How really works on the office of the office of Labor Statistics and why data revisions are not “a scam” as Trump claims

President Donald Trump said without evidence that the massive revisions of the last report on jobs were a “scam”, accusing one of the main supervisors of government statistics of cooking books in a vendetta against his presidency.

“In my opinion, today’s employment numbers have been faked in order to make the Republicans, and I look bad,” said Trump in a message on Truth Social Friday.

Consequently, he dismissed Dr. Erika Mcentarfer, the commissioner of the Labor Statistics Office.

Trump, in a message published on Truth Social Sunday, incorrectly affirmed that Mcentarfer “had the biggest calculation errors in more than 50 years”. In fact, the revisions that the BLs made to recent job reports were neither historical nor proof of corruption.

Created in 1884, the Bureau of Labor Statistics (BLS) is an organization operated independently within the American Labor Department. The labor secretary, member of the president’s office, has the supervision of the BLS, but he is led by a commissioner appointed to the Senate.

BLS collects data on a number of key economic concerns and regularly produces critical reports. These include data on prices, inflation, productivity, expenses, salary, work injury, employment and unemployment.

More than 2,000 people work for BLS, including a number of professional economists and candidates who regularly contact businesses and employees. Economists analyze data and produce reports for the public and the government.

BLS collects employment data in two separate surveys. The first is partly done with the old -fashioned door. The candidates for the investigation go home throughout the country to ask people their employment status and their demographic information.

A second survey, known as the current survey on employment statistics (CES), is collected from thousands of companies and government agencies using a variety of methods: by telephone, Internet surveys and – for large companies – thanks to an automated data transfer. According to the law, participation in the CES survey is voluntary, but the laws of states in New Mexico, Oregon and South Carolina require that companies produce data for the survey. A law in Puerto Rico also obliges territory companies to subject data to BLS.

Respondents in the CES survey submit monthly employment data, hours and profits for all workers paid to the BLS of their pay files. The data is collected for the pay period which includes the 12th of the month.

BLS economists first modify data to detect processing and ratio errors. If errors are detected, BLS employees contact the company to clarify.

Then, BLS staff prepare the data for a monthly report by estimating the use of America, the hours worked and the gains. To extrapolate data from the whole country, BLS economists add educated assumptions, depending on seasonal hiring trends. BLS also smooths data with calculations called seasonal adjustments to avoid huge points and data reductions each month.

BLS also protects the raw data and its estimates by processing it through statistical tests, so that individuals cannot have access to the data of a particular employer.

Each month, generally the first Friday, the BLS produces its summary of the job situation, known familiarly as the monthly report of American jobs.

The report is generated from the two surveys: the household survey provides demographic data and the unemployment rate. The commercial survey provides data on the remuneration, the hours worked and the number of jobs that the American economy has added or subtracted.

In addition to the information on the employment of this special month, the report also revises the total of the two previous months.

The BLS considers that its initial work numbers are preliminary during their first publication, because some respondents do not report their data on the payroll date. Low responses to the survey can make the report more difficult to estimate. But the BLS continues to collect payroll data as reported, and it revises the data accordingly.

The data is also revised due to seasonal adjustments. If the most complete data is well above or below the preliminary data, the revisions can be exacerbated by seasonal BLS adjustments, which must sometimes be recalculated.

The data is revised several times: in each of the two months following the initial report, then a preliminary annual revision in August and a final annual revision in February.

The July report, published on Friday, included revisions for May and June which were historically important, but they were not unprecedented.

The total mayor of May’s jobs was revised down 19,000, down compared to an initial estimate of 139,000 – a total revision of 120,000 jobs. For the total of June jobs, the BLS said on Friday that the American economy added only 14,000 jobs, down from a preliminary estimate of 147,000 – a revision of 133,000 jobs.

The BLS follows the revisions of 1979 each month, but the BLS introduced a new design of samples based on the probability for revisions in 2003. Between 1979 and 2003, the average monthly revision was 61,000 jobs. Since 2003, the average monthly revision has only been 51,000 slightly more precise jobs.

But you don’t need to look so far to find more revisions than those that BLs reported in May and June. In 2020 and 2021, job figures were everywhere on the map due to the pandemic: revisions in four months in these years were more important than revisions of the report of last month – including the largest revision of 679,000 jobs in March 2020, attributed to responses to the particularly mediocre survey during national locking.

There have also been greater revisions outside the pandemic, including a revision of 143,000 jobs in January 2009.

Trump complained about a preliminary annual revision which was published in August 2024, which showed that the American economy had added 818,000 less jobs in the past year than before. Trump on Friday in a social article of truth called to incorrect this revision a “record”: a revision of 902,000 jobs in 2009 was greater. And the final revision of 2024, published in February, showed that 2024 data were overestimated by 589,000 jobs. The BLS said that the difference between initial and final annual revisions was due to information received in American income statements.

Trump also properly noted on Friday that the BLS revised the initial total total jobs in August and September 2024 by 112,000 positions combined before the presidential election last year. But this revision was not out of the ordinary – several revisions were larger earlier in the year and in previous years. And the number of October jobs, reported a few days before the elections, was the worst month for jobs from the pandemic.

Many companies and government organizations are based on BLS data for decision -making regarding investment, remuneration and hiring decisions. The federal reserve, in particular, is based on BLS data to help guide its monetary policy and its rate set.

The president of the federal reserve, Jerome Powell, said last week that the whole economy was based on solid data.

“The right data does not only help the Fed, they help the government, but also help the private sector,” Powell said at a press conference last week. “It is very difficult to capture in real time the production of an economy of more than $ 20 billion in real time, and the United States has been a leader for 100 years, and we must really continue, in my opinion.”

In June, Powell told Congress that he was concerned about the “trajectory” of weaker data.

In addition to the Fed and the private sector, the BLS notes that its survey data is used by the National Bureau of Economic Research to determine whether the economy is in recession or not. It is used by the Conference Board to produce economic clues that help companies better understand the current economic cycle. And this helps BLS produce several other reports on jobs and the economy throughout the year.

The alternatives to BLS data have proven to be limited. The ADP payroll processing company produces a payable payment of pay which does not capture the hiring of the government and is notoriously outside synchronization with the BLS ratio. His inconsistencies have led economists to largely ignore the report.

Other surveys, such as a dismissal report of the Externalization and Placement Challenger, Gray and Christmas service company, are sometimes instructive but much less robust than the BLS ratio.

The BLS report also has its challenges. For example, the household survey which includes the monthly unemployment rate and demographic data is considered rather volatile due to its smaller sample size and the drop in response rates. But the investigation into business and government employment is widely considered by economists as the gold stallion.

“BLS is the most beautiful statistical agency around the world, its figures are reliable all over the world,” the former commissioner of Labor Statistics, William Beach in Kasie Hunt “State of the Union” said on Sunday.

“However, I believe that the president’s attack on the commissioner and the office undermines this infrastructure, could undermine this long-term confidence,” he added.