Some reflections on Monetary Policy past, present and future − speech by Michael Saunders

Given at the Resolution Foundation
Published on 18 July 2022
Michael Saunders is near the end of his term on the committee responsible for setting interest rates. In this speech, he reflects on his experience. Then he looks at the effect of demographic changes on the jobs market and gives his view on interest rates.


This will be my final speech as an MPC member, because my second (and last) term as an external member of the Committee will end after the August meeting. I would like to thank the staff at the Bank and my colleagues on the MPC for their help and support over the last few years. My successor on the MPC will be Dr Swati Dhingra of the London School of Economics, to whom I offer my congratulations.

In this speech, I want to review some of the challenges faced by the MPC in recent years, and discuss the implications of dissenting votes among MPC members. Looking ahead, I will highlight the extent to which demographic trends may pose new challenges in coming years and outline my perspective on the near-term monetary policy outlook. I will make four main points.

  • In recent years, a series of adverse shocks – Brexit, Covid, the war in Ukraine and the current energy price surge – have significantly reduced potential growth and lifted inflation.
  • It is not surprising that MPC members sometimes disagree on the appropriate policy decision, especially when (as now) the economic outlook is challenging. Dissenting votes reflect differing views on the economic outlook and key risks, rather than disagreement over the aims of policy.
  • The UK economy is starting to face increasing challenges from demographic trends, which seem likely to cause persistently low workforce growth in coming years and (all else equal) limit potential growth to well below even the modest post-GFC pace.
  • Economic growth has slowed. But, with excess demand and low potential growth, some further monetary tightening remains likely in coming months in my view, to ensure that inflation returns to the 2% target on a sustained basis. It is especially important at present to lean against risks that recent trends in inflation expectations, underlying pay growth and firms’ pricing strategies become more firmly embedded.

Let me start by briefly looking back over my six years on the MPC. It began just after the Brexit referendum and the subsequent monetary policy easing, and has encompassed the implementation of Brexit, the pandemic and the energy price shock associated with Russia’s invasion of Ukraine. I want to highlight three important features of that period.

First, as in the period during and after the 2008/09 recession, the economy has been hit by a series of adverse shocks (Brexit, pandemic, energy price shock). Brexit and the pandemic hit demand, but they also have reduced potential output. This has occurred through reduced inward migration (hence lower population growth and labour supply), as well as adverse effects on productivity (through lower investment and, from Brexit, reduced trade openness). The pandemic has also caused a marked drop in workforce participation, mainly reflecting increased long-term sickness and early retirement, exacerbating the drop in labour supply.footnote [1] As a result, potential output growth has undershot even the meagre post-GFC trend. At the same time, the currency depreciation triggered by the Brexit vote lifted inflation in 2017-18, and the energy price shock is playing a big role in the ongoing rise in inflation. Hence, while inflation has been close to the 2% target on average in recent years, it has (as during 2010-16) been volatile around that target.footnote [2]

Figure 1. UK – Potential GDP (indexed to 1997 Q1 – 100)


  • Note: The forecast path is consistent with the May 2022 MPR. Sources: ONS and Bank of England.

Second, as in 2008/09 and just after the Brexit vote, the MPC has acted to provide prompt and effective support to the economy when most needed. In particular, the monetary policy easing in 2020 played an important role in helping to underpin confidence and to reduce the risk of a vicious circle of heightened risk aversion and tightening financial conditions.footnote [3] I supported that easing at the time and still believe it was the right decision. A failure to act then could have greatly exacerbated the pandemic’s long-term scarring effects in terms of unemployment and business failures. The MPC’s ability to provide effective support at times of heightened uncertainty and strains relies on the credibility of the UK’s monetary policy framework. The need to maintain – and strengthen – that credibility is an important consideration in the current setting of monetary policy.

Third, in order to be well placed to fulfil its remit against a backdrop of a relatively low neutral rate and sizeable economic shocks, the MPC has continued to develop its unconventional policy toolkit. Asset purchases were introduced in 2009, and have become a well-established policy tool, used again (promptly, and on a large scale) in 2020. As well as expanding QE, the process of reversing it has begun. Early this year, the MPC began passive unwind (ie not reinvesting maturing bonds), and initiated a programme of corporate bond sales. Work is underway at present on whether to sell gilts held by the APF and (if so) the design of such a program. In addition, the BoE and financial system established scope to set a negative Bank Rate, although this has not yet been used. The Term Funding Scheme (which developed from the Funding for Lending Scheme launched in 2012) evolved into the TFSME, with extra incentives for banks to lend to SMEs. All this helps to ensure the MPC’s toolkit is fit for purpose, and thereby contributes to the credibility of the MPC’s ability to return inflation to the 2% target.

Another feature of recent years, and especially the last year, is that there have quite often been dissenting votes on the MPC. Without replaying each decision, I want to offer some general thoughts on the implications of dissenting votes.

The way the MPC operates is that monetary policy decisions are made by majority vote, MPC members are individually accountable for their votes, and the votes of each member are publicly disclosed in the minutes that accompany each policy decision. An inevitable result of this framework is that there will sometimes be split votes, and that dissents will be publicly announced. Many central banks have broadly similar approaches, but some do not – some have a single decision-maker, decisions reached by consensus, or through private votes that are not publicly disclosed.

There is, of course, no level of dissent that is inherently too high or too low. Since the MPC began in mid-97, roughly 12½% (one in eight) of votes by MPC members have dissented from the majority (on either asset purchases or Bank Rate), with some variation over time. The MPC’s rate of dissenting votes is above that of the US Fed, but fairly similar to other central banks which disclose votes of individual policy board members.footnote [4]

Figure 2. Selected Countries – Per Cent of Central Bank Policy Votes That Dissent from Majority


  • Note: The chart shows the percentage of policy board members that cast a dissenting vote on the policy rate and/or asset purchases. Data start in January 2006 for Magyar Nemzeti Bank, March 2009 for Central Bank of Iceland and from January 1998 for the others. Sources: Bank of England, US Federal Reserve, Bank of Japan, Sveriges Riksbank, Czech National Bank, Magyar Nemzeti Bank and Central Bank of Iceland.

It is worth stressing that dissenting votes do not imply any disagreement among MPC members on the aims of monetary policy. The MPC’s remit is set by the government, and all MPC members sign up to it. Nor, I think, is there any substantial inherent difference among MPC members on the appropriate trade-offs between output and inflation when the economy is hit by shocks.footnote [5]

Rather, in my experience, policy differences among MPC members have tended to stem primarily from different assessments of the economy’s current position and the outlook, and of the importance of various risks around the outlook. For example, over the last year, I have put more weight on risks that, as Covid restrictions eased and activity recovered, the UK would face persistent domestic cost and capacity pressures and that maintaining our previous degree of stimulus would create a damaging rise in inflation expectations. But it is perhaps not surprising that other MPC members might disagree about such issues, especially when the economic outlook is challenging and uncertain (as recently and now). These decisions are often easier with the certainty of hindsight.

I acknowledge that there may be times when the public explanation of dissenting views can give the impression of a cacophony of disagreement.footnote [6]

Even so, I regard the MPC’s current system (individual votes that are publicly disclosed) as greatly superior to the main alternatives (single decision-maker, a Committee with votes in private but not publicly disclosed, or a consensus-based approach).footnote [7]

A key advantage of having a Monetary Policy Committee rather than a single decision- maker is that the combined insight and knowledge of a group of experts will usually exceed that of any individual. And genuine individual accountability of MPC members is probably only possible if votes are publicly disclosed. The current system also helps to ensure that uncertainties and disagreements over the economic outlook and risks (and hence the appropriate policy stance) are brought to the surface and discussed. Arguments get tested in debate among Committee members and external speeches. This is likely to improve the quality of policy decisions and economic forecasts. Moreover, credibility would be eroded if central bankers disagree in private on the appropriate policy stance but feel they should hide such disagreements to give a false public appearance of unanimity.

Some external commentary has suggested that dissenting votes are a useful lead guide to the MPC’s future policy decisions.footnote [8] However, dissenting votes are certainly not used as a deliberate policy signal of the overall Committee’s intentions, and it is wrong to assume that dissents reliably predict future policy decisions. Sometimes they do and sometimes they don’t.footnote [9] One reason for this is that economic developments may affect inflation prospects from one meeting to another, and MPC members will (as you would expect) react to this. In addition, the MPC’s votes are always over whether to adjust policy at the current meeting, and require a distinct choice (whether to change policy and, if so, by how much). Such votes may not fully reflect the extent of agreement (or disagreement) among MPC members on the economic outlook and future policy direction. Moreover, the votes may not fully reflect the extent to which each MPC member’s decision is finely balanced (or not). Sometimes, a policy decision can be quite closely balanced even if the votes appears overwhelmingly in favour of one outcome.

It may be possible for the MPC minutes to convey more sense of the areas of consensus and disagreement among Committee members, and the nuances around the votes of individual MPC members. But, unless the MPC detail the views of every individual member at each policy meeting (and I would not favour that), such subtleties may not always be apparent to outside commentators.

So what should external commentators look at to judge the outlook for monetary policy? I would suggest that you read the Monetary Policy Summary. For most people, that may be enough. For those who want to dig deeper, start with the MPC’s remit, and then read the MPC minutes, the Monetary Policy Report and speeches of MPC members. Perhaps above all else, focus on the economic data, the outlook and risks around the outlook – because that is what the MPC themselves will be doing.

A turning point in demographic trends

Turning to the outlook, the economy in coming years will continue to be affected by the major shocks of recent years (Brexit, Covid and energy prices). The UK economy will also face increasing challenges from demographic change, because population ageing appears likely to produce persistently low workforce growth.footnote [10] As a result, potential output growth is likely to be weak in coming years, with the path of potential output falling further below an extrapolation of the trend to end-2018 (and also further below its pre-GFC trend).

I want to spend some time discussing this issue, given the extent to which it is now beginning to shape the economic outlook.

The UK population has been ageing for some time, with the share of the adult (ie 16+) population who are aged 50 years or over up from 39% in 1993 to 48% this year. However, until quite recently, the effects of this on labour supply were quite modest. This is partly because the growth of the 16+ population has been quite strong (averaging 0.6% YoY over the last 40 years).footnote [11] Moreover, from 1980 until around 2007, population ageing had little effect either way on workforce participation among the 16+ population. In that period, population ageing reduced the share of the 16-19 age population, who typically have relatively low workforce participation (because many are in full-time education). There was little change in the 65+ share of the adult population, who have relatively low participation.footnote [12] The share in prime working years (ie 20-64) – who have relatively high participation – actually rose between 1980 and 2007.

Figure 3. UK – Per Cent of Adult Population Aged 20-64 Years, and Effects of Population Ageing on Participation


  • Sources: ONS and Bank of England.

Over the last 15 years or so, as the share of the population aged 65+ has risen, the adverse effects of ageing on participation have increased, pushing down on participation among the 16+ population by about 2½ pp since 2008 (ie 0.2 pp per year on average). However, over that period, these effects were offset by two other factors that have lifted participation.

  • Education attainment has risen markedly over the last 20-30 years, as the expansion of secondary and tertiary education in recent decades has rippled through the adult population. Workforce participation is markedly higher among people with higher education attainment (more so among women than men). This is probably because, on average, people with higher education attainment have relatively high pay and a wider range of job opportunities. To give a sense of the effects, over the period 2001-2021, the share of the 25-59 year age population with degree level education (who have high participation) rose from 17% to 41%, while the share with education below secondary level (who have relatively low participation) fell from 52% to 32%. As a result, the overall participation rate for this age group rose from 81.6% to 85.8%, despite little change in participation for people of a given education attainment (see figure 4). In all, rising education attainment has lifted aggregate participation among the 16+ population by just over 2pp since 2008 (just below 0.2pp per year).
  • Since 2009, the female State Pension Age (SPA) has risen in stages from 60 to 65 years, and both the male and female state pension ages rose from 65 to 66 years since 2018. The delayed availability of retirement incomes appears to have significantly lifted participation in the 60-70 year age groups, and together have added about 0.6pp to aggregate 16+ participation since 2009.footnote [13]

Figure 4. UK – Workforce Participation Rate and Education Attainment of People Aged 25-59 Years


  • Note: In 2022 Q1, the proportions of people in each category were as follows: Degree 42%, Higher education 8%, Secondary education 19%, GCSE 17%, Other 7% and Unknown 7%. Sources: ONS and Bank of England.

Figure 5 shows a decomposition of these effects in terms of their effects on workforce participation in the 16+ population since 2008.footnote [14] Over that period, the large downward effects on participation from population ageing were roughly balanced by upward effects from increasing education attainment and the rising SPA. Other than these effects, there was little like-for-like change in participation over the period 2008-19, with a small decline during 2008-11 that reversed by 2019 (and may have reflected cyclical factors).footnote [15] Since 2019, there has been a like-for-like drop in participation of nearly 1pp, which (as discussed) appears to largely reflect higher rates of long-term sickness and retirement, especially among people aged 50-64 years.

Figure 5. Decomposition of Changes in the Workforce Participation Rate (for the 16+ population) since 2008, and Simulation to 2032


  • Sources: ONS and Bank of England.

Given these participation trends, even with population ageing, UK workforce growth averaged 0.8% YoY over 2009-19, similar to the prior 10 years (0.9% YoY). The resilience of workforce growth limited the decline in UK potential output growth over that period, partly offsetting weakness in productivity growth.

Looking ahead, the latest ONS population projections suggest that the downwards effects on 16+ workforce participation from population ageing will expand to about 0.25pp per year in coming years. This is because the share of the prime working age population (20-64 years) will fall more quickly, with a faster rise in the share of the 65+ population (and especially the 80+ population).footnote [16]

Figure 6. UK – YoY Change in Population by Age Group


  • Note: Population data and projections are from the ONS. Projections published in January this year. Sources: ONS and Bank of England.

At the same time, the upwards effect on participation from rising higher education attainment will slow. The prior rise in secondary and tertiary education has worked its way through most of the working age population. And, barring a big new expansion of further education, it is unlikely that education attainment will rise as much in coming years as over the last 20-30 years. In addition, most of the planned increase in the SPA (and its effects on participation) has happened.footnote [17]

Combining these effects, and assuming that half the drop in like-for-like participation during the pandemic unwinds, a simple simulation implies that participation among the 16+ population will fall by about 1½ pp over the next 10 years (and fall further beyond then).footnote [18] Using the ONS’s population projections, this implies that the trend in workforce growth will average about 0.3% per year over the next 10 years, less than half the average pace of 2009-19 (0.8% YoY).

Figure 7. UK – Actual Workforce Path and Simulations, Indexed to 2008 = 100