Take it slow and steady

It has long been evident that we are heading for a bumpy economic landing but the BoE’s pessimistic outlook delivered in this week’s Monetary Policy Report nonetheless came as a shock. According to the BoE, CPI inflation is set for a peak of 13% by the fourth quarter of 2022 whilst real output growth more or less flatlines over 2023 and 2024 and unemployment is projected to rise. All this is taking place at the same time as the Conservative Party is set to choose a new leader – frankly, this would be a good contest to lose given that politicians are almost certain to be blamed for the cost of living crisis heading our way. 

It has become fashionable in recent months to lay the blame for the inflation surge on the BoE. Liz Truss, the current front runner to replace the hive of inactivity that is Boris Johnson, argued recently that “the best way of dealing with inflation is monetary policy and what I have said is I want to change the Bank of England’s mandate to make sure in the future it matches some of the most effective central banks in the world at controlling inflation.” I am not sure which central banks she is referring to. In the US, inflation is already above 9% and in the euro zone it is within a whisker of this rate (according to Eurostat, it is likely to have hit 8.9% in July). Admittedly Japanese inflation is at 2.2% but this is after years of disinflation with real GDP growth averaging just 0.5% per annum since the turn of the century. 

Prominent Tory politicians, such as current Attorney General Suella Braverman, have suggested that “interest rates should have been raised a long time ago and the Bank of England has been too slow in this regard.” But this is to conflate a number of issues in the monetary policy debate. In my view there are a number of questions which should be tackled separately: (i) did central banks become complacent in the wake of the GFC by holding rates too low for too long; (ii) was the policy response around the time of the Covid outbreak appropriate and (iii) would higher interest rates have prevented the current inflation spike? 

The answer to (i) is undoubtedly yes. The decline in inflation in the years prior to 2008 buttressed central bank credibility and their actions to inject liquidity in the wake of the GFC without triggering a surge in consumer prices gave rise to the view that they really had conquered inflation. Modern macroeconomics also has a lot to answer for, with the dominant paradigm in academia and central banks having little to say about the inflation process (I touch on this below but a more detailed look is a topic for another day). Thus the recent spike in prices blindsided central bankers who had, frankly, grown complacent. 

With regard to (ii) there is a valid argument to suggest that the Covid-induced collapse in output was a supply side shock to which central banks responded by stimulating demand, which was inevitably going to lead to inflation as demand outstripped supply. Admittedly, this view has only emerged with hindsight but there is a ring of truth to it (even if it is understandable why central banks reacted as they did in 2020). As to whether higher interest rates would have prevented the inflation spike, the answer is unequivocally no. Monetary policy cannot hope to impact on the type of supply shock posed by a huge rise in energy prices.

What does the empirical evidence suggest?

All of this raises an uncomfortable question for central banks which is far less clear cut than they think it is. The standard response to rising inflation is that the credibility of central bank policy requires higher interest rates to bear down on inflation expectations. The economist John Cochrane has been looking at this in a US context (a shorter overview of some of the key points can be found in this blog post). His starting point is the classic 1972 paper by Robert Lucas (even after 50 years it is still heavy going) which demonstrated that money is neutral with regard to the real economy. But as Cochrane pointed out, “our central banks set interest rates. The Fed does not even pretend to control money supply. There are no reserve requirements. We need a theory of inflation under interest rate targets.” At the current juncture, there isn’t one.

Modern macroeconomic models rely on the Phillips curve to propagate inflation dynamics (this postulates there is a negative relationship between unemployment and inflation). In Cochrane’s words, “the Phillips curve has been a disaster, especially lately” (a subject I looked at briefly here). Indeed, Roger Farmer has argued that we should replace the standard Phillips curve with a ‘belief’ function in which nominal output in the current period depends only on what happened in the previous period (see my blog post on this issue). Cochrane concludes on the basis of his analysis that there is no need for central banks to overreact to the surge in inflation: “If the Fed does nothing, inflation may surge for a while, but it will not explode. Inflation will eventually come back on its own, so long as fiscal policy does not create more inflation. The Fed’s inaction does not spur more inflation or set off an inflation spiral.”

I find this conclusion rather comforting since most of the models I have ever used demonstrated only a tenuous relationship between interest rates and inflation. The chart above illustrates the impact of a simulation exercise conducted using my structural macro model of the UK in which Bank Rate is raised by 100 bps relative to baseline and held there for four quarters. After 12 quarters, output is reduced by 1.4 percentage points relative to baseline but the level of consumer prices only declines by 0.1% (note that the impact on wages is far larger).

Another way of looking at the relationship between interest rates and the macroeconomy is to look at vector autoregression (VAR) models which capture the linear dependencies amongst time series by modelling each one in terms of past values of all series in the system. VARs are designed such that we do not need to know the structure of the economy but instead capture how changes in one variable affect other variables in the system. This makes them sub-optimal as forecasting models but very insightful as a method of assessing how shocks percolate through the economy. The model used here contains series for real GDP growth, CPI inflation, Bank Rate and the nominal sterling exchange rate index. The impulse response functions shown in the chart above suggest that the impact of higher BoE interest rates on UK inflation is limited (purists will quibble about the specification of the model and the analysis may come across as a bit back-of the-envelope).

Key takeaway

Based on the evidence, those who argue that the BoE is too far behind the curve (the same applies to other central banks) and that huge interest rate rises are needed to curb the current inflation spike are misguided. The most likely impact of such action will be to crimp economic activity thereby making life harder for those already being hit by the cost of living crisis.

As Cochrane notes: “Why do we not know answers to such basic questions? I think we have been a bit guilty of studying the world as we wish it to be rather than the world we are in.” He goes on to point out: “How is it that we’ve been playing with interest-rate based models for 50 years, yet such basic questions are still unanswered? … As I look at the effort to build monetary models based on interest rate targets, we have been guilty of playing with far too complex models that we don’t really understand.” This is not an argument for central banks taking their foot off the pedal. However, it is an argument for due care in the monetary tightening process. If the economic downturn is as nasty as the BoE predicts, the calls for rate cuts in a year’s time will start to become louder.