A short guide to why we should not raise UK interest rates

Everyone expects the MPC to raise rates on Thursday. This would be a mistake. Discussion about interest rate changes in the press normally involve large amounts of data and charts about the state of the economy. Here I want to do the opposite: to present the minimum you need to know to understand that raising UK rates right now is the wrong thing to do.

Everyone should know that UK inflation is currently around 3% because of the Brexit depreciation. But because the impact of a deprecation on price inflation is temporary if wage inflation remains flat, the Bank said they would ignore this temporary rise. The key is to look at whether average earnings inflation is responding to higher consumer price inflation. The answer is they are not: average earnings growth has been slightly above 2% all this year, which is a little lower than the average for 2016.

But what about unemployment being at a 42 year low? Surely that means earnings growth is just waiting to kick off. The first point is that unemployment is not currently a good measure of labour market slack. A better measure is the Resolution Foundation’s underemployment index, which is still above levels before the global financial crisis. And before you say but that was a boom period, it wasn’t. UK core inflation was below 2% throughout, and earnings growth was consistent with this.

The other thing to say is that it is quite wrong to assume that we know what the level of labour market slack is that would lead to increases in earnings growth (what economists call the NAIRU). The NAIRU moves over time. As just one example of why it might move, a labour force that rents is likely to be more mobile than one that owns a house, and so the trend towards renting should reduce the NAIRU.

So looking at the labour market, there is no sign that we are close to a level where earnings inflation might pick up. And that is pretty well a precondition for inflation to exceed its target of 2% over the medium term. That is all you really need to know. If you want to know why the MPC probably will raise rates, read on.

What I suspect the Bank are worrying about is that Brexit has created what economists call a negative supply shock. In particular, both investment and productivity growth are much lower than the Bank were expecting before Brexit. They will point to various survey measures which show firms do not have any spare capacity. But this reasoning I think indicates a conceptual weakness.

Firms have two responses to lack of capacity: raising prices or investment. By choking off demand and raising rates when firms run out of capacity the Bank will discourage investment, and right now what the economy desperately needs is more investment and the productivity improvements that brings with it. The Bank shouldn’t worry about a bit of inflation that might come with higher investment, because 2% earnings growth is an anchor that will prevent inflation deviating from target for any length of time.

That should be enough, but there are two other reasons why the Bank should not raise rates. First, right now the downside risk on the demand side from Brexit surely exceeds the upside risk. Second, as the OBR chart here shows (look at orange bars), after a pause in 2017 austerity is planned to return in 2018 and 2019. Combining fiscal and monetary tightening in a boom would make sense, but we are currently in an economic downturn, with GDP per head growing this year at a third of its average pace since the recession of 2009. [1]

Finally, it is always important to consider risks. Suppose earnings growth does pick up sharply just after the MPC’s monthly meeting. The Bank always says it wants to be ‘ahead of the curve’, to avoid too rapid an increase in rates. This is the mentality that has led inflation to undershoot in the US and Eurozone since the recession, and if you take out the impact of depreciations by looking at the GDP deflator the same is true for the UK. The problem for the UK economy since the recession has not been too much inflation, but far too little demand.

[1] Specifically, average growth in 2017 is 0.1% per quarter, and averaging quarterly growth rates from 2010 Q1 to 2016 Q4 gives 0.3% per quarter          

Targeting wage inflation

I was pleased to see that David Blanchflower and Adam Posen have advocated using wage inflation as an intermediate target in their analysis of labour market slack in the US. Specifically they say

“Our results also point towards using wage inflation as an additional intermediate target for monetary policy by the FOMC, paralleling on the real activity side the de facto inflation targets on the price stability side.”

I have periodically argued for wage inflation targets in the case of the UK, but both their and my arguments are universal.

My own argument for targeting wage inflation has been a combination of theory and practicality. As I have often pointed out, there are good theoretical arguments for targeting alternative measures of inflation besides consumer prices. The way macroeconomists usually measure the cost of inflation nowadays is to score the distortion to relative prices created by the combination of general inflation and the fact that different prices are set at different times. The ‘ideal’ price index to target would be one that gave a higher weight to prices that changed infrequently, and a low weight to those that were changed often. Wages are just another price in this context, and they are changed infrequently.

The practical argument is that if we had been targeting wage inflation over the last few years, monetary policy would have worried less about the temporary inflation induced by shocks such as commodity price increases or sales taxes. Here is a chart of recent and expected wage inflation (compensation per employee) from the OECD. 

In normal times we would expect 2% price inflation to be associated with something like 4% wage inflation because of productivity growth. Wage inflation has not come close to that number in recent years in the UK, US or the Eurozone. It is difficult to see how the ECB could have raised interest rates in 2011 - as they did - if they had had wage inflation as an intermediate target.

The argument put forward by Blanchflower and Posen is rather different, because they associate wage inflation with the real side of the dual mandate in the US. To quote:

“wage inflation should be considered as the primary target of FOMC policy with respect to the employment stabilization side of the Fed’s dual mandate, at least for now. Unlike unemployment, the rate of wage inflation requires less judgment and is subject to less distortion by such factors as inactivity. At least four of the labor markets measures that Yellen cites as worth monitoring- unemployment, under-employment of part-timers, long-term unemployment, and participation rate- reveal their non-structural component by their influence on wage growth. And that is what the Fed should be trying to stabilize along with prices.”

To paraphrase, unemployment (or anything similar) can become distorted as a measure of labour market slack, but wage inflation is a good indicator of the true state of the labour market.

I would add one final point. The spectre that seems to haunt central bankers is the inflation of the 1970s. That has to be avoided at all costs. Yet the 1970s was associated with what was called a wage-price spiral: both price inflation and wage inflation rising rapidly, and a feeling that this was a contest between workers and firms that neither could win, but where society was a loser. If we want to avoid a wage-price spiral happening again, it is only logical that we look at wages as well as prices.

Adapting forward guidance

I have a personal form of forward guidance: that I try and wait a day between writing and publishing blog posts. So yesterday I wrote a post reacting to the previous day’s news that UK unemployment had fallen rapidly to 7.1%. That news led to some speculation that the Monetary Policy Committee (MPC) of the Bank of England might change the number for their unemployment ‘knockout’ (the point at which they might start thinking about raising rates) from 7% to, say, 6.5%. There has been similar speculation about US monetary policy. I wrote that I thought this would be unlikely, but rather than let guidance wither away, they would instead prefer to change the nature of their guidance.

So today, as that post laid waiting on my hard drive, I read that Governor Carney indicated that the Bank has decided not to revise its 7 per cent unemployment threshold. “We’re trying to get across that it’s all about overall conditions in the labour market . . . We wouldn’t want to detract from that focus by unnecessarily focusing on one indicator.” So I’ve lost my opportunity of showing that I can anticipate MPC thinking. Perhaps instead I can write about why they might be thinking this way, and what they might specifically do.

The place to start is with why unemployment has been falling much faster than expected. As Chris Dillow explains, it indicates that UK productivity continues not to grow. The Bank hoped that the return of output growth might be accompanied by a resumption in productivity growth, so that unemployment would come down more slowly. They can hardly be blamed for this. Zero productivity growth for four years during a recession was puzzling, but continuing flat productivity when there is a recovery in output growth is in macroeconomic terms just weird. 

So how might forward guidance change? Here we need to make one point, and then ask one question. The point is that forward guidance is all about providing information to the public about what policy might do if events deviate from forecasts. As a result, those critics of such guidance who use poor forecasting as an argument completely miss the point. It is not what I call forward commitment. This leads us to the question: what is it that makes the MPC relaxed about the unexpectedly rapid fall in unemployment?

The answer is in this chart, which shows year on year growth in private sector earnings (source:ONS).

  

The series can be erratic, in part because of bonuses. Indeed, to quote the Bank’s inflation report (pdf): “... growth was volatile in 2013 H1, rising from 0.1% in Q1 to 2.8% in Q2. That largely reflected some people taking advantage of the reduction in the top rate of UK income tax in April 2013, and deferring bonus payments and earnings they would have received in 2013 Q1.” So we can call this the ‘Osborne hiccup’. However smoothing this out, year on year growth has been gradually moving down towards a little above 1%, and there is no sign so far of any reaction to falling unemployment. (Public sector earnings are not growing at all.)

With earnings growth at 1%, and productivity flat, that means unit labour costs are rising well below the inflation target of 2%. If earnings growth stays at 1%, there is no reason coming from the labour market for raising interest rates. If private sector earnings do start increasing by more than 2%, then the focus will then shift to productivity growth. Only if this fails to match the increase in earnings will a rise in rates become a distinct possibility.

So the natural way to change forward guidance is to incorporate this thinking. The unemployment knockout could be replaced with one that says interest rates increases will not be considered as long as private sector earnings growth is not more than 2% above private sector productivity growth. And now I think I should post this, to avoid another rewrite. 

How to try and get more inflation quickly

The following is written in the UK context, but the proposal I make would work in the US and the Eurozone as well.

Wednesday will be Autumn Statement day in the UK: one of the two times a year the Chancellor makes big fiscal announcements. Yet this post will not be about fiscal policy: current policy looks too much like rearranging the deck chairs, and so I will just be repeating myself in saying that the strategy is all wrong. Instead this post is about monetary policy. How come - that is the MPC at the Bank of England surely? I think not - they are just rearranging the deck chairs too. The big picture on UK monetary policy is down to the Chancellor.

The MPC’s mandate, set by the Chancellor, is to hit 2% for CPI inflation ‘within a reasonable time period’. In the past I and others thought the phrase reasonable time period would allow the MPC in practice to minimise a combination of excess inflation and the output gap, which is what most macroeconomists assume monetary policy is all about. I think this was true, within limits. In the UK these limits were surpassed following the Great Recession, where we had a large negative output gap but above target inflation. We are at the ZLB, and have had Quantitative Easing (QE), because the MPC looked at the output gap as well as inflation, and thought that a large output gap would mean inflation would come tumbling down. It did not, so now they are just targeting CPI inflation (and rather conservatively at that).

What we need is for UK monetary policy to be as expansionary as it can be, and we are not going to get that under the current system, even with a new man running the Bank. I have talked to enough MPC members to know that they take the mandate they are given very seriously. They certainly believe they do not have the discretion to follow a different strategy while still paying lip service to the current set up. The system can only be changed by the Chancellor. Now some would like him to move to nominal GDP targets, and I’d be happy with that. However I cannot see him doing that overnight: the move to include output and go from changes to levels is probably too much for something that is not widely discussed in the UK . The best we can hope for (and hope is the word) is for him to launch some kind of inquiry into this (and other) possibilities - I have suggested in the past that the Treasury Select Committee could do this.

So for the next year or two we are almost certainly stuck with inflation targeting. He could announce an increase in the inflation target, which he has the discretion to do as part of the current system. But he will not, because it plays terribly in terms of politics. As Noah Smith laments, what we as economists understand as inflation is not what everyone else understands by inflation. The number of people in the UK who would be prepared to defend raising the current CPI inflation target to 4% (which I would) is probably less than the number of economics departments in the UK.

However, why not change the measure of inflation being targeted? The GDP deflator is the obvious alternative, but there is a better candidate if the aim is produce a significant but temporary rise UK inflation (and thereby to start a proper recovery in UK output). My suggestion is that 2% CPI inflation is replaced by a target for 4% growth in average earnings. In normal times 4% earnings growth and 2% CPI inflation would be quite compatible (because real wages grow with productivity), but currently average weekly earnings inflation is below 2%.[1] So moving wage inflation from 2% to 4% is a real challenge for monetary policy. To help with this challenge, the Chancellor could at the same time sanction unlimited QE, and could suggest this goes beyond just buying gilts.

If monetary policy still has some power despite the ZLB, then this move would have a significant effect in reducing real interest rates. It would also play well politically. “In the last few years hard working families have seen the real value of their earnings fall, as wage growth has been held back while the Bank of England has failed to keep consumer price inflation on target. It is time that we changed things so that the rewards from working increased rather than decreased over time. That is why I am announcing today ….” Rhetoric that is nonsense in terms of economics, but no worse than much of the rhetoric the government currently uses.

Now of course plenty of people will complain that this is returning us to the 1970s, losing all the gains in inflation credibility that we have carefully built up etc etc. But their arguments can be easily countered. If CPI inflation does rise, this helps get round the ZLB, and so should stimulate a recovery. Yet it cannot ‘open the floodgates’, because all that has happened is that the definition of the inflation target has been changed. There is no further change in definition that can lead to yet higher inflation. In the longer term 4% wage inflation is pretty compatible with 2% price inflation, unless you believe that UK productivity growth can never grow again at levels that were thought normal before the recession. Finally I do not think there is anything in the academic literature which says that the CPI index is a significantly better measure to target than an earnings index.

In an ideal world, would I be suggesting this change? Almost certainly not. Do I think it is better than nominal GDP targeting? Probably not. However, we are not in an ideal world, and we need some action right now. Otherwise there is a great danger that we continue down a road of self-fulfilling pessimism, with all the misery and loss of resources that this entails. 


[1] That wage inflation is currently well behind price inflation plus underlying productivity growth is not just a UK phenomenon - it applies in the US and Eurozone as well. To that extent this proposal is not just of relevance to the UK, although elsewhere central banks have the power to make this change, whereas in the UK they do not.