The UK pension system has come to have a strong reliance on ‘means testing’, ie, where state benefits are made conditional on not exceeding a given level of income or wealth, and which is causing serious damage to private pension provision.
The last 10 years have seen a dramatic increase in means testing in UK pensions. But is this a bad thing? If you asked an economist, you would not get a clear-cut answer. The reason is that while means testing has many bad points - it distorts incentives to save and to work and is complex and awkward to administer - it also has some significant good points: it helps focus public assistance on the most needy and therefore keeps overall tax rates low, improving labour incentives.
Indeed, some of the strongest support for means testing in the academic literature comes from those who intellectually we would associate with the conservative movement. Milton Friedman (the Nobel laureate and leading monetarist economist) advocated the use of means testing due to the positive effect on incentives, in that it could lead to lower income tax. For any desired level of benefits (or minimum income level), means testing clearly serves to lower costs and public expenditure relative to a universal benefit, thereby enabling payroll taxes, which are a disincentive to work, to be reduced. Friedman hence concludes that means tested systems are superior to universal benefit systems on grounds of economic efficiency.
Another prominent conservative economist, Martin Feldstein of Harvard University, extended Friedman’s analysis by incorporating such factors as the stigma associated with means tested benefits and the consideration that some members of the population may not adequately understand the mechanics of the system – two factors that have been identified as key weaknesses in the government’s policy.
Yet, even taking these factors into account, Feldstein still finds (for the same reasons as Friedman) that means tested systems are generally superior to those with universal benefits. He also argued that a lower tax burden improves incentives for those outside the means test and in most cases this is sufficient to offset the negative impact, on incentives, on those brought into a means tested system. On balance, unless the proportion of workers drawn into (or near) the means test is large, the effect on economic efficiency is positive because of the lower tax rate.
It is not means testing alone that is problematic, but the likely extent of it and the way in which it has been implemented, which are the roots of the problems we foresee. In a recent study1, we show that the pension system in the UK is evolving towards the following unsatisfactory outcomes:
q Unsustainable. The system will increasingly lead to economically implausible outcomes and even greater disincentives for saving
q Confusing. The system has not been explained well by the government and this is reflected in the lack of understanding amongst the general population of the means-tested benefits that they are eligible to receive;
qUnimaginative. The system is designed in an old-fashioned and unimaginative way, which relies on static data. It borrows from past failures in British welfare history, rather than thinking through what is possible and preferable, in a modern period where dynamic data are available. Implicit within the system are unnecessary disincentive effects such as a minimum income guarantee well above the level of the full basic state pension.
In the UK system the means testing is achieved by means of a minimum income guarantee, whereby state benefits are topped up to the desired level. This would clearly remove any incentive to save for those on small incomes (since any additional income from savings would simply reduce the state top-up), so it has been modified by the introduction of a pension credit. The effect of this is to credit back 60% of the income from savings that would otherwise be lost.
A particularly striking finding of our study was that the present pension system will result in a large proportion of the retired population receiving means-tested benefits. Also, as a result of the fact that the basic state pension is indexed to prices while the pension credit to average earnings, many more retirees will be drawn into the system in the future. Currently it is estimated that around 52% of adults over age 65 are in households eligible to receive the pension credit, and this is forecast to increase to 73% over the next couple of decades as a result of the indexing rules.
Unfortunately, this edifice has confused the population to a considerable extent and people on modest incomes cannot clearly see a link between making voluntary savings while working and achieving a higher standard of living in retirement. Some Watson Wyatt/You Gov survey research found that in 2003 only 3% of the population were “very clear” about the Pension Credit, and 61% were “very unclear” (see table 1).
It is not that people are always confused about retirement. Our Watson Wyatt/YouGov research shows that they have a fairly good understanding of the basic state pension. Over 60% of individuals were able to correctly identify the value of the full basic state pension for single individuals and this percentage was close to 70% among individuals expecting to receive a majority of benefits from the state (see table 2).
So, we are now in a situation where individuals do not understand a very complex pension system and are not making sufficient provision. The situation is not something which can be easily rectified by improving understanding either - even if individuals did understand the system they might not make provision as there are significant design flaws such as setting the minimum income guarantee well above the level of the basic state pension.
Unfortunately all this is happening at a time when occupational pension provision is set to decline. Recent surveys carried out by Watson Wyatt show an unmistakeable trend of employers closing final salary plans and replacing them, for future entrants, with significantly less generous defined contribution plans.
Some means-testing is unavoidable in pension systems but we are concerned about the degree of means testing in the UK, particular features of its design and the confusing way it has been implemented.
Paul Thornton, Jonathan Gardner, Mike Orszag are with Watson Wyatt
1Reference: The Pensions Predicament – Means Testing, the Savings Trap and the Labour Market by Paul Thornton, Jonathan Gardner, Mike Orszag. Published by Politeia, 2004