Like a prophecy from the Delphic oracle the US jobs report for January was ambiguous enough to fit whatever an economist wanted to believe about the economy.

Analysts had to make two decisions about a perplexing report. First, they could focus on the survey of employers which pointed in a different direction to the survey of households used to derive the unemployment rate. Second, they could choose which of a range of seasonal and statistical effects they wanted to adjust for.

Markets normally pay most attention to the employment report which, this month, reported the creation of only 36,000 jobs. That is down from 121,000 in December and far below the 200,000 or so that is needed to keep up with population growth and steadily bring down unemployment.

But the employment report appeared to be particularly badly affected by blizzards that engulfed the US during the week of the survey in mid-January. Airports such as Atlanta were closed; 886,000 workers said that they could not show up for work because of the weather.

Analysts at UBS argued that bad weather “subtracted somewhere around 150k from the change in January payrolls” which would have transformed the number from weak to strong job creation of nearly 200,000. But other economists, such as Steven Ricchiuto at Mizuho Securities, argued that the numbers were compiled consistently and needed to be taken at face value.

The household survey, which at first glance showed a very strong improvement with the unemployment rate dropping to 9 per cent, offered its own set of conundrums. The fall appeared to reflect a large drop in the number of people in the labour force, but in January the Bureau of Labor Statistics updated its figures for the size of the population.

“The labour force, at 153.2m, was unchanged after adjustment for updated population controls,” said Keith Hall, commissioner of the BLS.

Economists argued about whether it was fair to ignore the statistical change, in which case the unemployment rate really did fall to 9 per cent because of strong growth in jobs, or whether to leave it in.

Stripping it out, Paul Ashworth at Capital Economics in Toronto said that “on that basis the drop in the unemployment rate is very encouraging”. Steve Blitz at ITG Investment Research said that, adjusting for people leaving the labour force, the true rate would be 9.2 per cent – still positive but not as dramatic.

Given all the confusion, a number of economists advised clients to ignore the employment report altogether and look at a wider range of data that paint a more promising picture of the US labour market.

Those data include a gentle decline in new claims for unemployment insurance, low levels of new job losses, and optimistic results from employer surveys. The latest survey of purchasing managers at manufacturers, released this week, said that many more were now creating jobs rather than cutting them.

Deep in the guts of Friday’s employment report there are encouraging signs that the other indicators are telling the truth and job creation is picking up steam.

For example, the number of jobs added in the manufacturing sector in January was 49,000, which is very high by recent standards. There were 27,500 new jobs in retail trade and an extra 31,000 in professional and business services.

Those were offset by a drop of 32,000 in construction and 38,000 in transport. That is exactly what you would expect to see if snow dominated the January figures.

Get alerts on US downturn when a new story is published

Copyright The Financial Times Limited 2018. All rights reserved.

Comments have not been enabled for this article.

Follow the topics in this article