Andrea Napolitano is Chief Operating Officer at Iromed Therapeutics, which develops ophthalmic technologies. Danny Wagemans is an Engagement Manager at McKinsey & Company, a management consulting firm. Both write in a personal capacity.
Across Europe, two important debates unfold in parallel: one about declining fertility and ageing populations; the other about the threat of AI to employment. Strikingly, these conversations happen in silos, as if unconnected. But analysing these phenomena together reveals a crucial insight: that given Europe’s demographic winter, the reduction of jobs driven by technological progress might well be a blessing, not a curse.
In the last 75 years, Europe’s fertility rate plummeted from 2.7 to 1.4. Simultaneously, life expectancy rose from 62 to 79 years. A population that contracts and ages so rapidly puts immense strain on the welfare system, as every worker has to support ever more non-workers, and for longer. Today, the EU population aged over 20 is roughly split into one worker per non-worker. If fertility rates remain at today’s levels, and life expectancy gains plateau, the number of non-working adults per working adult will increase by ~35% in 2050, and more than double in 2100. Ceteris paribus, this means that we’ll have to either double welfare taxes or halve welfare spending – a tragic dilemma.
There aren’t many ways to salvage welfare states as we know them. Having more babies is an obvious solution. Yet, no country in the world has so far managed to bring fertility rates back above extinction levels – and not for a lack of trying. Baby bonuses, subsidised childcare, extended parental leaves… no measure has moved the needle significantly, and those policies that did have a meaningful impact came at great financial cost and risk of economic distortions.
Relying on immigration from countries with younger demographics is no real answer, either. Besides the societal frictions that this strategy might cause, fertility rates are declining virtually everywhere, including in hotspots of migration to Europe like North Africa and the Middle East. This, coupled with the fact that migrants tend to adopt fertility rates similar to the host country’s, makes immigration at best a temporary fix.
Another option is to make people work for longer. But most EU states already increase retirement ages periodically and many have linked them directly to life expectancy. Furthermore, measures that delay retirement are extremely unpopular and, in an ever-older Europe, will be ever harder to pass.
There is only one path that Europe has not really pursued yet: boosting labour productivity without fearing job losses, so that the extra output per worker compensates for the shrinking pool of active citizens.
True, Europe failed to exploit the productivity bonanza created by the internet revolution of the late 1990s. However, decades ago, the Union’s demographic outlook was not as bleak, and the Draghi report has placed productivity at the centre of European policymaking.
Today’s wave of technological advances offers Europe another chance. According to many studies, AI and automation can boost productivity by as much as 3.4% annually – that is five times the average annual productivity increase Europe experienced in the last decade. Capturing only a third of that would be enough to offset the workforce shrinkage and maintain current welfare levels.
South Korea, the country with the lowest fertility rate in the world, realised this quickly and, after introducing specific robotics subsidies and tax breaks, is now also the only country where as much as 10% of the workforce is made up of robots. Japan has also pursued similar policies recently. Europe should follow these examples.
But boosting productivity alone won’t suffice – two fundamental things must also happen. First, labour surpluses and shortages must be matched precisely. Investing in retraining is essential so that displaced workers – say, a 30-year-old taxi driver made redundant by automation – can re-skill to replace retirees. Second, states must ensure effective value distribution: productivity gains must actually accrue to the welfare system. For example, when driverless-vehicle firms capture extra economic value, this benefit must be taxed – via corporate taxes or through higher wages for remaining workers – to offset lost revenue from retiring taxpayers.
To be clear, these are enormous challenges. But we can’t afford to be picky. Europe must start connecting the dots between its grim demography and the productivity promise of the AI revolution. Doing so will reveal that we should embrace new technologies, not despite, but precisely because they will entail fewer jobs.
