A monetary policymaker faces uncertainty − speech by Catherine L Mann

Speech

Monetary policymakers are facing myriad uncertainties these days. As I see it, the key balance of uncertainties is between near-term inflationary pressures that are becoming embedded in domestic expectations and realisations on the one hand, and on the other what is projected to be a dramatic deterioration in the real purchasing power of people’s incomes in the near and medium term. The challenge is to deploy monetary policy tools to short-circuit the inflationary expectations dynamic now, so as to prevent inflation from remaining above target for longer, and by doing so to moderate the hit to purchasing power later on.

Today, I’d like to put some structure on how I think about these uncertainties and how they relate to my policy decisions in recent meetings. I pay close attention to expectations of firms, consumers, and financial markets. Uncertainties about price expectations and price realisations are particularly important, being influenced by current inflation and, in turn, affecting future price inflation. Relatedly, but its own source of uncertainty into the medium-term, is how price realisations affect the evolution of real income and demand, which in turn, will either ratify the price expectations, or not. Another uncertainty, particularly important for the medium term impact of policy, relates to the transmission mechanism from monetary policy (domestic and external) through financial intermediaries to influence inflation and output. A third uncertainty relates to the long-term equilibrium policy rate, which evolves with fundamentals, among which is productivity growth, but also has endogeneity to the monetary policy decision.

Before getting into it, I thought it would be useful to remind ourselves of the objective of monetary policy. By remit, the Monetary Policy Committee is tasked to maintain inflation as measured by the Consumer Price Index at 2% and that this target applies “at all times”. Therefore, the MPC does not “make up” for past misses of its target – bygones will always be bygones. However, the remit also recognises that shocks will push inflation away from 2%, and that appropriate monetary policy can temporarily tolerate such a deviation. This is because the monetary policy path needed to achieve 2% at any specific point in time could cause undue volatility in economic activity. Instead, monetary policy ought to be forward-looking and be concerned principally with ensuring that medium-term inflation expectations are well-anchored at 2%. In order to achieve this aim, our decisions at each meeting are based on a wide range of data, models, and judgments on how shocks, financial intermediation, and microeconomic decisions will affect macroeconomic outcomes of inflation, employment, and output in the near, medium, and long term.

Different uncertainties at different horizons

In the normal course of events, the monetary policy decision-maker faces various uncertainties at different time horizons, but often they can depend on some aspects of the transmission mechanism to be relatively stable and similar to historical experience. However, these days we are continually learning about how economic and financial relationships may be changing on account of the size and set of shocks associated with Brexit, Covid, and the Russia-Ukraine war. Not only is there uncertainty about the nature of the shocks hitting the economy in real time, we are also uncertain about how those shocks affect economic variables after the fact. The appropriate monetary policy decision and its path necessarily will evolve as economic relationships and outcomes become clearer over time.

A first uncertainty particularly important to me is the degree to which shocks of the recent past have generated a transitory inflation surge versus ushering in a period of persistently higher inflation. Importantly, the degree of persistence depends on the reaction of monetary (as well as fiscal) policy. The same shock can play out differently conditional on the chosen policies. However, firms and workers in the real economy, and also actors in financial markets, observe our policy choices and economic outcomes to form their expectations. Therefore, they, too, play a key role in determining the degree of persistence. The way these expectations are formed may have been changing in the face of the sequential shocks that the UK economy has been experiencing in recent years. But in any case, to the extent that monetary policy decisions in the short term can influence expectations over future price-setting, it can influence medium-term inflation and purchasing power, and therefore its own appropriate future stance.

A second uncertainty is the relationship between monetary policy, financial intermediaries, and aggregate demand – consumption by workers, investment by firms, and international trade flows. Monetary policy, both by domestic and external central banks, works through financial intermediaries to change financial conditions which in turn steers aggregate demand. Typically, the transmission of monetary policy to financial conditions takes place at different time horizons and at different speeds than does the transmission of financial conditions to consumption, investment, and trade flows. The combined channels of transmission determine how monetary policy affects macroeconomic outcomes in the medium term.

A key question is the stability of these relationships between monetary policy, the real economy, and the inflation process. With regard to the first linkage: Has the transmission of monetary policy signals to financial conditions changed, with liquidity from past monetary accommodation perhaps affecting transmission now? To the extent that previous accommodation absorbs the transmission of monetary policy to overall financial conditions, then the signals provided by monetary policy actions to the real economy could be muted. If so, then bolder monetary policy action might be needed to affect financial conditions, to then cascade to affect real decisions.

With regard to the second linkage: Has the relationship between tightness in labour and product markets and downstream wage and price pressures changed on account of the shocks that have hit the economy? Since the adjustment of the real economy to monetary policy signals is often estimated to last some 12-24 months, it is difficult to know the effects of policy in real time.^{footnote [1]} We can use econometric methods to calculate an average effect of past policy in order to infer the likely effect of current or future decisions. But then we need to judge whether those estimated relationships stay stable over time. The potential for a change in these relationships and/or the lagged response to policy needs to be incorporated into monetary policy decision-making on a rolling basis at each meeting – along with any new shocks.

Finally, a third uncertainty is the equilibrium real interest rate at which the economy settles to its long-run growth rate consistent with the 2% inflation target. Monetary policy should consider demand, supply, and asset markets when thinking about the so-called terminal Bank Rate. The conventional wisdom has been that monetary policy is broadly neutral for the long-run resting point of the economy. My colleague Huw Pill explained in his latest speech that policymakers may choose to abstract from these effects as they make current policy decisions.^{footnote [2]} However, to the extent that monetary policy affects the inputs to the long-run equilibrium, the terminal rate will to some degree be endogenous to the economy’s path to it.

Collectively, these uncertainties are important in my monetary policy decision, but the relative importance of the uncertainties to my decision may change from meeting to meeting. The first uncertainty about the role of expectations in determining the degree of inflation persistence was particularly relevant for me at the February meeting. For the March meeting, the second uncertainty was particularly important as I evaluated the balance between a ratchet effect pushing up on inflation in the near-term against the purchasing power loss moderating inflation in the medium-term. Because of the inter-relationships between the terminal rate and the policy path, thinking about the terminal rate is important as the longer term starts to come into focus. Clearly, these days there is uncertainty at all horizons, and among all relationships, and of course, shocks continue to affect both financial intermediation and the real economy.

Near-term horizon: Uncertainties and inflation expectations dominate December to February

Lift-off from the lower bound is now behind us, but I would like to review why I thought that the next policy move, at the February meeting, should have been 50 basis points. Key for me was that both the DMP and the agents’ survey reported continued strengthening of expectations by workers and firms for wage increases and price increases, which could generate a domestically-driven wage-price dynamic with higher inflation embedded in forward-looking pricing decisions. In my speech at the beginning of the year, I illustrated that if the 2021 wage settlements and price hikes were repeated in 2022, inflation would remain well above target for longer than initially predicted.^{footnote [3]} For me at the February meeting, research argued that monetary policy needed to lean against these expectations with a front-loaded policy path.

For wages, important evidence for the role of expectations for wage settlements came from the agents’ survey and wage equations which included inflation expectations. The decomposition of underlying pay growth in our February Monetary Policy Report implies that, over the course of the pandemic, the increase in inflation expectations was about as important as the effect of economic slack over the same period.^{footnote [4]} Therefore, even given the February projections, which showed a negative output gap from year 2 onwards, inflation expectations today likely will continue to play a significant role in wage settlements – keeping them stronger than consistent with the inflation target for longer.

But I was mainly concerned about price expectations. The DMP survey and agents’ commentary reported that firms’ pricing expectations for 2022 – at that time running north of 4.5% – sought to recover margins eroded from input price increases and higher energy prices in 2021. In addition, while import prices strongly influenced firms’ pricing strategies, so too did domestic costs. I thought that even when global goods-price inflation started to decelerate, the erosion of international competition facing UK domestic firms could strengthen their ability to pass-on, or even top-up import prices. Data show a Brexit-related sluggish trade rebound and reduced variety of imported products^{footnote [5]} and there remains a significant gap between domestic goods prices and international goods prices (Chart 1a).

Chart 1: Manufacturing output PMIs ^(a) and UK CPI doppelgänger ^(b)

Diffusion index (LHS) and index 2015 = 100 (RHS)

Footnotes

• (a) Source: Refinitiv Eikon. Latest observation: March 2022.
• (b) Source: ONS and Bank calculations. Latest observation: 2021 Q4.

A doppelgänger pricing model^{footnote [6]} suggests that UK consumer-facing prices are higher than those of comparable economies (Chart 1b), which cannot be fully explained by the post-referendum Sterling depreciation. Indeed, the gap between actual and counterfactual consumer prices widened again through 2021 – a time when Sterling, on a trade-weighted basis, strengthened vis-à-vis trading partner currencies. Naturally, much of this divergence may be due to the heterogeneous impact of the energy price shocks which hit Europe much harder than elsewhere.

Other research points to a possible inflation ratchet – a situation where price increases beget price increases in a set of seemingly irreversible steps. In my previous speech, I illustrated how a single or even two successive shocks to the price level have a “transitory” effect on inflation, since these shocks mechanically wash out after either the one year or the two years have passed. Considering UK inflation history, one could make the case that the inflation of 2011 was such a one-off shock which died out quickly. In that case, the MPC was right in holding interest rates low even when headline inflation overshot the target.

The situation we face now is different. The economy – firms, workers, consumers – is confronted by repeated shocks on the same side – Brexit, Covid, supply bottlenecks, energy prices – that on balance increase the prices of goods and services in the UK. Over the past year, these shocks have caused the price level to ratchet up and rather than letting each shock die out with inflation returning to target, they have piled on top of each other increasing inflation to historic highs.

This has not gone unnoticed. We know that consumer expectations for aggregate inflation are built from observed inflation of salient purchases, such as energy and food.^{footnote [7]} Since household inflation expectations are part of wage-setting, goods inflation can become embedded into overall cost inflation, albeit with a lag. Monetary policy that fully “looks through” these price changes may inadvertently allow for a ratchet-up in costs and prices, sustaining aggregate inflation even after those prices come down or their growth rates mean-revert.

Moreover, evidence continues to mount that an inflation ratchet may make it more likely that externally-driven inflation in just a few items may spread into more firms’ pricing strategies, yielding higher overall inflation that persists without monetary policy action. The analysis of a volatility-based measure of underlying inflation that my colleague, Lennart Brandt, and I have conducted suggests that high inflation is no longer limited to components that are typically quite volatile, but now has seeped into those that typically are rather stable.^{footnote [8]} This may herald a regime change to where underlying inflation of the low volatility categories on average runs above the inflation target for a longer period of time – as it had before the Financial Crisis, when it was offset by lower inflation for other categories.

It could be that long-run inflation remains low and all we are seeing are jumps in the price level. Since we do not target the price level, even if prices stay high forever but inflation returns to target, this is completely consistent with our remit. However, since expectations matter and are at least to some extent backward-looking we must be mindful of the length of time away from target, the breadth of inflation, and recognise that inflation salience informs inflation expectations. Additionally, specific mechanisms of the UK energy market, namely the Ofgem price cap, are designed in a way that prolongs the effect of the current increase in wholesale energy prices. All of these points led me to believe a hike of 50 basis points was warranted in February.

Chart 2: Monetary and Financial Conditions Index ^(a)

Cumulative change in index points

Footnotes

• (a) Source: Bank calculations. Notes: See Bank Overground (2021) “How do we monitor UK financial conditions?” for more information on the construction of this index. Latest observation: 02 February 2022

An additional concern I had in February was that except for the yield curve, financial markets were not pricing in credit risks commensurate with either the aggregate demand or the inflation projection at that time. In other words, the transmission of the rise in the Bank Rate in December and forward guidance that “some modest tightening of monetary policy over the forecast period” would be necessary was muted – in particular into
high-LTV mortgages and high-yield corporate bonds. Chart 2 shows the cumulative change in the Banks’ Monetary and Financial Conditions Index through 2021 up until the February decision. We can see how the appreciation of Sterling and the higher yield curve had tightened financial conditions but were more than offset by looser spreads and to some extent by buoyant equity prices. Whether because of previous accommodation or a delay in incorporating the trajectory of tightening monetary policy, the resulting continued accommodative financial conditions could have tempered any effect of policy tightening on aggregate demand, which would therefore have required a more robust signal from policy – a 50 basis point hike.

Research on the dynamics of inflation under uncertainty supported a front-loaded path. The research considers two kinds of uncertainty – uncertainty about whether inflation is persistent or transitory, and uncertainty about the underlying model of inflation. Each of these cases can be compared to a full information world – the all-knowing central banker!

In the first type of uncertainty, if the shock is truly persistent, but is misperceived as transitory, the policymaker initially under-reacts relative to the full information case. Even as they learn about and respond gradually to the true nature of the shock, the inflation overshoot is both larger and more persistent than under full information. Vice versa, if the shock is mistakenly assumed to be persistent but turns out to be just transitory, the policymaker over-reacts leading to a more negative output gap. But, the literature suggests that if there is uncertainty about the degree of inflation persistence, it is better to assume a high degree.^{footnote [9]} The costs of making a mistake if the true inflation process is more persistent are larger than if the true inflation process is less persistent.

Finally, suppose we are working with full information, but with a distribution of inflation outcomes. If there is a risk of more upside persistence to inflation – the mass in the right tail is more than the left tail – even if the upside risk is not realised, a tighter monetary stance is warranted relative to the case where there is no such asymmetric tail risk. As we look at the distribution of firms’ price expectations today, presented below, we will see such an asymmetric distribution.

Front loading the policy reaction did not necessarily imply a higher Bank Rate in the
long-run – my focus was on the slope of the policy path. A bold tightening followed by a hold, or even a modest rate reversal appeared to me to be appropriate to address inflation expectations. A rate reversal at some future point would not necessarily be a mistake, nor a harbinger of recession, as it is sometimes viewed, particularly in the US. The optimal path for the policymaker evolves depending on the responsiveness of inflation expectations and the transmission of policy through financial markets – the key factors in February – versus the medium-term adjustment of the real economy. Under these types of uncertainties, staying below the long-run neutral rate for the entirety of the cycle could be considered mistaken. Indeed, according to the research cited above, the front-loaded policy path would yield smaller deviations from the inflation target with no additional loss in terms of output volatility.

In sum, I argued that a front-loaded tightening path was needed to send a strong signal of commitment to the inflation target so as to forestall the embedding of inflation expectations in wages and prices. A front-loaded tightening also would have emphasised the importance of financial market outcomes to the transmission of monetary policy.

Medium-term horizon: Uncertainties on the real side cloud March decision

In March, all the factors supporting an inflation ratchet present in February were still in place, if anything accentuated. But I voted for only a 25 basis point increase rather than another dose of front-loading. What changed? First, of course, Bank Rate already was 25 basis points higher from the February meeting. More importantly, though, I turned my focus to the new shocks and the second uncertainty – the relationship between financial markets, the real economy, and the inflation process – as the dominant concern.

Around the March meeting, financial markets remained quiescent and were not a new source of uncertainty. But, energy prices, already in train and looking forward, were expected to have a dramatic impact on aggregate demand. I focussed on evidence that the ratchet effect to push-up inflation would be moderated as the purchasing power loss over non-energy goods and services pushed-back on price increases for those products. The key ingredient to this balance was the potential magnitude of the purchasing power loss, over what time horizon consumers would either actually feel or start to prepare for the pinch, and how firms would react in setting both current prices and expectations. Because there was not a full forecast for the March meeting, considering historical experience with very large energy price shocks was important as an input to judging prospects for both inflation and demand. So too was looking at real-time data and forward-looking indicators on consumer behaviour.

The first issue with demand was that the negative shocks to real income were coming at a time when the level of real average disposable income had stagnated for the 2 years of COVID, rather than rising at some 0.8% per year as in the decade before the pandemic. Given the robust employment and wage growth coming out of Covid, why should the level of aggregate income matter? With soft initial conditions in the levels, despite strong growth, there may be little margin to absorb the new negative shocks. In fact, aggregate employment remained below where it was before the pandemic. Important as well was the concentration of the projected price shock in low-elasticity products – energy and food. Given these lower elasticities of demand, the available real spending on other goods and services was likely to be squeezed even more than aggregate income or GDP.

A back-of-the-envelope calculation of the impact of a doubling of consumer-facing energy prices – about what the October price cap might be relative to before-Covid – generated a reduction in real aggregate non-energy consumption of roughly 2%. This would come on top of any reduction in demand caused by the 2% drop in real post-tax labour income that was projected for 2022 as of the February forecast.

Intuitively, consider that households must decide how to allocate their income to the different goods and services they would like to or need to consume. Before Covid, UK households were – on average – spending about 2.5% of income on their electricity and gas bills. If these items get more expensive but households cannot meaningfully reduce the real amount that they consume of them, their nominal spending on these bills will have to rise. This spending will then need to be financed either by dis-saving, borrowing, or by reducing the consumption of everything else, implying a shift in aggregate consumption shares. Of course, not only have prices risen for gas and electricity, but the price of
non-energy consumption has gone up as well, albeit not by as much. Overall, the hit to aggregate demand is large, but for non-energy goods and services there is reason to believe it will be relatively larger.

The average 2.5% energy share masks considerable heterogeneity across the income distribution. At the lower end, households spend over 10% of their income on gas and electricity. Since they cannot easily substitute away from this spending (i.e. their price elasticity is low) and they cannot typically smooth consumption by much (i.e. their marginal propensity to consume is high), they likely will reduce their consumption of non-energy goods and services nearly one-for-one. These households make up a non-trivial share of total consumption: Before Covid, the bottom three deciles, i.e. those containing that third of households with the lowest income, had a combined consumption share of 15%.

Even at the top of the income distribution there would be a significant loss of real income. Although the loss to demand for core goods and services would be less dramatic given that those households are more able and more likely to smooth consumption. These households also have accumulated the most amount of liquid savings throughout the pandemic^{footnote [10]} and will therefore be able to draw on savings in order to sustain their consumption.

Another key uncertainty for me was whether the projected drag on domestic demand for non-energy goods and services would be sufficiently large as to temper price realisations for firms producing and selling these products. What evidence can we marshal on these two uncertainties? First, on consumer demand: Is there evidence related to either the actual or projected pinch in real income? Second, on pricing strategies: Do we know how firms’ pricing strategies might respond to a moderation in demand?

On demand conditions: As of the March meeting, notwithstanding then-current robust nominal wage increases and employment growth, retail sales were at best holding up (or even falling in real volumes) and consumer confidence had weakened, particularly the forward-looking balances. Notably, as can be seen in Chart 3b, consumers’ assessment of their own financial situation deteriorated by more than at the worst point of the Covid crisis. Their assessment of the change in their own situation is consistent with the dramatic drop in projected real average weekly earnings as of the February MPR (Chart 3a). So, on early assessment, it looked like consumers were forward-looking, which would translate into a period of softer demand growth, perhaps even retrenchment.

Chart 3: Real average weekly earnings ^(a) and GfK consumer sentiment indices ^(b)

GBP at 2015 CPI (LHS) and diffusion index (RHS)

Footnotes

• (a) Source: ONS, Bank of England, and Bank calculations. Latest observation: February 2022 (actual), 2024 (forecast). (b) Source: Refinitiv Datastream. Latest observation: March 2022.

But what about firms, and their sales and price realisations and expectations? Analysis of the DMP survey shows an asymmetric relationship between sales and price growth: Stronger sales growth is associated with higher price growth, but prices are less flexible downward when demand falls. This builds in a ratchet. For the March meeting, given that the energy price hikes were just being incorporated into the Ofgem pricing formula and Russia’s invasion of Ukraine had just started, more insight might come from looking at firms’ expectations for sales and prices. Arguably, the big hit to consumer purchasing power and demand was still ahead.

The monthly DMP survey asks firms’ executives about sales growth at their firms over the past year and for sales and price forecasts over the next year, in the context of scenarios and probabilities. In the last survey, firms reported a jump in their uncertainty over both sales and, particularly, price expectations (Chart 4a). Looking deeper into the subjective sales-forecast distribution, the uncertainty for sales (and also for prices) is driven particularly by strong positive subjective forecast distributions (Chart 4b). In contrast to the spike in uncertainty in 2020 where the bottom fell out of the subjective sales-forecast distribution (associated with Covid lockdowns), the current spike in uncertainty is strongly on the upside. Firms think sales have upside risks throughout 2022. If a negative demand shock hits over this horizon, firms would not only be surprised, but they presumably would reassess sales prospects, which would also be the signal to reassess prices.

Chart 4: DMP uncertainty measures ^(a) and expected sales growth ^(b)

Index 2019 = 100 (LHS) and % year-on-year (RHS)

Footnotes

• (a) Source: Anayi et al. (2022). Notes: Solid lines show 3-months moving averages, dotted lines show single-month data in 2022. Latest observation: March 2022.
• (b) Source: Bunn et al. (2021). Latest observation: 2022Q1.

The DMP survey also gives guidance on the drivers of price expectations as well as on how price uncertainty is related to forecast revisions and future price realisations. A large share of the increase in firms’ price expectations over the last year can be attributed to past increases in both their own prices and aggregate inflation.^{footnote [11]} Also evident from analysis of the survey data, when firms do adjust their prices, they adjust them by less than one-for-one with their forecast errors.^{footnote [12]} These two observations from the DMP survey suggest strong upside bias for prices.

At the March meeting, I evaluated the uncertainty in the purchasing power hit against these uncertain, but very strong sales and price forecasts. With a full forecast coming in May, and energy prices gyrating, raising the Bank Rate by 25 basis points was appropriate, to bide time to see how some of the uncertainties would resolve.

In thinking about this balance of uncertainties for the May meeting, several additional issues come into focus. First is the time profile of the consumption response to the hit to real incomes. Will it be forward-looking, which would reduce demand in 2022, or will it take time for consumers to react, softening consumption relatively more in 2023? More information will come in before the May meeting on credit and borrowing, possible bonuses and top-up especially for lower wage workers, precautionary savings, and changes in consumption habits, for example, in product quality. Any muting of the demand signals would likely keep firms’ pricing decisions robust.

The latest DMP survey on the distribution of firms’ price expectations is revealing, particularly in recent historical context (Chart 5). During COVID, firms’ expected prices for 2020 and 2021 were rising somewhat as was aggregate price uncertainty. But as 2021 unfolded, the mass of the expected pricing-distribution for 12 months ahead shifted rightward and the distribution widened notably; these distributions were reflected in increased average expected prices and price uncertainty, as mentioned earlier.

Chart 5: Distribution of expected price changes 12 months ahead ^(a)

Probability density

Footnotes

• (a) Source: Bank of England Decision Maker Panel and Bank calculations.

The survey data for the first quarter of 2022 shows a continued shift right, but also an increasingly fat right tail. On the one hand, this points to upward risks to our goods price forecast; that is, if firms in the tail are able to push through these price increases. On the other hand, if the hit to household purchasing power is as large as it is projected to be, there are likely also to be large forecast errors. If their past reaction is any guide, firms will revise downward their price expectations and, going forward, their actual price realisations. This change in price realisations in the face of the aggregate demand response to the energy price shock is critical to stem the ratchet effect, and to moderate aggregate CPI inflation. In the case of a fat right tail on the distribution of prices, the research discussed earlier indicated that a more robust monetary policy path is appropriate.

I am watching these surveys and other data closely to see whether and when firms receive the demand signal that would alter their pricing expectations. In March I was willing to wait to see more information on the size and timing of the demand slow-down, even as the ratchet-effect on inflation was clear to see. If the consumption hit is moderated by other policies or by savings and other smoothing behaviours, it may be well into 2023 before firms receive the demand signal. The consequent embedding of current overall inflation into firms’ own-pricing is a concern because it may point to a situation in 2023 where pricing remains robust even as demand remains weak.

Longer-term horizon: Uncertainties and prospects for the long-term equilibrium interest rate

Even though settling at the Bank Rate appropriate for long-term equilibrium in the economy may be somewhat distant, it is not too far in the background of the policy-path decision and, most importantly, is relevant for longer-term decision-making by private sector investors, firms, and consumers. The long-term real rate is underpinned by the rate of growth of potential output, key elements of which are aggregate productivity, capital stock and investment, and the demographics and institutions in labour markets.

Traditional economic theory says that monetary policy has no effect on potential output and that structural policies (e.g. productivity outcomes) dominate.^{footnote [13]} However, to the extent that monetary policy affects real business investment then there is a connection between monetary policy and potential output, with business investment raising the capital stock. To the extent that monetary policy can increase labour force participation by running the economy hot, it can affect potential output through the labour component. However, these days, Covid, Brexit, and the Russia-Ukraine war appear to dominate the role for monetary policy. That said, continued sluggish investment and workers reporting they cannot join the labour force due to illness, if sustained, would lower the long-term equilibrium rate.

Two factors associated with productivity growth may be particularly relevant at present when thinking about the long-term equilibrium rate. First, the more sluggish recovery in UK trade relative to G7 peers, which is associated with Brexit, portends downside risk to productivity growth, and therefore a lower long-term real rate. Erosion in competition, fewer varieties, and less new technology imported into the domestic UK market, and the loss of economies of scale and market knowledge associated with export expansion translate into lower firm-level and aggregate productivity growth.^{footnote [14]} On the other hand, my colleague Jonathan Haskel has suggested that Covid-related new investments associated with work-from-home as well as more efficient use of home resources and less unproductive time commuting could represent a positive productivity shock.^{footnote [15]} It is too early to tell which of these productivity dynamics will dominate, although the external one has been in place for longer, and also affects business investment. So I am keeping an eye on both in terms of looking forward to the appropriate level of Bank Rate in the long term.

Conclusion

For the May meeting, key topics for me are an assessment and judgment on how much and when the expected consumption drag materialises, and whether we start to see any indication of price forecast revisions in the DMP survey. If they do, this potentially would short-circuit the expectations-formation process underpinning the domestic inflation ratchet, which has been my central concern. Tracking these price expectations and forecast revisions is of paramount importance since inflation ultimately is due to firms systematically able to raise their prices.

Unfortunately, more modest price outcomes for some products will not bring headline inflation near to target for 2022 or even for most of 2023 because the energy price shock is so large and the mechanism of the Ofgem price cap could drag it out. But, without some moderation in some prices, the underlying inflation ratchet associated with lagged CPI in firms’ pricing expectations will imply more persistence in keeping inflationary pressures above target.

Should the impact on aggregate demand of the energy price shock end-up being more modest than currently foreseen, should wage and price expectations and outcomes remain as strong as they currently are, and should financial markets return to being copacetic on private credit and duration risk, a reassessment of the pace of tightening would be warranted. Monetary policy needs to keep inflation expectations anchored; by doing so now, less tightening will be required later, when demand may still be weak.

I would like to thank Lennart Brandt, Lena Anayi, Andrew Bailey, Phil Bunn, Jenny Chan, Federico Di Pace, Bob Hills, Giacomo Mangiante, Riccardo Masolo, Huw Pill, May Rostom, Michael Saunders, Fergal Shortall, Matt Swannell, Katie Taylor, and Chris Yeates for their comments and help with data and analysis.