From the Luddites destroying woollen machines in 1800s to doctors or bankers today being disrupted by robots, it is now clear that technological progress has a major impact on the wage distribution in an economy. As firms, heterogeneous in their capabilities, continuously innovate and adopt new technologies, the relative demand for different types of workers changes. In the process, the returns to these types of workers changes differentially, such as for skilled and unskilled workers (Faia et al. 2020).
Over the last few decades, technological change has induced significantly higher wage and employment growth for skilled workers as compared to unskilled workers – a phenomenon broadly referred to as skill-biased technological change by researchers. And this has spurred extensive discussions on its welfare consequences (OECD 2011, UNDP 2013, ILO 2016, World Bank 2016). For instance, the proportion of routine (low-skilled) labour in the US declined from 39% in 1968 to 23.6% to 1968 in 2013, while non-routine (skilled) labour saw an increase from 24.4% to 33.6% (Eden and Gaggl 2014). Further evidence confirms these findings for several countries with different levels of technology (e.g. Srour et al. 2013, Marouani and Nilsson 2016, Gaggle and Wright 2017).1
Studies of this disproportionate demand for skilled relative to unskilled workers focus mainly on structural changes driven by an increasingly connected global economy and its interaction with innovation. As also highlighted by OECD (2013, 2014), public policies, such as patent policies, are one of the main elements to influence the relationships between technology and labour in both developed and developing countries. Academic work has also shown how globalisation, trade, and technological change have contributed to the rise of wage inequality both within firms across worker groups as well as across firms, with some firms attaining ‘superstar’ status by cornering a large share of the total worker compensation in the economy.
To better understand this process from the vantage point of firms as the predominant economic units that drive employment and determine the wage structure, it is important to ask how incentives for technological change and innovation affect the compensation structure of a firm. A major challenge in answering this question is to find an economic event (i.e. a natural experiment) that influences the incentives for innovation for firms without simultaneously being subject to the influence of the wage structure or organisational architecture. Our new paper (Bhattacharya et al. 2021) takes up this exercise and investigates how a change in the Indian intellectual property regime (IPR) brought about by landmark legislation, the Patents Amendment Act (2002) affected the wage share of managers (vis-à-vis workers or non-managers) for a cross-section of firms in the Indian manufacturing sector. To the best of our knowledge, our work is the first to look at how a change in IPR affects wage inequality. Specifically, we study how a large cross-section of Indian manufacturing firms responded to this Act in terms of changes in compensation structure.2
Prior to the Act of 2002, India only recognised patents for new production processes (and not new products), which allowed for easy imitation of new products with mildly altered processes. Highlights of the Act of 2002 were the introduction of product patents in all fields of technology, an increase in the term of patents from 14 to 20 years, limiting the scope of the government use of patented products, and streamlining the process of patent grants. The Act also significantly broadened the scope for the implementation of the TRIPs-complying IPR regime that India was committed to adopting as a signatory to WTO in 1995. By strengthening the protection of patent rights, the Act significantly increased firms’ incentives for innovation.
Higher investments in innovation and/or higher technology adoption by a firm can increase the relative demand for managers through at least two channels (Aghion et al. 2020). First, innovation involves a whole range of activities that are intensive in skilled-workers or managerial talent: research, conceptualisation and development of new products, branding and marketing the product and so on (Teece 1996). Second, technology adoption pushes existing processes closer to the technological frontier through use of more R&D expenditure, technology transfer, import of capital goods, and so on (Acemoglu et al. 2006). If one conceptualises production as a sequence of problems and managers as problem-solving experts (Garicano 2000), innovation presents a firm with more complex problems, thereby raising the value of managers. These reasons indicate that managerial effort and capital (both technology and knowledge) are complementary inputs in innovation. Therefore, an across-the-board increase in incentives to innovate will not only lead to expansion in demand for managers relative to non-managers (within-firm inequality), such expansion will be larger in firms that have a larger stock of capital (between-firm inequality) (Bloom et al. 2021).
In our study, we employ PROWESS, a comprehensive database of Indian manufacturing firms which reports detailed labour compensation, divided into managerial and non-managerial components (Chakraborty and Raveh 2018), details of technology such as R&D investment and technology transfer, as well as other firm characteristics on a panel basis. The panel format of the data enables a dynamic specification in which technological investments and other firm decisions can potentially affect the demand for managers.
One of the main aims of our study is to establish a causal link between innovation incentives and compensation structure both within and between firms. A benchmark hypothesis in our study is that the increase in wage inequality between managers and non-managers as a response to the IPR shock is higher in the firms that were technologically advanced at the time of the shock. We split the firms into two groups – those above the median in technological investments in each industry pre-2002 (the ‘high-tech’ firms), and those below the median (the ‘low-tech’ firms) – and compare the response to the IPR shock of these two groups of firms.
Panel A of Figure 1 plots the normalised technology adoption expenditure for the sample of Indian firms for the period 1990-2006. The two groups show similar trends before 2002, but after 2002 we see opposite effects: technology adoption expenditure doubles for high tech-firms and drops for low-tech firms. Panel B of Figure 1 does the same for the normalised share of managerial compensation in total compensation. The figure clearly shows that while there was not much difference between the two groups in the early 1990s, but once India started to liberalise from the mid-1990s the difference started to grow. It became significantly stark once stronger patent reforms were passed in 2002. This figure suggests a causal association between technology adoption, innovation, and demand for managers. The regression results show that the change in manager’s share of total compensation due to the reform is 1.3% to 8.3% higher in the high-tech group of firms than the low-tech group.
Figure 1 Innovation expenditure and managerial compensation: High-tech and low-tech firms
Further in-depth analysis shows that the response to IPR shock is strongest in the marginally high-capital stock firms (i.e. those in the 5th to 8th decile within the respective industries), while the effect is insignificant in the below-median firms and superstar firms. This result that is reminiscent of a ‘snail shape’ is consistent with the idea of wage dynamics and innovation effort being driven by competition for patents between firms within a given product group or industry. Figure 2 plots the coefficient estimates for each of the deciles.3 The figure shows the snail-shape clearly; no effect for below-median and 9th decile firms and highest for marginally big firms. The broader picture that emerges is that it is not only the firms within high-tech industries, but it is the marginally large firms within each industry that drives wage inequality between firms in the economy. Thus, a strengthening of IPR and consequent technology adoption produces wage inequality virtually across the entire economy, and not just within the high-tech industries, also within and across firms.
Figure 2 2002 patent reform and managerial compensation: ‘Snail-shaped’ effect
These results provide strong indications of the kind of changes developing economies can go through with increasing formalisation and integration with the global economy. Given that the TRIPs+ provisions are soon to be implemented in the least developed countries, one must recognize that wage polarisation throughout the economy is an important trade-off associated with the structural transformation of these economies leveraging technology. While our paper provides an early indication, there should be more careful examination of the effects of technology adoption or innovation on wage inequality across nations such as Brazil, Chile, or China using employer-employee datasets to bring in more systematic evidence on the welfare effects of innovation and technological changes across the world.
References
Acemoglu, D, P Aghion, and F Zilbotti (2006), “Distance to frontier, selection, and economic growth”, Journal of the European Economic Association 4(1): 37-74.
Aghion, P, A Bergeaud, R Blundell, and R Griffith (2020), “The innovation premium to soft skills in low-skilled occupations”, VoxEU.org, 17 June.
Berman, E and S Machin (2000), “Skill-biased technology transfer around the world”, Oxford Review of Economic Policy 16(3):12-22.
Bhattacharya, S, P Chakraborty, and C Chatterjee (2021), “Intellectual property rights and wage inequality”, Journal of Development Economics, forthcoming.
Bloom, N, T Hassan, A Kalyani, J Lerner and A Tahoun (2021), "How disruptive technologies diffuse", VoxEU.org, 10 August.
Chakraborty, P and O Raveh (2018), “Input-trade liberalization and the demand for managers: Evidence from India”, Journal of International Economics 111: 159-176.
Eden, M and P Gaggl (2014), “The substitution of ICT capital for routine labor: Transitional dynamics and long-run implications”, World Bank Policy Research Working Paper Series No. 7487.
Faia, E, S Laffitte, M Mayer, and G Ottaviano (2020), “Automation, globalisation, and vanishing jobs: A labour market sorting view”, VoxEU.org, 2 January.
Gaggl, P and G Wright (2017), “A short-run view of what computers do: Evidence from a UK tax incentive”, American Economic Journal: Applied Economics 9(3): 262-94.
Garicano, L, (2000), “Hierarchies and the organization of knowledge in production”, Journal of Political Economy 108(5): 874-904.
ILO (2016), “Global Wage Report 2016/17: Wage inequality in the workplace”, International Labour Organization: Geneva, Switzerland.
Kremer, M and E Maskin (2006), “Globalization and inequality”, Mimeograph, Harvard University.
Marouani, M and B Nilsson (2016), “The labor market effects of skill-biased technological change in Malaysia”, Economic Modelling 57: 55-75.
Maskin, E, (2015), “Why haven’t global markets reduced inequality in emerging economies?”, The World Bank Economic Review 29(1): S48-S42.
OECD (2011), Divided we stand: Why inequality keeps rising, OECD Publishing.
OECD (2013), Supporting investment in knowledge capital, growth, and innovation, OECD Publishing.
OECD (2014), Science, technology and industry outlook 2014, OECD Publishing.
Srour, I, E Taymaz, and M Vivarelli (2013), “Skill-biased technological change and skill-enhancing trade in Turkey: Evidence from longitudinal microdata”, IZA Discussion Paper No. 7320.
Teece, D (1996), “Firm organization, industrial structure, and technological innovation”, Journal of Economic Behaviour and Organization 31(2): 192-222.
UNDP - United Nations Development Programme(2013), Humanity divided: Confronting inequality in developing countries.
World Bank (2016), World development report 2016: Digital dividends, Technical report.
Endnotes
1 This trend appears to be common in both developed and emerging countries, contrary to the principle of comparative advantage (Berman and Machin (2000); Kremer and Maskin (2006); Maskin (2015)).
2 By ‘compensation structure’ we mean total labour compensation of firms. In our case, compensation is equal to wages plus incentives.
3 Since the coefficients below the median are not significantly different from zero, we treat them as zero; as for the others above the median, they are as per the regression estimates.
