US economy

It’s still wrong to panic on inflation

This article is an on-site version of Martin Sandbu’s Free Lunch newsletter. Sign up here to get the newsletter sent straight to your inbox every Thursday

Last week, US inflation numbers for December were released. They showed year-on-year inflation at 7 per cent, the highest rate in four decades. This caused much alarm. Much of it is based on an ill-conceived reading of the data, for three reasons.

Chart showing annual change in CPI

First, please ignore year-on-year measures. Whatever is driving price dynamics, these forces are liable to change fast and unpredictably.

We know this because they have been changing fast and unpredictably already. Take this example: last February, consumer prices were a subdued 1.7 per cent higher than a year earlier. That was no good guide to the big jumps in prices in the months that immediately followed, which drove the year-on-year rate above 5 per cent (and will keep it there for some time even if monthly inflation comes right back down this month).

It would have been wiser to look at the monthly rate of change (mea culpa: I did not). It was the fourth consecutive monthly price acceleration, showing a rapid change in inflation pressures, and it hit a rate of change that, if sustained, would on its own eventually push yearly inflation above 4 per cent.

Because of this, the head-turning inflation data should not have been December’s, but that of September and October, when monthly inflation soared back up, high above sustainable levels, after decelerating from the early summer peak (see chart below). After that, high year-on-year numbers were arithmetically baked into the December release, and anyone professing to be shocked by the December numbers without pointing this out is not enlightening you (nor perhaps themselves).

Line chart of Seasonally adjusted Consumer Price Index (all urban consumers, all items in US city average), month-on-month percentage change showing Monthly inflation falling again

Second, what do we learn from the December release that is new?

Most importantly, that overall inflation has been moderating fast. The (seasonally adjusted) monthly rate has fallen by half in the past two months. The same pattern holds for all the main subcategories — including food prices, and energy inflation has even turned negative — but one. The exception is commodities excluding energy and food, for which prices are growing at a stable but high rate.

In particular, there is no pick-up in broad services inflation, which has also halved in the past two months. Non-energy services price inflation is stable; energy services price inflation is negative.

Those who, like me, expect inflation to fall by itself do so because we do not think it is caused by excessive aggregate spending, but by the enormous and unprecedented shift of spending away from services towards goods. Nothing in the December numbers gives a reason to give up that belief. Those who fret about inflation are on the lookout for price pressures spreading from goods to services. There is not much to support that fear either.

If we think that people will eventually allocate their spending between goods and services more or less like they did before the pandemic (or that global capitalism will soon enough adjust to any modest, permanent shift in their shares), then we should expect goods price inflation — which is all there is at the moment — to go away too.

The important data release to wait for, then, is not for inflation but for gross domestic product: on January 27 we will get a sense of whether growth in the fourth quarter continued the tentative trend away from spending on goods and back towards services, which could be gleaned in the third quarter. As I have said before, reasonable disagreement on inflation must logically come down to different views on whether this will happen.

A colleague has pointed out that 0.3 per cent month-on-month services inflation could still make for 3.7 per cent price growth over a year if sustained. That is true — but it is right on par for services inflation. For decades, overall inflation near 2 per cent has meant services prices rising at 3 per cent or more, and goods prices falling enough to bring the average down.

Chart showing US consumer price index by sector from 2015 to 2021

My colleague Rana Foroohar made an important point in her column this week: if durable goods prices stop falling like they used to, this averaging will no longer work. But why should they? The factors Rana mentions — reshoring, 3D-printed manufacturing and decarbonisation — could shift up the cost of goods. But apart from a one-off shift, why should these reduce productivity growth in how we make goods? 3D printing, in particular, should boost productivity. In any case, once consumption patterns settle, so should normal sectoral inflation rates.

Third, the fear of rising inflation is usually expressed as a fear of a wage-price spiral. So another string of US data releases — of wage data, showing average hourly earnings rising about 6 per cent year on year — have got people worried. Here too, things look different with the monthly changes — overall wage growth has come down sharply. But it is true that wages for ordinary workers (“production and nonsupervisory employees”) keep rising strongly.

The fear is that as businesses face rising costs, they jack up prices. But wage pressure is not the same as cost pressure. The best measure of how much labour compensation is costing businesses is instead the Employment Cost Index, which includes non-wage compensation costs (such as benefits and health insurance). That did rise strongly in the third quarter, the latest available (look out for Q4 on January 28) — by 1.3 per cent in three months. Year on year the figure was 3.7 per cent.

But we should note that this comes after a dip in compensation growth early in the pandemic: on average, employees only cost employers 6.2 per cent more than two years earlier. It also comes after decades of employee compensation failing to match productivity growth. To judge whether growth in employment costs constitute “pressures”, we should compare it with labour productivity. And if we merely match the (weak) growth in real output per hour worked of the past decade, which averaged 1.1 per cent a year, employers can, with no real cost pressure, absorb yearly labour compensation growth of 3.1 in a world of 2 per cent inflation. That is a conservative estimate; it ignores both that compensation has lagged behind productivity growth — so there is room to catch up — and that a high-pressure economy can help lift productivity growth.

All this reasoning is based on the state of the US economy. Other economies have seen rising inflation but for different causes. In particular, Europe has not experienced the same extraordinary swing from services to goods spending, as OECD chief economist Laurence Boone has just highlighted in a recent blog post (see her chart below).

Very strong demand for goods, in particular durable, in the US

There are other sources of inflation — Europe’s energy squeeze, above all — but there are clear signs that policymakers all over the world are now grappling with a goods price inflation very much made in America.

Other readables

Numbers news

  • My FT colleagues have put together a marvellous postcode tool where you can see how your very own high street fared in the UK’s pandemic.

  • UK inflation, too, is at a record high.

Unhedged — Robert Armstrong dissects the most important market trends and discusses how Wall Street’s best minds respond to them. Sign up here

City Bulletin — Our pre-market update and commentary. Sign up here


Leave a Reply

This website uses cookies. By continuing to use this site, you accept our use of cookies.