Schumpeterian cycles with pro-cyclical R&D

Recent empirical work finds that R&D expenditures are quite pro-cyclical even for firms that are not credit-constrained during downturns. This has been taken as strong evidence against Schumpeterian-style theories of business cycles that emphasize the idea that downturns in production may be good times to allocate labor towards innovative activities. Here we argue that the pro-cyclicality of R&D investment is in fact quite consistent with at least one of these theories. In our analysis we emphasize three key features of R&D investment relative to other types of innovative activity: (1) it uses knowledge intensively, (2) it is a long-term investment with uncertain applications, and (3) it suffers from diminishing returns over time.     

What is the optimal rate of R&D investment to maximize productivity growth?

Governments in modern economies devote much policy attention to enhancing productivity and continue to emphasize its drivers such as investment in R&D. This paper analyzes the relationship between productivity growth and levels of R&D investments. The econometric analysis shows that more than 65 per cent of productivity growth variance is due to its dependence on gross domestic expenditure on R&D expressed as percentage of GDP (GERD). Economic analysis shows that productivity growth = f(GERD) is a concave function downwards due to diminishing returns to research investments. In addition, the research shows that the range of GERD between 2.3 per cent and 2.6 per cent maximizes the long-run impact on productivity growth and it is the key to sustained productivity and technology improvements that are becoming more and more necessary to modern economic growth.     

Nonparametric quantile regression analysis of R&D-sales relationship for Korean firms

This paper applies the nonparametric quantile regression estimation procedure to the analysis of the innovation-firm size relationship using Korean manufacturing firms data. Due to the high asymmetric distribution of R&D expenditure, the mean regression does not capture properly the stylized facts of R&D behavior; hence it underestimates the sales elasticity. Comparing the parametric estimates and nonparametric estimates allows us to see that there exists a nonlinear relationship in innovative activity and sales. Dividing the data into three groups according to the sales volume, the elasticity in the medium-sized firms is the biggest for scientific firms. This result conforms that the findings of Scherer (1965) coincide with findings from Korean manufacturing firms data in the sense that R&D expenditure tends to increase faster than firm size with size up to a point and then more slowly among larger firms. For the non-scientific firms, it steadily increases showing increasing returns to scale in innovative activity in large firms.     

Property rights protection and corporate R&D: Evidence from China

Following the recent literature on institutions and economic growth, we examine the effects of property rights protection on corporate R&D. Using a unique 2003 World Bank survey of over 2400 firms in 18 Chinese cities, we obtain the following findings: (1) property rights protection is positively and significantly related to corporate R&D activity (for both process and product R&D); (2) government services and helping hand are conducive to corporate R&D, while informal payments to government officials are not; and (3) government ownership of firms and direct appointment of CEOs are negatively associated with corporate R&D activities. We also find that corporate R&D is positively related to firm size, and access to finance, but negatively related to product market competition and firm age.     

Evaluating R&D performance of EU countries using categorical DEA

Over the coming decade, Research and Development (R&D) performance will be the key component of bringing innovation and the determinant of global competitiveness of nations. Therefore, this paper presents categorical Data Envelopment Analysis (DEA) for evaluating R&D performance of European Union (EU) countries. We utilise the output-oriented constant returns to scale (CRS) and variable returns to scale (VRS) DEA models with categorical data, namely, CAT-O-C and CAT-O-V models. In addition to DEA based framework, to examine the relationship between R&D performance and political-regulatory-economic situation of the countries; three research hypotheses are stated and their results are analysed. Policy implications about R&D activities can be derived for EU countries from the findings of this study.     

International R&D spillovers and institutions

The empirical analysis in “International R&D Spillovers” [Coe, D., Helpman, E., 1995. International R&D Spillovers. European Economic Review, 39, 859-887] is first revisited on an expanded data set that we have constructed for the purpose of this study. The new estimates confirm the key results reported in Coe and Helpman about the impact of domestic and foreign R&D capital stocks on TFP. In addition, we show that domestic and foreign R&D capital stocks have measurable impacts on TFP even after controlling for the impact of human capital. Furthermore, we extend the analysis to include institutional variables. Our results suggest that institutional differences are important determinants of TFP and that they impact the degree of R&D spillovers. Countries where the ease of doing business and the quality of tertiary education systems are relatively high tend to benefit more from their own R&D efforts, from international R&D spillovers, and from human capital formation. Strong patent protection is associated with higher levels of total factor productivity, higher returns to domestic R&D, and larger international R&D spillovers. Finally, countries whose legal systems are based on French and, to a lesser extent, Scandinavian law benefit less from their own and foreign R&D capital than countries whose legal origins are based on English or German law.     

R&D productivity: A global multi-industry comparison

This study considers the productivity of R&D expenditures on an international and multi-industry basis. Using 1992 data reported by Business Week on 117 companies in the United States, Europe, and Japan, we examined two measures of innovative output, patents and impact-adjusted patents, in relationship to R&D spending. Our results clearly show a decreasing returns to scale of R&D expenditure. Although this pattern holds true regardless of industry or global region, the level of return and rate of decreasing returns vary by both industry and international region.     

Are there diminishing returns to R&D?

Semi-endogenous models and, to some extent, also Schumpeterian models are based on the assumption of diminishing returns to R&D. This paper shows that the null hypothesis of constant returns to R&D cannot be rejected for the OECD countries.     

Research and development and labour productivity: do high-tech firms exhibit labour- or capital-saving technical change?

Employment and output in the advanced technology sectors have generally exhibited above-average growth for more than two decades. While this industry accounts for a relatively small share of total employment, the majority of private sector research and development (R&D) expenditures in the US is concentrated within seven sub-sectors. However, little attention has been paid as to whether high-tech productivity exhibits Hicksian capital or labour ‘savings’ bias or tendency to displace either factor input over time. Biased technical change can occur as economies transition between growth regimes. An augmented production function is employed to analyse the additional impact of R&D activity on firm-level labour productivity. A panel data set comprised of high-tech firms located across the advanced economies, China and India from 1990 to 2013 is used in the analysis. Labour-saving technical change was present across the advanced technology sectors and most countries. The expanded models of labour productivity that used fixed effects with lagged regressors confirmed the prior results as well as finding that R&D per employee, relative R&D intensity and firm market share contribute to firm-level labour productivity growth across countries and sectors. Additional support was found for diminishing returns to scale but not for R&D spillover effects.     

Too little or too much R&D?

According to the first generation models of endogenous growth based on expanding product variety, the market economy unambiguously generates too little R&D. Later, by disentangling returns to specialization from the market power parameter, it was shown that with sufficiently low returns to specialization too much R&D can occur. The present paper takes a step further, disentangling the market power parameter from the capital share in final output. At a theoretical level this helps finding too much R&D as well. On the other hand, in view of the empirically realistic order of magnitude between the parameters, disentangling market power and capital share tends to diminish the scope for excess R&D. Finally, by differentiating between net and gross returns to specialization we demonstrate what drives the differing inefficiency results in this literature.     

Product differentiation, process R&D, and the nature of market competition

We investigate the relationship between process and product R&D and compare the incentives for both types of R&D under different modes of market competition (Bertrand versus Cournot). It is shown that: (i) process R&D investments increase with the degree of product differentiation and firms invest more in product R&D when they can do process R&D than when they cannot; (ii) Bertrand firms have a stronger incentive for product R&D whereas Cournot firms invest more in process R&D; and (iii) cooperation in product R&D promotes both types of R&D relative to competition whereas cooperation in both types of R&D discourages R&D relative to cooperation in just product R&D.     

Technology variation vs. R&D uncertainty: What matters most for energy patent success?

R&D is an uncertain activity with highly skewed outcomes. Nonetheless, most recent empirical studies and modeling estimates of the potential of technological change focus on the average returns to research and development (R&D) for a composite technology and contain little or no information about the distribution of returns to R&D – which could be important for capturing the range of costs associated with climate change mitigation policies – by individual technologies. Through an empirical study of patent citation data, this paper adds to the literature on the outcomes of energy R&D by focusing on the behavior of the most successful innovations for six energy technologies, allowing us to determine whether uncertainty or differences in technologies matter most for success. We highlight two key results. First, we compare the results from an aggregate analysis of six energy technologies to technology-by-technology results. Our results show that existing work that assumes diminishing returns but assumes one generic technology is too simplistic and misses important differences between more successful and less successful technologies. Second, we use quantile regression techniques to learn more about patents that have a high positive error term in our regressions – that is, patents that receive many more citations than predicted based on observable characteristics. We find that differences across technologies, rather than differences across quantiles within technologies, are more important. The value of successful technologies persists longer than those of less successful technologies, providing evidence that success is the culmination of several advances building upon one another, rather than resulting from one single breakthrough. Diminishing returns to research activities appear most problematic during rapid increases of research investment, such as experienced by solar energy in the 1970s.     

Technology variation vs. R&D uncertainty: What matters most for energy patent success?

R&D is an uncertain activity with highly skewed outcomes. Nonetheless, most recent empirical studies and modeling estimates of the potential of technological change focus on the average returns to research and development (R&D) for a composite technology and contain little or no information about the distribution of returns to R&D – which could be important for capturing the range of costs associated with climate change mitigation policies – by individual technologies. Through an empirical study of patent citation data, this paper adds to the literature on the outcomes of energy R&D by focusing on the behavior of the most successful innovations for six energy technologies, allowing us to determine whether uncertainty or differences in technologies matter most for success. We highlight two key results. First, we compare the results from an aggregate analysis of six energy technologies to technology-by-technology results. Our results show that existing work that assumes diminishing returns but assumes one generic technology is too simplistic and misses important differences between more successful and less successful technologies. Second, we use quantile regression techniques to learn more about patents that have a high positive error term in our regressions – that is, patents that receive many more citations than predicted based on observable characteristics. We find that differences across technologies, rather than differences across quantiles within technologies, are more important. The value of successful technologies persists longer than those of less successful technologies, providing evidence that success is the culmination of several advances building upon one another, rather than resulting from one single breakthrough. Diminishing returns to research activities appear most problematic during rapid increases of research investment, such as experienced by solar energy in the 1970s.     

Organization Design and Information-Sharing in a Research Joint Venture with Spillovers

The paper presents a non-tournament model of process innovation with spillovers in the R&D process when firms engage in Cournot competition in the product market. It is shown that careful modelling of information-sharing and coordination of research activities leads to the conclusion that a Research Joint Venture (RJV) will economize on scarce R&D resources. There is an analysis of the effects of R&D cooperation, in the form of an RJV, on the organization of R&D, i.e. the efficient number of research labs. R&D expenditure, which precedes production, results in lower unit costs. R&D is modelled as a two-stage process: in the first stage, firms incur expenditure that will generate new knowledge, while in the second stage this knowledge is employed to reduce unit costs. A distinction is made between single and complementary research paths. It is shown that the RJV will operate one lab in the case of a single research path exploiting its coordination advantage. In the case of complementary research paths the number of labs the RJV will operate crucially depends on the stage of the R&D process at which diminishing returns occur: it will operate both labs when diminishing returns occur at the first stage (creation of knowledge), while it will be indifferent as to the number of labs, one or two, when diminishing returns occur in the second stage (cost reduction).     

Technology Strategy And Product Diversity

This paper considers a three-stage game of a differentiated oligopoly: firms first make their entry decisions, then they choose production technologies and in the third stage of the game they decide product prices. The technology choice can be understood as selecting one from a pool of those recently available as well as developing a new technology through innovative activities. The resulting market equilibrium is then compared with the social optimum. The main conclusions are that a monopolistically competitive market will typically undersupply both product variety and production scale. R&D competition in a free entry differentiated oligopoly will lead to insufficient R&D investment at firm and industry levels.     

An empirical analysis of firm process and product patenting

This paper presents an empriical study of the determinants of firm patenting. Since industrial research and development (R&D) encompasses a variety of activities, this study distinguishes between patents on process innovations and patents on product innovations. The property rights provided by a patent may differ between process and product patents, which suggests that the determinants of process innovations and product innovations may difer as well.

Several recent studies have distinguished between research directed toward process innovations and research directed toward product innovations. Scherer (1983a) included measures of process R&D spending and product R&D spending in regressions for inter-industry differences in productivity growth. Link and Lunn (1984) found that the returns to process-related R&D activity exceed the returns to product-related R&D activity. Link (1982) found inter-firm differences in the allocation of R&D spending for process innovations and product innovations. Lunn (1986) found differences in the determinants, as well as the effects, of process patenting and product patenting at the industry level. This paper examines whether the determinants of product and process patenting differ at the firm level.     

The struggle to survive in the R&D sector: Implications for innovation and growth

By allowing for investment activities by research and development (R&D) firms to prevent product obsolescence, we show that if legal patent protection is too strong, a higher R&D subsidy rate delivers insufficient investments for survival in the R&D sector, depressing innovation and growth in the long run.     

Non‐Scale Endogenous Growth with R&D and Human Capital

This paper examines the conditions under which increasing knowledge, encapsulated in ideas for new technology through R&D and embodied in human capital through education, sustains economic growth. A general model is developed where, consistent with recent literature, growth is non‐scale (not increasing in population size) and endogenous (generated by factors within R&D and education). Recent models feature the counterfactual assumption of constant returns to existing knowledge and restrict the substitutability of inputs within R&D and education. We find that non‐scale endogenous growth is possible under less stringent conditions. The findings reconcile sustained economic growth with evidence of diminishing marginal returns in education and R&D, which suggests an ambiguous role for R&D policy.     

The Causality Between R&D And Investment

The aim of this paper is to investigate the relationship between R&D expenditure and investment in machinery and equipment in order to test for causality. New growth theory emphasises the role of R&D in creating blueprints needed to produce new capital goods implicitly assuming causality running from R&D to investment. Other recent studies using firm level data have investigated the relationship between innovative activity and investment in fixed capital. In this paper we use aggregate data from the US economy on R&D expenditure in the industrial sector and aggregate investment in machinery and equipment. Standard Granger causality tests, together with the Hsiao version, are then performed, showing that causality runs from R&D to investment. In addition we perform a cointegration analysis allowing a test of possible long-run feedbacks. This dynamic representation shows that any feedback between investment and R&D is only significant in the long run.