How to fund a happier retirement

A US government plan for civil servants is the unlikely blueprint for a stakeholding pension scheme.

For a few months during 1996 new Labour flirted with Will Hutton and his ideas about "stakeholding". It soon turned out, however, that this meant becoming like Germany, and the brief affair fizzled out. Nevertheless, one token of new Labour's early affection for stakeholding remains: the government's plans for so-called "stakeholder" pensions formed part of Alistair Darling's pensions green paper, published just before Christmas.

Its proposals herald a profound shift in the balance between public and private welfare provision. Though private welfare became steadily more important under the Conservatives, its growth was relatively slow. "Free market" welfare services - financed and provided by the private sector and purchased voluntarily by individuals - grew from 25 per cent of total (public and private) welfare spending in 1979/80 to 29 per cent by 1995/96.

The green paper envisages that this gradual shift in favour of the private sector will continue. At present, it tells us, 60 per cent of pensioners' income is derived from the state and 40 per cent from private sources. But under the government's plans, by 2050 the situation will have reversed, so that private sector support for pensioners will be half as large again as state support. Whether we like it or not, the majority of us are going to have to rely more heavily on private sector pensions - or Lisas (lifetime individual savings accounts) - under the government's newest rubric.

Twenty-five years ago the then Labour government was moving in precisely the opposite direction. The postwar scheme, based on Beveridge's blueprint, had been wholly flat-rate. It provided a universal minimum income which, while sufficient for the needs of the poorest, left workers accustomed to higher standards reliant on the private sector to supplement their retirement income.

Labour's 1975 legislation was intended to rectify this dangerous over-reliance on the private sector, as it was then perceived, by extending earnings-related pensions to those parts of the population that did not have access to an occupational scheme. The new state scheme, Serps, would provide secure, defined-benefit pensions, protected from market fluctuations and the ravages of inflation. As Richard Crossman described the policy, the new scheme would transfer "the privileges of the minority into the rights of the majority".

Darling's green paper proposal to turn Serps into a flat-rate State Second Pension scheme brings policy full circle. While not explicitly stated, it is clear that the government believes the state's role should be confined to ensuring a minimum income standard. If Darling's proposals are followed through, in the future earnings-related insurance will once again be a private affair. The job of the state will be confined to ensuring everyone receives flat-rate benefits at the minimum level. People who want to retire on an income above this minimum will decide for themselves how much they want to pay into a pension and which private provider offers them the best deal. The government will be responsible for regulation, rather than provision or finance.

There is, however, a fly in the ointment. The pensions mis-selling scandal, a legacy of the Conservatives' reforms of 1986, has soured the relationship between private pension providers and the public. As is now well-known, the introduction of personal pensions led to hundreds of thousands of people being sold a new product when they would have been better off staying with their existing pension arrangements. This mis-selling mostly affected people who could have joined or stayed with an occupational scheme and who forfeited employers' contributions by choosing to opt out, a fact that commission-hungry sales staff conveniently ignored. But it also showed up the very high costs and charges attached to personal pensions, with up to a quarter of an individual's savings being lost through management and administration fees. Low and moderate earners in particular do badly, because of the high front-end, flat-rate charges that are often made. Forcing people to rely on existing personal pensions would be bad economics and worse politics.

So far there have been two general types of proposals for dealing with the problems of private pensions. The first argues that the problem is excessive regulation, which, it is claimed, drives up compliance costs and drives down the number of providers. The second suggests that the problem is not enough regulation, and that the government should go actively into the business of setting appropriate fees for pension provision.

Both of these arguments are, in serious ways, flawed. First, the argument for less regulation does not take into account the problem of what economists call "asymmetric information". That is, the providers of a product have much more information about its quality than consumers, a problem often faced in health care, for example. In such a case, consumers can either spend a substantial amount of time attempting to close the gap between what they know and what the providers know, or they can choose randomly, basing their decisions upon advertisements. This leads to a market where providers compete more on marketing than on reasonable charges, which is exactly what happened with personal pensions.

The second argument is equally flawed. The government does not have the information to determine what price is fair or reasonable for financial products, nor does it have any reason to prefer certain providers over others. Such a decision would require not only that it have comprehensive knowledge of the market as it exists today, but also of the adaptations that will occur in the future. Even if it is sold as part of a "partnership between government and business", detailed regulation of charges and providers would put the government in the untenable position of setting prices and closing out competitors. In other words, it would be back to "old Labour" command and control standards.

The government's proposals for stakeholder pensions do, to some extent, finesse the problem of whether to regulate more or less. In essence the government's idea is that only schemes meeting particular "benchmark" standards will be able to call themselves stakeholder schemes. According to the green paper these standards will include allowing individuals to transfer funds between schemes without penalty; sending out regular information about the value of pension entitlements that members have accrued; and, most importantly, ending flat-rate charges. In return the government will lighten the regulatory conditions attached to taking on new clients, relying instead on an annual administrative audit to ensure that the provider is still up to scratch.

While this is an improvement on the personal pensions regime, such a strategy is still fraught with problems. In particular there is a danger that private firms and their paid lobbyists will twist the government's ear about the impossibility of providing a product more cheaply or better. Weak consensual regulation could all too easily result, ensuring that the benchmark tests will leave most providers able to carry on their business as normal.

We think that there is a better option available, based not on theory but on existing practice in the United States: the US government's own plan for civil servants, the Thrift Savings Plan (TSP). Suitably adapted for a different purpose, it would admirably deal with all the considerations the government finds so nettlesome.

The TSP is remarkably simple in its basic structure. The US government contracts with an outside provider to set up and manage, at arm's length (and off the government's accounts) three indexed accounts, one for stocks, another for bonds, and a third for short-term government securities. All are managed passively - they do nothing more than mechanically purchase a standard basket of financial assets, such as the S&P 500 or the Lehman Brothers Long Bond Index. Overhead costs are therefore kept very low, with administrative expenses being covered by an annual deduction of approximately one-tenth of 1 per cent of members' accounts. In comparison, most personal pension providers currently charge around ten times this level.

The advantages of having a single provider, with very narrow investment choices, all managed passively, are enormous. First, a single provider can attain remarkable economies of scale, spreading basic management costs over a large contributor base. Second, narrow choices reduce transaction costs for those who either don't know or don't want to learn the ins and outs of financial products. This all but eliminates the need for outside (and expensive) financial advisers. Finally, a simple, centralised system can accommodate a wide range of contribution levels, in the TSP as low as ten dollars every two weeks.

The most important advantage, however, is that active managers cannot, on average and over any reasonable period of time, beat passive investing, for reasons that are quite obvious but usually overlooked. An index simply measures the average of what all stocks are doing. By definition, all investors collectively cannot beat an average, since taken together they are the average. Once you subtract the costs of investment research, trading and managers' payments, active managers will always under-perform the index. In fact, according to a study by Barclays Global Investors, approximately 80 per cent of active managers trail the index. They are attempting, collectively, the impossible: it should be no surprise that they usually fail, or that they fail more dramatically the harder (and more expensively) they try.

For all these reasons, the ideal stakeholder pension would incorporate all of the major components of the US government's Thrift Savings Plan. But it would be a mistake to have the government monopolise or otherwise limit providers. The better option is for the government to provide the "default option", but to allow any pension company that provides a comparable product (no front-end load, charges assessed on an annual basis, and so on) to enter the market as well. This is desirable in part for political reasons: completely shutting out the pension companies would be a huge obstacle for a centrist government. But there are principled reasons as well. Most of what our knowledge tells us is that an indexed, centrally run system will out-perform most private options. Even so, governments work better when they recognise their own fallibility, and that circumstances may change over time.

A monolithic system would tend to discourage innovation; when financial service companies have to compete to survive, though, they promote the development of new and better products. The very low charges of the state system will inevitably drive down costs in the private sector, squeezing out many low-quality providers, and forcing the rest to focus on simplicity and value for money, rather than marketing. In this way the stakeholder pensions system could act as a catalyst for a revolution in the way pensions are managed and sold.

Our Thrift Pensions proposal cuts through the Gordian knot of regulation, achieving all the government's objectives without reducing competition, but by actually expanding choice. Instead of a messy scheme that introduces unnecessary and possibly damaging cosiness between the government and the financial system, our scheme allows the government and markets each to operate within the sphere of their own competence.

In sum, the scheme is based on a tested and well-liked model which deals well with all the political and economic problems of pensions provision for the low-paid. Increasing competition through expanding choice represents the best route forward for stakeholding.

Steven M Teles is research assistant professor at Boston University and is completing a book on the politics of pensions privatisation in the US and UK. Phil Agulnik is a research student in the ESRC Centre for Analysis of Social Exclusion at the London School of Economics