The UK economic outlook and monetary policy - speech by Michael Saunders

Speech

In this speech, I want to discuss recent developments and the outlook for the UK economy and monetary policy. I intend to make four main points.

• Recent strength in energy prices is likely to lift CPI inflation further in coming months. This effect is likely to be temporary. But there are also clear signs of pressures – in terms of capacity use, inflation expectations, firms’ pricing strategies and pay growth – that threaten to keep CPI inflation above the 2% target even once energy effects fade, unless restrained by monetary policy.
• Guided by our remit, the MPC has the tools to return inflation to the 2% target on a sustainable basis. The MPC has already raised Bank Rate to 0.5%. And, as the MPC has noted, if the economy develops roughly in line with the central forecast in the February Monetary Policy Report (MPR), some further modest tightening in monetary policy is likely to be appropriate in the coming months to return inflation to target.
• I suspect that risks are on the side of stronger and more persistent inflation pressures than implied by the February MPR forecast. As a result, at the February meeting, I favoured a 50bp rate hike, in order to move more rapidly to a more neutral monetary policy stance.
• My preference for a 50bp hike at the February meeting does not necessarily imply that I will vote for 50bp steps in the event that rates have to rise further. Nor does it necessarily imply a higher peak in rates than the yield curve used for the February MPR forecasts. All else equal, prompt tightening now could, in my view, help limit the total scale of tightening that will be needed to return inflation to target.

***

Let’s start with the background for the recent increases in Bank Rate in December and February.

Before the pandemic, in late 2019, the economy was in reasonable balance. Unemployment was just below 4% (the lowest for decades), underlying pay growth was 3-3½% YoY, and CPI inflation was close to our 2% target. Economic activity was restrained by Brexit uncertainties, given that the UK’s trading relations with the EU had not been agreed. Against this backdrop, our policy rate, Bank Rate, was at 0.75%, having risen a little in 2017 and 2018. Estimates of the neutral rate are inherently uncertain, but that 0.75% level was still what I would regard as an accommodative policy stance, with Bank Rate below the trend in the neutral rate.^{footnote [1]}

In early 2020, when the pandemic hit and the first lockdown was introduced, activity fell sharply. CPI inflation fell well below the 2% target and indeed was close to zero in Q4 2020. Credit spreads widened markedly for both mortgages and business loans. The MPC loosened monetary policy significantly in order to support the economy and prevent a persistent inflation undershoot, and to limit the extent of lasting damage to the economy through high long-term unemployment and business failures. The Bank’s FPC also eased macro-prudential policy, while the government put in place substantial fiscal support.^{footnote [2]}

That policy support could not, of course, prevent GDP from falling sharply in 2020. But it did help to limit the second-round damage. Despite a large drop in GDP, unemployment rose less than expected during 2020 and has since fallen back, the initial spike in credit spreads was reversed and the flow of credit to households and businesses has been maintained. Aggregate household and corporate balance sheets have improved, with substantial increases in holdings of liquid assets.

With the vaccination program expanding, restrictions were eased during last year. As a result, and also reflecting the fiscal and monetary policy support that was in place, the economy recovered. The level of hours worked had regained most of the lost ground in Q4 last year and GDP in that quarter was similar to Q4-19.

It appears that the Omicron wave has caused modest weakness in activity in December and January. Much of the effect on activity came through reduced labour supply, and as a result the BoE Agents’ measure of capacity use in firms was unchanged in January. The adverse effects on activity of the Omicron wave seem to have been much less than the prior waves. And these adverse effects are likely to be fairly temporary, given that the various restrictions introduced against Omicron have been largely removed since mid-January.

The recent recovery in output to roughly the same level as Q4-19 does not mean the economy is in the same shape as it was then. Since Q4-19, the economy has been hit by a series of adverse shocks – Brexit, Covid and an energy price surge – which have worsened the relationship between the level of activity and inflation. The energy price surge is directly lifting inflation and reducing real wages. Brexit and Covid have reduced both demand and supply. Initially, Covid’s adverse effects on demand exceeded the impact on supply, and hence capacity use in the economy fell. Over time, policy support has helped offset the adverse effects on demand, but has not (and probably cannot) eliminate the adverse effects on potential output through lower investment and lower workforce growth. It is also possible that Brexit may have steepened the wage and price Phillips curves in the UK, such that wages and prices respond more to any given level of excess demand (or excess supply).^{footnote [3]} This is because Brexit has reduced the economy’s openness (in trade and labour mobility) and thereby lowered the extent to which capacity pressures can be eased by imports and inward migration.^{footnote [4]}

As a result of these developments, the recovery in output has generated widespread capacity pressures, and gone alongside a marked rise in inflation. These effects are now being reflected in higher inflation expectations and domestic inflation pressures, including pay growth. I will discuss these in turn.

Capacity pressures. In the period before the pandemic, the economy’s potential output grew by around 1½% per year. If potential output had continued to grow at that pace, then a return to the pre-pandemic level of GDP would still leave the economy with considerable spare capacity. However, since Q4-19, the pandemic and Brexit have reduced the economy’s potential output significantly relative to that rising trend. The workforce has fallen by over 1% since Q4-19 and is about 2¾% (0.9-1.0 million people) below the pre-pandemic trend, because of the outflow of foreign workers and lower participation among those in the UK – especially the over 50s.^{footnote [5]} This is the first significant drop in the UK workforce for around 30 years.^{footnote [6]} Potential output may have been further reduced by mismatch – lifting the medium-term equilibrium rate of unemployment (otherwise known as the NAIRU) – caused by shifts in the composition of demand and labour supply in terms of skills, sectors and geography.

With lower potential output, the return to the pre-pandemic levels of activity has pushed the economy into excess demand, with severe capacity pressures. Unemployment has fallen faster than the MPC had expected and is now similar to pre-pandemic levels, and job vacancies are at a record high. Under-employment also has fallen markedly.^{footnote [7]} The BoE Agents’ index for capacity use in firms, and their index for firms’ recruitment difficulties, are both the highest since those series began about 25 years ago.^{footnote [8]} Capacity pressures are widespread: not just in sectors where global demand has been strong (e.g. manufacturing), but also in sectors (including consumer services) for which activity has not fully recovered the lost ground.

Inflation has continued to rise faster than expected. The YoY rates of headline and core CPI inflation (5.5% and 4.4% respectively) are both the highest for several decades. The QoQ annualised gains in both measures are even higher, at about 8% for the headline CPI and about 6% for the core measure. Recent inflation outturns have generally exceeded Bank of England and external forecasts.

Energy prices continue to play a major role in lifting inflation. Non-energy consumer goods prices also are still rising rapidly, reflecting the ongoing strength of global spending on goods coupled with Covid-related disruptions to supply chains for manufacturing output around the world. But part of the inflation pickup, especially since around mid-2021, reflects higher services inflation, which probably stems from the build-up of domestic cost and capacity pressures.

The trend in inflation expectations is uncomfortable. Longer-term inflation expectations (measured from financial markets and household surveys) have risen to around the top end of the ranges of the last 15-20 years. The strength of firms’ expectations for their selling prices evident in the DMP and other surveys suggests that – across a wide range of sectors – firms believe they can pass on rapid cost increases to prices, and do not appear to be constrained by the 2% inflation target.^{footnote [9]}

Underlying pay growth has picked up notably, and is likely to rise further. Pay growth for new hires was very strong during last year. Moreover, the BoE Agents report that a substantial number of firms made an unusual mid-year “top-up” to pay of existing staff during last year, so that the average of pay deals in 2021 was around 3½%, significantly above firms’ expectations at the start of last year (which was for pay growth of around 2¼%). Bank staff estimate that underlying AWE growth was around 4-4½% YoY in late 2021, above the pre-pandemic pace and higher than one would expect given the usual inverse relation between unemployment and pay growth.^{footnote [10]} Underlying pay growth is in line with models that include inflation expectations as a driver of pay, which suggests that the pickup in inflation expectations may be contributing to the pickup in pay.

There are signs of a significant further step up in pay settlements for this year. The BoE Agents report that on average, firms expect pay growth of 4.8% this year (weighted by firm size), an increase well above recent years.^{footnote [11]} The pickup in pay growth is widespread across industry sectors. The key factors that firms cite in pushing up pay deals relate to the tight labour market, rising headline inflation and the upcoming further rise in the National Living Wage.

A norm for pay deals of 4½%-5%, given trend productivity growth of roughly 1% YoY, would imply unit labour cost growth that is well above the pace consistent with the inflation target. In theory, one might take the view that this pickup in pay would be self-correcting. If firms cannot pass increases in labour costs on to prices, then margins will fall and eventually firms will cut back on hiring such that unemployment rises and pay growth slows again. But there is little sign of this restraining effect at present. As noted, many firms believe they can pass on these cost increases to their selling prices. Indeed, the BoE Agents report that around 70% of firms have passed on (or expect to pass on in the next year) some or all of the increase in pay to their selling prices.

So, in my view, the big picture is that, while the surge in energy prices accounts for quite a lot of the inflation overshoot, it is also the case that the economy is in significant excess demand and inflation expectations are not as well anchored as I would like.

What can Monetary Policy do?

In setting monetary policy, our focus is usually mainly on the inflation outlook roughly two years ahead, because it typically takes more than a year for changes in monetary policy to have their main impact upon economic activity and thus on prices. As our remit recognises, there is little monetary policy can do to address factors that have only temporary effects on inflation – affecting the near-term outlook, but not the outlook for inflation at the policy-relevant horizon.

The energy price surge is an example of such a temporary shock. It will lift inflation (and lower real wage growth) for a period. But, unless energy prices go on rising or inflation expectations are destabilised, it is unlikely to generate a sustained inflation overshoot. It would not make sense (and would not be consistent with our remit) to tighten policy so much as to try to return inflation to the 2% target in the next few months, when the temporary effect of energy prices is at its peak. This does not mean the MPC has abandoned its commitment to low inflation. It is a recognition of the fact that 2% inflation could only be achieved in the next few months by tightening policy very sharply in order to weaken the economy enough to quickly push non-energy inflation well below a target-consistent pace. This would cause undesirable volatility in output and jobs. And, given that the full effect of those monetary policy changes would occur after the energy effect fades, it would leave inflation well below the 2% target further ahead.

So the MPC’s focus is on factors that affect the medium-term inflation outlook, especially domestic capacity and cost pressures, and inflation expectations. With signs that these are building, the MPC has raised Bank Rate (in two steps) to 0.5%, withdrawing part of the support implemented during 2020. But even after these moves, monetary policy remains relatively loose. Interest rates are still below neutral, and lower than directly before the pandemic, even though capacity strains and domestic inflation pressures are significantly stronger than then.

The Economic Outlook

Let me turn to the outlook.

The MPC’s latest forecast was published in early February.^{footnote [12]} In that forecast the rise in the Ofgem price cap, plus continued strength in core inflation, lift CPI inflation to around 7¼% YoY in April. With high inflation eroding real incomes, consumer spending is weaker than the November forecast. Household energy prices stop rising next year, and hence stop adding directly to CPI inflation. But the level of energy prices remains high two and three years out, and therefore continues to depress the level of household real incomes and spending.

This softer path for demand is reinforced by tighter monetary policy (reflecting the market rate curve prevailing at the time) and fiscal policy (in line with the government’s stated plans). At the same time, potential output recovers somewhat through higher labour supply as participation reverses some of the recent decline (although most of the decline in potential output versus its trend before Covid and Brexit persists). Hence, the jobless rate rises slightly in Q2 this year and continues to rise thereafter, exceeding 5% three years ahead, with a widening margin of excess supply two and three years ahead.

In the MPR forecast, underlying pay growth picks up to around 4¾% in Q4 this year, consistent with the BoE Agents’ survey and, as in 2021, higher than one would normally expect given the jobless rate. But, this is offset by the unwind of composition effects in the mix of employment, such that AWE growth this year is roughly stable at 3¾%, similar to the average for the 2019 calendar year. Unit wage cost growth this year (3%) is projected to be slightly below the 2018-19 pace (3¼% YoY).^{footnote [13]} AWE growth slows to around 3% in Q4 next year and 2¼% in Q4 2024. This slowdown in pay growth is partly because of the gradual rise in unemployment. But, in the MPR forecast, a key factor is a decline in inflation expectations, such that the relation between unemployment and wage growth returns to something like that seen in 2010-19. CPI inflation is still a little above the 2% target two years ahead but falls below the 2% target three years ahead.

The MPR also includes a projection based on Bank Rate remaining at 0.5%: that forecast shows CPI inflation above the 2% target over the forecast period.

Since the MPR forecast was finalised, the Ofgem price cap has been announced and was close to the assumption in the MPR forecast. The net effect of the government support measures announced at the same time, which were not in the MPR forecast, will be to lift household incomes this year, partly offset by a lower path in subsequent years. All else equal, this fiscal support would slightly lift economic activity over the three year forecast period relative to the MPR.^{footnote [14]}

Risks around the MPR Outlook

There are obvious uncertainties around Covid. The MPR forecasts assume no further new Covid-related severe restrictions or lockdowns in the UK, and that the direct effects of Covid on activity in the UK will fade from this month (ie March). But there are still risks that new variants could affect the outlook, and (as with Omicron) this might affect both demand and supply.

There are also uncertainties around energy prices. The MPR forecast, in line with the MPC’s usual practice, assumes that wholesale energy prices follow the path implied by forward markets at the time for the next six months and are stable thereafter. This leads to a further rise (of about 10%) in the Ofgem price cap in October this year, with prices roughly stable thereafter. As a result, gas prices stay well above the 2021 level even three years ahead, creating a lasting squeeze on the level of households’ real incomes. The MPR includes an alternative scenario, which assumes that wholesale energy prices follow the path implied by forward markets (at the time of the MPR) over the whole forecast period. In this scenario, the Ofgem price cap falls by about a third during 2023 and 2024, returning close to the end-2021 level. Hence, the squeeze on household real incomes is more short-lived. Relative to the MPR forecast, CPI inflation is lower two and three years ahead (and real activity is higher). That would not be an unreasonable assumption in my view, and hence (at the time of the MPR) there was a downside risk to the MPR inflation forecast (and upside risk to real activity) two and three years ahead from this factor.

Apart from energy prices and Covid, my view is that (conditioned on the yield curve used in the February MPR) risks to inflation are on the high side of the MPR projection two and three years ahead, reflecting more persistent domestic capacity and cost pressures. In my view, the key risk is that, unless restrained by monetary policy, the interplay between excess demand, relatively high wage growth and relatively high inflation expectations will become more entrenched, leading to stronger growth in average earnings, unit wage costs and inflation.

In the near-term, I suspect that the MPR forecast may understate the extent to which the pickup in pay reported by the Agents will feed through to higher unit wage cost growth this year. The MPR forecast that composition effects will depress AWE growth rests on further job growth being tilted towards low-paying jobs. However, composition effects on AWE growth often do not unwind quickly and, given the changes in the workforce since Q4-19, may not unwind fully.^{footnote [15]} And, in so far as the previous boost to AWE reflected the decline in employment in low-paying roles in the worst of the pandemic, my hunch is that any unwind of this effect has largely occurred given that the labour market is now tight. In any case, if further composition effects do depress AWE growth, these same effects would probably also lower productivity growth below its trend. Either way, it seems likely, in my view, that unit wage cost growth this year will exceed the MPR forecast and rise above the 2018-19 pace.

More broadly, for the next 2-3 years, I put more weight on a scenario in which unemployment remains below the MPR forecast, and the 2010-19 wage Phillips curve does not fully reassert itself, such that pay growth runs above the MPR forecast and firms continue to pass on those costs to prices.

In my view, risks are on the side of a further tightening in the labour market during this year, in contrast to the MPR view that unemployment is likely to start rising next quarter.

Vacancies are at record levels and surveys suggest firms’ hiring intentions remain strong. The high level of labour demand does not seem to rely only on expectations of strong demand growth in the future. Rather, it appears that firms need more staff to sustain the current level of activity, as evident in the high level of capacity use and the high share of firms who report output is already constrained by labour shortages. This pent-up demand for labour implies that labour demand may not decline rapidly enough to lift unemployment even as high energy prices erode real incomes and limit spending.

Moreover, risks are on the side of slower growth in labour supply than the MPR forecast, because of a smaller (or later) recovery in participation. Since Q4-19, the number of people (aged 16-64) who are outside the workforce and do not want a job has risen by 470,000 (1.4% of the workforce), and much of this (about 300,000 people, 0.9% of the workforce) reflects people with long-term sickness.^{footnote [16]} These are substantial increases, especially given that the total rise in the numbers of people aged 16-64 who are outside the workforce has been 287,000 since Q4-19. Consistent with this, the drop in participation (among the 16-64 age group) has been concentrated in people with disabilities, reflecting both a rise in the numbers of people with disabilities and a lower participation rate among people with disabilities.^{footnote [17]} There has been little aggregate change in participation among people (aged 16-64) who are not classified as having disabilities. It is possible that much of this reflects effects from the pandemic, for example Long Covid and the backlog in non-Covid hospital waiting times.^{footnote [18]}

There are signs that vaccinations help protect against the incidence of Long Covid^{footnote [19]} and that some people recover from Long Covid over time. In this case, there might be some recovery in participation over time, as effects of Long Covid gradually decline. Nevertheless, there is little sign of this at present. Indeed, among those aged 16-64 who are not in the workforce, the number that would like a job is at a record low.

There may also be an upside risk to CPI inflation from a further pickup in the rents component. The shift in consumers’ preferences for more housing space, perhaps as a side effect of increased working from home, has driven a sharp rise in house prices and rents on new rental agreements over the last year.^{footnote [20]} House prices do not directly enter the CPI, but rents do. The CPI rents index measures the average change in rents on all outstanding rental agreements (private and public), and hence adjusts gradually to changes in rents on new agreements.^{footnote [21]} It is likely to pick up further in the coming months, and add materially to inflation in the year ahead. A scenario in which pay growth overshoots the MPR forecast also may support continued strong gains in rents. This effect may not be fully captured in the MPR inflation forecast which, beyond the next couple of quarters, is constructed from a top-down perspective and does not include specific housing indicators.

An additional risk is that inflation expectations may drift higher, if the high upcoming inflation rates coincide with a continued accommodative monetary policy stance. Household inflation expectations have recently risen by more than would be expected given their usual relationship with CPI inflation. It is clear from business surveys that many firms believe they can pass on rapid cost growth to prices at present. I worry that the very high inflation figures that are likely this year will have a lasting imprint on inflation expectations among households and businesses in coming years, unless accompanied by clear actions that demonstrate the MPC’s willingness and ability to return inflation to target.

The combination of these risks – higher unit wage cost growth this year, continued tight labour market, elevated or rising longer-term inflation expectations, continued relatively high pay growth – would, if they materialise and are not accompanied by tighter monetary policy, probably threaten to cause a more persistent inflation overshoot than the MPR forecast.

Of course, high pay growth accompanied by high productivity growth is good news, both at a macro and micro level, and does not pose a threat to the inflation target. And, in a market economy, there are always shifts in relative prices and wages, such that some individual sectors or skills may see relatively high or low growth in pay and prices even if overall CPI inflation is in line with the 2% target. I have no problem with that. But a situation in which across the economy as a whole, people expect high inflation to continue, aggregate nominal wage gains far outstrip productivity, and firms believe they can pass on high cost growth to their selling prices, would concern me, because it would probably not be consistent with returning inflation to the 2% target.

Implications for Monetary Policy

As I said earlier, monetary policy cannot prevent the surge in energy prices from reducing real wages and profits (for non-energy producing sectors). Nor can monetary policy alone lift productivity growth and hence generate a faster trend in real wage growth once the effects from energy prices fade. In line with our remit, the role of monetary policy is to ensure that inflation expectations are well anchored, and that domestic cost and capacity pressures are consistent with a return of inflation to the 2% target over time as the effects of the energy price surge fade.

Given the economy’s recent trends and the outlook, I supported the decision to tighten monetary policy at the February MPC meeting. And, in line with other MPC members, I expect that we will need some further monetary tightening in the coming months to return inflation to target. Reflecting my view of the balance of risks around the MPR forecast, my preference is to move quite quickly towards a more neutral stance in order to prevent the recent trend of higher inflation expectations and rising pay growth from becoming more firmly embedded.

When we loosened policy in 2020, my view was that, with the economic risks we faced then, it was better to err on the side of providing too much stimulus rather than too little. I supported that easing then, and still believe it was the right decision. That easing was effective in loosening financial conditions and supporting the economy because of the credibility of the UK’s inflation-targeting framework. For me, that willingness to ease promptly during the worst of the pandemic has always come with a willingness to tighten again if, as now, the economy is in excess demand with above-target inflation. Risk management considerations resulting from the proximity of Bank Rate to the Effective Lower Bound (ELB) – which could argue for a relatively slow and delayed tightening – have carried less weight for me in recent months.^{footnote [22]} In part this is because, with some scope to adopt a negative Bank Rate since August last year, the ELB is further away than previously. In addition, I put increasing weight on other risk considerations, namely that maintaining a relatively loose policy stance under current conditions would be likely to produce a further undesirable rise in inflation expectations. Such an outcome would be costly to reverse and could limit the scope for prompt monetary easing the next time the economy needs support.

I am not in favour of aiming to restore the lost potential output by “running the economy hot”. The deterioration in potential output due to Covid and Brexit probably cannot be reversed by cyclical pressures. For example, among the people who have left the workforce because of long-term sickness, decisions on whether to re-enter the workforce are likely to depend more on health considerations than on the availability of jobs. Rather than lift potential output, I suspect that a policy of “running the economy hot” would simply produce an even more persistent inflation overshoot and further lift long-term inflation expectations.

At the MPC’s meeting last December, my view was that the economic outlook and risks around the outlook justified some monetary policy tightening, but that uncertainties around Omicron argued for a relatively small move. Hence, I supported the 15bp hike agreed by the Committee. Had the Omicron wave not occurred, there would have been a case for a larger hike at that meeting in my view.

At the February meeting, with Omicron uncertainties having declined, my preference was to raise Bank Rate by 50bp, rather than the 25bp hike supported by the majority.

But I want to stress that my vote for a 50bp hike in February does not necessarily imply that I would vote for a 50bp hike in the event that further tightening is required. My policy votes will depend on the economic outlook, and risks around the outlook, at the time. All else equal, the case for policy to move in a larger step probably is greater when Bank Rate is clearly further away from the approximate level that, if maintained, would return inflation to target and keep it there.

Nor does my vote for a 50bp hike in February necessarily imply that I believe that the level of rates one or two years ahead will be higher than the yield curve used for the February MPR. It is important to distinguish between the pace of tightening and the level of rates at the end of a tightening cycle. Indeed, to the extent that quicker tightening early on could help ensure that the nexus of inflation expectations, pay growth and firms’ pricing strategies return to being consistent with the 2% inflation target, this might help limit the overall tightening cycle (relative to what would otherwise be the case).

Whichever way the economy develops, the MPC will, as always, remain focussed on returning inflation to the 2% target in a way that supports output and jobs, in line with our remit.

The views expressed here are not necessarily those of the Bank of England or the Monetary Policy Committee. I would particularly like to thank Lee Robinson and Katie Taylor for their help in preparing this speech. I have received helpful comments from Andrew Bailey, Thomas Belsham, Alan Castle, Harvey Daniell, Michael Goldby, Jonathan Haskel, Catherine Mann, Andre Moreira, Fiona Shaikh, Silvana Tenreyro and Lukas von dem Berge, for which I am most grateful.

References

Broadbent, B (2015), ‘Compositional shifts in the labour market’, speech given at ‘Understanding the Great Recession: from micro to macro’ conference, Bank of England.

Carney, M (2017), '[De]Globalisation and Inflation', speech given at the 2017 IMF Michel Camdessus Central Banking Lecture, Washington.

Daher, M, Fazio, M and Harper, G (2021), ‘How much of the recent house price growth can be explained by the ‘race for space’?’, Bank Overground.

Evans, C, Fisher, J, Gourio, F and Krane, S (2015), ‘Risk management for monetary policy near the zero lower bound’, Brookings Papers on Economic Activity, Economic Studies Program, The Brookings Institution, vol. 46(1, Spring), pages 141-219.

IMF (2022), ‘United Kingdom: 2021 Article IV Consultation-Press Release; Staff Report; and Statement by the Executive Director for the United Kingdom’, IMF Staff Country Reports, IMF, Country report No. 2022/056.

Mann, C (2022), ‘On returning inflation back to target’, speech given at Official Monetary and Financial Institutions Forum (OMFIF).

Saunders, M (2020), ‘Covid-19 and monetary policy’, speech given at online webinar.

Saunders, M (2021), ‘The Inflation Outlook’, speech given at online webinar.

UKHSA (2022), ‘The effectiveness of vaccination against long COVID: A rapid evidence briefing’, published 15 February 2022.