When deciding about the strength of patent protection, policymakers face a trade-off between increased innovation incentives and higher market power of patent holders. This column studies the adoption of a TRIPS-compliant patent regime in India and finds that firms with a higher exposure to stronger patent protection increased patenting, quality-adjusted patent counts, and R&D expenditure. Firm-product level markups increased, but primarily due to lower marginal costs rather than higher prices. The results indicate that cost-reducing innovations and output expansion contributed to these cost savings.
While exclusive rights through stronger patents could incentivise firms to invest in research and development (R&D), they can lead to greater market power, which can reduce consumer surplus, decrease labour demand, and distort investment incentives (Arrow 1962, Bloom et al. 2019). The net gains from stronger patent protection remain theoretically ambiguous and have been widely debated (Boldrin and Levine 2013). Empirical evidence on whether patent protection boosts investment in innovation is mixed and often limited to specific markets. In addition, whether and how market power is affected on account of stronger patent laws remains underexplored.
In our recent paper (Gupta and Stiebale 2024), we bring cross-industry evidence on the effect of stronger patent protection in India on both innovation activities and market power, i.e. the ability of firms to set prices above marginal costs, of domestic firms. To explore how stronger patent protection affects markups – the wedge between prices and marginal costs – we bring evidence on several potential underlying mechanisms. This includes reduced price competition allowing firms to charge higher prices, a reduction in marginal costs due to cost-saving innovations that are not passed onto consumers, and improved products that can alter demand elasticity. The findings contribute to a growing literature examining specifically the effectiveness of different elements of patent protection such as patent screening (Schankerman and Schuett 2020), patent publication (Hegde et al. 2022) and so on, and to a broader literature examining the role of government intervention in boosting innovation (Moretti et al. 2019, Santoleri et al. 2020). Our paper also speaks to the literature on the determinants of rising market power (De Loecker et al. 2020, Döpper et al. 2024, Miller 2024).
TRIPS in India
Prior to joining the WTO in 1995, India had weak patent laws characterised by fourteen years of patent protection, and restrictions on product and process patents in certain technology classes. In the wake of a balance of payments crisis in 1991, India found itself obliged to become a member state of the WTO, and consequently moved towards a stronger patent regime in compliance with the Trade Related Intellectual Property Rights (TRIPS) agreement.
The adoption of stronger patent rights was met with staunch opposition in the Indian Parliament and early attempts to comply with obligations under TRIPS failed. This created prolonged uncertainty about the timing and nature of policy change (Reddy and Chandrashekaran 2017), and the eventual adoption of reforms for pharmaceuticals and chemical industries in 2000, and all other industries in 2003 occurred unexpectedly, and is akin to a quasi-natural experiment. The reform relaxed several restrictions on product and process patents in the pharmaceutical and chemical industries. It also extended the duration of patent protection in all industries to twenty years and, among other things, allowed for methods and processes of manufacturing to be patented.
Empirical strategy
Our main data source is the Prowess database compiled by the Centre for Monitoring of the Indian Economy (CMIE). Along with accounting information, a unique feature of the data is that it contains the prices and quantities of disaggregated products produced by a firm. Following De Loecker et al. (2016) and Cairncross et al. (2023), this information enables us to estimate markups at the level of the firm and firm- product and decompose them into prices and marginal costs. Additionally, access to firm names allows us to merge patent data from the Indian Patent Office, and thus obtain a firm-level measure of patenting.
We evaluate the impact of the patent reform on innovation and market power using a difference-in-differences design. For identification, we exploit variation in the extent to which different industries, within broad sectors, rely on patents to protect their innovations (EPO 2013). Consequently, we use the pre-reform product mix of firms from CMIE Prowess to calculate its exposure to the patent reform. We find that our measure of exposure to the patent reform is not statistically significantly correlated with pre-reform firm performance and other policy reforms, such as India’s trade liberalisation. Nonetheless, we show that our results are robust to accounting for policy changes introduced in India during the 1990s and 2000s, to controlling for several time-varying firm characteristics, unit-specific pre-trends and to using alternate measures of exposure to the reform.
Effect on innovation and market power
Figure 1 shows that in the years prior to the reform, innovation activities of firms producing patent intensive products was not growing at a differential rate. Thus, firms did not seem to anticipate the patent reform. However, as early as one year after reform implementation, there was a positive and statistically significant increase in both R&D expenses and patenting of firms most exposed to the reform. Both outcomes gradually increase over time, and the effect persists for at least eight years after the reform. Relative to firm active in product-markets with the lowest patent intensity, firms in product-markets with mean patent intensity increase patenting by approximately 13.5% after 8 years. Moreover, three indicators of patent quality – the number of patents renewed by a firm, the number of inventors per patent, and the number of patents filed internationally – show that the surge in number of patents was accompanied by an increase in high-quality patents.
Figure 1 Effect on patenting and R&D
Notes: The dependent variables Patent applications and R&D expenditure are the inverse hyperbolic sine of the number of patent applications and R&D expenditure of a firm in a given year, respectively. PatentIntensity is the firm-level exposure to the reform defined using the pre-reform product mix of firms. X-axis shows time relative to the year of the reform.
Regarding market power, as measured by markups, Figure 2 shows that the reform led to a gradual and persistent increase in markups a few years after the initiation of the reform. The increase, however, was mainly driven by a decline in marginal costs, with no significant changes in prices in any period after the reform. This holds both within firm-products, and within firms.
Figure 2 Effect on markups, prices, and marginal costs
Notes: The dependent variable Markups is the log of firm-product markups, Prices is the log of the unit price reported in CMIE Prowess for each product sold by a firm, and Marginal Costs is the difference of log of prices and log of markups. PatentIntensity is the number of patents per 1000 employees for each four-digit industry. X-axis shows time relative to the year of the reform.
Mechanisms behind declining marginal costs and rising markups
We provide evidence supporting two potential explanations for the observed decline in marginal costs (Figure 3). First, by disaggregating patents into product and process patents using a keyword search in patent abstracts, we observe an increase in both categories. Since process patents are theoretically linked to improvements in production efficiency, the rise in such patents may help explain the reduction in marginal costs. Alternatively, when classifying patents as cost-saving – those aimed at reducing material usage or energy consumption – we also detect a gradual and positive effect on these patents following the reform.
Second, we observe that sales of patent-intensive products increased after the reform, particularly for firms operating in industries characterised by economies of scale. Higher output levels can lead to reduced marginal production costs, further contributing to the cost decline. Moreover, we find that the incomplete pass-through of marginal cost reductions to prices explains much of the increase in markups observed in our sample. We do not find evidence that alternative channels explaining rising markups, such as improvements in perceived product quality or a reduction in price competition due to stronger patent protection, played an important role.
Figure 3 Process and cost-saving patents
Notes: The dependent variables Process patents and Cost-saving patents are the inverse hyperbolic sine of the number of the type of patents filed by a firm in a given year, respectively. PatentIntensity is the firm-level exposure to the reform defined using the pre-reform product mix of firms. X-axis shows time relative to the year of the reform.
Conclusion
To conclude, the findings suggest that – at least in the case of India – stronger patent rights have fostered innovation with limited price effects for the average industry. However, the gains have predominantly accrued to producers in the form of higher markups rather than consumers and buyers, thereby presenting important implications for policymakers that aim to redistribute gains from technological growth. The mechanism underscores the role of cost-saving innovations in shaping the technology driven rise in markups.
Source : VOXeu
