Jon Cunliffe: Monetary policy one year on
The views expressed in this speech are those of the speaker and not the view of the BIS.
My appointment to the Bank of England was announced in July 2013, at about the same time we first had official data in showing the UK economy growing strongly. I joined almost exactly a year ago on 1 November, shortly before the inflation data for October came in showing a 0.5pp drop to 2.2%, the biggest fall in over 18 months.
But as always in economic policy making one has to distinguish between correlation and causality. The latter is nearly always uncertain. Indeed, as an economic policy maker, especially a forward looking monetary policy maker, one must accept one is almost always making policy under conditions of considerable uncertainty. Perhaps that is why you so often hear from monetary policy makers about puzzles and how "the data are unusually difficult to read this month".
Our job on the MPC is precisely to make judgments about the future on the basis of imperfect information. To decide when we have enough to act and when it is better to wait. And to adapt those judgments as the economy evolves.
Against that background, and on the occasion of my first anniversary at the Bank, I want tonight to review how the economy has evolved over the last year against what was expected when I joined the MPC. And to look at the judgments and uncertainties we now face. Needless to say, as always, the "data are difficult to read".
When I arrived on the MPC, we were seeing signs of a strong recovery - stronger than anyone had forecast - that had started in the spring of 2013. Over the middle 2 quarters of that year, GDP had increased at an annualised rate of over 3%.
The key drivers of the recovery seemed to be renewed consumer and business confidence due in no small part to the de-escalation of the Euro crisis and an easing in credit conditions. These in turn released pent up demand in the economy.
The evidence suggested that this strong growth would continue into 2014. Forward looking economic indicators were very strong. The business surveys were pointing to continued growth at that pace. The squeeze in living standards from higher energy and food prices had started to ease.
But, growth seemed to be almost entirely driven by consumer spending and the emerging recovery in the housing market. That strength in consumer spending did not come from growth in real incomes but from a large, two percentage point, drop in the savings rate. Business investment still appeared to be dragging on activity. So there were real question marks about whether this strong growth could be sustained.
The MPC took an optimistic view. We forecast business investment would increase as the surveys suggested and play a bigger role in supporting growth. We forecast also a return to productivity growth - a real casualty of the recession - that would drive higher pay and hence provide a more sustainable base for consumer spending. And, as a corollary of higher productivity we thought the growth in output would be accompanied by a slower fall in unemployment.
So overall, our expectation at the time I joined was that the recovery would broaden out in 2014. But over the year the boost from pent up demand would fade and as a result growth would begin to fall back from very strong rates to a pace around its historic averages.
Three other MPC judgments from my first meeting and Inflation Report are also worth recalling.
First, we did not think the UK economy would get much help from the rest of the world. A combination of relatively weak growth in the parts of the word that mattered most for UK exports, particularly the Euro area, and the disappointing performance of UK exports in recent years led us to believe that the UK's net trade would not provide a material contribution to growth.
Second, while we expected productivity growth to pick up in 2014 we did not forecast that the UK would recover any of the very substantial amount of productivity growth, 14% relative to the pre-crisis trend, that had been "lost" in the five years since the crisis.
This so called "productivity puzzle" - why the UK's productivity had fallen so far and had not recovered - was a major part of the uncertainty the Committee faced about how fast the UK could grow without generating inflation pressure. If some or all of that lost productivity growth could be recovered, the economy could grow faster and for longer without putting the inflation target in jeopardy. But looking forward the Committee took a cautious view and based its view of the future on the expectation that productivity growth would recover only slowly and that none of the "lost" productivity would be recovered.
Third, the MPC was operating under the "forward guidance" it had agreed in August last year and to which I had subscribed on joining the Committee. The underlying concern behind forward guidance was that the market might react too quickly and too simplistically to strong GDP growth numbers and forget the amount of spare capacity still in the economy. If the market predicated the Bank's reaction to strong growth on a pre-crisis basis and, incorrectly, began to price in a tightening of policy the recovery could be choked off.
To head off that risk the Committee had, earlier in the year, made clear its intention, through "forward guidance" not to consider tightening policy at least until there was a substantial reduction in spare capacity as evidenced by a fall in the unemployment rate to below 7%.
So, over the subsequent year what have we learned? Have the puzzles and uncertainties resolved themselves? Have they been transformed into new puzzles and uncertainties? And what does it all mean for policy going forward?
The good news is that the recovery has remained strong. Indeed, after the revisions to the past data in this year's blue book we now know the economy passed its pre-crisis size in the third quarter of 2013. We can talk now of an expansion rather than a recovery.
On the latest data, growth in the four quarters to September now stands at 3% - a little stronger than expected back last November. Growth was particularly strong in the first half of the year and the business surveys suggest that may well be revised upwards.
In the latest quarter of this year, growth is reported to have fallen back to 0.7%. The burst of pent up demand that propelled the initial recovery was a little stronger and lasted a little longer than we had been expecting last November. But it does now seem to be fading.
Moreover, growth has become more broadly based due to the marked pick up in business investment. The shift of the balance of growth towards business investment that the surveys were pointing to last November, has I am pleased to say come through more strongly than expected. Business investment now looks to have driven around a third of the increase in GDP we have seen over the past year. That has made the recovery more broadly based and more sustainable.
And unemployment has continued to come down at a much faster pace than was expected, dropping from 7.6% in the data that came out when I arrived in the Bank last November, to 6.0% in the latest data - a rate of decline not seen since the late 1980s. Around half a million people have been taken out of unemployment. That has to be good news.
But, as always in economics, there is a cloud to the silver lining.
Output has grown strongly over the past year but unemployment has dropped faster than output has grown. So though the UK economy is producing more, it is using even more people to produce it. In other words, productivity growth has not made even the modest recovery to 1%, half of its pre-crisis trend, that we forecast last November. It was negative over last year and has achieved only 0.5% growth over the past 4-quarters. This is around half the pace the MPC had expected, though it picked up a bit in the latest quarter.
And, needless to say, the productivity puzzle remains a puzzle. Upward revisions to the official data for the immediate post crisis years have reduced the size of the puzzle a bit. But the level of UK productivity is 14% down relative to pre-crisis trend. So with productivity recovering very slowly and still growing well below the pre-crisis trend, it is looking less and less likely that that "lost" productivity will be recovered any time soon.
And there is now a new puzzle: why, despite this rapid fall in unemployment to 6%, have we seen such weak pay growth in the official data? On the average weekly earnings measure, annual private sector pay growth is running at just 1%, still below the rate of inflation. Had the MPC been told last November that unemployment would fall to just 6%, we would have expected pay growth to have been very considerably stronger. What has happened to what, hitherto in the UK, has been a relatively stable relationship between pay and unemployment?
Part of the explanation for the weak growth in pay is surely the very low increase in productivity - if workers are producing less output per hour, there is less scope for pay rises. But even adjusting for low productivity by looking at pay relative to the output of the average employee, we find a very weak picture. Growth in "unit labour costs" has been negative over the past year. Not only have we been producing more but with relatively more people and hence lower productivity. We are also paying less labour cost per unit of output than we were before.
There are, as always, possible seasonal and one-off factors: some patterns of pay seem to have been distorted by the change in income tax rates in April last year, which cloud annual comparisons. Even adjusting for that effect though, unit labour costs look broadly flat - much below the 2% rate of increase we need to meet the inflation target in the medium term.
The big surprise, therefore, for the MPC - and it is a positive, if puzzling, development - has been the extent to which employment has been able to grow without generating clear inflationary pressure. Understanding why that has happened and how long it will persist is, in my view, now key to deciding policy.
One possible explanation is simply a longer-than-usual lag between falls in unemployment and pay pressure emerging. Once you take into account productivity, the unexplained weakness in wages only really materialised over the last three quarters, a period over which unemployment was falling very rapidly.
Some survey measures suggest the pay of newly hired employees is growing at a much faster rate, of between 2% and 6%, rather than the 1% rate seen across employees as a whole in the official data. On this view, a rapid pick up in average pay levels is in prospect, once more and more workers start to shift jobs and as pay of existing employees is adjusted in annual pay rounds. The parable of the rock pulled by the elastic band is a very relevant one in economic policy making: the rock may not move for a long time, but eventually it will. And the longer it takes to move the faster the move is likely to be. On this view, the UK labour market may be reacting more slowly, but it is fundamentally not very different from previous recoveries. Inflationary pressure is building in the pipeline. And given the lags in the impact of monetary policy, that pressure will be more difficult to curtail if the Bank does not act now.
However another possible explanation is that a combination of factors has caused labour supply - the amount of hours of labour available to the economy - to be much stronger than in previous recoveries.
Participation in the labour force was expected to continue to fall, in line with its pre-crisis trend, mainly as a result of an ageing population. This does not seem to be happening. Participation in the labour force for the over 60s has increased by some 1.5 percentage points since the crisis. An increase in longevity is causing people to work longer in a number of countries so as to maintain the same standard of living over their retirement.
That shift seems to happening in the UK as well. It has been reinforced by a structural change in policy - the increase in women's state pension age from 60 to 65. Indeed, it is women in the 60-65 age bracket that have increased their participation rate most - by just under 6% since 2010 when the increase in the pension age started to be phased in.
Alongside this, changes to the incapacity benefits regime may well also have had the effect of bringing more workers into the market. The number of people under the age of 65 who are inactive because of incapacity is down by around 200,000 since mid-2008.
For much of the post-war period, there has been in the UK a steady decline in the hours people have wanted to work. As income grew, the trend was to reduce hours and maintain income rather than to maintain hours and increase income. Not surprisingly, the largest fall in real pay since records began has reversed that trend; the evidence suggests that, as the economy has recovered, with lower real pay, many have sought to increase hours to maintain - or recover - income. That of course may well be a temporary effect and as and when real pay rises the pre-crisis trend may reassert itself. But it may be that this effect will last for some time yet.
Another explanation for the weakness of pay despite the fall in unemployment might be that the pressure to catch up lost income is much weaker post recession than it has been in the past. Based on previous episodes in the UK, we might have expected a reduction in real incomes to result in increased inflationary pressure from the labour market.
The big lesson from the 1970s was that weakness in real incomes and price movements, which drive a wedge between consumer prices and output prices, lead to pressure for increases in pay to "catch up" on the lost income: or as it is sometimes called "real wage resistance". The end result is more inflationary pressure for a given amount of unemployment. Put differently, to keep a lid on inflation, unemployment has to be higher following periods of pressure on real incomes.
The current reduction in real incomes has been as big as we've seen for nearly a century. Real wages are still almost 9% below their pre-crisis peak. That is a fall of around 20% relative to the pre-crisis trend. Following a year of strong growth and a very rapid fall in unemployment, this would suggest workers would now demand some catch up in pay.
However, there appears to be little evidence of real wage resistance - the sharpness of the recession and the years of austerity that have followed it, appear to have caused a shift in the psychology of UK workers.
There appears, for the present at least, to be an acceptance that pre-crisis pay levels are no longer achievable. Real wages have proved to be remarkably flexible and, as a consequence, unemployment has been able to drop quickly without inflationary pressure on pay emerging.
To complete the picture, the fall in unemployment has included a high number of long-term unemployed, who probably drag less on pay. Taken together with workers' apparent willingness to accept lower real pay levels, the non inflationary rate of unemployment, or NAIRU, looks to be at a lower level than we might have expected.
The MPC already built into its August forecasts the possibility of some positive labour supply shock. It took the judgement then that, contrary to previous assumptions and the experience of previous recessions, the labour participation rate would be less affected by an ageing population. And, having seen little evidence of pay pressure, it took the view that the NAIRU had fallen to 5.5% - not far off the 5% rate that prevailed in the UK before the crisis.
But even relative to last August's judgement on labour supply, pay growth still looks weak relative to models calibrated on the pre-crisis period. At the same time, unemployment is still falling more rapidly than expected.
The MPC is left with a judgement whether those falls in unemployment presage an imminent sharp pick up in pay growth. Or whether the continued weakness in labour costs means we can afford to let falls in unemployment continue to run for some time yet.
As always with economics, it is possible to make a good case on either side. We will know more when we see the extent of pay pressures in the big annual settlement rounds in the first quarter of next year. But there are grounds to continue to suspect that labour supply may be behaving differently to the past, to be cautious about moving to tighten monetary policy on the basis of falling unemployment alone and to wait for clearer signs of strength in pay growth. Even after adjusting for one off effects, pay growth is currently still some 2 percentage points below the pace consistent with achieving the inflation target in the medium term. That weakness in pay might go some way in explaining why inflation has been undershooting the MPC's predictions of late.
The instinct to be cautious about tightening monetary policy too early is further reinforced by the constraints that come with Bank Rate remaining close to its effective lower bound. We've now seen evidence of a slowing in UK growth, and prospects for the global economy have deteriorated. With the scope for tightening monetary policy substantial, but the scope for loosening it much more limited, the risk of a surprising pick-up in inflationary pressure may be more manageable than the risk of the expansion stalling and inflation dropping further. The latter risk, of course, has and will continue to become less acute the longer the expansion continues. But it remains, for me at least, a concern.
Uncertainty remains, old puzzles are unresolved and new ones appear. That is the normal world of the forward looking monetary policy maker. The question as always is whether the risk of not waiting until the picture is clearer is outweighed by the risk of not acting in time.
To sum up, some of what the MPC, at the point I joined it, thought would happen has happened. We have gone from recovery to expansion. Growth is slowing to something close to its pre-crisis average. Investment has made and is making an increasing contribution to growth. But the behaviour of unemployment, productivity and, most recently, pay continue to surprise and puzzle. More "normal" relationships between these variables may reassert themselves quickly and the pay rounds next year in particular should illuminate some of the picture. But there is also reason to believe that some changes, as in labour supply, may be longer lasting.
When I started at the Bank this time last year, the MPC was half way through the process of putting together its latest set of forecasts for inflation and growth. That process is now again in train. The softening in the pay and inflation data, together with the weaker external environment, for me implies that we can afford to maintain the current degree of monetary stimulus for a longer period than previously thought. That of course will be a subject for the MPC to debate in the coming weeks.