The last monthly report of the labor department on hiring and unemployment includes revisions for the previous months that have to give a more accurate picture of the American labor market – but that can also cause confusion.
If the data is released on Friday, a significant degree of employment will be revised. Another will be revised. Some historical figures will be revised, but others will not. And the updates, although part of a routine process, will take place in a political environment where both parties have sometimes expressed skepticism about the government's economic statistics.
“There will be an enormous amount of confusion,” says Wendy Edelberg, director of the Hamilton Project, an economic policy of the Brookings Institution.
This is what economists say that you need to know about the revisions to understand the figures.
The revisions are part of a long -term annual process.
The monthly task figures are based on two surveys, one of the employers and one of the households. Those surveys are generally reliable – they include a number of interviews that are much larger than a survey of the presidential elections, for example – but they are not perfect. And so once a year the government reconciles the figures with less timely but reliable data from other sources. Similar processes are present for revising other government statistics, such as gross domestic product and personal income.
“Revisions are how statistical agencies reach both timeliness and accuracy,” said Jed Kolko, who supervised the economic statistics at the Commerce Department during the BIDEN administration. “Nearly real-time data such as the job report will be revised later to match other data sources that are more accurate, but take longer to collect and publish.”
The revisions that were released on Friday were planned far in advance and will use methods that were announced in advance, so that economists, including Mr Kolko and Mrs. Edelberg, can publish detailed predictions about what the new figures will show.
That transparency must give people confidence in the revision process, Mr. Kolko said. When statistical agencies start making inexplicable or unannounced updates, he added that “trust in government statistics could undermine.”
Recent job growth will be revised.
In August, the labor department published provisional data, based on records of the Unemployment Offices of the State, which shows that employers added about 818,000 fewer jobs in 2023 and early 2024 than was initially reported. On Friday, the department will release an updated version of those figures and include them in the official baneng data.
The final revision, which influences the totals of the work for each month since March 2023, can be greater or smaller than the provisional estimate. But it will almost certainly be the largest in recent years, possibly the largest downward adjustment since 2009. That will look weaker during the Biden administration than reported before.
However, the updated figures will probably not change the basic story of a solid labor market. They could even make the recent delay in accepting more muted, because they will make the big job profit look like in 2023.
Work estimates will be misleading.
The other series of revisions, if there is something, is even more complicated.
Data on the labor force – including estimates of employment and unemployment – come from a monthly research among households, formally known as the current population survey. The answers in that survey are weighed to match the estimates of the population that are produced annually by the Census Bureau.
But for various reasons, the Census Bureau is struggling to fully explain the increase in immigration in recent years, which underestimates the number of population growth. In December, the agency released new figures with the help of an updated methodology that the experts believe that they record better recent immigration – and which showed much faster population growth in 2023 and 2024.
The job report on Friday is the first to use those new estimates. But in accordance with its earlier practice, the government will not revise any of the historical data on household survey. Instead, the new population numbers will appear as a huge increase of one month in almost every measure based on it: Mr. Kolko estimates that the labor force will add about two million people, and that unemployment totals will also jump.
Of course, the workforce did not really grow with two million people in January. That growth has taken place in recent years. As a result, the figures will not be directly comparable to earlier estimates in January – for example, it will not be possible to say how many people have joined the labor force since the pandemic. (At least not based on official data from the labor department – Mr. Kolko and others are planning to produce their own unofficial revisions of historical figures.)
Fortunately, measures based on relationships – such as the unemployment rate and the participation percentage of the labor force – must usually not be influenced by the population updates. These are the measures on which most economists concentrate anyway.
The revisions will help solve a mystery.
At first glance it may seem strange that one degree of employment will be revised and another will be down. But the new data should actually help to resolve a disagreement between the two sources, which have sent conflicting signals for years.
According to the investigation among employers, the US economy had 7.2 million more jobs in December than on the eve of the Pandemie, in February 2020. However, the research among households shows a profit of only 3.1 million jobs in the same period. The two sources often move in different directions for short periods, but it is rare that an opening is so large or lasts so long.
The revisions on Friday must be a long way to close that gap. Mrs Edelberg estimated that after the updates the discrepancy between the two surveys should shrink from more than four million to less than half a million, more than within a normal level.
They will also dispel a myth.
Recent job reports have shown that employment has grown with immigrants, but who falls under indigenous Americans. But that data is misleading.
The household research has correctly established that immigrants have made a larger part of the workforce in recent years. But the size of that labor force was based on the estimates of the Census Bureau, which underestimated recent population growth. So the data released last year by the labor department have underestimated employment among immigrant and indigenous employees.
Because the labor department will not revise the data from the historical survey about households, it is not possible to see exactly how employment has changed in recent years at native or employees born abroad. But the new figures must make it clear that employment has also grown among indigenous employees, not just immigrants.
