Getting the right directions − speech by Alan Taylor

Introduction

What are central bank mandates for? And how should they evolve?

Mandates are, at their most abstract, but also at their best, devices that align a specialised public institution with the aims that society cares about. When it comes to the key responsibilities of central banks, the public’s aims are straightforward to state but hard to deliver: price stability in support of economic prosperity.

In terms of macroeconomic measurement, those objectives often translate into low inflation and unemployment, robust and sustainable growth, and financial stability and an absence of deep recessions.

A sceptic might call this a tall order. Macroeconomic history, however, shows that while we have fallen short at certain times, we have also had periods of considerable success, reminding us that the right institutional scaffolding can make a meaningful difference.

A more specific and useful way to think about mandates is as a structured set of choices about goals, targets, and instruments. Jan Tinbergen spoke to this^{footnote [1]}; so before him did the fundamental theorem of linear algebra. They taught us that we need at least as many instruments as targets. Given that, the art of policy design is to find a configuration in which the instruments chosen are well matched to the desired ends.

Yet the devil remains in the details. What counts as “low” unemployment? Can GDP growth be “too high” if it comes with instability? Is there an “optimal” inflation rate and, if so, how tight should our tolerance be? And how do we calibrate “too much” versus “too little” risk in a financial system that is both the conduit of growth and a catalyst for crises? Those questions determine how credible and workable a mandate will be when tested by real shocks.

For the Bank of England, the post‑1997 regime offers some constructive clarity. After the government legislation enshrined the Bank’s operational independence, price stability was set in law and conveyed annually through the government’s remit. A statutory committee, the MPC, was created to be accountable for its implementation.

Since then, our primary goal has been set as price stability, currently defined as 2% CPI inflation. That clarity has profound advantages: it unifies the committee around a shared end point, disciplines our analysis, and provides a clear yardstick for accountability.

But central banks do not live by inflation alone.

Monetary stability is of enormous importance, but, after the global financial crisis of 2008, societies rediscovered – through harsh experience – the great significance of financial stability.

Around the world, that recognition led to an expansion in the nature and form of explicit mandates and in the tools for macroprudential policy and oversight. It was an evolution, and these lessons re-learned were welcome developments, complementing monetary policy rather than substituting for it, but requiring both sets of policies, and policymakers, to work together in a harmonious way to achieve multiple objectives.

The UK was no exception, and its institutional framework evolved accordingly.

Indeed, the government refined the MPC’s 1998 remit in 2013 to clarify the inherent real‑world trade‑offs between the speed of returning inflation to target and the volatility that this adjustment can impose on output, incorporating explicitly forward-looking language and including additional allowances under “exceptional circumstances”.^{footnote [2]} It also created in 2013 a new mandate and a new policy committee, the Financial Policy Committee (FPC), to guard against systemic risks and reduce the likelihood and severity of crises. The current architecture thus reflects a simple proposition: when objectives multiply, the instrument set and the governance model must also adapt.^{footnote [3]}

But given my habitat on the MPC, let me turn back and focus a little bit more just on monetary policy to frame what follows.

Let me note that designing a monetary policy mandate forces several first‑order choices, which we have seen throughout central bank history for centuries. Do you announce a numerical target publicly, or do you guide expectations more implicitly? If you target, do you target ultimate macroeconomic outcomes – like inflation or unemployment – or intermediate variables such as the exchange rate or broad money and credit aggregates? Do you target levels or growth rates? Do you specify a single target at all times, a corridor, or a set of targets that are flexible according to circumstances? Each decision has implications for communication, credibility, and the practical alignment of instruments with ends. These, in turn, shape how policy behaves through the cycle and under stress.

In this talk I want to offer two perspectives.

First, some broad historical perspectives on how central bank objectives, and the Bank of England’s in particular, evolved from the past to the present. Second, a narrower lens on how these choices have influenced my own policy decisions over the past eighteen months.

A look back at history will give us a view of the larger map; the recent period is the bumpy terrain that provides a road test. I will conclude with some thoughts for the future.

Why mandates came about: independence and accountability

At the simplest and broadest level, the mandates that govern modern central banks are the latest expression of what societies have long expected their monetary authorities to do: safeguard economic stability through a coherent framework that integrates monetary and financial policy.

As history shows, these mandates did not emerge fully formed. They are the legacy of centuries of experimentation: periods in which central banks shifted between fixed exchange rate commitments, money supply rules, lender‑of‑last‑resort practices, and, at times, direct financing of government. Across this long arc of history, and while political constraints gradually shifted, the objectives changed as experience taught policymakers what worked and what did not. Examples include decisions to adopt or abandon the gold standard, shifts in how balance sheets were used and how liquidity support was provided, and more broadly the various judgements to be made about the balance between stability and flexibility, rules and discretion.

In their earliest incarnations, the predecessors of today’s mandates were implicit rather than explicit. Central banks were not typically independent of the fiscal authority and often played a significant role in public finance. For example, before the creation of the Debt Management Office in 1998, it was the Bank of England that was responsible for gilt issuance, a state of affairs that dated back to the Bank’s creation in 1694.^{footnote [4]}

Over time, and particularly through episodes of severe macroeconomic stress, societies learned hard lessons about what happens when money and credit are not managed with care. Extreme volatility, financial crises, and inflation swings demonstrated that monetary policy frameworks that lacked clear objectives or hampered by political pressures had a tendency to produce poor outcomes.^{footnote [5]}

This issue came to a head, and was most especially visible, in the inflation crisis of the 1970s and early 1980s (Chart 1), the episode that set the stage for the emergence of independent central banks with inflation targets that we see today.

In a pattern that repeated across most advanced economies, what began as an inflation burst driven by an energy‑price shock, something that no central bank could have fully prevented then or now, was allowed over time to propagate through the economy for many years, unchecked by effective policy mechanisms. As experience showed, once the second-round effects get going and inflation expectations become unanchored, regaining control requires much more drastic and very painful action later on.^{footnote [6]} After that unhappy episode, something had to change. The monetary status quo was not an option: unmoored from the fixed exchange rate system of Bretton Woods, the world had yet to find a new nominal anchor.

Chart 1: Global energy prices – oil and gas (LHS) and Inflation rates across countries (RHS)

Percent

Thus, after the dust settled, by the end of the 1980s and into the early 1990s, a broad consensus had formed across advanced economies: monetary and financial policy needed clearer objectives and greater insulation from the short‑term incentives of political cycles. If, on occasion, central banks were going to have to take difficult decisions, those that impose near‑term costs for long‑term gain, then they could not be subject to day‑to‑day political pressures.

Indeed, this conclusion was grounded in empirical evidence. Research by Alesina and Summers (1993) found that the monetary discipline associated with greater central bank independence had delivered lower and more stable inflation in practice. Therefore, the modern solution to this problem was central bank independence coupled with clear numerical (inflation) targets. Moreover, as Debelle and Fischer (1994) argued, inflation performance is likely to be better if a central bank has a mandate for monetary stability, without goal independence, but with instrument independence. That is, the goal should be a societal choice, a democratic aggregation of preferences, depending on how averse society is to inflation volatility as well as to large output fluctuations. However, the central bank should then have the independent instruments needed to achieve that goal.

Independence, crucially, was never conceived as autonomy without accountability. Instead, it created a need for explicit mandates – targets and governance structures that ensured central bankers were empowered to act but, as unelected bureaucrats, compelled to justify and explain their actions to the public and their elected representatives.^{footnote [7]}

In the UK, this transformation took concrete shape with the Bank of England Act of 1998, which formally established the Monetary Policy Committee (MPC). The MPC was given responsibility for setting monetary policy with the primary goal of maintaining price stability, defined through the government’s inflation target. The more detailed parameters of the mandate were, and still are, set out annually in the Chancellor’s remit letter.

For nearly three decades, two features of this mandate have been especially important:

• The primacy of the inflation target, defined currently and for most of this period as 2% annual growth in the Consumer Price Index.
• A strong requirement for accountability, both through formal reporting (e.g. the open letters to the Chancellor), informal public statements (such as speeches), and through scrutiny by Parliament.

These twin pillars have not only anchored monetary policy but, in my view, also supported the legitimacy of central bank independence itself.

But accountability is not a static concept confined to statutory reporting, speeches, and testimony. It is increasingly about direct engagement with the public. An important, and I must say rewarding, part of my job on the MPC is going around the UK to speak with businesses, workers, and citizens in the course of visits arranged through the Bank’s network of Agents.

I try to make one of these Agency visits every few weeks. This month I will do my 12th visit, traveling to Cardiff where upcoming meetings will involve firm visits, roundtable discussions, and participation in a Citizens’ Panel. These meetings are important as they provide real‑time insight into the economy as it is on the ground, complementing formal data with qualitative evidence. But this kind of direct engagement also strengthens the legitimacy of central banking and ensures policy judgements are grounded in a clear understanding of economic reality across the country and how it affects the people whom we serve.

How mandates evolve: learning from the past

As I have already mentioned, the MPC’s mandate today is not simply about inflation alone. Instead, it recognises that there will be circumstances in which returning inflation to target as quickly as possible (subject to the lags in monetary policy transmission) could cause undesirable volatility in economic activity and employment.^{footnote [8]}

This issue is not hypothetical, and was raised in the current interest rate cycle as one of the factors the MPC has been taking into account after that large shock of 2022. And, in my view, this refinement does not weaken our commitment to the inflation target but instead provides more clarity and guidance for the MPC as to how we should best navigate large shocks without undermining credibility.

The history of mandates, therefore, is also a history of institutional humility and the accumulation of learning from experience.

Monetary policy obviously cannot, except in the very simplest of models and not in the real world, deliver inflation at exactly 2% in every month of the year. Under realistic conditions, with uncertainty about the shocks that have hit the economy and uncertainty about the effects of monetary policy itself, measured inflation will almost always not be exactly at the target.

Beyond that, large shocks can introduce structural change that doesn’t become apparent until well after the shock has occurred. For instance, one recent point of contention on the MPC has been the degree to which labour market dynamics may have changed through the post-Covid era. This may or may not turn out to have been the case, but I think it is currently impossible to know with any degree of certainty. We have seen quite exceptionally large revisions to labour market variables and productivity so we will continue to learn about the past for a number of years. So policymaking in real time requires a balancing of all the evidence, to judge the current position and then determine the best action within the parameters of the mandate, but with a clear focus on the future outcome, well beyond the current state of the economy.

In the MPC’s mandate, this great, cross-domain uncertainty is reflected both in its forward-lookingness and in its recognition of near-term trade-offs between output and inflation. It would be a fool’s errand to try and steer prices in the shops today or even this month. Any policy decision I, or we, make, will take some 12 to 18 months to fully pass through, given the lags in monetary policy transmission.

Instead, the MPC tries to set policy such that we can be reasonably confident that the rate of price changes would return to target reasonably quickly absent new shocks. If that weren’t the case, Bank Rate would have to swing around wildly, plausibly leading to undesirable swings in employment and aggregate demand.

Forward-lookingness allows the MPC to set policy in a gradual and predictable manner protecting the real economy from costly adjustments to transitory shocks. For example, even in the best of cases, with perfect foresight of Russia’s invasion of Ukraine and its economic ramifications, it likely would not have been desirable to set monetary policy such as to try to fully offset the inflation surge from energy prices.^{footnote [9]}

The next set of charts show an update of a thought experiment first done by my predecessors on the MPC, Ben Broadbent and Silvana Tenreyro^{footnote [10]}. It asks what a policymaker who had perfect foresight of the future might have done in 2020 (and beyond) to return inflation to target more quickly.^{footnote [11]}

The aqua lines are realisations of CPI inflation, the output gap, and Bank Rate from 2019 until now. We tell our synthetic policymaker to take these paths as given and then design an alternative policy stance. We further tell our synthetic policymaker to only care about inflation and to pay no heed to trade-offs or output stabilisation. To me, this exercise is instructive because it shows the limits of monetary policy and the need for careful judgement about the costs of inflation stabilisation.

First, we see on the right-hand chart that the selected policy stance is much tighter than what happened in reality. Instead of peaking at 5¼%, Bank Rate climbs to over 10% in 2023 and would still be about 7% in early 2026. The proximate effect of this policy is a deep and lasting recession. The output gap opens up further, even beyond its Covid-era trough. It also never recovers. This might seem abstract, but we should be aware that this would be a truly catastrophic recession: as deep as the recession after the Global Financial Crisis with millions of unemployed and wide-spread deprivation but longer-lasting.

And for what? Inflation in this thought experiment still peaks at over 7% in 2023 and returns to target not much sooner or not much more sustainably than in the event. We actually under-shoot the target more often than not. Now, this is in some sense an extreme scenario and some people may think that a policy path somewhere between the aqua and the orange would have been optimal. But to me it shows that the path actually taken – by looking through much of the relative price shock and letting the price level adjust while keeping an eye on the medium-term anchor – ex post, was the right choice.^{footnote [12]}

Chart 2: Inflation stabilisation under perfect foresight – post-Covid inflation

Percent

This point has been well understood within analysis undertaken at the Bank, and the point is more general: similar findings have been produced for the case of the United States and the path of Fed policy.^{footnote [13]}

Now, this is all very much in the past, and I was not part of the committee that made these decisions. But I was and am part of the committee that had to grapple with the much smaller but still vexing inflation hump of 2025. So let me repeat the above exercise also for the more recent period, which I think underscores the point.

We now take the second quarter of 2024 as our jumping off point – for everything before that we let bygones be bygones. Thereafter, so when I joined the committee, we again assume perfect foresight of the data until today and further follow the latest MPR forecast from February 2026. And we set the objective function such that it ignores trade-offs and only looks at inflation. The question now becomes: With this impossible knowledge of the future, would I or should I have set policy differently, in order to aggressively reduce the deviation of inflation from the target?

And again, to me, the answer is no. The ‘Alter Alan’ in this exercise again engineers a markedly tighter policy path than in actuality with Bank Rate rising through 2025 and thereafter cutting only slowly. Under this alternative policy path, Bank Rate would still be a full percentage point higher than it is today.

In turn, the output gap would be much wider, and unemployment would be much higher than it currently is. Meanwhile, inflation would return to the target perhaps a quarter earlier than we now think likely. But thereafter it would significantly and persistently undershoot. Again, I believe that our policy of looking through the hump while keeping an eye on the medium-term anchor was a good one. Personally, and I’ve said this many times before, I would have preferred a looser path for Bank Rate to prevent the output gap from opening and unemployment from rising further, but I recognise that real-time uncertainty means that reasonable people can reasonably disagree.

Chart 3: Inflation stabilisation under perfect foresight – inflation hump

Percent

So, again, humility is an important keyword. But humility also goes hand in hand with accountability. We have a mandate, but we also know ex ante we cannot hit the target at all times, and from which some deviations may result from exogenous shocks, like energy prices, which are large and unpredictable.^{footnote [14]}

That will require some explaining. So explain it we must. And that ties back to my earlier point about being held accountable for our policies, through our speeches and communications, and importantly our direct engagement with the public.

Translating the mandate into policymaking

Mandates provide the principles; policymakers must apply them. This translation from principle to practice is where judgement becomes indispensable and unavoidable.

The MPC must weigh incoming data, evaluate risks, and consider the trade‑offs inherent in every policy decision. Some periods are defined by clear signals and smooth adjustment; others, especially those following large shocks or global turbulence, are marked by uncertainty, non-linearities, and the distinct and very real possibility of temporary over‑ or undershooting of the target.

The MPC must continuously interpret the world through the lens of the mandate: assessing supply shocks, labour market conditions, dynamics in inflation expectations, and the risks on both sides of the distribution. Setting policy is not a mechanical exercise and requires constant deliberation and judgement.

Some may raise the spectre of groupthink, but if we succumbed to that it would be detrimental to a holistic judgement of all these risks. Suffice to say that the MPC’s decisions and voting patterns of the past few years put paid to that thesis. And I cannot stress enough that the role played by external members on the MPC is central to the analysis and debate that forms the basis of sound policymaking. We all take our job seriously.

Moreover, the MPC mandate’s medium‑term orientation is absolutely essential. It recognises that returning inflation to target too quickly can impose unnecessary costs on output and employment, while returning too slowly can risk un‑anchoring expectations. The art of policy is managing the balance – tightening or loosening at the right pace, guided by evidence, models, and experience.

In each of my twelve votes so far on the MPC, I have had to reflect on this balance. I think there are several useful examples here of how the mandate feeds into policy deliberations. At least for me:

1. Signal, noise, and speed of adjustment: At the end of 2024, we were approaching the inflation target, and in the last half mile.^{footnote [15]} But that didn’t mean that new shocks couldn’t appear which might knock us off course again. The inflation hump of early 2025 was just such a shock, from new energy and food shocks and a raft of changes to employment taxes and administered prices (the blue bars in Chart 4). But the mandate, based on common sense not to mention economic theory, urges us to look through those one-off shocks, to extract the signal from the noise, and think about where underlying inflation sits and where it is headed. I was confident then, as I am now, that the disinflation process was intact, and that since 2022 it had followed a predictable downward course. There is another similar shock for the UK in 2026, with new interventions now pushing down on prices (salmon-coloured bars in Chart 4), but symmetrically, we also need to avoid getting unduly distracted by that too.

Chart 4: Inflation decomposition

Contribution of deviation of inflation from target (pp)

2. Level target versus rate-of-change target: One of the key aspects of that underlying inflation path is the destination of wages and services prices, which are always slowest to adjust after a shock works its way through the system. But we have an inflation, not a price-level target in the mandate. So the adjustment is a natural consequence: real wages are pushed down by an exogenous positive price shock, but some of the price shock persists, and nominal wages eventually catch up. I saw in late 2024 that real-time wage settlements were coming off at pace, and that continued throughout 2025, so that now these settlements sit near 3%. These developments would be obscured if we only observed backward-looking annualized data, leaving us behind the curve. Given staff estimates of productivity growth, and wage settlements near 3%, wage growth will likely converge to target-consistent rates this year, and services inflation (which is labour-intensive) can then achieve normalization. In fact, I have recently noted that high frequency 3-month CPI inflation data very much suggest that normalization is close to complete (Chart 5).

Chart 5: Pay settlements and high-frequency inflation rate

Percent

3. Trade-offs: As laid out above, the mandate requires us to consider the trade-off between pushing inflation back toward target quickly versus incurring too much output volatility, following a large shock to the economy. That was the case after the inflation shocks of 2021 and 2022, and especially after the Russian invasion of Ukraine caused energy prices to spike. The economy proved resilient at first, as rates were hiked to combat second-round effects, defying the pessimistic projections of many observers. We did not see the high sacrifice ratios of the 1980s this time. In fact, there seemed at first to be little trade-off at all. I think this was as a result of anchored inflation expectations, a definite payoff to the kind of mandate-based inflation targeting regime we have had in place for 30 years. But resilience in the face of tight monetary policy may not last forever, and in the last 12 months sustained restrictiveness has been bearing down: It has struck me that, over 2025, and into 2026, broadly speaking, inflation has been weaker than expected in successive MPR forecasts, the unemployment rate higher, and wage growth lower (Chart 6). Given the mandate, attention to trade-off considerations will often recur in our deliberations, as it should. But I judge that we will soon find ourselves largely outside of trade-off territory – and even at risk of entering the familiar realm of deficient demand.

Chart 6: Revisions to the MPR forecast

Forecasts for inflation, unemployment rate, and wage growth

Finally, to embrace the spirit of humility, let me briefly make the case for a different outlook. What if I am wrong about all of this? In what world might we see further, persistent inflationary pressures over the next few years, when one-off fiscal measures have washed out? In my view, if we were to end up in such a world, it would likely be one in which productivity growth in the UK persistently surprises to the downside and remains below its long-run growth rate.^{footnote [16]} I am a growth optimist, especially given the prospects for AI and technology, but this alternative world would require a tighter monetary policy stance to achieve target-consistent levels. This is not a question of central bank mandates but rather a question of keeping an open mind about fundamentally uncertain determinants of inflation.

Continuity versus change: looking to the future of mandates, and beyond

In summary, central bank mandates should be strong yet flexible. History demonstrates that as the economy evolves, central bank strategies have adapted over time to reflect this reality.

We also need to acknowledge that central banks and their mandates can never fully solve every type of inflation problem, including the big shocks of recent years. As an economic historian, I think when future scholars look back at the macroeconomic shocks of last fifty years and not just the past five, one of their main takeaways will be the outsized role of energy shocks in causing disruptive spikes in overall CPI inflation.

Prices are set in markets, and price shocks disturb that system and its dynamic equilibrium. Large energy shocks move faster than inflation-targeting central banks can respond, and after they propagate into the system, they leave an imprint, even if we can do better to contain them (now versus the 1970s, for example).

No amount of tweaking in central bank mandates will fully insulate us from these exogenous risks, which as of now derive from the present and very time-specific nature of our dependence on energy and the patterns of geopolitical risk. Tackling those issues requires thinking and actions well beyond the domain of central banking, so I will leave it at that.

Whatever form mandates take, two essential principles must endure:

• Clear objectives focused on price stability and financial stability, with appropriate recognition of trade-offs.
• Robust accountability, making sure independence aligns with democratic values and serves public interests.

Ultimately, a mandate offers a guiding structure for policy amid uncertainty. While it cannot promise specific outcomes, it enhances the likelihood of policies being effective, consistent, and legitimate. As history illustrates, when mandates are well-defined and institutions answerable, societies can better cope with the inevitable volatility of economic life.

Acknowledgements

Thanks to Lennart Brandt and Vitor Dotta for help preparing this speech, to David Elliott for conducting the policy counterfactuals, and to Andrew Bailey, Olly Bush, Francesca Diluiso, Rich Harrison, Neha Jain, Lien Laureys, Clare Lombardelli, Dave Ramsden, Martin Seneca, and Fergal Shortall for their comments.