What’s in an index?

As the government reportedly plans to freeze most social security benefits for two years, before uprating them in line with wages rather than prices, Professor Deborah Mabbett investigates the politics of indexation.

The BBC reported earlier this month that ‘sources’ suggest that the government is planning to freeze social security benefits (except old age pensions) for two years, and then uprate them in line with wages rather than prices. It could be that the ‘sources’ are unruly elements in the Conservative party, speaking under the influence of an extended campaign about the relationship between benefits and earnings. But it could be that the government thinks that there is more room for cutting benefits in a stealthy way through an opportunistic choice of index. It has worked the trick once, changing from RPI to CPI indexation (of which more below). Can it do it again? I think not. The art of indexation in retrenchment politics is to achieve savings through a process which runs automatically, without evident government interference. Tweak the index too much, and the process is no longer automatic. Each up-rating starts to look like a political decision. To paraphrase Lady Bracknell, for the government to change the index once is a misfortune; to change it twice looks like carelessness.

Why do we have indexation arrangements at all? In some political systems, where legislation is not easy to get through Parliament, indexation may be a way of locking in an agreement that the parties have struggled to reach. Lock-in has not been an important motive for UK governments, because they can generally pass legislation readily. However, its close cousin, agenda control, does play a role. Governments want to be able to pick their fights. Consigning a policy to run automatically leaves the government able to spend its time on other policies with more political rewards.

When the Conservatives came to power in 1979, only some parts of the social security system were subject to legislation on indexation: principally old age pensions. They were meant to increase by prices or earnings, whichever was the higher. The Thatcher government changed this to price indexation. While there was some opposition to this change, its potential impact was not apparent to many. Opponents found the case for earnings indexation difficult to communicate. Ministers emphasised that price indexation maintained the real value of the pension, and insisted that the formula was not a major assault on state provision.

The Thatcher government’s policy on other benefits appeared more radical. In 1980 it applied a 5% ‘abatement’ to increases in benefits other than the state pension (this was a period of very high inflation, so benefits still rose in nominal terms). However, the government soon lost its appetite for high-profile fights about benefit levels. The convention that benefits would increase in line with prices became established, and was backed by statutory provisions in 1986 and 1992, although the government never tied its hands completely. Indexation for the Tories had an element of ‘stop me before I kill again’: having made some deep cuts, they accepted a commitment to maintain benefits relative to prices. Some on the right of the party may have found this too soft, but the status quo was set nearer the centre-right, where fiscal planners knew what Outraged of Tunbridge Wells did not: that the formula would see benefits fall steadily relative to earnings, provided the economy grew.

By the time New Labour came to power, benefits and pensions had fallen substantially relative to earnings. In a different constitutional system, it might have made sense for the new government to try to lock in a better deal for benefit recipients with new legislation, but legislation does not lock in policies in the UK. For Labour, it was more attractive to maintain the existing indexation conventions and then claim credit for selectively increasing benefits by more than inflation. The clearest example was Child Benefit, which the Conservatives had refused to index. Labour not only increased Child Benefit substantially; it also enhanced provision for children in the system of in-work benefits. This system was greatly expanded in scope and generosity under the new name of Tax Credits. In short, price indexation left a lot of scope for new credit-claiming policies. Labour had no reason to tie its hands with more generous indexation rules.

In both these periods, we can see how indexation was used to control the agenda: holding a defensible position against those who would cut benefits further under the Tories, creating space for more generous but selective policies under Labour. However, to work in this way, indexation must lock in a sustainable position. Otherwise, pressure for a change will build up, not only among affected recipients but also among policy specialists. Whereas the general public pays little attention to indexation rules because the year-by-year changes they produce are small, indices are the bread and butter of policy analysis. Policy specialists make projections and discern the cumulative effects of indexation.

Earnings rose more than prices throughout the 1980s and 90s, so pensions and benefits, indexed to prices, fell substantially relative to earnings. Was this sustainable? Yes, if you believed that income maintenance by the state could be allowed to wither away, with private savings and insurance taking the place of public provision. No, if private provision failed to fill the gap, and more and more people fell into poverty. Under Labour, tax credits were meant to provide a route from poverty for working-age people, but that still left pensioners out in the cold. By the time Labour came to power, pensioner poverty had come onto the political agenda. The indexation arrangements were a target, especially when the index produced an embarrassingly small increase in pensions of 75p in 1999.  Pensioner organisations sought restoration of the earnings link, by then much better understood than it had been in 1980.

New Labour fended off the pressure by improving the means-tested supplementation of the basic pension, variously named ‘Pension Credit and ‘Guarantee Credit’. This minimum pensioner income standard was indexed to earnings rather than prices.  However, the advance of means-testing in old age pension provision began to produce disfunctionalities of its own: more and more people were put in the position that they would not benefit from saving for their retirement, as they would not achieve an income above the means-tested level. It was against this background that the Pensions Commission proposed that the basic state pension should be improved and the earnings link restored. Price indexation of pensions was no longer seen as ‘good policy’ by specialists. Yet, for the reasons explained above, neither the Conservatives nor Labour wanted to give up the room for manoeuvre that price indexation gave them. It fell to the Liberal Democrats to advocate a change, in the form of a ‘triple lock’: pensions should be uprated in line with wages, prices or by 2.5% at minimum. This policy was written into the Coalition Agreement.

Thus George Osborne started the Spending Review with a policy that would work in quite the wrong way for his austerity plans. Public pensions had been on a path that would see them taking a diminishing share of GDP; now savings had to be found elsewhere. However, the blow was lessened by a new reason to take benefits off automatic control: price indexation was no longer producing steady savings. Since the financial crisis struck, prices have been rising faster than wages. In mid-2012, real wages were back where they had been in 2005 (since 2005, the Consumer Price Index has risen 23%; average weekly earnings have risen 22%). With a period of low growth in store, price indexation looked less attractive than at any time in the previous three decades.

At first sight, there might seem to be little to stop the government switching to manual: freezing benefits, or choosing the amount to increase benefits at its own discretion. But automaticity is valuable to a coalition government: the less that comes up for debate the better. The trick is to find a mode of indexation that really takes the issue off the agenda. Switching the price index to one that would produce lower increases was evidently very attractive. The government announced a switch from RPI to CPI in 2010, although, out of deference to the pensioner vote, it postponed its implementation for pensions for a year.

Several organisations have sought to explain the differences between RPI and CPI and I won’t try to replicate their efforts here (one of the best explanations is offered by Jill Leyland of the Royal Statistical Society: link). What is interesting is how statisticians have responded to the government’s manouevre. Back in the 1980s when the convention of annually up-rating benefits in line with inflation became established, there was a serious discussion about the appropriate composition of a cost-of-living index for benefit recipients, bearing in mind that their consumption patterns are not necessarily the same as for the population at large. The Royal Statistical Society has made it clear that it feels that the government has avoided this discussion with its latest change. It points out that the CPI was established for different purposes than benefit up-rating, and implies that the government’s adoption of the measure is opportunistic and unprincipled.

Does this matter? Probably not: the government will get away with a change that will save it £10 billion per annum by 2015. The statisticians will keep nibbling away and there may be changes to the CPI that reduce the savings it generates, but the switch is a big gain for austerity. Few people realise that changing from RPI to CPI is the largest single cutback in the government’s expenditure plans. Much more high-profile measures, such as the introduction of the ‘benefits cap’, which limits the maximum benefit a family can receive to a designated fraction of average earnings, are small beer by comparison (the benefits cap is supposed to save about £270 million a year).

Now the siren voices on the right of the Conservative Party are calling for further changes. Having seen real wages fall, they want to know why benefit recipients are protected. The answer is pretty obvious: because they are already on the breadline. This is where Lady Bracknell comes in. While the political climate remains hostile to benefit recipients, it is not clear that it would stay that way if their benefits fell much more. Furthermore, the government has a coalition to maintain, which points to taking its winnings now and getting benefit levels off the agenda. And finally, real wages will not fall forever, so the government could find itself adopting the wrong index for its preferences. To have to re-open benefit indexation once was a misfortune; to do it twice would be careless.