Lies, damned lies, and year-on-year inflation rates

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

We’re coming up to a series of important central bank meetings: the European Central Bank announces its latest decisions today, the Federal Reserve and the Bank of England next week. They matter because everyone can feel we are coming up to a pivot point, and when that happens could determine everything from market behaviour to whether our economies deliver growth rather than stagnation this year. Today’s column consists of a warning about misreading the current inflation data: disinflation has come a lot longer than what many headlines might lead you to believe.

Before that, though, let me bug you once more about our wonderful charity for financial literacy and inclusion — do check out the charity auction where just a few days remain to bid for lunch with yours truly, or one of my amazing colleagues.

Over recent weeks, the most noticed inflation news in the US, eurozone and the UK has been that after falling consistently for many months, headline inflation rates have ticked up, sometimes going against expectations for further falls in the rate.

This has played into the fear that disinflation has stalled, or its slight variant, and that there remains a lot of hard work to do before our latest inflationary episode is safely behind us. As my colleague Chris Giles described last month, this is the debate on whether the “last mile” really is the hardest.

Worries about a premature end to disinflation dovetail with the argument of many central bankers that they cannot let up their vigilance. The name of the game of monetary policymakers is now to postpone any celebration of victory. Both ECB president Christine Lagarde and Dutch central bank head Klaas Knot have been trying to discourage observers from thinking that the fight against inflation has been won. “Once bitten, twice shy” is how one of my colleague Martin Arnold’s sources describes the ECB. Similarly, one Fed governor insisted last week the US central bank would “take our time”. And the BoE’s Andrew Bailey said in December there was “still some way to go”.

But should we really be worried? It’s worth taking a slightly more detailed look at the alleged wobble in disinflation, beyond the uptick in headline year-on-year inflation in these three economies. Note that core inflation (excluding energy and some other volatile prices) did not pick up in the UK, unlike the headline figure:

and it continued to fall in the eurozone:

In the US, overall inflation as measured by the consumer price index has stabilised at just over 3 per cent year on year ever since last July — but the core CPI has kept falling:

Besides, if you look at the inflation measure the Fed is actually aiming to bring down, the personal consumption expenditures index (only available up to November at the time of writing), both the headline and the core version have kept nicely falling from month to the next:

There is a simple common point to observe across all of these measures: while the December uptick in some year-on-year rates is apt to trigger media headlines, those inflation measures more reflective of sustained price dynamics contain no sign whatsoever that the disinflation that’s been at work for more than a year in Europe and a year and a half in the US is fizzling out.

And, in fact, even the headline rates don’t tell the story you might think.

There is a mechanical aspect to year-on-year inflation measures — that is to say, the percentage by which this month’s price level is above that of the same month a year ago — that creates a risk of bias towards being too worried about persistent inflation. The fact that central banks all target a year-on-year inflation rate leads to an asymmetry between inflation and disinflation that isn’t sufficiently taken into account in our debate. At worst, it can fuel a bias towards keeping interest rates too high for too long.

The asymmetry is this. Suppose you start from completely stable prices (or constant inflation at the 2 per cent target). Then something causes prices to accelerate — say a megalomaniac dictator invading a neighbouring country or cutting off natural gas exports. The year-on-year inflation rate will go up immediately, reflecting the ongoing inflationary pressures. But the day prices stop rising (or stop rising faster than the target rate), inflation will remain above target for a full year, simply because year-on-year inflation measures the changes that have happened in all of the past 12 months and not just what is happening now. Our most used inflation measure will pick up inflationary pressure immediately but only recognise the end of disinflation one year late. The chart below represents a stylised version of this phenomenon.

The graph shows a situation where prices rise faster than the normal 2 per cent annualised rate for 18 months, indicated by the shaded area, leaving prices about 18 per cent above where they would otherwise be. You can see that the year-on-year inflation picks up this inflationary pressure immediately, but only gradually returns to 2 per cent a full year after disinflation is complete (disinflation back to 2 per cent happens right after the 18 months in this constructed example).

The point four-to-five months to the right of the shaded area in that graph essentially illustrates where we are today. In the UK, the consumer price index is basically unchanged since September. In the US, the PCE index is actually lower (in November, the last available reading) than in September. So is the eurozone price index targeted by the ECB. And just for laughs, why not look at US prices with the index the ECB uses (the “harmonised index of consumer prices”): on that measure, prices have been outright falling and are down almost 1 per cent since September.

Don’t make too much of this. Things could still change, and the devil is in the detail of how you define specific inflation measures. But understand what it means in terms of judging incoming data. The point is that there is no observable price inflation right now. The arithmetic of year-on-year inflation means the current above-target numbers do not reflect any ongoing price rises; they simply capture inflationary pressures that were at work more than about five months ago. So, for all we can see, the inflationary episode ended many months ago. As Paul Krugman puts it, there is no last mile.

Central bankers know this, of course. It’s basic arithmetic. But you will not have heard any of them say in public that the required disinflation was complete by last autumn. The honest thing to say — if you really want to keep interest rates high — would be that the job is done but could come undone, so it’s better to keep rates high out of an abundance of caution. The reason to do this, but not to say it, is surely that it’s hard to defend in public — and central bankers’ fear of having to reverse course. That may be a sensible plan to avoid losing face. But it’s not an encouraging basis for public policy.

Other readables

• EU business leaders would like to see more common infrastructure spending at EU level, it emerges from my colleague Peggy Hollinger’s interview with the head of the European Round Table for Industry. As I wrote last week, that is where a “grand bargain” on the next EU budget could lie.

• A new study shows how European politics is not just divided between left and right, but by which crisis matters most in different countries.

• It’s a Free Lunch article of faith of sorts that there are lots of unused opportunities for greater productivity out there, which are not exploited when it is too cheap to use a lot of labour rather than sophisticated machinery (I have written about this as the car wash parable). Today’s exhibit: a nice Wall Street Journal video about how modular construction needs many fewer hands and makes for faster building. Also, note Leo Lewis’s note on how high-productivity Japanese housebuilders are muscling in on the US market.

Numbers news