The Biden-Harris Labor Department seems to exist in the land of make-believe.

Month after month, they released blockbuster jobs reports, only to have the numbers revised down in subsequent releases.

The problem has become so persistent that members of Congress are wondering if something nefarious is going on behind the scenes.

The latest data show that initial estimates of job growth for the year ending in March 2024 were grossly overestimated by 1.2 million—a third of all supposed job growth.

Senate Republicans have had enough of these phony numbers and are demanding answers.

While there’s not yet a smoking gun, some suspect government statisticians are committing lies of omission.

The Labor Department released its preliminary annual benchmark last week, something that’s normally done each year as better data becomes available in the labor market.

As part of normal operating procedure, these figures are used to adjust monthly jobs numbers.

But what makes this year’s benchmark so abnormal is the size of the downward revision: At 818,000, it’s the largest since 2009, when the labor market was bottoming out in the depths of the Great Recession.

At that time, economic conditions were deteriorating so quickly that the Labor Department’s statistical models were no longer producing accurate results.

Assumptions that made perfect sense in 2005 ended up producing erroneous results in 2009—garbage in, garbage out.

The same problem seems to be happening today, and it’s been going on since the spring of 2022.

Again, the problem here is not simply that economic data are being revised; that’s routine.

Rather, the alarm bells should be ringing because the jobs numbers are being so consistently revised down by such large amounts.

This new benchmark data is bad enough with a drop of over 800,000 jobs—but the situation is even worse when you realize that it comes atop the existing downward revisions to the monthly jobs reports.

Soup to nuts, the total overestimation of job growth was 1.2 million in a 12-month period.

Wiping out a third of all job growth over the course of a year is the equivalent of losing four entire months on the calendar.

Even more frustrating is that this was no surprise to those paying attention to the labor market.

In the manufacturing sector, for example, countless private data have been pointing to job losses for over a year, even as the Labor Department has been logging job gains.

It turns out that the sector has in fact been losing jobs, down over 100,000 since the beginning of 2023.

Much of the problem with the wild overestimation of job growth stems from the “birth-death model” the Labor Department uses to estimate employment data.

This statistical model tries to account for firms that are being created and firms that are going out of business month by month—and it’s laughably off the mark.

In brief, the birth-death model has been wildly overestimating the number of firms in the American economy, assuming nonexistent businesses and the jobs that go along with them.

Before the COVID pandemic, the economy supported a robust creation of both firms and jobs, but under current conditions this is no longer true.

The statistical assumptions that made sense in 2019 are worthless today.

But even the robust business creation the Labor Department logged immediately after the COVID lockdowns ended was not representative of reality.

It turns out that was a mirage, associated with widespread fraud as swindlers “created” businesses to get government handouts—but those firms never actually employed people.

By assuming artificially robust business creation today—not reflective of actual economic conditions—the birth-death model is still habitually overestimating job growth.

And the problem remains unaddressed: Year to date, it accounts for 908,000 jobs added in the non-seasonally adjusted data.

Over the last 12 months, the model “added” 1.3 million jobs—more than half of the 2.5 million new jobs the Labor Department has claimed.

To be clear, some of the jobs gains imputed from the birth-death model are indeed real, but most of them are likely a statistical error.

Decision-makers from Wall Street to Congress to the Federal Reserve rely on this data to make weighty calls, so it’s imperative that the numbers be as accurate as possible.

The Labor Department owes Congress, and the American people, a full explanation of what’s wrong with its models and methodologies—as well as why nothing has been done to fix their obvious problems.