Retirement in a changing demographic: What the future holds for pensions
This content is AI generated, click here to find out more about Transpose™.
For terms of use click here.

With an ageing population, low private savings for many, and widening differences in pension provision between sectors, the challenges for retirement are stark. But there is a path forward – if we act soon.
Protecting the State Pension foundation
Introduced in 2010, the triple lock guarantees that the State Pension increases each year by the highest of earnings growth, price inflation, or 2.5%. It created a rules-based way of raising the pension without annual political wrangling. Reform will be needed in time, but for many – particularly women and those without substantial private pensions – it remains a vital foundation. Until private savings improve significantly, we can’t take our foot off the gas on the State Pension.
An ageing population inevitably means pension spending will take up a greater share of GDP. The levers are limited: pay it for fewer years, pay less, pay fewer people, or raise more tax. None are politically easy. The better trade-offs are to increase economic participation among working-age adults and raise State Pension Age gradually with long notice periods. Means-testing would deter voluntary saving and undermine the system.
Closing the gap for the “missing middle”
While auto-enrolment has brought millions into workplace saving, contribution rates remain low and many started too late. The group most at risk are people now in their 40s and 50s who missed out on final salary pensions and only began saving through auto-enrolment in the past decade. Without change, many will face a sharp drop in living standards at retirement.
The gap between public-sector defined benefit (DB) schemes and private sector defined contribution (DC) pensions is widening. Rather than levelling down, the aim should be to improve private sector outcomes through higher contributions, greater scale and collective approaches. Public sector schemes have already undergone significant reforms, and closing the gap should mean levelling up.
Housing and later-life risks
Housing costs are another critical – and often overlooked – factor in retirement security. More people are carrying mortgages into later life or renting privately. A mortgage balance at 68 can wipe out a DC pot; private renting can double the income needed for a decent retirement. Pension and housing policy need to be better joined up if people are to enjoy stability in later life.
There are also risks in later retirement that the system is ill-equipped to address. Women often face widowhood, and today’s pension structures don’t provide the survivor benefits they once did. Defaulting more annuities to joint life, and ensuring women build up strong pensions in their own right, would help reduce this vulnerability.
Making the system work for everyone
Broadly, the new State Pension is in the right place. The bigger challenge lies in the second tier – fragmented DC pots, variable investment performance, and too little risk pooling. Contributions must be made to work harder, and the proliferation of small pots needs to end. Letting pots “follow the member” would keep savings together and easier to manage.
Good defaults will do more for most people than hours of education. Engagement matters, but the system must work for those who never actively engage with their pensions.
Despite the pressures, there are reasons to be optimistic. The original Pensions Commission proved that long-term, cross-party consensus is possible. The new Commission brings together experienced voices with the potential to deliver another durable blueprint. If it takes a genuinely holistic view, it could help secure the future of retirement provision for decades to come.
Beyond Curious with LCP: Retirement in an ageing society
Listen nowSubscribe to LCP podcasts
Dive into expert insights with LCP’s podcast range: Energy, Investment, Insurance, and megatrends facing businesses today. Listen & subscribe across all major audio platforms.