Abstract
The main objective of this study is to investigate the impact of corporate research and development (R&D) activities on firm performance, measured by labour productivity. To this end, the stochastic frontier technique is used on a unique unbalanced longitudinal dataset comprising top European R&D investors over the period 2000–2005. In this framework, this study quantifies technical inefficiency of individual firms. From a policy perspective, the results of this study suggest that if the aim is to leverage firms’ productivity, the emphasis should be put on supporting corporate R&D in high-tech sectors and, to some extent, in medium-tech sectors. On the other hand, corporate R&D in the low-tech sector is found to have a minor effect in explaining productivity. Instead, encouraging investment in fixed assets appears important for the productivity of low-tech industries. Hence, the allocation of support for corporate R&D seems to be as important as its overall increase and an ‘erga omnes’ approach across all sectors appears inappropriate. However, with regard to technical efficiency, R&D intensity is found to be a pivotal factor in explaining firm efficiency and this turns out to be true for all industries.
Similar content being viewed by others
Notes
See Sect. 2 for an overview of the relevant literature on the subject.
In fact, for high-tech firms the R&D elasticity was found to be highly significant ranging from 0.125 to 0.176, while for the remaining firms the R&D elasticities were either not significant (although positive) or systematically lower (ranging from 0.090 to 0.096). These results were based on different estimation techniques.
In this regard it is commonly argued that by carrying out R&D activities within a certain company tacit knowledge is accumulated (useful also beyond the corresponding R&D project), which is assumed to reduce technical inefficiency (avoiding wastes). Moreover, the same effect may be due to raising the awareness and understanding of cutting-edge technologies.
For example, Hunt-McCool et al.(1996) and Stanton (2002) on finance; Adams et al. (1999), Fernandez et al. (2000) and Lozano-Vivas and Humphrey (2002) on banking; Wadud and White (2000) and Zhang (2002) on agriculture; Reinhard et al. (1999) and Amaza and Olayemi (2002) on environmental economics; Perelman and Pestieau (1994) and Worthington and Dollery (2002) on public economics; Pitt and Lee (1981) and Thirtle et al. (2000) on development economics.
The analysis is based on accounting data using 117 agricultural enterprises and 43 light manufacturing industries for the period 1985–1991.
These two hypotheses are tested using a panel of manufacturing industries across six European countries over the period 1980–1997.
The study by Bos et al. 2010b is based on 80 countries over the period 1970–2000. The model explicitly accounts for inefficiency, augmented with a latent class structure, which allows production technologies to differ across groups of countries. Membership of these groups is estimated instead of being determined ex ante.
Bos et al. 2010a model both the technology clubs and the parameters within each club as a function of R&D intensity. This framework makes it possible to explore the components of output growth in each club, potential technology spillover and catch-up issues across industries and countries.
For the DTI scoreboards, see www.innovation.gov.uk/rd_scoreboard (various editions available).
Although the DTI databases contain information from 14 European countries (Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, the Netherlands, Norway, Spain, Sweden, Switzerland and the United Kingdom), British firms are over-represented.
Measurement of R&D investment is subject to accounting definitions for R&D. For UK companies, the definition given in the Statement of Standard Accounting Practice (SSAP) 13 ‘Accounting for research and development’ is applied. For non-UK companies, R&D investment is defined in accordance with the International Accounting Standard (IAS) and corresponds to the R&D component of the accounting category 38 ‘Intangible assets’. Both figures are based on the OECD ‘Frascati Manual’ definition of corporate R&D and are therefore fully comparable.
In this study 28 of the original 39 DTI sectors were retained. Sectors with fewer than five firms were excluded (see “Appendix” Table 5).
For the definition of K, see below. Note that the Grubbs test, also known as the maximum normalised residual test, assumes normality (which is a desirable property anyway). Accordingly, normality tests were run on the relevant variables and this assumption was never rejected. Results from both Grubbs and normality tests are available from the authors upon request.
Merger and acquisitions were treated as a new entry and the firms that merged were labelled as ‘exit’ from the dataset.
For the detailed ICB sectoral classification, see http://www.icbenchmark.com.
Compared with the OECD classification, low-tech and medium–low-tech sectors were grouped together in order to have enough observations in each sectoral group. Out of the total of 1,787 observations, 516 fell into the low-tech sector, 671 into medium-tech and 600 into high-tech (see “Appendix” Table 5).
Note that these thresholds are significantly higher than those adopted by the OECD for the manufacturing sectors (2 and 5%, see Hatzichronoglou 1997). This is the obvious consequence of dealing with the top European R&D investors.
Only two sectors (automobiles and food) were upgraded (see “Appendix” Table 5). This is due to dealing with top R&D investors alone.
Technology club refers to the technology parameters characterising the corresponding efficient production frontier.
Durlauf and Johnson (1995) endogenised the division rule by applying a regression tree analysis in order to identify multiple technology clubs of cross-country growth behaviour. In their approach, both the parameters and the number of clubs result from applying a sorting algorithm to the whole sample, incorporating cost into sample splits to avoid over-parameterisation. However, for testing the hypotheses outlined above the more general approach suggested here may serve the purpose, since given the particular context of our study, the technological group as such and not the individual firms in it is what matters most.
Alternative approximations of the impact variables were tested (such as variables capturing level or ratios of the relevant inputs). However, the final restricted model in stocks avoids severe problems in modelling lags and also captures the 'learning by doing effect'.
See Table 5 in the “Appendix” for a detailed overview of OECD to ICB sectoral conversion. German sectoral figures were applied to Swiss firms because of the unavailability of corresponding OECD data.
Physical capital also embodies technology, and rapid technological progress makes scrapping more frequent.
Different depreciation rates were applied while estimating the panel data models as sensitivity checks in order to ensure robustness to the analysis. No discernable changes were seen in the results.
The original DTI dataset selects top UK and foreign R&D investors on the basis of aggregate rankings independent from sectoral representativeness. This implies a sectoral bias, presumably more severe in low-tech industries because only outstanding firms are included here. In general, the cut off point is a moving target (depending on the amount of companies the Scoreboard is envisaged to report on). For example, in the case of the 2001 DTI R&D Scoreboard, 500 international companies and 597 UK firms were included in descending order, based on their absolute R&D figures.
See, for example, Coelli et al. (1998) for a fairly general introduction to efficiency and productivity analysis.
For this purpose 'time' trend was introduced as a shift variable in the production function (Hicks-neutral technological change) and was found to be significant. 'Time' trend was also tested as explanatory variable in the inefficiency term (but found to have an insignificant impact in this regard). See Sect. 5 and the discussion of the empirical results for more details.
An alternative way to introduce determinants of inefficiency is to make the mean of u a function of exogenous variables, thereby using a truncated normal distribution. The variance of u in this formulation can also be a function of exogenous variables. The model with truncated normal distribution in which both the mean and variance are functions of exogenous variables did not converge. The model we used here is more parsimonious and easy to get convergence.
See Sect. 3.4 of Kumbhakar and Lovell (2000) for an extensive discussion on these extensions and problems in ignoring them while estimating inefficiency.
For more details see Kumbhakar et al. (2009) or request these results directly to the authors.
Note that compared to more comprehensive samples the individual firm heterogeneity in our study tends to be limited. This is primarily because of 'picking the winner' which is due to the nature of selecting firms for a Scoreboard (i.e., including just the 'top R&D investors' and the 'top VA performers'). This limitation in terms of 'between firm variability' holds even more when we further split the sample into sub-samples (high-, medium-, low-tech firms). On the other hand, the use of stock variables rather than flows reduces the 'within firm variability'.
Note that although these variables were not used as input variables for the production frontier, they were used as explanatory variables for firm inefficiency.
Similar to the analyses of technical efficiencies at the sectoral level as presented in Sect. 6.1, the explanatory variable 'R&D intensity' for a certain firm in a given year represents R&D/VA, normalized by the corresponding industry average (in a particular year). Thus, it indicates over-or under-proportional R&D spending.
The marginal effects for the variable z were calculated from \( \partial E(u)/\partial z \) (see Wang 2002, for details).
More specifically, the marginal effect of R&D can be interpreted as the percentage change (when multiplied by 100) in (labour) productivity for a 1 point change (in a scale of 100) in its R&D intensity.
The correlation between TE and marginal effects of R&D intensity was found to be rather low (0.28, 0.21 and 0.24 for high-, medium- and low-tech, respectively). This indicates that the lower mean TE and the higher marginal effects of R&D intensity found for low-tech sectors compared with other industries are not an effect of the very nature of this sectors. Rather, this seems to be a result of the particularly high heterogeneity across the industries and companies grouped together as ‘low-tech’.
Although the variation of mean TE across sectors is substantial, for some sectors the estimated average, minimum and maximum TE scores should be treated with caution due to the small number of firms in that sector. For example, the oil equipment, services and distribution sector has a mean TE of 13.4% (minimum 4.1% and maximum 20.6%), but these figures are based on estimates comprising only 7 companies.
Interestingly, examples of this can be found across the three sectoral groups (see Table 4). For example, the marginal effects for aerospace and defence (high-tech) is 0%; general industrials (medium-tech) 0%; and construction and materials 1.4% (low-tech).
References
Acemoglu D, Zilibotti F (2002) Productivity differences. Quart J Econ 116:563–606
Adams R, Berger A, Sickles R (1999) Semiparametric approaches to stochastic panel frontiers with applications in the banking industry. J Bus Stat 17:349–358
Aigner DJ, Lovell KC, Schmidt P (1977) Formulation and estimation of stochastic frontier production function models. J Econom 6:21–37
Amaza P, Olayemi JK (2002) Analysis of technical inefficiency in food crop production in Gombe states, Nigeria. Appl Econ Lett 9:51–54
Barro RJ, Sala-i-Martín XX (2004) Economic growth. MIT Press, Cambridge, MA
Basu S, Weil D (1998) Appropriate technology and growth. Q J Econ 113:1025–1054
Becker B, Pain N (2002) What Determines industrial R&D expenditure in the UK?. National Institute of Economic and Social Research, London
Bischoff CW, Kokkelenberg EC (1987) Capacity utilisation and depreciation-in-use. Appl Econ 19:995–1007
Bönte W (2003) R&D and productivity: internal vs. external R&D—evidence from West German manufacturing industries. Econ Innov New Technol 12:343–360
Bos JWB, Economidou C, Koetter M, Kolari JW (2010a) Do all countries grow alike?. J Development Econ 91:113–127
Bos JWB, Economidou C, Koetter M (2010b) Technology clubs, R&D and growth patterns: evidence from EU manufacturing. Eur Econ Rev 54:113–127
Cameron G, Proudman J, Redding S (2005) Technological convergence, R&D, trade and productivity growth. Eur Econ Rev 49:775–809
Coelli TJ, Prasada Rao DS, Battese GE (1998) An Introduction to efficiency and productivity analysis. Kluwer Academic Publishers, Boston
Crépon B, Duguet E, Mairesse J (1998) Research, innovation and productivity: an econometric analysis at firm level. Econ Innov New Technol 7:115–158
Cuneo P, Mairesse J (1983) Productivity and R&D at the firm level in french manufacturing, NBER Working Paper No. 1068, National Bureau for economic research, Cambridge
Dosi G (1997) Opportunities, incentives and the collective patterns of technological change. Econ J 107:1530–1547
Durlauf SN, Johnson PA (1995) Multiple regimes and cross-country growth behaviour. J Appl Economet 10:365–384
European Commission (2002) More research for Europe. Towards 3% of GDP, COM (2002) 499 final, Brussels
European Commission (2010) Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, Europe 2020 Flagship Initiative Innovation Union, SEC (2010) 1161
European Council (2002) Presidency conclusions, Barcelona European Council, 15 and 16 March 2002, Brussels
Farrell MJ (1957) The measurement of productive efficiency. J Royal Statist Soc 253–281 (Series A, General 120, Part 3 (1957))
Fernandez C, Koop G, Steel MFJ (2000) Multiple-output production with undesirable outputs: an application to nitrogen surplus in agriculture. J Am Statist Assoc Appl Case Stud 97:432–442
Freeman C (1982) The economics of industrial innovation. Pinter, London
Fried HO, Lovell CAK, Schmidt SS (2008) The measurement of productive efficiency and productivity growth. Oxford University Press, New York
Griffith R, Redding S, Van Reenen J (2004) Mapping the two faces of R&D: productivity growth in a panel of OECD industries. Rev Econ Stat 86:883–895
Griliches Z (1979) Issues in assessing the contribution of research and development to productivity growth. Bell J Econ 10:92–116
Griliches Z (1986) Productivity, R&D and basic research at the firm level in the 1970s. Am Econ Rev 76:141–154
Griliches Z (1995) R&D and productivity: econometric results and measurement issues. In: P. Stoneman (ed) Handbook of the Economics of innovation and technological change. Blackwell Publishers Ltd., Oxford, pp 52–89
Griliches Z (2000) R&D, Education and productivity. Harvard University Press, Cambridge
Griliches Z, Mairesse J (1982) Comparing productivity growth: an exploration of French and US industrial and firm data, NBER Working Paper No. 961, National Bureau of economic research, Cambridge
Hall BH (2007) Measuring the returns to R&D: the depreciation problem, NBER Working Paper No. 13473, National Bureau of economic research, Cambridge
Hall BH, Mairesse J (1995) Exploring the relationship between R&D and productivity in French manufacturing firms. J Economet 65:263–293
Harhoff D (1998) R&D and productivity in german manufacturing firms. Econo Innovat New Technol 6:29–49
Hatzichronoglou T (1997) Revision of the high-technology sector and product classification. OECD, Paris
Heshmati A, Kim H (2011) The R&D and productivity relationship of Korean listed firms. J Prod Anal. doi:10.1007/s11123-010-0206-y
Hulten CR (1991) The measurement of capital. In: ER Berndt, Triplett JE (eds) Fifty years of economic management, University of Chicago Press, Chicago
Hunt-McCool J, Koh SC, Francis BB (1996) Testing for deliberate underpricing in the IPO premarket: a stochastic frontier approach. Rev Financial Stud 9:1251–1269
Janz N, Lööf H, Peters B (2004) Firm level innovation and productivity—is there a common story across countries? Probl Perspect Manag 2:1–22
Jones CI (2005) The shape of production functions and the direction of technical change. Quart J Econ 120:517–549
Jorgenson DW (1990) Productivity and economic growth. In: Berndt ER, Triplett JE (eds) Fifty years of economic growth, Chicago University Press, Chicago. pp 19–118
Kim S, Han G (2001) A decomposition of total factor productivity growth in korean manufacturing industries: a stochastic frontier approach. J Prod Anal 16:269–281
Klette J, Kortum S (2004) Innovating firms and aggregate innovation. J Political Econ 112:986–1018
Koop G (2001) Cross-sectoral patterns of efficiency and technical change in manufacturing. Int Econ Rev 42:73–103
Koop G, Osiewalski J, Steel MF (2000) Modelling the sources of output growth in a panel of countries. J Bus Econ Stat 18:284–299
Kumbhakar SC, Lovell CAK (2000) Stochastic frontier analysis. Cambridge University Press, Cambridge
Kumbhakar SC, Ortega-Argilés R, Potters L, Vivarelli M, Voigt P (2009) Corporate R&D and firm efficiency: evidence from Europe’s top R&D investors. In: Forschungsinstitut zur Zukunft der Arbeit, IZA DP No. 4657, Bonn/Germany, 12/2009
Kwon H, Inui T (2003) R&D and productivity growth in Japanese manufacturing firms. ERSI Discussion Paper Series No.44, Tokyo
Lichtenberg FR, Siegel D (1989) The Impact of R&D Investment on Productivity — New Evidence Using Linked R&D-LRD Data. NBER Working Paper No. 2901, National Bureau for Economic Research, Cambridge
Limam YR, Miller SM (2004) Explaining economic growth: factor accumulation, total factor productivity growth and production efficiency improvement, Department of Economics Working Paper Series, 2004–20, Connecticut University, USA
Lööf H, Heshmati A (2006) On the relation between innovation and performance: a sensitivity analysis. Econ Innov New Technol 15:317–344
Los B, Timmer MP (2005) The appropriate technology explanation of productivity growth differential: an empirical approach. J Dev Econ 77:517–531
Lozano-Vivas A, Humphrey DV (2002) Bias in Malmquist index and cost function productivity measurement in banking. Int J Product Econ 76:177–188
Mairesse J, Mohnen P (2001) To be or not to be innovative: an exercise in measurement, NBER working paper No. 8644, National Bureau of economic research, Cambridge
Mairesse J, Mohnen P (2005) The importance of R&D for innovation: a reassessment using French survey data. J Technol Transfer 30:183–197
Mairesse J, Sassenou M (1991) R&D and productivity: a survey of econometric studies at the firm level, NBER working paper No. 3666, National Bureau for economic research, Cambridge
Malerba F (2004) Sectoral systems of innovation. Università Commerciale Luigi Bocconi, Milano
Marsili O (2001) The anatomy and evolution of industries. Edward Elgar, Northampton
Meeusen W, van den Broeck J (1977) Efficiency estimation from Cobb-Douglas production functions with composed error. Int Econ Rev 18:435–444
Musgrave JC (1986) Fixed reproducible tangible wealth series in the united states, 1925–91. Surv Curr Bus 66:51–75
Nadiri MI, Prucha IR (1996) Estimation of the depreciation rate of physical and R&D capital in the US total manufacturing sector. Econ Inq 34:43–56
OECD (2005) OECD science, technology and industry scoreboard 2005. OECD Publishing, Paris
Orea L, Kumbhakar SC (2004) Efficiency measurement using a latent class stochastic frontier model. Emp Econ 29:169–183
Ortega-Argilés R, Piva M, Potters L, Vivarelli M (2010) Is corporate R&D investment in high-tech sectors more effective? Contemp Econ Policy 28:353–365
Pakes A, Schankerman M (1986) Estimates of the value of patent rights in european countries during the post-1950 period. Econ J 96:1052–1076
Parisi M, Schiantarelli F, Sembenelli A (2006) Productivity, innovation creation and absorption and R&D: microevidence for Italy. Eur Econ Rev 8:733–751
Pavitt K (1984) Sectoral patterns of technical change: towards A Taxonomy and a theory. Res Policy 13:343–373
Perelman S, Pestieau P (1994) A comparative performance study of postal services: a productive efficiency approach. Annales d’Économie et de Statistique 33:187–202
Piesse J, Thirtle C (2000) A stochastic frontier approach to firm level efficiency, technological change, and productivity during the early transition in Hungary. J Comparative Econ 28:473–501
Pitt MM, Lee LF (1981) The measurement and sources of technical inefficiency in the Indonesian weaving industry. J Dev Econ 9:43–64
Reinhard S, Lovell CAK, Thijssen G (1999) Econometric estimation of technical and environmental efficiency: an application to Dutch dairy farms. Am J Agric Econ 81:1 (February), 44–60
Rogers M (2006) R&D and Productivity in the UK: evidence from firm-level data in the 1990s, Economics series working papers No. 255, University of Oxford, Oxford
Sanders M, Bos J, Economidou C (2007) R&D over the life cycle. Tjalling C. Koopmans Research Institute discussion paper series No. 07–18
Scarpetta S, Tressel T (2002) Productivity and convergence in a panel of OECD industries: do regulations and institutions matter?, OECD economics department working papers No. 342
Stanton KR (2002) Trends in relationship lending and factors affecting relationship lending efficiency. J Bank Finance 26:127–152
Thirtle C, Bhavani S, Chitkara S, Chatterjee P, Mohanty M (2000) Size does matter: technical and scale efficiency in indian state tax jurisdictions. Rev Develop Econ 4:340–352
Tsai KH, Wang JC (2004) R&D productivity and the Spillover effects of high-tech industry on the traditional manufacturing sector: the case of Taiwan. World Econ 27:1555–1570
Verspagen B (1995) R&D and productivity: a broad cross-section cross-country look. J Prod Anal 6:117–135
von Tunzelmann N, Acha V (2005) Innovation in ‘Low-Tech’ industries. In: Fagerberg J, Mowery DC, Nelson RR (eds) The Oxford handbook of innovation, Oxford University Press, New York. pp 407–432
Wadud A, White B (2000) Farm household efficiency in Bangladesh: a comparison of stochastic frontier and DEA methods. Appl Econ 32:1665–1973
Wang HJ (2002) Heteroskedasticity and non-monotonic efficiency effects of a stochastic frontier model. J Prod Anal 18:241–253
Winter SG (1984) Schumpeterian competition in alternative technological regimes. J Econ Behav Organisation 5:287–320
Worthington AC, Dollery BE (2002) Incorporating contextual information in public sector efficiency analyses: a comparative study of NSW local government. Appl Econ 34:453–464
Zhang Y (2002) The impacts of economic reform on the efficiency of silviculture: a non-parametric approach. Environ Development Econ 7:107–122
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Kumbhakar, S.C., Ortega-Argilés, R., Potters, L. et al. Corporate R&D and firm efficiency: evidence from Europe’s top R&D investors. J Prod Anal 37, 125–140 (2012). https://doi.org/10.1007/s11123-011-0223-5
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11123-011-0223-5