[Ed Note: The following post is part of the TLF Editorial Board Test 2020-21. It has been authored by Sankalp Jain, a fourth year student of NALSAR University of Law.]
This post discusses policy responses to the threat of automation on India’s labour, contextualising the same amidst a major technological shift and efforts to revive India’s economy after a nation-wide lockdown crippled it.
The ‘future of work’ has been the subject of hot-debate across the world as machines increasingly replicate and replace human labour. The current wave of automation has been characterised as the ‘Fourth Industrial Revolution’ i.e., the convergence of technologies such as robotics, artificial intelligence, machine-learning, etc., and experts predict that this time will be different. The ramifications of this wave would depend upon the socio-economic and labour profiles of different countries.
Many studies on the effects of automation focus on developed economies where such technologies are being increasingly adopted (Illustratively: a widely-cited 2013 study and a 2019 study). However, some experts have argued that the threat of automation is far greater to developing countries. While there have always been grim predictions of mass-unemployment, the displacement of labour, either temporary or long-term, is certain.
First, I will give a brief overview of India’s labour-force and its susceptibility to automation. Second, I will introduce the possibility of a tax-policy response to counter this threat. Finally, I conclude by enumerating some ‘non-tax’ measures which will have to be deployed for a comprehensive response.
India’s Labour Profile and Automation
India’s labour-force is characterised by a very high rate of informalisation – around 80 percent is engaged in informal and casual work, most of it being either under or low-skilled, with high-skilled labour comprising a meagre 5 percent. More worryingly, the government’s skilling missions have also failed to produce employment. Further, since 1991, there has been an alarming increase (page 5) in contract labour which implies a lack of job-security and other safety-nets. This has serious implications as it is jobs in primary and secondary sectors – involving routine-processing tasks and manual work which are more susceptible to automation than ones that are analytical and non-routine – mostly associated with the service (tertiary) sector. It is important to note here that primary and secondary sectors contribute to the employment of most of India’s labour-force.
According to a germinal report by the McKinsey Global Institute, around 52-62 percent of jobs in India are potentially automatable (2017, Page 48). Further, it has been shown that women may be disproportionately affected by automation-induced labour displacement. However, an important distinction must be made here – between ‘Automation Potential’ and ‘Automation Adoption’ (Page 9). While the former gives a measure of tasks and operations which are automatable from a technical standpoint, the latter depends on actual market factors, costs of adoption, labour costs, government incentives, etc.
It is also expected that COVID-19 will have a significant impact on the future of work. Its effects are already being felt in some industry sectors due to the resultant infusion of technology in response to adaption to changing social and health norms. The logic that robot-workers are more cost-efficient than human-labour in performing routine tasks may soon dominate Indian industry. Therefore, policy responses must tackle ‘Automation Adoption’ with an eye on ‘Automation Potential’.
Tax Policy to Guide Automation
In 2017, Bill Gates voiced support for a ‘tax on robots’. Soon after, South Korea implemented its own version of a ‘robot tax’ where tax incentives given to firms which automate were slashed.
The ILO set up a Global Commission on the ‘Future of Work’, which published its report in 2019. One of the three pillars the report suggested for its ‘human-centred agenda’ was “Increasing investment in decent and sustainable work” comprising a focus on inter alia ‘shifting incentives’, and underscored the need for tax systems to be ‘equitable’.
Tax policy serves dual aims, first, it is a tool for redistribution of resources, and second, it promotes some behaviours over others. The role of the State as a facilitator of economic progress has been one of aligning itself towards incentivising automation-adoption. Economic and Tax policies support investment in capital assets like machinery and computer hardware by giving subsidies, easing credit for their purchase and/or tax incentives like deductions in the form of accelerated depreciation (AD).
In the U.S.-context, Abbott and Bogenschneider demonstrate how the tax system favours robot-workers and have argued for ‘tax neutrality’. They show how automation allows employers to, one, avoid costs like wages-taxes and social security contributions towards human-workers and, two, claim AD on capital costs. While contexts differ, insomuch as India raises more revenue from indirect taxes compared to the U.S. where direct taxes form the bulk of tax receipts, tax deductions in the form of AD are a heavy burden on the Indian government’s budget. Indeed, loss of revenue due to AD was projected as ₹ 56,585 crores in FY 2019-20 (only slightly higher than actual loss of ₹ 55,022 crore in FY 2018-29). These policies have been criticised and even the Government has admitted that larger companies are claiming higher deductions and incentives on account of these policies (Page 28).
Incentivising ‘Collaborative’ Automation
To avoid framing this as a ‘luddite fantasy’, I submit that a tax on ‘automation’ simpliciter i.e., a general tax on the use of robots – is not feasible. Not only would this disrupt and sink industry confidence, it would also be difficult to enforce. Further, cost-based labour-protection measures act as double-edged swords as they create the risk of increasing the cost of employing humans, further incentivising employers to look for robot replacements. Finally, it cannot be denied that automation improves overall efficiency and, in many cases, has removed humans from hazardous, menial and undignified jobs. Therefore, one must take a nuanced approach to look at what kind of automation is desirable.
While automation is usually seen as replacement of human-labour, a case could be made for incentivizing the adoption of automation that does more towards collaborating with humans rather than replacing them entirely. This case for co-operation between humans and robots involves realigning our desired outcomes from automation technology, towards ‘socially-responsible automation’. A tax policy could be modelled on environmental taxation where ‘tax credits’ are provided to businesses which employ labour-complementing automation, thereby reducing negative externalities (labour displacement).
Goh and Ooi’s proposal for an ‘automation tax’ is a good place to begin as it accounts for this distinction between labour-replacing and labour-complementing automation. They favour a backward-looking and empirical approach to measure labour-complementing automation where tax deductions are given by correlating a company’s net employment with its capital expenditure. To offset these incentives, the rate at which tax deductions are claimed on labour-replacing automation should be decreased. However, this may complicate the tax system, increasing compliance and regulatory costs for businesses which calls for a focus on simpler design. The end result of the tax policy would be a decrease in the amount of revenue forgone on account of deductions for labour-replacing automation.
Conclusion
The revenue raised through tax measures could be spent to finance labour re-skilling initiatives to bridge the ‘skill-deficit’ in India. The government’s existing programmes have been criticized as they create more informality, therefore, these programmes need to be overhauled. The creation of ‘safety-nets’ by way of Basic Income guarantees could also be looked into to smoothen the transition of labour-force towards sectors which are more resistant to automation. Finally, developing countries must collaborate and increase R&D in automation technologies which suit our own needs and contexts.