Written evidence submitted by Professor Tony Yates (MON0027)
Submission for Treasury Committee inquiry on the ‘effectiveness and impact of post-2008 monetary policy’
Tony Yates, Professor of Economics, University of Birmingham
This written evidence is submitted in a personal capacity.
A Introductory/summary remarks
B In support of the broad overall MPC strategy since the financial crisis
1 I take the broad strategy that MPC adopted to have emerged from the view that the best contribution to the macroeconomy is to ensure inflation is as stable as possible around a fixed, low target, and, weighed against this, to lean against what would be otherwise inefficient and detrimental volatility in the real economy. [See, e.g., the mandate review document written by the Treasury in 2013].
2 The MPC achieve this objective by moving interest rates around as conditions demand: if activity falls below potential, or inflation falls below target [or either look like doing so in the future] the interest rate is cut, and vice versa. The cut in nominal rates leads a temporary fall in real rates, which boosts borrowing and spending and asset prices and subsequently inflation. This view of the transmission mechanism is confirmed by a great deal of empirical research organised around attempts to identify monetary policy ‘shocks’ and measure their effects. Simplified versions of these interest rate strategies, like ‘Taylor Rules’, (named after a famous paper by John Taylor published in 1993), do a good job of stabilising the inflation rate and the output gap in estimated macroeconomic models.
3 Once the financial crisis broke proper in the late summer of 2008, the BoE, like all central banks, soon started to cut rates sharply, going all the way to what the MPC judged to be the effective lower bound of 0.5% by March 2009. The interdiction in credit markets had caused a precipitous drop in confidence, and, absent this stimulus, would otherwise have led to disastrous fire sales of assets, hasty deleveraging, default, falls in employment income and a collapse in demand. The effect on inflation in the UK was masked by the fall in Sterling, and by the fact that the credit crunch affected supply as well as demand. After the effective lower bound was reached, the MPC judged rightly that more stimulus was needed and quantitative easing was begun. The decision to look through the initial surge in inflation, which reached 5.5% in the UK, was probably vindicated ex post by the subsequent 2 pp undershoot; and expedient once the Eurozone crisis, and its impact on the UK became clear.
4 Standing back somewhat, the extraordinarily low rates which have persisted for some time can be viewed differently, and as a predominantly global phenomenon.
5 Several forces have been pushing down on equilibrium real interest rates – something ultimately beyond the control of central banks – for years. Namely: demographics, and the bulge of those in middle-age saving into uncertain funds for retirement; the import of capital from wealth-accumulating emerging economies, anxious to save outside the reach of their own countries’ unreliable regimes and vague property rights; the effect already mentioned of those realising they overextended themselves paying down their debt, and less willing to borrow the abundant savings.
6 If central banks had failed to accommodate this fall in equilibrium real rates with a cut in nominal rates, deflation and depression would have ensued.
C A counter to some critiques of MPC and other central bank low interest rate and QE policies
There have been several critiques:
i) Low interest rates and money-printing will lead to hyperinflation
7 This critique has proven false, ex post. Ex ante it failed to understand that the relationship between money and prices away from the floor to interest rates does not hold at it.
ii) Low nominal and real rates, and quantitative easing, defrauded old age savers who depend on the returns to those savings.
8 This criticism has to be taken more seriously, but I think it fails on several counts.
9 First, savers profit from real, not nominal returns, and central banks only set the real rate over short horizons, and over longer horizons they can only accommodate the global forces – mentioned above – that ultimately determine it. This is one aspect of the long-run neutrality of monetary policy.
10 Second, we should remember that if indeed deflation and depression had ensued, following nominal rates that were set higher in an attempt to protect savers, the value of the principal [ultimately, the capital stock of UKPLC] would have been greatly reduced, and the profits/returns, out of which those pensions are ultimately funded by companies, would have been depleted, until more stabilising policy had ensued. In general, failing to use policy to counter the business cycle would increase risk and reduce resources savers would have at the point where they choose to cash their wealth in.
11 Third, even supposing monetary policy had harmed savers at the expense of others: given the mandate the central bank had it would have been for the government to effect redistributive policy to correct. And in that context we need to note the highly advantageous situation that the older generation are relative to others regarding i) possession of defined benefit pensions that were underfunded given pessimistic estimates of life expectancy / optimistic estimates of returns, ii) housing costs and iii) government policy regarding the protection of the value of state pensions.
iii) Inflation is low because, not in spite of low central bank rates [dubbed recently ‘Neo Fisherianism’]
12 This argument has been expounded recently by Steve Williamson, John Cochrane, Noah Smith and others. The analysis starts from a view more widely shared, that in the long run, [and in this case because of the need by investors to compensate themselves for a nominal debt being eroded by inflation], nominal interest rates and inflation rise one for one. It can be shown that the same simple monetary policy models also have it as an equilibrium that this is true in the short run too. However, this holds only under the implausible assumption that people in the model are sophisticated enough to understand exactly how the model works and how shocks that hit the economy play into future inflation.
iv) QE/low rates worsened inequality, for example, by bidding up asset prices.
13 It’s a feature, and not a bug, that for both interest rate and QE policy the aim is to boost asset prices. So there is a case to answer here. For example, take QE: the hope was that buying gilts raised their price, encouraging former holders into private securities, bidding up their price, and lowering the cost of finance for the issuers.
14 There are several things to be said in support of what the Bank did despite this charge.
15 First, as the BoE’s Broadbent (2016) pointed out, and compressing the story recounted there somewhat, neither inequality of wealth nor income has risen since the onset of the crisis, so there is nothing for monetary policy to account for.
16 Second, and to re-emphasise, without the action taken by the MPC the UK would have experienced a much deeper recession and much worse unemployment. Unemployment is experienced disproportionately by those at the lower end of the income/occupational scale. Reducing this business cycle risk is, indirectly, the best contribution monetary policy can make to correcting what could otherwise be argued to be a form of economic injustice.
17 Third, one must keep in mind the low frequency global forces pushing up on developed economy asset prices: these are the same forces that are pushing down on global equilibrium real interest rates discussed above. [Global demography, capital importing, flight to safety]. These are not the fault of monetary policy, and nor are they something monetary policy can or should fight against.
v) Low interest rates have weakened productivity, keeping alive ‘zombie’ companies and inhibiting ‘creative destruction’
18 There is a case to be answered here. Productivity has been surprisingly low during the most recent recession. And, as the BoE’s Barnett et al (2014) note, dispersion in productivity across sectors was high during the crisis (a sign that there is misallocation); insolvencies were surprisingly low.
19 Notwithstanding this potential cost, a number of points can be made in defence of the persistent stimulus.
20 First, the counterfactual policy of not loosening sufficiently to control inflation and activity would have been a much deeper recession, a period of protracted deflation, possibly, and a period, ultimately, presuming policy is forced to respond eventually, of much prolonged lower interest rates.
21 Second, the charge against policymakers presumes that the underlying depression averted by policy would have been the working out of efficient market forces. The weight of evidence against such a view of how the macroeconomy is close to overwhelming.
22 Third, there are other explanations for weak productivity: including, for example, that the pronounced fall in real wages encouraged firms to switch out of capital and into labour [‘capital shallowing’], lowering labour productivity in the process.
D Aspects of the MPC/BoE monetary policy strategy where I think criticism is due.
QE
23 QE had several flaws.
24 At its instigation, there was little or no evidence that purchasing only government securities was likely to be a good instrument of first recourse. Yet this is what the Bank relied on, almost exclusively. Subsequently, event-study analysis of QE programs by the BoE and the Fed looked encouraging, showing significant and sizeable effects on the yields of securities included in QE programs. [See, for example, Williams (2011) for a summary].
25 This might seem to have vindicated the MPC in its course of action. But I differ for several reasons.
26 First, although ex post the impact effects of QE look good, these weren’t known to the MPC ex ante, and in that respect it was decidedly risky to focus on one policy response, which up to that point did not look promising.
27 Second, although QE had a clear impact on government security yields, it is much less clear that they had a beneficial impact on the spread between government and private security yields – which was the more important indicator of success, since this is one of the ways we see that the private cost of finance.
28 Third, the event study analysis, the only really convincing approach to disentangle the effects of QE from the effects of the economic developments that prompted it, could not tell us whether the effects were durable. Time series studies which try to do just that are not persuasive, because they struggle to identify autonomous changes in QE.
29 Fourth, the BoE executive took upon itself the responsibility for deciding which assets to buy, leaving for the MPC only the decision about ‘how much’. The decision seemed innocuous because of the mistaken view taken at the time that what was important was the amount of reserves created, and not what assets were bought. [In fact a widespread view of economists is that at the zero bound, reserves creation is immaterial; QE could have been implemented just as well by issuing short term debt to finance asset purchases]. Since the composition of purchases is key to their impact, it would seem to have been vital that MPC take a view on this.
30 A more robust course of action – ie one likely to work well across a range of outcomes regarding how well instruments work – would have been to throw the kitchen sink at the problem. Namely: commit to low interest rates, and buy both public and private sector assets in bulk from the outset. As evidence came in about these policies’ effects, the program could have been tuned appropriately.
31 A final criticism of QE relates to the coordination of QE with the DMO. One danger with QE was that as the central bank bought up long bonds, raising their price and lowering the yield, the DMO might be encouraged to tilt its issuance towards those bonds and lower the cost of finance. This would have two costs for monetary policy. First, it would undo the effect on yields somewhat. Second, it might fuel suspicions that QE was monetary financing by subterfuge. Larry Summers argued that the US Treasury tilted issuance in just this way. It’s hard to assess accurately whether this was true for the UK, for the UK since one does not know precisely what issuance in the counterfactual scenario in which there had been no QE would have been.
Forward guidance
32 I think the MPC’s forward guidance [FG] – which I am a supporter of in principle – in late 2013 was badly conceived and executed.
33 The momentum for the BoE’s FG policy was begun by the commission by the Treasury in its review of the Bank’s mandate of March 2013 for the BoE to examine the case for embarking on forward guidance. At that time, the recovery was weak, QE was thought by some to have run its course. Why not commit to lower rates to impart more stimulus?
34 The MPC reported on the matter in August 2013, shortly after Governor Carney took up his post, and the MPC’s version of forward guidance was launched.
35 So what were the problems with what the MPC did?
36 First, the Governor and the MPC tied themselves in knots over whether the new policy was meant to impart stimulus or not. The settled position of MPC as a whole was that FG was not additional stimulus. But this jarred with some descriptions of it, e.g. that it was about ‘securing the recovery [sic]’.
37 Second, thus far, all variants of FG executed by others had been done so to provide stimulus. It was thus confusing to introduce a variant that wasn’t. Not least because one might wonder whether it was worth introducing a major change in policy communication at that point when there was no need for further stimulus [at least as MPC saw it].
38 Third, the FG was described as implying no shift from the way MPC would normally respond to news, but a device to clarify what they would do. This was difficult to square with the fact the new monitoring framework involving the unemployment rate and other indicators was a big change from what had gone before. If things were no different, why did they require such a quantum change in presentation? This difficulty hindered efforts to locate the MPCs policy regarding whether it was ‘clarifying’ or ‘commitment’ forward guidance.
39 My own position on this is that the MPC should be engaging in continual and quantified forward guidance about future interest rates [sometimes referred to as Delphic forward guidance], voting on complete interest rate plans. It should also keep in reserve the possibility of engaging in commitments to keep rates lower for longer [sometimes referred to as Odyssean forward guidance], implying subsequent inflation target overshoots. I worry that the confusion around what the MPC actually did will have made the subsequent use of this policy harder to communicate.
5 Prospects for the future; suggested reforms
40 The amount of monetary and fiscal room for a response to a future crisis will be considerably diminished relative to 2008/9.
41 Even if a crisis were not to hit until Bank rate had climbed back to its likely rest point, that rest point is probably between 2-3%, on account of the highly persistent global, real forces pushing down on real rates, offering half as much conventional interest rate stimulus as was available in 2008/9. It might take many years [>10] to unwind the Asset Purchase Facility, so there could well be diminished room for stimulus via QE, assuming safe limits to the central bank balance sheet, and/or diminished returns to QE. It is not that implausible that no progress will have [nor should have] been made to take government debt back to 40% of GDP.
42 A few reforms to create more room for stimulus, or provide for other sorts of it are:
i) Raising the inflation target to 4%. So long as the next crisis does not hit before this higher target is reached, this would make room for larger interest rate cuts as the resting place for the central bank rate tends to rise one for one with the inflation target via the ‘Fisher effect’. This is about future crisis prevention, not cure for current woes, and in fact would better wait until the current target is demonstrably achieved. Such a move is not without costs; of being seen to change the inflationary goalposts; of setting the authorities up for failure if they can’t raise inflation further. But these seem outweighed by the countervailing risks of losing control on the downside in the event of another crisis.
ii) Some form of limited delegation of the macro-stabilisation element of fiscal policy that kicked in at the lower bound to the central bank interest rate. A minimal form of this I have advocated is the following. The MPC indicates that the lower bound is likely to imply a missing stimulus, and quantifies this in terms of interest rate changes. HMT is invited to consider a fiscal measure of its own choosing to implement, but reserves the right to ignore the advice, publishing its response. If it takes the advice, the OBR evaluates whether the measure and its planned unwinding is consistent with fiscal sustainability. This delegation would provide institutional cover for fiscal stimulus that was lacking in the last financial crisis, and perhaps led to unnecessarily tight fiscal policy in recent years. It would also maximise the expectational benefits of stimulus. In 2010, almost the reverse happened, with greater austerity threatened than was actually implemented subsequently. A limited delegation of this form also offers some of the benefits, without generating the disquiet about the loss of democratic control of fiscal tools that a full scale ‘fiscal council’ would entail.
iii) Pre-announcement of the preparedness to use Odyssean forward guidance, and tolerate post crisis inflation target overshoots [something the current committee have studiously and I think spuriously avoided].
iv) Sharpening of the rules DMO use to tilt gilt issuance, so that it can be more readily verified ex post that any MPC QE reduces average maturity of private sector holdings relative to a counterfactual of not intervening.
v) Routine voting on, and disclosure of, interest rate plans.
vi) A preparedness to widen and scale-up the purchase of [and loan against] private sector assets as a response to the next major stimulus-requiring event. This policy requires overcoming several obstacles, not least: the small size of the corporate bond and other asset markets; managing exposure to credit risk; establishing and maintaining infrastructure and competence within the BoE. I think the Bank could be working on an overall plan for private asset intervention that i) was transparent, to help maximise expectational benefits, ii) included estimates of expected costs and stimulus benefits, iii) was integrated into a plan alongside interest rate policy so that observers could form a view about what would be purchased, and how much, under what kind of circumstances.
vii) A serious, open consideration of the costs and benefits of undertaking some form of ‘helicopter money’ as a measure of absolute last resort and whether there exist institutional designs that preserve some separation between monetary and fiscal policy and distinguish it from such proposals as ‘people’s QE’.