Saturday, 31 August 2013

Krugman, Phillips and Carney

A recent post by Paul Krugman defending the Phillips Curve and the unemployment/wage trade-off. I'd never dream of challenging someone so much better at economics than me but I've long believed that this is a defunct trade-off. The labour market is global, surely, and that means that workers have little bargaining power even if unemployment falls. There are exceptions in some industries, of course  but my underlying view is that the world has moved on. Not only that but here in the UK at least, consumer price inflation has been incredibly sticky and resistant to downward pressure from recession. Wage growth and inflation have become increasingly unrelated, or so it seems. 

Professor Krugman is much wiser than me and heaven forbid me from actually disagreeing, so I have spent part of the week playing with data. What I ended up concluding was: 1) The wage/unemployment trade-off in the US is OK but the inflation unemployment trade-off is all but invisible. 2) If you want to see a properly operating Phillips curve, go to Japan. And 3) if you want to see somewhere where the inflation/unemployment trade-off is far, far worse than in the US, come to the UK. Which makes the decision to put inflation and unemployment at the hart of the MPC's monetary policy framework, a little odd. And indeed the risk that we face is that although growth remains weak, and wages remain under downward pressure, inflation will remain high, eroding the credibility of the MPC's 'forward guidance'. 

In the interest of understanding the Krugman piece, I did my best to recreate the chart he uses to show that there is indeed a decent wage/unemployment trade-off in the US, using annual data from 1985 to 2012. That is below.
So far, so good! But if I then plot the same scatter chart for the Fed's favourite inflation measure, the core PCE deflator, I get much less helpful results.... 

I guess there's a relationship of sorts but really? This may not matter to someone who is defending the Philips curve, but the Fed is targeting the core PCE deflator, and so that is to a large degree what matters to policy. 

There are all sorts of reasons why this matters for the US (not least  don't be surprised to see unemployment fall while inflation remains very well behaved) but I really wanted to compare what is happening there to what is happening elsewhere. So look at the same charts for Japan...

Now that's a Phillips curve trade-off. A less internationalised labour market? And bizarrely, this is good news for Japan. In many countries, news that inflation is going up and unemployment falling would be a source of concern but if what you want to do is escape a 20 year deflationary disaster, this is great. Inflation expectations are going up and the so far, Abenomics is delivering. I think monetary reflation can work.

OK, now for the horror show.... here are the same charts for the UK....

So... since 1985 there's been very little relationship between wage growth and unemployment but insofar as one exists, higher unemployment correlates with faster wage growth. There's a far better correlation between inflation and unemployment but unfortunately the sign is wrong.  The upward slope means that low inflation co-exists with low unemployment. This says nothing about causality but it does suggest that the idea of a wage/unemployment trade-off in the UK, is perhaps a bit out of date. Maybe the economy grows faster with low unemployment, or maybe British workers are being forced to compete with workers elsewhere to prevent jobs leaving the country?

The unemployment/inflation trade-off is used as an easier way of looking at a trade-off between inflation and economic slack. We view falling unemployment as a sign that the economy is getting closer to its potential growth rate. Likewise, we look at inflation as a measure of slack, a sub-par economy taking inflation rates lower. But when so much of what drives inflation has nothing to do with the strength of the economy, that doesn't quite work. Transport, education,  water, gas, and electricity prices are all regulated to a greater or lesser degree. Rail fares and utility prices through a formula agreed with the government in order to encourage much-needed investment. The price rises to pay for cross-rail, a new generation of nuclear reactors, or repair to Victorian water and rail network, have nothing to do with the strength of the economy. Indeed, when the economy is struggling, any suppliers of services that are subsidised by the government (like local authorities), see their subsidy fall and have to respond by pushing price up faster.

I wish the MPC hadn't put inflation and unemployment explicitly into the mix for their forward guidance. I know they have left themselves so much wiggle room that they can ignore persistently high inflation, falling unemployment and focus on soft growth, but this just looks like a  policy import that makes no sense over here.

I hope to find find time to write about this, a cracking argument for capital controls from a London Business School professor, tomorrow evening....

No comments:

Post a Comment