(1991). Trade, knowledge spillovers and growth. European Economic Review, 32(2–3), 517–526.

15.

Herrerias

M. J.

, & Orts

(2011). Imports and growth in China. Economic Modelling, 28, 2811–2819.

16.

Jurowetzki

, Lema

, & Lundvall

B. A.

(2018). Combining Innovation systems and Global Value Chains for development: Towards a research agenda. The European Journal of Development Research, 30(3), 364–388.

17.

Kambampati

U. S.

(1996). Industrial concentration and performance: A study of structure, conduct and performance of Indian industry. Oxford University Press.

18.

Kathuria

(2002). Liberalisation, FDI and productivity spillovers: An analysis of indian manufacturing firms. Oxford Economic Papers, 54(4), 688–718. https://www-jstor-org.web.bisu.edu.cn/stable/3488772

19.

Katrak

(1985). Imported technology, enterprise size and R&D in a newly industrialising country: The Indian experience. Oxford Bulletin of Economics and Statistics, Department of Economics, University of Oxford, 47(3), 213–229.

20.

Keller

(2002). Trade and the transmission of technology. Journal of Economic Growth, 7, 5–24.

21.

Kumar

(2001). Determinants of location of overseas R&D activity of multinational enterprises: The case of US and Japanese corporations. Research Policy, 30(1), 159–174. 10.1016/S0048-7333(99)00102-X

22.

Lee

(2013). Schumpeterian analysis of economic catch-up: Knowledge, path creation, and the middle-income trap. Cambridge University Press.

23.

Lucas

R. E. J.

(1988). On the mechanics of economic development. Journal of Monetary Economics, 22(1), 3–42.

24.

Lundvall

B. A.

(1992). National systems of innovation: Towards a theory of innovation and interactive learning. Pinter Publishing.

25.

Lundvall

B. A.

(2016). The learning economy and the economics of hope. Anthem Press.

26.

Lundvall

B. A.

, & Johnson

(1994). The learning economy. Journal of Industry Studies, 1(2), 23–42.

27.

Mahambare

, & Balasubramanyam

V. N.

(2005). Trade liberalisation and India’s manufacturing sector. Development and Comp systems 0505010, University Library of Munich, Germany, https://www.researchgate.net/publication/23742484

28.

Malik

S. K.

(2015). Conditional technology spillovers from foreign direct investment: Evidence from Indian manufacturing industries. Journal of Productivity Analysis, 43(2), 183–198.

29.

Mondal

, & Pant

(2018). Firm capabilities and productivity spillovers from FDI: Evidence from Indian manufacturing firms. In Siddharthan

N. S.

, & Narayanan

(Eds.), Globalisation of technology (pp. 91–125). Springer.

30.

Mukharjee

, White

, & Wuyts

(1998). Econometrics and data analysis for developing countries (1st ed.). Routledge.

31.

Narayanan

(2004). Technology acquisition and growth of firms: Indian automobile sector under changing policy regimes. Economic and Political Weekly, 39(5), 461–70.

32.

Nelson

, & Winter

(1982). An evolutionary theory of economic change. The Belknap Press of Harvard University Press.

33.

Pack

(1988). Industrialization and trade. In Chenery

, & Srinivasan

T. N.

(Eds.), Handbook of development economics (vol. 1, pp. 334–373). NorthHolland.

34.

Parameswaran

(2002). Economic reforms and technical efficiency: Firm level evidence from selected industries in India (Working paper 339). Centre for development studies (CDS).

35.

Penrose

(1959). The theory of the growth of the firm. M.E. Sharpe.

36.

Pietrobelli

, & Rabellotti

(2011). Global value chains meet Innovation systems: Are there learning opportunities for Developing countries. World Development, 39(7), 1261–1269.

37.

Pradhan

J. P.

, & Singh

(2009). Outward FDI and knowledge flows: A study of the Indian automotive sector. International Journal of Institutions and Economies, 1(1), 156–187.

38.

Rijesh

(2015). Technology import and manufacturing productivity in India: Firm level analysis. Journal of Industry, Competition and Trade, 15(4), 411–434.

39.

Romer

P. M.

(1986). Increasing returns and long-run growth. Journal of Political Economy, 95(5), 1002–1037.

40.

Rosenberg

(1982). Inside the black box: Technology and economics. Cambridge University Press.

41.

Satpathy

L. D.

, Chatterjee

, & Mahakud

(2017). Firm characteristics and total factor productivity: Evidence from Indian manufacturing firms. Margin: The Journal of Applied Economic Research, 11(1), 77–98.

42.

Sharma

(2010). Does productivity differ in domestic and foreign firms? Evidence from Indian machinery industry. Indian Economic Review, 45(1), 87–110.

43.

Siddharthan

N. S.

, & Pandit

B. L.

(1998). Liberalisation and investment: Behaviour of MNEs and large corporate firms in India. International Business Review, 7, 535–548.

44.

Stiglitz

(2002). Globalization and its discontents. W.W. Norton and Co.

45.

Veugelers

(1997). Internal R&D expenditures and external technology sourcing. Research Policy, 26(3), 303–315.

46.

Veugelers

, & Cassiman

(1999). Make and buy in innovation strategies: Evidence from Belgian manufacturing firms. Research Policy, 28(1), 63–80.

47.

Williamson

O. E.

(1981). The economics of organization: The transaction cost approach. American Journal of Sociology, 87(3), 548–77.