BMI View: Assessing the amount of slack in the UK economy remains a contentious issue and one which has created an air of uncertainty over the shape and speed of the recovery. Far from being just an abstract concern, excess capacity has a direct bearing on growth, inflation and the design of fiscal and monetary policies. While survey-based measures and our own model estimates suggest that there is only limited excess capacity in the economy (at most), this will not necessarily translate into near-term policy tightening. Instead, much will still depend on price stability and the vulnerability of mortgagors to a rise in lending rates.
Since the financial crisis uncertainty over the amount of spare capacity in the UK economy has proved to be an elusive 'known unknown', which has confounded policymakers and clouded analysis of the macro environment. Equally bewildering is the range of perspectives on spare capacity, ranging from those that believe no slack exists to those that see a gap so large as to justify open-ended quantitative easing in a bid to readdress the balance. Much depends on whether changes in operating capacity during the recession are viewed as structural or cyclical, and whether future growth potential has been irreparably damaged by the financial crisis.
While some of the economic concepts concerning capacity are abstract, the implications are not. The amount of slack in the economy has a direct bearing on the pace of economic growth and the trajectory of inflation, and is a central element in the design of fiscal and monetary policies. Moreover, with the Bank of England's poorly executed foray into forward guidance and unemployment targets disintegrating into a more vague objective of keeping rates low until economic slack is eradicated, the issue of spare capacity is firmly in focus.
Assessments of slack can be derived from model estimates based on measures such as potential output and equilibrium unemployment, as well as survey estimates of capital and labour usage. Below we attempt to measure slack in the UK economy using both methods, and consider the impact on inflation, lending rates and the short-sterling futures market, which is our preferred medium for expressing our view of monetary policy re-pricing.
Output Gap Measures
The broadest measure of spare capacity, and the one with the strongest economic interpretation, is the output gap, which measures the difference between potential and actual GDP. Potential GDP is the sustainable level of output that can be produced in the long term using the optimal combination of factor inputs. It is consistent with a full-employment equilibrium and stable inflation. This can be considered the natural trajectory, which the economy periodically fluctuates around.
| Potential GDP Profile Has Weakened |
|UK - Real & Potential GDP, Constant Prices, GBPbn|
Broadly speaking estimates of the output gap can be statistical, fundamental or a combination of the two. Statistical measures exploit the mechanics of the GDP time series by using algorithms to isolate the cyclical from trend components. Aside from the relative simplicity of estimation, statistical detrending methods have a neat economic interpretation by identifying an underlying trend which is responsive to long-term rather than short-term dynamics. The chart above shows our estimate of the trend series using the Hodrick Prescott filter technique, and suggests that the output gap has closed.
| Economy Operating At Potential? |
|UK - Output Gap, % of Potential GDP|
There are three main drawbacks to using a filtering technique such as this when applied to the business cycle. First, the estimate of trend GDP is not completely synonymous with potential GDP. One is a statistical concept, the other is an economic one based on assumptions about factor inputs such as capital and labour usage. Second, the nature of the estimation algorithm gives rise to end-point bias whereby the trend series is significantly influenced by the most recent data points. However, this can be partially (but imperfectly) mitigated by extending the series with forecasts for future periods to improve the credibility of today's output gap estimate. Third, the estimation procedure requires a subjective assessment of the appropriate smoothing parameter, for which there is little broad consensus.
Fundamental estimates have a stronger grounding in economic theory and therefore have an advantage over purely statistical techniques. However, what is gained in economic sophistication is lost in the difficulty of estimating fundamental models. Indeed, modelling potential output in this way can involve hard-to-measure variables such as the aggregate capital stock, or estimation of unobserved components such as the level and progress of technology. Our method involves specifying a production function for the UK economy and incorporates separate estimates of equilibrium unemployment (which we discuss later).
| Current Growth Unsustainable? |
|UK - Output Gap, % of Potential GDP|
Our estimated series in the chart above similarly suggests that the wide output gap in the immediate aftermath of the financial crisis has now closed. Although there are clearly difficulties in estimating potential output using either statistical or fundamental approaches, we nonetheless believe that our estimates are broadly in line with the additional evidence presented below.
Unemployment Gap Measures
A related concept to the output gap is the unemployment gap, which measures the difference between actual and equilibrium unemployment. It is a measure of equilibrium specific to the labour market rather than the whole economy, although there is a degree of comparability. The standard equilibrium measure is the non-accelerating inflation rate of unemployment (NAIRU), which is the lowest level of joblessness that can be sustained without triggering a breakout in inflation. To that end, it is not necessarily the change in unemployment which impacts inflation, but rather the change in unemployment relative to an unobserved full-employment equilibrium level, which changes over time.
We can again differentiate between the statistical and fundamental estimation techniques, and in the case of former reuse the HP filter to extract the underlying unemployment trend. Similar to our output gap measures, the estimated series suggests that the unemployment gap is now mildly negative (which theoretically would be inflationary).
| Closed Gap |
|UK - Equilibrium Unemployment Rate (HP Filter), %|
Our fundamental model estimate exploits a state-space framework, which allows us to estimate an unknown quantity (equilibrium unemployment) by using data that we can observe (such as unemployment, inflation and exogenous supply-side price shocks). The model is built around a 'measurement' component which describes the interaction of observed and unobserved elements, and a 'transition' component which considers only the relationships of unobserved elements. The chart below shows the result of this estimation, and suggests that actual unemployment is now close to equilibrium, or equivalently that the unemployment gap has nearly closed.
| Unemployment Close To Equilibrium |
|UK - Unemployment & Equilibrium Unemployment (NAIRU) Rate, %|
A major benefit of our method is that it produces a time-varying measure of equilibrium unemployment, rather than a static estimate which does not consider changes in the labour market over time. Indeed, a higher equilibrium rate may have been justified in the stagflationary environment of the 1970s, while a lower natural rate would have been likely during the Great Moderation of the 1990s and 2000s. Our framework also has appealing economic foundations in that we estimate an equilibrium unemployment rate consistent with stable inflation. However, like our fundamental model estimate of the output gap, which requires a subjective assessment of the appropriate smoothing parameter, our model for equilibrium unemployment requires an a priori assumption of the signal-to-noise ratio, which determines the volatility of the unobserved series (NAIRU).
Whereas estimates of the output and unemployment gaps carry a significant degree of uncertainty, survey based measures are more reliable and less controversial. That is not to say that such measures are error free as surveys still rely on the sometimes subjective assessment of capacity limits and current utilisation rates, which may be relatively straightforward for manufacturers to estimate but more complicated for the financial services industry, for example.
| Brushing Up Against Capacity Limit |
|UK - Capacity Utilisation Rate (%) & Pre-Crisis Long-Term Average|
Our first port of call is the manufacturing capacity utilisation rate series compiled by Eurostat. As the chart above shows, the capacity utilisation rate is slightly above the long-term pre-crisis average and aligns with our previous model estimates for excess capacity. Similarly, the proportion of firms operating below capacity has fallen sharply and is now at levels not seen since 1997.
| Pressure Slowly Building |
|UK - Proportion Of Firms In Industry Operating Below Capacity, %|
The final chart below shows the results of survey responses concerning the motivation of capital expenditure over the coming 12 months, where there has been a noticeable increase in the proportion of respondents citing capacity expansion as the reason for fixed investment. The previous three charts all suggest that more firms could be brushing up against the capacity limit as demand rises.
| Signs Of Incipient Recovery In Investment |
|UK - Reasons For Capital Expenditure In Industry Over The Next 12 Months, %|
However, this may, in part, reflect a lower production maximum. Indeed, during and after the financial crisis, the collapse in demand and financial distress would have forced many firms to close down operating facilities and sell off assets, resulting in a loss of capital and labour inputs that can subsequently create a capacity shortfall when demand recovers. As such, evidence of emerging capacity constraints may just mean that demand is overwhelming current lower production capacity, which may have been less of an issue with the larger production capacity that existed before the financial crisis. This would tie in with the assumption that the future trajectory of potential GDP is now lower.
Even if firms were able to initially maintain existing production facilities in the face of recession and a constriction in credit availability, persisting weak demand (below potential) may have still underpinned a natural reduction in the production maximum. Indeed, in much the same way that the labour market can exhibit hysteresis as a result of rising unemployment (i.e. skills deteriorate or even become obsolete, making it more difficult to secure employment), capital that is not being maintained will eventually deteriorate. At the aggregate level this can be seen in the fixed investment data. Gross fixed capital formation has been one of the biggest casualties of the financial crisis as firms have been on a de facto investment strike in light of demand uncertainty. However, the existing capital stock still deteriorates through depreciation. The chart below shows that while there has been a hiatus in new fixed investment, the consumption of fixed capital (depreciation) has continued unabated. This suggests that the aggregate capital stock is both ageing and shrinking.
| Investor Strike Underpinning Capital Stock Deterioration |
|UK - Gross Capital Formation & Depreciation, Constant Prices, GBPbn|
Further Evidence From The Labour Market
Additional labour market data further suggest that economic slack is being gradually used up. The headline unemployment rate has fallen sharply in recent months despite the steady increase in the labour force (largely the result of European migration), which is reflective of the rapid pace of job creation. Much like the initial increase in unemployment following the financial crisis, the recent improvement is mainly cyclical - in other words there is a natural drop in unemployment alongside the recovery in demand. As the chart below shows, long-term unemployment (over 12 months) remains high, which could suggest fundamental damage to the labour market, especially as hysteresis kicks in. Then again, it may be the case that the longer-term unemployment rate will eventually improve with a lag given that firms will soak up appropriately skilled workers in the first instance before considering lower skilled candidates in the pool of long-term unemployed. If the long-term unemployment rate does not eventually fall in line with broader unemployment, this could be a sign of persisting labour market slack.
| Signs Of Some Damage To The Labour Market |
|UK - Unemployment By Duration, mn|
In line with the improvement in headline employment data and evidence of emerging capacity constraints, the Bank of England's recruitment difficulties indicator has edged back into positive territory. This suggests that the labour hoarding of recent years may now be creating excess demand on the margin for skilled workers.
| Recruitment Difficulties Emerging |
|UK - Recruitment Difficulties Index|
While the improvement in unemployment and emerging recruitment difficulties may be taken as evidence of a reduction in economic slack, this overlooks the existence of underemployment - the idea that not all workers are being fully utilised. There are several signs of this, from the proliferation in part-time and temporary work, to the fact that employment has increased much faster than aggregate output (suggesting a fall in productivity). The Office for National Statistics publishes micro level data from the Labour Force Survey showing the proportion of part-time workers wanting to work full-time, which has increased sharply since 2008. Similarly, the aggregate work force report a desire to increase average working hours.
| Underemployment Provides Evidence Of Labour Market Slack |
|UK - Underemployment, %|
Finally, we present evidence from the Beveridge curve, which shows the relationship between unemployment and vacancy rates in the UK over the last decade. Rightward shifts in the curve can indicate an increasingly inefficient labour market as a result of skills mismatching and labour mobility freezing up. Furthermore this could suggest that deterioration in the labour market has become structural rather than cyclical. In the case of the UK, there has been a clear worsening in labour market functioning during the post-financial crisis period, but this has since turned the corner. Indeed, further falls in the unemployment rate would suggest that the labour market could soon return to conditions that prevailed pre-crisis.
| Labour Market Recovery? |
|UK - Beveridge Curve (X axis = unemployment, Y axis = vacancy rate)|
On the back of our broad analysis of capacity usage across the economy, we judge that if there is any slack left in the economy it is limited and being used up quickly. Survey data is generally the most supportive of reduced slack, while the more mixed results from our model based estimates suggest a modest degree of excess capacity at best. We have long questioned the existence of an extraordinarily large output gap following the financial crisis, which would signal an almost unchanged trajectory for potential GDP. Indeed, some of the largest estimates that we have seen are clearly based on an extrapolation of the pre-crisis trend in real GDP. Not only was the pre-crisis trend inflated by a credit bubble, but the post-crisis trend has been lowered as a result of deterioration in the aggregate stock of capital and labour. It would be foolhardy to suggest that a rise in long-term unemployment, idling assets and a lack of investment in innovative technologies and modernising the capital stock, would not materially influence the operating potential of the economy.
It is not just misjudgements of potential output which are at fault, as we have also argued for a long time that the official GDP data has probably underestimated the health of the UK economy. While it is clear that the economy has struggled in recent years, we have pointed to the steady rise in employment and continued expansion in service sector output (the backbone of UK PLC) as being at odds with official data that previously suggested a flat lining economy. A spate of extraneous events, such as adverse weather and additional bank holidays as a result of Royal events, have further compounded the accuracy of GDP estimates. Finally, we have also highlighted the impact of the construction sector which, despite representing less than 7% of GDP, has had an outsized impact - a negative drag - on the headline macro data. Taken together, some deterioration in growth potential and arguably a stronger economy than official data suggests, would point in the direction of a more modest output gap. Moreover, with the UK economy suddenly bursting into life in 2013 and on track to be one of the fastest growing in the developed world this year, spare capacity is being quickly used up.
What, then, are the implications of economic slack? Much depends on the initial amount of excess capacity and the persistence of the output gap. In the absence of a systemic financial crisis of the sort experienced in 2008-2009, a sudden sharp drop in GDP would be quickly corrected as the economy returns to its equilibrium path, with limited implications for inflation and medium-term policymaking. However, when a negative output gap persists over the medium term - due, for example, to a breakdown in credit transmission channels following a banking crisis - the economy runs below potential, which leads to hysteresis in factor inputs, fuels deflationary pressures and has a direct bearing on fiscal and monetary policies.
In terms of fiscal policy, the output gap has implications for the size of the structural (or, cyclically adjusted) budget balance, which is the part of the fiscal accounts which is independent of the output cycle. If the economy is operating at the full employment equilibrium with a closed output gap, any fiscal deficit should be largely structural rather than cyclical, and so would require austerity measures to eradicate. In the case of the UK, we believe that equilibrium unemployment is higher than before the crisis, and that synonymously potential GDP is lower. If we are correct in assuming that the output gap has been largely eliminated, this would suggest that the economy is closer to operating at the full employment equilibrium, for which there are two key implications for fiscal policy. First, at the full employment equilibrium, any deficit in public finances is more likely to be structural rather than cyclical, which will require additional austerity measures in future budgets. Second, if the economy is operating close to capacity, there is less scope for 'catch-up growth' as the economy cannot sustain demand above potential for a sustained period. A situation in which growth slows relative to current levels, and additional austerity is required, will pose a major headache for future chancellors of the exchequer attempting to anchor the national debt to a sustainable trajectory.
On the monetary policy front, persistent excess capacity fuels deflationary pressure and poses a threat to central bank inflation targets, which in turn necessitates the need for monetary easing. As such, the central bank's judgement of the output gap will influence its assessment of medium-term inflationary pressures and determine the need for monetary intervention today. As it stands, our view that there is some limited economic slack left in the economy suggests that the unemployment rate can keep falling without exerting immediate upside pressure on wages. We would expect this to be the case throughout this year. By the end of 2014, we believe that excess capacity will have been used up, which would then set the stage for a build up in demand-side inflationary pressure.
| Inflationary Pressures Contained, For Now |
|UK - Probability Of Inflationary Pressures In Six Months Time, %|
To provide a useful benchmark to gauge the likelihood of inflationary versus deflationary pressures in the future, we have built a probit model which estimates the probability of different scenarios for consumer prices, and incorporates our estimates of the unemployment gap. The chart above compares the result of our six-month forward probability model and inflation and suggests that deflationary pressures have now subsided sharply alongside a build up in inflationary pressure. While we warn that the UK (like the rest of Europe) is still vulnerable to a deflationary shock, in the absence of such an exogenous shock we would expect demand-side price pressures to build.
| Short-Sterling Still Pricing In Aggressive Tightening Cycle |
|UK - Short Sterling Implied Rate (%), 95% Confidence Interval Bands|
In terms of the end result for monetary policy, we still favour a rate cut in 2016 by which point we expect headline inflation to test the top end of the Bank of England's target. As such, we are still relatively more dovish than the short-sterling market which has priced in a December 2014 hike and expects a far more aggressive tightening cycle, which we believe is unlikely given the fragility of the recovery and the risk posed to heavily leveraged mortgagors. This in turn plays into our view that there will be some re-pricing of monetary policy over the medium term which will favour bullish positions in the 2015-2017 section of the short-sterling curve.