Pension flexibility and the long-term public finances

At Budget 2014, the Government announced a number of tax measures that increase the flexibility with which individuals can access their defined contribution (DC) pension assets. The current system allows some flexibility for individuals with small or very large pots, but others are subject to a 55 per cent withdrawal tax and most still buy an annuity. From April 2015, everyone with a DC pension pot will be able to withdraw their funds as they choose, from the age of 55, subject to their marginal rate of income tax at the time.

In our March EFO, we noted that offering more flexible access to pension pots was likely to affect the composition of households’ financial and non-financial assets. Some people may redirect funds from annuities into other financial assets or housing. Some may increase their pension saving temporarily to benefit from tax-free lump sum withdrawals. Some may use their funds to increase their consumption, although we judged that this effect was likely to be small. In the EFO, we highlighted the fact that the size of these effects are very uncertain, but we concluded that the principal impact of the change would be on the composition of household assets, rather than on the aggregate flow of saving or spending.

This box considers two possible sensitivities around our March judgement that have been the subject of public interest. Specifically, we consider: first, the possibility that there will be more money flowing into the housing market; and second, that people may spend their pension pots relatively early in retirement, leading to greater reliance on income-related benefits.

What if pension flexibility leads to significant amounts money flowing into the housing market?

The IFS estimated the size and characteristics of the population affected by the pensions flexibility measure in a recent report.^a It found that around 40 per cent of individuals aged 55 to 59 have some money in a DC pension, and a quarter of these are already able to draw down funds ahead of retirement because they have sufficiently high income or a sufficiently small pot. As such, 30 per cent (around a million people), with average DC assets of around £180,000, are estimated to acquire new flexibility as a result of the measure.^b

There is no way of knowing how much extra money will be invested in housing given the new pension flexibility, particularly over the projection horizon of this report. Other investments, including annuities, may become more attractive as financial conditions normalise. And the greater flexibility may also mean other would-be investors in the housing market will choose pension saving instead. The amount of money flowing into housing might also be more than the capital deployed from pension pots, if it were leveraged through mortgage borrowing.

But for illustrative purposes, we can consider what the effect on house prices and the public finances might be if the entire annual flow of new annuity business, estimated by the Treasury to be around £11 billion a year,^c was to be diverted into housing. This is an extreme illustration, as it includes the whole population of retirees. But assuming a one-for-one impact on house prices, no housing supply response and commensurate growth in DC pension pots, this would add ¼ percentage point to house price inflation every year. A simple fiscal ready-reckoner suggests this would add a relatively small £25 million to £50 million a year to our receipts forecast.^d Even cumulated over 50 years, the effect would only remain around 0.1 per cent of GDP.

If the effect on house prices was matched in rents, there could be offsetting effects from higher spending on housing benefit and lower VAT receipts as a higher proportion of consumer spending went on housing costs that are not subject to VAT.

What if people spend their pension pots too quickly and need support from the state?

Some commentators have argued that the new flexibility could result in a greater proportion of people using their pension pot for short-term consumption, perhaps because they discount the future too heavily (i.e. they place a greater weight on enjoying a Lamborghini today than on living comfortably in, say, 10 years’ time). More pensioners might therefore become dependent on income-related benefits in the future.^e People may also simply underestimate their cohort life expectancy and may exhaust their pension pot, causing them to fall back on state support.^f Depending on what this extra consumer spending purchased, it could boost VAT receipts. But as this would bring forward the use of pension assets, rather than increasing them, that would be a temporary timing effect rather than a lasting boost to receipts.

To the extent that pension pots were exhausted more quickly, the main fiscal impact on the spending side would be likely to come from the housing-related elements of income-related benefits, notably housing benefit and council tax support, rather than pension credit, the straightforward income support for pensioners. Eligibility for housing-related benefits extends further up the income distribution than it does for pension credit, especially following the removal of the savings credit element of pension credit as part of the single-tier pension reforms. Our central projections do not assume greater recourse to income-related pensioner benefits as a result of myopic or misjudged use of pension pots. But the sensitivity is modest: by way of illustration, we project that spending on these income-related benefits for pensioners will amount to around ¼ per cent of GDP in 2063-64, so even a 5 per cent rise in caseloads would increase spending and the primary deficit by only 0.01 per cent of GDP in that year.