Universal Basic Income in the World of AI?

If AI replaces most tasks and jobs for the general population, can we fund a universal basic income at least in poor countries?

In my last substack piece I discussed the need for voice of labor in influencing the R & D decisions of companies in shaping the pattern of innovations in a labor-absorbing direction—otherwise increasingly more powerful AI is likely to make most workers redundant in their current jobs and tasks. In the latter eventuality how will people survive in that not-too-distant future? The Big Tech entrepreneurs of Silicon Valley and elsewhere—which include some avowed libertarians (though being libertarian has not usually stopped them from lobbying for large government contracts) and some open supporters of political parties with neo-Nazi roots—have often suggested a simple solution: Universal Basic Income (UBI).

I have been intrigued by this suggestion. If more of the current types of labor-replacing AI which Big Tech is rushing to bring about are in our inexorable future, and in that future if most people have no jobs and thus no income (nor any income taxes to pay), how will UBI be financed? In the US, for example, a level of UBI for everyone at even the current dismal official poverty line of the country will exhaust more than two-thirds of the federal budget, leaving very little for anything else. I have yet to hear about any ringing endorsement among the tycoons supporting UBI for the high taxes on the rich that will be necessary to fund a decent level of UBI.

Then there is the question of dignity and life purpose that jobs often provide to a worker, and just a monthly check of UBI from the government may not be a meaningful substitute. I am actually less bothered by this issue than that of funding, as we can think of, if not anything else, a whole set of community or public service jobs that may be assigned to all able-bodied adult recipients of UBI (with enough scope for initiative, choice and volunteering).

As the funding problem for UBI in rich countries will be formidable for quite some time and the UBI-supporting tycoons have to sort this out, let me turn to the issue of UBI for developing countries where the job displacement by AI at their average level of skills may be no less serious. As it is, substantial numbers of people there are under-employed or employed in the informal sector (without benefits) or outside the income-earning labor force altogether (as in the case of many women). Then there are others who are trapped in extremely ‘bad’ jobs or stigmatized occupations (like manual scavenging or sex work) and who are looking for an escape ladder that UBI can provide, giving them the respite and opportunity to look for better jobs or to get training or more qualifications for such jobs.

I have been personally an advocate of UBI in India for the last two decades, long before I started seriously thinking about the impact of AI on jobs. In my book, A World of Insecurity (Chapter 8), I have worked out a back-of-the-envelope calculation by which a modest level of basic income supplement looks financially feasible in a low-income country like India, if the leadership has the political courage to eliminate and redirect much of the current government subsidies to the relatively better-off sections of the population and to impose some additional taxation (mostly capital taxes) on the ultra-rich. This is proposed without touching the existing skimpy welfare programs for the poor (and also allowing for some necessary increase in public expenditure on infrastructure, education and health). There is, of course, some scope for reducing the waste and fraud in the current implementation of these programs, and to that extent it may be possible to support even a somewhat higher level of basic income supplement than I have envisaged.

One question that is often asked about any UBI proposal is: why is it ‘universal’. “Do you mean the filthy rich tycoon will get as much as any other citizen?” My answer is “yes”. This answer is based on the conception of a basic right for minimum economic security for every citizen. (It may be thought of as linked to the right to social security as part of the United Nations Universal Declaration of Human Rights of 1948).

Just as a tycoon has a basic citizen’s right to physical security or police protection, even when they can well afford such protection with a full army of bodyguards and gatekeepers, and do not really need help from the state; similarly, they are entitled to a basic right to minimum economic security in the form of UBI, even though they do not need it. If they decide to waive it, I’d not object; but that is on practical or pragmatic grounds, without yielding on the basic rights issue. (In any case, some of the UBI paid to them would come back to the government in the form of taxes). My approach to UBI is thus qualitatively different from that of many of my fellow development economists who look upon UBI as an anti-poverty policy. They then have to grapple with cost-benefit comparisons with policies that are more targeted to the poor. Looking upon UBI as part of a citizen’s basic rights is also a way of resisting the populist politician’s all-too-common tendency to treat cash transfers to individuals or groups as part of patronage distribution to supplicants or of clientelistic vote-buying.

Considered as a right to minimum economic security, some form of UBI is thus not entirely unaffordable at least in poor countries, provided one can cope with the inevitable political resistance of the wealthy to the necessary restructuring of the taxes and subsidies.