Newspaper Article

Auto-enrolment has changed the pensions landscape but big questions still remain

Date: 11 December 2017
Publisher: News International
Published in: Times

Seven years ago I completed a review of automatic enrolment for the new coalition government. The idea that most workers should be automatically enrolled into a pension had been proposed by Adair Turner in 2006, taken up by the Labour government and followed through by the coalition. Amid concerns that too many younger people would end up with an inadequate income in retirement, the hope was that harnessing people’s inertia would increase pension coverage in the face of a long-term decline in employer-provided pensions.

Along with the prevailing consensus, my review certainly supported the notion of auto-enrolment. It made a number of detailed recommendations.

Since then, the world has changed rather a lot. Another review is due to be published within the next couple of weeks. Richard Thaler, one of the founding fathers of nudge theory and arguably an intellectual father to the policy, has just won the Nobel prize in economics. So what have we learnt in the intervening years?

The first thing to say is that, so far and within its own terms, pension auto-enrolment has been a roaring success. The vast majority, something like 90 per cent, of those eligible are now members of pension schemes. The power of the default option, and in particular, perhaps, the power of inertia, has rarely been so well demonstrated. If you don’t put people into a pension automatically, only a minority will decide to join; if you do put them in, the vast majority will stay put. Pension coverage has risen dramatically. At least in terms of numbers enrolled in an employer pension, if not in amounts being contributed, decades of decline have been much more than reversed.

This success has been accompanied by two big changes in the economic context that I did not envisage back in 2010. The first has been the decade long stagnation in earnings. Back then, we all imagined that earnings growth soon would get back to normal, pre-recession levels. That has failed to happen to a spectacular degree. What is worrying for government is that earnings are again rising less quickly than prices, while automatic contribution rates are due to rise — from 1 per cent of pensionable pay for employees and employers today to 3 per cent from employers and 5 per cent from employees by April 2019.

Whether present coverage rates of 90 per cent or so will survive these increases remains to be seen. If they do, one consequence will be an even bigger squeeze on earnings and on take-home pay. The power of inertia could see income, consumption and living standards suffer more than was ever envisaged by the architects of the scheme.

The second big and unexpected change in the labour market has been the growth in self-employment and the gig economy. There are many explanations for this, but among them is the very different treatment of employees and the self-employed when it comes both to tax and to regulation. Not only do companies pay a lot less in national insurance contributions if they contract with a self-employed person, they also avoid all the cost and hassle of auto-enrolling them into a pension. Because this creates a distinctly unlevel playing field and because this means that millions potentially are missing out on pension provision, I would expect this year’s review to pay a lot more attention to the issue than I did.

Issues about quite what will happen when the generation of auto-enrolees reach retirement also remain to be determined. The expectation when the policy was implemented was that most would convert their pension pot into an annuity; they would have had no choice. Recent legislative changes provide a lot more choice; there is no compulsion to buy an annuity. How a generation that has relied on its own inertia until retirement will deal with complex choices at retirement we won’t discover for a while.

For all the success of this particular policy so far, questions clearly remain both about it specifically and about the use of this sort of policy more generally. They need careful consideration.

First, if inertia really does prove to be as powerful as it seems, then extreme care needs to be taken in determining the correct default. It may well be desirable that young people, perhaps unwittingly, are sacrificing present income that they might have put towards a house purchase in favour of more money in retirement. But it’s not obviously right for them all.

Second, we need to be very careful about places where we have perhaps unintentionally created defaults. One obvious example is the choices students make at 16 and 18. The default, and by the far the easiest route, is to do A levels and go to university. It requires far more effort, for example, to find an apprenticeship. This almost certainly leads to many bad decisions. The answer is not to make going to university harder, by the way, it is to level the playing field by making other routes easier.

Third, deciding on a default is only a first step. Nearly all those who default into a pension also end up in the default fund choice. Regulation and arguably education become more, not less, important in the face of disengaged consumers, especially if they are expected to become engaged at some point later on.

Finally, it is striking that more than a decade after the consensus was built around the desirability of auto-enrolment, it remains easily the biggest example of the use of behavioural economics in policy. Nothing else on even remotely the same scale has been contemplated.

There are, as ever, no magic bullets here.

Paul Johnson is director of the Institute for Fiscal Studies. This article was first published by The Times and is reproduced here with permission.