One, the last computed salary of the worker being replaced is a function of regular renegotiation, so there would be differences in opinion over how much the relevant salary would have risen or fallen. The potential for corruption in such an interpretative and discretionary approach — familiar to Indians facing the taxman — are clear. Second, not all robots will replace human jobs; some may perform jobs humans never did before. In that case, questions arise over the reference salary level to tax. The third problem is a philosophical one: Assuming the human being was operating a machine before his displacement by a robot, why should the latter be subject to an income tax when the machine was not? To this, Mr Gates had suggested a lump-sum tax that would slow the pace of automation. This could, however, incentivise producers to bundle artificial intelligence with regular machinery and defeat the purpose.
The intensity of debate on either side of the Atlantic, especially in Donald Trump’s America, underline the undeniably disruptive impact of robotisation and the urgent need for some sort of stabilising policy response. Suggestions have varied from lowering payroll taxes, which fall disproportionately on workers, to financing a UBI by taxing income from dividends and capital gains earned by the top one per cent. India demands a radically different response. Unlike China and Japan, where robotisation is being driven by falling working age populations, India does not face a demographic crisis — half its population is under 24 years of age and the cost arbitrage should be a compelling advantage. For example, the 1.5 million engineers India produces every year have a per-hour labour cost of less than a dollar. Yet many corporations are leapfrogging to artificial intelligence in India because of the critical lack of a skilled labour force. Most good engineers, for example, migrate to the US and a quarter of those who remain are deemed near-unemployable — the import of Chinese telecom engineers and the basic inability of Indian Uber drivers to read a map point to the problems.
In that sense, the skilling initiatives of successive governments have been woefully inadequate, and commentators suggest that the target of training 400 million people by 2022 is way too distant. Thus, Mr Gates’ robot tax may not be applicable to India just yet, but the significantly lower levels of job creation in the future as a result of growing robotisation, with all its implications for social unrest, suggests that the initiatives for meaningful training and skilling need to be accelerated by several orders of magnitude.
