«Published Annually Vol. 6, No. 1 ISBN 978-0-979-7593-3-8 CONFERENCE PROCEEDINGS Sawyer School of Business, Suffolk University, Boston, Massachusetts ...»
The model contains supply side control variables of profitability (PBT). Apart from the expected returns, financial factors form another basic set of considerations relevant to R&D expenditure decision making. The relationship between financial factors and R&D investment has been extensively explored in literature. Hall (1992) explored different means to fund R&D investment by firms including cash flow, debt, and equity. Firms prefer internal source including retained earnings or internally generated funds to finance R&D expenditure. Since, R&D expenditure involves extreme riskiness, moral hazard problem, and extensive transaction costs firms rely on internal resources for funding innovative activities. Gosh (2010) has explored different funding avenues for the R&D expenditure for Indian industry and finds evidence supporting Hall (1992).
Technology Related Factors
The technology related factors include availability of technology in the market through embodied form in capital imports (CI) and disembodied form through licensing (TECH) (Ray and Bhaduri 2001). The emerging economies like India can either create novel technology or buy the same in the international market for technology (Lall 2000). The technology bought may complement in-house research efforts or could be a substitute to the same. According to Ray and Bhaduri (2001), the capital imports necessitate Conference papers © Knowledge Globalization Institute, Pune, India, 2012 further R&D to adapt and absorb the technology whereas licensing imply direct import of technology reducing indigenous efforts.
Market Structure The relationship between the market structure and innovation has been extensively explored in literature. Schumpeter (1942) argue that the market power provides incentive to the producer to innovate. Furthermore, the ex-post monopoly power that a successful innovation may generate also entice firms to invest in R&D. The vast empirical literature on this issue remains inconclusive (Cohen and Levin 1989). The present study use concentration ratio (CR4) to capture the market structure. At the same time, the higher product differentiation in the market is also expected to influence the firms to invest in the innovative activity to introduce newer products (Aggarwal 2000). In order to capture this aspect, the model includes advertisement intensity (ADV).
Appropriability problem, that is the ability of innovator to realize reasonable rate of return on their innovative activity, is the major reason for market failure associated with R&D expenditure. I have used two approaches to capture the patent policy changes for the Indian industry. The study uses an index based on the text of the patent policy that is most widely used in literature that is Ginarte and Park (1997) index (hereinafter GPI) updated by Park (2008). The authors have calculated index of patent rights for 110 countries for period 1960-2000 with a gap of five years. It is calculated as an unweighted average of five categories of patent laws: (i) extent of coverage, (ii) membership in international agreement, (iii) provision for loss of protection, (iv) enforcement mechanism, and (v) duration of protection. Each category is assigned value between 0 and 1 and index ranges between 0 The sensitivity analysis by authors show that the ranking of countries is insensitive to weights tested. and 5 with higher value indicating strong level of protection. Considering that the index will remain same for all the industries and to check for the robustness the paper includes three dummy variable. DUMM1 = 1 if 1994; DUMM2 = 1 if 1999; DUMM3 = 1 if 2002 based on the years in which the major policy changes were introduced in the Indian patent policy.
Data and Empirical Strategy
The data is collected from CMIE Prowess that provide firm level data for 25346 companies listed on India's major stock exchanges and others including the central public sector enterprises. The data is aggregated for 54 industries at 3 digit National Industrial Classification from 1990-2010. SAL and PAT are logarithmically transformed as other variables are in ratio form.
A panel data model poses the problem that the error term may be non-spherical due to omitted country and/or time-specific effects. The existing literature suggests two types of models to take care of country and time specific effects that are fixed effects model (FEM) and random effects model (REM). The choice between the two models depends upon whether we treat country and/or time specific effects as „fixed‟ or „random‟. FEM that allows unobserved effects to be correlated with the included variables estimates individual effects as parameters. FEM, however, considerably reduces degrees of freedom. If country and/or time specific effects are strictly uncorrelated with the independent variables then these are to be modeled as randomly distributed (Greene 2005).
The choice between classical regression model on panel data and FEM is based on F test. Breuch and Pagan (1980) (BP) test based on ordinary least square residuals is used to test for REM against classical regression model. Under the null hypothesis, test statistics has chi-squared distribution with one degree of freedom. The alternative hypothesis suggests REM specification for panel data. The choice between FEM and REM is based on Hausman (1978) test. Under the null hypothesis REM estimator is BLUE, consistent and asymptotically efficient. REM estimator is, however, inconsistent if the null hypothesis is not true. FEM estimator is unbiased and consistent whether the null hypothesis is true or not. The test statistic has chi-squared distribution with k-1 degrees of freedom.
The macro-panel data with large N and T also poses the problem of unit toot for the series. There are different test to check the problem of unit root in the panel data (Baltagi 2008) including test by Levin-Lin-Chu (LLC) (2002), Im-Pesaran-Shin (2003), and Choi Fisher-type (2001). The study uses Fisher type test as Maddala and Wu (1999) show that it is superior to LLC. The null for the Fisher test is that all panels contain unit roots. Each variable is checked for unit root.
Conference papers © Knowledge Globalization Institute, Pune, India, 2012
Another problem in the model is that of endogenity as with respect to most of the independen t variables there may exist two-way relationship with research intensity. For instance, the research intensity that may determine the competitiveness of the firms may positively influence exports. With respect to market structure variables as well any successful innovation will also influence the concentration ratio and the extent of product differentiation. Moreover, the intensity of technology imports (both embodied and disembodied) will influence the research intensity. In order to control for the endogenity the variables are taken in lag form (Kumar and Aggarwal 2005). Table 1 gives the summary statistics as well as the correlation
matrix of the variables. The final model estimated is:
RDIit f (SALit, EXPit, PBTit,CIit,TECHit,CR4it, ADVit, PATt)
The results of the unit root test reveals non-stationarity of RDI and SAL, accordingly I firstdifferenced the series and rechecked for unit root. The test rejected the null for the differenced series. The result of the Hausman test show that REM is the appropriate model as the null is not rejected with the (χ2 = 10.66; p= 0.15). Further, BP test statistic rejects the null (χ2 = 16.26; p= 0.0001) that also suggests preference for REM. Accordingly, I report the results of REM. The results given include robust standard error in order to account for heteroscedasticity.
The results are given in the Table 2. At the industry level only two factors have a considerable impact on the research intensity. In line with the existing literature the study finds significant impact of the growing demand on the research intensity (SAL). The demand growth provide incentive to the producers to innovate new products and devote more resources to R&D. The variable is significant in all the models.
The changes in the patent policy over a period of time has a positive and significant impact on the research intensity of the industries. Column 3 and 4 show that the results whereby the dummy variables captures the policy changes. The coefficient of the dummy showing change after 1994 is significant in both the model. After 1994, India as a founding member of WTO was required to make legislative changes in the patent law to comply with TRIPs agreement. It appears the anticipation has lead to significant increase in the research intensity. The coefficient of the dummy after 1999 is not significant when the changes were made. However, the coefficient of the dummy for 2002 is significant and positive. After 2002 product patent was allowed for food and pharmaceutical industry that has an impact on the decision of the producers to invest in R&D.
The model does not find any significant impact of the CI and TECH. It appears that the relationship between the research Conference papers © Knowledge Globalization Institute, Pune, India, 2012 intensity and technology is more complex. It may require further investigation probably in terms of simultaneous model that is not warranted in the given space. Market structure variables and profitability have no significant impact on the RDI.
The relationship between the patent protection and innovation is conditioned by industry-specific characteristics. In order to check the robustness of the results to such characteristics I introduced a dummy for industries that are more sensitive to protection based on Cohen et al. (2000). Interestingly, the coefficient of the dummy was not significant and it did not affect the coefficient of other variables.
The theoretical literature on the impact of patent policy on innovation advocate ambiguous influence. The appropriability thesis states that the firms will invest more in R&D as they will be able to appropriate the sunk costs if the patent protection is strong. The disclosure requirements will ensure that the technology is available in the public domain to conduct further research instead of being hidden. Alternately, patent will lead to competition that may lead the firms to reduce the R&D expenditure. Furthermore, defensive patenting may have a negative influence on the innovation in an industry. Therefore, the question of the influence of patent protection on innovation is essentially empirical. This paper fills the gap and studies the impact of the patent policy changes on the R&D expenditure of the Indian industry using data from 54 industries at 3 digit NIC classification for the time-period 1990-2010. The results show that the policy changes have a positive impact on the innovation. The results also reveal that the growing market demand also induce firms to spend more R&D activities.
Aggarwal, A. 2000. “Deregulation, Technology Imports and In-House R&D Efforts: An Analysis of Indian Experience.” Research Policy 29 (9): 1091-1093.
Allred, B.B., and W.G. Park. 2007. “Patent Rights and Innovative Activity: Evidence from National and Firm-level Data.” Journal of Conference papers © Knowledge Globalization Institute, Pune, India, 2012 International Business Studies 38: 878-900.
Baltagi, B.H. 2008. Econometric Analysis of Panel Data. 4th ed. West Sussex: John Wiley & Sons.
Chen, Y., and T. Puttitanun. 2005. “Intellectual Property Rights and Innovation in Developing Countries.” Journal of Development Economics 78 (2): 474-493.
Cohen, W.M., and R.C. Levin. 1989. “Empirical Studies of Innovation and Market Structure.” In Handbook of Industrial Organization, Vol. 2, ed. R. Schmalensee and R. D. Willig. North-Holland: Amsterdam.
Cohen, W.M., R.R. Nelson, and J.P. Walsh. 2000. “Protecting their Intellectual Assets: Appropriability Conditions and Why US Manufacturing Firms Patent (or not).” NBER Working Paper No. 7552, Cambridge, MA: National Bureau of Economic Research.
Cohen, W.M., and D.A. Levinthal. 1989. “Innovation and Learning: The Two Faces of R&D.” The Economic Journal 99 (397): 569Evenson R., and K.J. Joseph. 1997. “Foreign Technology Licensing in Indian Industry: An Econometric Analysis of the Choice of Partner, Terms of Contract and the Effect on Licensees' Performance. ” Economic Growth Centre, Yale University, Working Paper no. 273.
Evenson, R., and L. Westphal. 1995. “Technological Change and Technology Strategy.” In Handbook of Development Economics Vol. 3A, ed. J. Behrman and T. Srinivasan. Amsterdam: North Holland Publishing Company.
Geroski, P.A., and C.F. Walters. 1995. “Innovative Activity over the Business Cycle.” Economic Journal 105 (431): 916-928.
Ginarte, J.C., and W.G. Park. 1997. “Determinants of Patent Rights: A Cross-National Study.” Research Policy 26 (3): 283-301.
Ghosh, S. 2010. “R&D in Indian Manufacturing Enterprises: What Shapes it?” Economics of Innovation and New Technology 18 (4): 337-352.
Grabowski, H.G. 1968. “The Determinants of Industrial Research and Development: A Study of the Chemical, Drug, and Petroleum Industries.” The Journal of Political Economy 76 (2): 292-306.
Greene, W. H. 2005. Econometric Analysis. 5th ed. New Delhi: Pearson Education.
Griliches, Z. 1990. “Patent Statistics as Economic Indicators: A Survey.” Journal of Economic Literature 28 (4): 1661-1707.
Hall, B.H. 1992. “Investment and R&D at the Firm Level: Does the Source of Financing Matter?” NBER Working Paper No. 4096, Cambridge, MA: National Bureau of Economic Research.
Hausman, J. 1978. “Specification Tests in Econometrics.” Econometrica 46 (6):1251-1271.
Kanwar, S., and R.E. Evenson. 2003. “Does Intellectual Property Protection Spur Technological Change?” Oxford Economic Papers 55 (2): 235-264.
Kaufer, E. 1989. The Economics of the Patent System. New York: Harwood Academic Publishers.
Kortum, S., and J. Lerner. 1998. “What is Behind the Recent Surge in Patenting?” Research Policy 28 (1):1-22.
Kumar, N. and A. Aggarwal. 2005. “Liberalization, Outward Orientation and In-House R&D Activity of Multinational and Local Firms:
A Quantitative Exploration of Indian Manufacturing.” Research Policy 34 (4): 441-460.