By: Alex Coad, CENTRUM Católica Graduate Business School, Santiago de Surco, Lima, Peru, email@example.com
Nanditha Mathew, Sant’Anna School of Advanced Studies, Italy, firstname.lastname@example.org
Emanuele Pugliese, Institute of Complex Systems, CNR, Rome and Department of Economics, University of Bath, email@example.com
For Citation: SITE4Society Brief No.10-2018
Related to SDG Goals and Indian National Programmes: #SDG8 (Decent Work and Economic Growth) #SDG9 (Industry, Innovation and Infrastructure) #IIGP (India Innovation Growth Programme) #AIM (Atal Innovation Mission)
Economics sub-disciplines: Industrial organization, Development economics, Economics of innovation.
SITE focus: Innovation
Country focus: India
Based on: Alex Coad, Nanditha Mathew and Emanuele Pugliese, “What’s good for the goose ain’t good for the gander:cock-eyed counterfactuals and the performance effects of R&D” LEM Working Paper, No. 2017/21.
Context: In developed and emerging countries, government policies target firms in the manufacturing sector to encourage them to engage in innovation. Even though there is a consensus at a macro level that innovation is a main driver of economic growth, at a micro level, the relation between innovation and growth is not clear. Previous empirical studies at the micro-firm level have revealed confounding results concerning this relationship, and thus, there is no consensus on whether or not R&D undertaking improves firm performance. Such lack of clarity questions the policies that assume that innovation is beneficial for all firms that engage in innovation. Moreover, most of the studies deal with firms in developed countries and it is not clear whether such results hold for emerging country firms. Hence, in this work, we investigate the innovation-firm performance relationship of Indian manufacturing firms, keeping in mind the findings of previous studies that have highlighted the firm characteristics that distinguish ‘innovative’ from ‘non-innovative’ firms. Finally, with respect to Indian manufacturing firms, we explore the counterfactual story: what would have been the performance effects of firms that did not invest in R&D, if they actually did and vice-versa.
Research Questions: Which kinds of firms spend on research and development? Do all firms benefit from research and development spending?
Motivation for Research Questions: It is indeed crucial to disentangle and understand clearly the relationship between innovation and performance at a micro level, given the interest in the topic from both firm managers and policy makers. It is noteworthy that several studies investigating the relationship between R&D and firm performance in terms of growth, productivity or profitability have shown conflicting results including – a positive relationship (Hall, 1987; Singh, 1994; Del Monte and Papagni, 2003), no relationship (Geroski et al., 1997; Bottazzi et al., 2001; Stam and Wennberg, 2009) or a negative relationship (Brouwer et al., 1993; Freel and Robson, 2004).
Furthermore, studies have identified that innovators have some characteristics that are distinctly different from non-innovators (Hall et al., 2010; Audretsch et al.,2014), but ironically, much of the studies have not properly empirically accounted for them and this could be the reason for contradicting results observed by previous studies.
Indeed, the effectiveness of technological investments such as R&D investment is likely to depend on innovative capabilities that are correlated with both the R&D decision and the expected performance benefits. Firms without these capabilities would be ill-prepared to invest in R&D and would probably not benefit from R&D if they did. The irony of the standard econometric approach, however, is that performance benefits from R&D are evaluated by assuming homogeneous effects for innovators and non-innovators. In this work, we demonstrate and emphasize that a non-R&D investor need not be a good counterfactual for an R&D investor.
Methodology: The Prowess database, provided by the CMIE (Centre For Monitoring Indian Economy Pvt. Ltd.) was used. Firm performance was measured using two variables: sales growth of the firm and firm profitability. The R&D expenditure of firms was taken as their investment in innovation.
The innovation performance linkage was evaluated by performing an OLS and Fixed Effect estimation. This gave us inconsistent results: the R&D dummy did not have any significant effect on firm growth, and furthermore had significant negative effects on relative profitability.
However, as we pointed out before, innovators and non-innovators could be different from one to the other, and considering them the same could have been the reason for the poor results.
We empirically investigated this using econometric techniques suited for dummy dependent variable, like Linear Probability model (Logit), Probit and a random effects Probit estimation and confirmed this to be the case.
Hence, we faced two issues: first the issue of selection, i.e, the firms that invest in R&D are different from those that didn’t, and second, the issue of censored data, i.e R&D spending can only be a positive value. To tackle these issues, we applied the endogenous switching regression which allows to compute the counter-factuals in order to calculate the effect of investing in R&D for firms that actually did R&D and the effects of R&D investment for those that did not. In other words, using this method we could confirm the benefits of R&D investment for the two firm clusters separately. The results we obtain with this method are clear and sharp.
- Firms that invest in R&D and those that don’t differ significantly in terms of several observed firm characteristics. Firms performing R&D are bigger, younger, exporters, with high capital investments and financial leverage (i.e. total borrowings over total assets).
- This leads to a self-selection bias, i.e. when a firm is bigger, exporting and has a lot of capital equipment – it probably has to engage in R&D.
- Once controlled for the selection bias, it is possible to make a causal interpretation of the results. Our results indicate that both firms that invest in R&D and those that do not are making a rational decision.
- Indian firms investing in R&D would have had less growth and less relative profitability if they had not invested in R&D and firms that did not invest in R&D would not have had better performance if they had done so. This implies that even if policy makers observe performance benefits of innovative activity among innovative firms, nevertheless this should not be taken to imply that there might be any benefits from encouraging non-innovative firms to invest in R&D.
- Our approach and results provide a possible explanation for why there is no consensus in the existing literature on the performance benefits of innovative activity. This could have been either because of the nature of the information available on firms or because the studies did not distinguish between low and high R&D investors.
- Our study demonstrates that firms investing in R&D can be making a rational decision and that firms that are not investing in R&D can also be making a rational decision. This should remind econometricians that firms are not as irrational as we might sometimes think, and that their decisions are based on much more information than are included in standard econometric specifications.
In academic and policy debates, it is often assumed that for all firms, R&D spending is beneficial in all circumstances. But this is not true – what is good for the goose might not be good for the gander!
For example, our study indicates that in the Indian manufacturing sector, R&D spending is good for firms, which are younger, exporters, with high capital investments and low financial leverage, while R&D spending might not yield much benefits for firms that have never done R&D before. In short, R&D investment is not for everyone!
Hence, firms that do not do R&D could be having good reasons to do so and vice-versa. This would indicate that a homogeneous economy wide tax breaks on R&D might not be an optimal policy. Innovation policies should be sector and technology specific, and any incentive should be shaped considering the characteristics of the firms involved. This is important to maximize the effects of the incentives, to ensure that firms do not undertake R&D for mere tax benefits, without any innovation output.
Sources of images: