The recent UK census has sparked significant concerns about the sustainability of the UK welfare state. Specifically, with over one sixth of the UK’s population now over retirement age and rising, the working population’s capacity pay enough in tax to fund the entitlements of the retired is diminishing. This concern is not limited to the UK. The demographic challenge is more pronounced in Europe and China, but even in the US, social security is projected to become underfunded. Countries are resorting to higher retirement ages to stave off the problem, but simply put, we are running out of workers.
Predictably, centrist pundits have taken a break from chastising poor people for having too many children to chastise poor people for not having enough children. Particularly unhinged was a suggestion in the Times that we should tax people who do not have children. Leaving aside the gall of these people who cheered on cuts in child tax credits and child benefit, there is a larger issue for our modern anti-Malthusians.
What jobs will these children do? Technological advancement is occurring at an extremely rapid pace, with labor-saving technology driving down demand for workers. Effects include rising inequality, but its future is likely to be even more pronounced. As automation advances, and technology becomes sufficient dexterous and imaginative to replace humans in all jobs, we may face a future without work. In the starkest possible terms, we are running out of jobs.
We are running out of workers. And we are running out of jobs. Only capitalism could contrive such a stark contradiction.
Policymakers therefore face a dilemma. Do nothing about the coming demographic squeeze, and our welfare states will crumble. With them will go any semblance of a good life for the majority under capitalism. Force a demographic rebalancing, and we will bring into the world a generation who will suffer immensely. (There is a reason why people are not having children!) They will compete for what little, low-paid work remains.
Whenever there is an ostensible dilemma, it is often a good idea to stress-test the assumptions we’re basing our conclusions on. Perhaps it is not necessarily the case that an ageing population makes our welfare states unsustainable? Table 1 shows the scale of the demographic problem. For selected developed countries, I show the projected increase in the working age and retirement age population from 2020 to 2050. In all cases, the increase in the retirement age population is an order of magnitude higher than the projected increase in the working age population.
| US | Canada | UK | Germany |
Working Age | 8.33% | 10.79% | 0.88% | 16.85% |
Retired | 54.1% | 67.37% | 48.27% | 32.30% |
Required Annual Wage Growth | 1.18% | 1.38% | 1.29% | 1.56% |
Table 1: Projected Demographic trends. Source: United Nations, author’s calculations.
Let’s assume that we want to maintain our current welfare state generosity without increasing the tax burden on working-age populations, and that labor market participation is constant over time. Then these numbers can only be squared by real wage growth of between 1.18% and 1.56%. To get a sense of how large a difference from the status quo this implies, I estimate that since 1982, real average wages per worker in the US have fallen by 1.4% per year. Real median wages amongst full-time workers rose an average of 0.6% per year over the same period.
Could we be bailed out by technological change? It does not appear so. The new technology we are seeing is a substitute, not a complement, for labor. Hence the displacement of workers and falling wages. Far from higher wage growth being on the horizon, the form of technological change we are experiencing is driving wages lower. Neither are we likely to be saved by rising corporate tax revenues. Although it is the case that corporate profits ought to increase in an environment of falling wages and accelerating technological growth, corporate tax is less than 10% of government revenue in OECD countries (and lower in the US).
We face the inescapable conclusion that if we want to maintain our welfare states in their current form, we are running out of workers. Furthermore, technological trends reinforce, rather than undermine, this conclusion.
What about the claim that we are running out of jobs? It is worth noting that at any other point in capitalism’s history, those who thought that labor-saving technology would lead to the permanent collapse of demand for labor would have been grossly underestimating capital’s dynamism. If it is indeed true that AI and robotics will replace human labor, then we are truly in a unique moment.
It is the contention of many that we are. The nature of the fourth industrial revolution is not just that labor becomes a less essential input into the production process. Rather, a combination of exceptionally cheap energy in the form of renewables and the importance of information in production leads to a zero marginal cost economy. To give some examples, reductions in the cost of renewables accelerate at an increasing rate as they are used (the “experience curve”). The use of AI and online platforms to create Massive Open Online Courses (MOOCs) mean that education has never been more democratized or cheaper to produce (though someone should tell universities this). Gene-editing of plants and lab-grown meat mean that we will soon be able to produce food at a rate unparalleled in human history, with minimal negative impact on ecosystems. And if Susskind and Susskind are to be believed and AI is set to save the labor even of lawyers and doctors, even expensive and labor-intensive industries such as healthcare will see their labor replaced with computers, and their marginal cost will be related only to the ever-lower cost of energy. While previous technological revolutions have only slackened scarcity constraints, the coming one runs a serious risk of all but eliminating them.
Here is another contradiction: the faster technological advancement, and the more abundant our resources become, the poorer we will get. As technology replaces even higher-wage labor trading one’s leisure for stable levels of consumption becomes increasingly impossible. And as wages and employment collapse under the weight of competition from technology, the revenue needed to fund our welfare states will dry up.
To repeat, we will simultaneously run out of workers, and jobs for them to do. And we will do it at a time when our productive capacities have never been greater.
These apparent contradictions are a mirage. It cannot be the case that our ability to produce becomes historically unrivaled while our collective access to consume that production falls. This mirage occurs as a result of applying an accounting system designed for a world of scarcity to a world of abundance. Specifically, we try to allocate consumption to people based on 1) labor earnings and 2) profit. Neither of those will work when labor is not needed and marginal costs tend to zero.
Let’s take labor earnings first. The designers of the postwar welfare state were rightly concerned that giving unconditional entitlements to consumption would undermine incentives to participate in production. They baked in unequal citizenship based on one’s labor contribution to society. Welfare capitalism assures unconditional access to a small number of goods (such as healthcare, very basic levels of consumption and legal rights) while making access to others (pensions, some social insurance benefits) contingent on either selling enough of one’s labor or being faithfully married to someone who sells enough of theirs. Accordingly, the funding system is that those who currently work will pay for the consumption of those who currently do not. In some welfare systems, particular “social insurance” taxes are earmarked for spending on benefits, as a direct recognition of this. Clearly, technological change which removes labor from the production process collapses this arrangement. An accounting system which balances labor earnings with entitlements cannot function while earnings trend downwards and entitlements increase.
One potential solution to this problem is to shift the burden of funding the welfare state from labor to capital. If the economy is hiring less labor, but its production is constant or increasing, then one way to ration consumption is to decouple consumption entitlements from earnings, and correspondingly require funding by the owners of capital. I believe that this is behind much of the resurgence in the popularity of the Universal Basic Income (UBI). Firms could simply employ machinery, pay a portion of their profit in taxes to fund a UBI, and then consumers buy their products. In this world, jobs would not be required, because employment would not be a precondition of full citizenship. Since all payments are funded from corporate taxes, we would not need to worry about the supply of workers.
This would be a radical departure from capitalism as we know it. Wage slavery, regulated and enforced by corporate and state power, has been a key aspect of capitalism for its entire history. However, there are three reasons to dislike this proposal. Firstly, it would be a grossly unequal setup. Capital owners would be entitled to significantly more consumption under this system, and would almost certainly exist in a relationship of domination with most consumers. A UBI fails to fundamentally reconstitute political relationships within an economy. While we might be free of being required to sell our labor to eat with a UBI, we are still dependent for our consumption choices on the whims of multinational companies. Secondly, and relatedly, with economic domination comes political domination. In such a political economy, the ability of large corporations to dictate government policy would likely increase, to the point where it is unlikely that policy which benefits the majority could be sustained.
Most importantly, however, it does not solve the fundamental problem, which is the attempt to price abundance with mechanisms designed for a world of scarcity. Prices are a way to ration consumption, and revenue a measure of how many and how scarce the goods a firm has been able to sell. If there is no natural scarcity, firms will have to invent it, by restricting supply. Thus, the true potential of our emerging technology cannot be realized under even this radical version of workless capitalism. It will take something much more imaginative to unlock the possibilities of our future. Some have made valiant attempts at sketching what this future might look like, although the likelihood is that the future is as unimaginable as our present would be to a medieval peasant.
What is the conclusion we should draw from this? Along with the myriad other contradictions of capitalism, its attempt to handle a shift to post-scarcity technology and an ageing population is leading to a grotesque scenario where we will shortly simultaneously run out of both workers and jobs for them to do. No amount of policy tinkering can avoid this fact. Social democracy and liberalism have nothing to say here. And that this is due to capitalistic accounting is little consolation when these are rooted in the material conditions we insist on continuing to live in. Only a radical solution which dissolves the structure which leads to these contradictions can solve the problem.