15 November 2108
RIP RPI
Change the inflation measure for everything including gilts.
by Frank O’Nomics
Is this just another government “cake” event (as in having it and eating it)? The announcement that the inflation index to be used for National Savings products will be changed to CPI from RPI is hardly the stuff of headlines. Moving to an index that is, on average, around 0.7- 1% per annum lower, will reduce the returns on such products for the 500,000 who have invested in them but, given that they are no longer being offered, the impact will become less significant over time. The move does, however, re-stimulate the debate as to whether all government receipts and payments should be moved to CPI for reasons of fairness and consistency. At present it may seem that the government has the best of both worlds, indexing state and public sector pensions, unemployment benefit and now National Savings to CPI (having cake), and repayments on student loans and rail fares to RPI (eating it). Such allegations are, however, a little unfair. The real accusation is of inefficiency. A universal move to CPI on both sides of the balance sheet could actually save money for the government.
At the heart of this debate lies the question of what the right measure of inflation should be. We are all subject to subtly different inflation rates depending on our life choices (school fees inflation has been rampant while food inflation has been moderate for some time) and our time of life; pensioners spend their income on a very different basket of goods and services to those in work. Picking something that is fair for all is difficult. Some argue that RPI overstates inflation, others that CPI understates it. The difference between the two lies largely in RPI being an arithmetic average and CPI geometric, but there are also issues with what is in each index, with RPI including housing costs which are left out of CPI to a large extent. Clearly there is a need to choose one or the other and, since the government changed the measure for the inflation element of the triple lock on pensions to CPI in 2011, this would appear to be the one they think is the fairest. Given that RPI focuses purely on the expenditure of private households, while CPI covers the whole economy including students, those in nursing homes and overseas visitors, it is becoming harder to argue with them.
So why have they not moved to CPI for everything? Student loans, for those who started their courses since 2012, are charged interest at RPI +3%, which means that they are currently accruing debt at 6.3%. Over time compounding at 1% less per annum would make a big difference to the size of the outstanding amount. Why hasn’t this been changed? The cost of doing so would be nothing like the figure you would get by applying the difference to the full amount of student debt, given that it is expected that many will not earn enough to trigger repayments before the debts are written off after 30 years. Similarly, rail fares are increased at RPI, yet the transport secretary, in an effort to reduce costs, will now subsidise rail industry costs at a rate linked to CPI and not RPI. The disparity means that both the consumer (paying RPI linked fares) and the rail companies (receiving a CPI base subsidy) are taking a hit. There will be other areas which might result in a cost for the government, such as linking alcohol and tobacco taxes and air passenger duty to the lower measure.
All of this reduction in revenue could be more than countered by a much bigger saving if the inflation measure used for government bonds (gilts) were to be changed to CPI from RPI. The market value of government debt is close to £2 trillion, of which around 23% (£660 billion) pays interest at a rate linked to RPI. Changing this to CPI could save the government £3-4 billion per year. There are good reasons why this hasn’t happened, in that it would instantly wipe a huge amount off the value of the debt (as much as £150 billion), most of which is owned by UK pension funds – who certainly don’t need a further setback. It would also be difficult to make the change for existing debt from a legal point view, given that RPI is the measure stated in each prospectus. However, there seems little reason not to link all new issues of debt to CPI.
A consultation in 2011 concluded that such a move would not be cost effective, partly because the market would demand a cheaper price for a product that would be a small part of government debt for some time, with limited liquidity and a lower return. The Debt Management Office did state that the future issuance of CPI-linked gilts was not precluded and, given the almost insatiable demand for inflation-linked debt from pension funds, the time could be ripe. Clearly, if funds are paying pensions linked to RPI there will be a mismatch and there may be a need to allow funds to switch to using CPI on their payments. Recent evidence highlights some headwind. To some extent changes are outside government control, with a judge recently ruling against BT using CPI instead of RPI when determining pension payments. A move to CPI would hit many pensions (government estimates suggest it would cost the average pensioner £12,000), but this could be countered by the enhancement to pension fund solvency. Being paid 5% of something is better than getting 6% of nothing.
Rather than gorging on baked confections, the government is in a position to shift to a system that is consistent, ultimately fairer and cost effective. Making this sound a simple process is dangerous. Adjusting the calculation of pensions is a very sensitive area and one that would produce some significant opposition. Making a start by issuing a modest amount of CPI-linked debt makes sense, as does easing the debt-servicing burden on students. On the one hand it could save them a huge amount of money in debt servicing costs, on the other it would introduce a fairer measure for student loans repayments. An even wider cohort would welcome the additional benefits of cheaper rail and airfares.
