A sensible approach to pensions

The answer to most questions in economics is exactly the same – yes and no. The imprecise nature of the subject allows reasonable people to disagree and for most economists to keep their jobs, even when they get it hopelessly wrong. This basic rule was in evidence in the comments on George Osborne’s Budget proposal to link public sector pensions to the Consumer Prices Index (CPI) rather than the traditional Retail Prices Index (RPI). It was right to claim that this could lead to lower pension payouts – but it was also wrong. And there are compelling reasons for saying that neither measure is really appropriate.

Until the start of 2004, there was very little argument about index linking. As chancellor, Gordon Brown set the inflation target for the Monetary Policy Committee (MPC) in terms of the RPI, but then switched to the CPI. Both are shopping baskets purporting to track the impact of price movements on the average household. But there are certain technical differences, chief among which is that the RPI includes an allowance for housing costs (logically so, in a nation of owner-occupiers). The CPI does not. There was also an RPI measure which excluded mortgage interest payments. Leaving interest changes in the index clearly means that the policy weapon used to control inflation could actually create inflation.

In setting interest rates from 2004 to 2007, the MPC had to track a basket of goods which excluded one of the key factors pushing inflation upwards – asset prices. The CPI, therefore, produced a lower measure of inflation and thus lower interest rates, very conveniently for a government planning an election 18 months later. The switch added fuel to an already hot housing market.

So, given the pivotal role of housing costs and mortgage interest payments and in household expenditure, it is logical to argue that the CPI-linked payments will be lower than those related to the RPI. Except that it isn’t. From September 2008 and right through 2009, the CPI was higher. For eight conse cutive months last year, moreover, the RPI was negative, implying index-linked payments should be reduced. This clearly reflected movements in interest rates and house prices. And with some analysts predicting another bout of house-price deflation, it is by no means certain that the 2009 experience was a one-off.

Which measure will produce the fastest rate of inflation, therefore, will depend on the price movements of components in the index and their relative weights. To claim pensioners will suffer lower payments as a result of linking to the CPI is to claim far more than can be justified by recent experience. And there are other arguments supporting chancellor Osborne’s proposal.

When Gordon Brown changed the MPC target, he only did half the job. It seemed illogical to fix interest rates according to one measure of inflation yet link benefit payments to another. But then politics is rarely logical. Perhaps he realised the CPI at the time would be lower than the RPI, so interest rates could be kept down for homeowners – while payments could be maintained at the higher rate. Good politics maybe, but bad economics. As long as the RPI was higher than the CPI, and as long as the RPI was used as a benchmark by government for benefits and unions for wage claims, it was going to contribute to inflation. Just one measure of inflation makes much more sense.

But for pensioners, there might be an exception. By the time they retire, most people have paid off their mortgage and many will have benefits, such as concessionary travel, free prescriptions and senior citizen rates at cinemas. But they will spend more on food and heating. So, the typical pensioner shopping basket will be very different from that of the working family and the CPI index should be re-weighted to take this into account. Is it time to revive the old Pensioner Price Index, which seems to have fallen into disuse recently?