Hamish McRae: The decision to hold back on increasing interest rates may turn out to be wrong

Economic Life: The danger would be that the rise in short-term rates next year would have to be even greater than it otherwise would

So it looks like August. Given the run of iffy data in recent weeks, including a shading back of confidence about growth in the all-important service sector, there is a perfectly decent justification for holding back on increasing interest rates. Some of us think that this may well be shown as a policy error but the case for "when in doubt, do nowt" is hard to argue against. Given the data as published the Bank of England's decision makes sense. Given what is probably happening to the economy it may well turn out to be wrong.

Click HERE to upload graphic (241k jpg)

First, what can we sensibly say about the economy? Then, what is the likely profile of rates? And finally, what are the risks?

When you are trying to assess how an economy is performing you look at a range of data and try to make some sort of judgement on the basis of these, rather than looking at any single number, especially the official GDP figures, because these are invariably revised. Indeed, there seems to be some consistent downward bias in these. As Goldman Sachs points out in a recent paper, of the 31 occasions between 1975 and 2008 when the initial estimate of quarterly GDP was negative, the average upward revision was nearly 1 per cent, or 3.9 per cent annualised. It would be consistent with that experience were that minus 0.5 per cent figure for the final quarter of last year to turn out to be a plus.

The contrast between published GDP data and other indicators is shown in the first graph, which shows a range of estimates based on purchasing managers' indices, the CBI, the Bank of England agents, and employment growth. As you can see the GDP estimates during the downturn have been much weaker than other indicators. I find the employment data particularly encouraging. It is rising strongly. Over the past year total employment climbed by nearly 400,000, despite a reduction of 120,000 in public sector jobs. In the three months to February the pace of job creation speeded up, with net extra 143,000 jobs being created, a rise in labour participation rates and a fall in unemployment from 8.0 per cent to 7.8 per cent.

That really is not consistent with the economy being stagnant. Goldman Sachs thinks the GDP numbers will be revised upwards and that is surely right. The bank is also forecasting 2 per cent growth this year and 2.7 per cent next, both figures higher than the consensus. But that will take several years before the final figures are agreed on what is happening now and, of course, policy decisions can't wait.

Interest rates are going to go up – there is no doubt about that. The issue is the profile of that rise. As a general guide to what is the "right" short-term rate, a formula was developed by the Stanford economist John Taylor. The idea is that you look at inflation and employment and set a interest rate that is consistent with full employment and stable inflation at the targeted rate. As you can see from the second graph the Bank has pretty much stuck to what the Taylor rule would suggest, going right back to the 1980s, but with two twists.

One twist is that rates did not dip below zero during the recession – they couldn't, or at least negative interest rates are hard to engineer. Instead the Bank introduced quantitative easing: it could not cut the price of money so it increased the supply. You can see in that chart an intellectual justification for recent policy.

The other twist is the impact of a widening of bank spreads – the gap between the amount banks pay on deposits and what they charge for loans. The Taylor rule, calculated conventionally, would have called for an increase in rates before now and would lead to a rise to between 2 per cent and 3 per cent by the end of next year. But if you allow for the increase in spreads (and there is a whole different debate about that phenomenon) then you can argue that the correct Taylor rule is appreciable lower. On this adjusted basis, according to Goldman Sachs, the first rise in rates should come in the final quarter of this year and the climb next year should be slower.

You can have a huge argument about all this but I suspect it does not matter that much either way when the first increase comes through. There have been times when monetary policy has been clearly wrong – some of us argued that the Bank should have leant harder against soaring house prices – but this is not clear now as there are risks either way. But while the corporate lobby has been calling strongly against a rise in rates, some commentators, for example the economics team at Nomura, are worried that by delaying a rise now, rates may have to go up by a greater amount in the future.

That leads to an assessment of the risks. The risks of a too-early increase have been widely discussed and essentially hinge on the hesitant state of the recovery. Fiscal consolidation, despite all the shouting, has only just begun. The forward-looking indicators from the private sector are mixed. The housing market remains very flat in most of the country. Personal incomes are being squeezed. Against this background the first rise in rates has to be managed carefully in public relations terms. There will be some hit to confidence, though since it is widely expected, maybe not that serious a one.

Against this there are other risks. One is the exchange rate, the weakness of which has to some extent been concealed by the even greater weakness of the dollar. Sterling is close to a one-year high against the dollar but at a 14-month low against the euro. (The ECB is clearly more concerned about inflation than the Bank so expect another rise in euro rates in July.) Another is what might happen to long-term sterling rates.

At the moment it costs the British Government a little more to borrow for 10 years than it costs the average between Germany and France, about 0.3 per cent in fact. For what it is worth, Goldman suggests that a year from now the gap could rise to 0.8 per cent. If that were to happen you could almost say that because we had a laxer policy than the eurozone we would have to pay half a percentage point more than France and Germany for 10-year money.

Given our still huge fiscal deficit that would not be great. For a start it would mean that other borrowers, people seeking fixed-rate mortgages, for example, would pay more too. And then there is the danger of a more serious loss of credibility of UK monetary policy, with an even weaker pound and even higher long-term rates. The danger then would be that the rise in short-term rates next year would have to be even greater than it otherwise would.

Let's see how the economy develops. It cannot be said too often that the early stages of any economic recovery always feel fragile. Look at hard data such as employment numbers, the balance of payments, house prices and tax receipts. Discount all those dire warnings from self-interested lobby groups. Ignore political point-scoring of course. And be as distrustful of unexpectedly strong data as well as unexpectedly weak stuff. It is the big picture that matters and that is one of steady modest growth.