Gabriel Heller Sahlgren is Research Fellow at the Institute of Economic Affairs, and Affiliated Researcher at the Research Institute of Industrial Economics in Stockholm, Sweden.

The past decades have been a remarkable success story in terms of the trajectory of our well-being in old age. We live longer and our health has improved, which of course should be causes for celebration. But, unfortunately, there’s a dark side to it. This is because the improvements in longevity and health have not been matched by increasing employment rates in old age. In fact, labour force participation rates have slumped compared to the situation in the mid 20th century.

This situation has put immense pressure on the pay-as-you-go state pension system. With fewer babies born and old people living longer, the situation is becoming more untenable every day. This is why consecutive governments have committed to the idea of raising the state pension age, which is now set to reach 68 in the mid 2030s.

But this is not going to be enough. Despite the changes, the age dependency ratio is set to increase by 19 per cent in the next fifty years, while annual state pension expenditures are set to increase by 329 billion pounds, figures that don’t take into account people who retire through other de facto retirement programmes, such as disability insurance. Further changes are clearly necessary.

How, then, can we solve the problem? In my paper Income from Work – The Fourth Pillar of Income Provision in Old Age, released today by the IEA and the Age Endeavour Foundation, I analyse the impact of public policy on retirement worldwide and outline several solutions to the conundrum.

A key reason why people retire is the state pension programmes. A large number of studies find that people tend to retire whenever the state pension age allows them to do so. Consequently, much research shows that people stop exiting the labour force once the state pension age is increased.

But changes to the state pension system are insufficient. This is because there’s strong evidence that people also use alternative government programmes as de facto retirement pathways when regular ones aren’t available. Key examples here are disability insurance and unemployment benefits. Reforms to such programmes have also been shown to increase labour force participation among the elderly. While studies on age discrimination laws are somewhat mixed, general employment protection legislation has been found to reduce labour force participation rates by pushing older people into retirement.

This picture clearly shows the necessity of a much broader-based reform programme than the government’s current approach – and we need to go both further and faster. Here’s what I suggest: from November 2018, the state pension age should be increased two months every quarter, thus reaching 68 in 2023. Thereafter, it should be linked to life expectancy. In the long run, the state pension system should be means-tested, and generally replaced by a privatised pension programme based on compulsory savings. This is similar to the system put in place by Australia’s Labour government in 1992, an arrangement that encourages people to work longer while also making it possible for them to retire early without threatening the viability of the system long term.

Meanwhile, state disability insurance eligibility should be tightened and benefit levels should be reduced so that work isn’t discouraged, while more active strategies should be used to assist people back to work. To counter the effects of employment protection legislation, no-fault compensated dismissals of workers who are recruited within five years of the state pension age should be permitted. Given the mixed evidence on age discrimination laws, the government should introduce a large-scale pilot scheme in which some firms are exempted from them. Once the results are in, we can decide whether to scale it up or not.

Carrying out the proposed reforms isn’t going to be easy politically. But that doesn’t make them less necessary. The ticking demographic bomb needs to be addressed, whether we like it or not. Kicking it in to the long grass is not an option, and ad hoc increases in the state pension age are not sufficient. We need comprehensive reform. And we need it soon.