Two Centuries of Data, One Clear Lesson

Seeing the Economy More Clearly

Perhaps you’ve seen a series of photos taken of Pluto over different decades. At first, the planet appears as a blurry dot. Over time, as telescopes and technology improved, those fuzzy images gave way to sharper, more detailed views.

Did Pluto change? Of course not. It has been the same for billions of years. What changed was our ability to see it more clearly, thanks to better technology and the work of generations of astronomers building on each other’s discoveries.

The same idea applies to the economy. The way it functions at it’s core has always been there as it is driven by human behavior. But our ability to measure, track, and understand it has improved as data collection methods and technology advanced.

Which is a reminder to us not to confuse better visibility with change.

A Brief History of U.S. Economic Data

Over two centuries, U.S. economic statistics have evolved from sparse questions in the 1800s into the most comprehensive, transparent data system in the world.

• 1810: The first economic questions were added to the U.S. Census, asking about manufacturers and goods.

• 19th Century: Over time, questions expanded to include mining (1840), commerce, and industry. Though data was only collected every 10 years as part of the general population census.

• Early 20th Century: Economic data was separated from population data in 1905 in the Census of Manufacturers. This new standalone economic census instead looked solely at manufacturing, retail, wholesale, and services. In 1915, the Bureau of Labor Statistics (BLS) launched with surveys on employment and wages, later expanding to cover more industries with nationwide samples.

• 1920s–1930s: New surveys and reports, such as the Survey of Current Business (1921), set the stage for national income accounting. During the Great Depression, Simon Kuznets developed the first national income estimates. This later evolved into today’s GDP.

• Post-World War II: Economic data became more systematic. The modern Economic Census, conducted every five years since 1967 (and intermittently before), now covers nearly every corner of business activity. Regular monthly and quarterly indicators such as CPI, employment, and GDP, were developed to give timely snapshots of the economy.

💡 Did You Know?

The U.S. collects more economic data than any country in the world. That’s why global investors rely on U.S. releases, even if they’re about U.S. jobs, inflation, or growth.

Why Do Data Revisions Happen?

Revisions are a normal and essential part of producing reliable statistics. Early releases aim to be timely, while later updates make them more accurate. Common reasons for revisions include:

• Late responses: Businesses and households often report after the initial release, adding new information.

• Benchmarking: Estimates are adjusted against more complete records, such as tax data or the Economic Census.

• Seasonal adjustments: Models are updated as more data becomes available.

• Methodological improvements: Agencies refine how they collect and process information to keep pace with changes in the economy.

Revisions aren’t signs of manipulation. They’re the result of continually improving the accuracy of a vast and complex $30 Trillion economic system. Markets and investors rely on both the early estimates for speed and the revised figures for accuracy and long-term planning.

Why So Many Downward Revisions Recently?

Recent downward revisions in jobs data reflect both the economy itself and the way surveys work:

• Lower survey response rates: Smaller firms often respond later, and their data has shown weaker hiring.

• Economic slowdowns: Weakness in small businesses or sensitive sectors may be missed in early estimates.

• Statistical modeling limits: Models like the “birth/death” adjustment can overestimate job creation during downturns.

• Benchmark corrections: More complete tax and payroll records later reveal earlier job gains were overstated.

A pattern of repeated downward revisions usually signals a softening economy, not faulty data.

Understanding Jobs Data Revisions

• Most revisions are small. Typically under 0.1% of the workforce. Even the recent large revision of 258,000 jobs equals only about 0.2% of U.S. full-time employment.

• Accuracy has improved. Over the last two decades, percentage revisions have generally gotten smaller thanks to better methods.

• Bigger revisions happen during shocks. Periods like the Great Recession or sudden disruptions naturally lead to larger adjustments as conditions shift quickly.

While revisions can make headlines, they rarely change the big picture of employment trends.

Ultimately, consistent downward revisions are often an early warning sign of a softening labor market.

The Bigger Lesson

As of mid-2025, the U.S. has about 135 million full-time workers. In that context, even a quarter-million job revision is a minor adjustment. What matters isn’t perfection in the data, but the continuous effort to refine it.

Economic data will never be flawless, but it doesn’t have to be. The U.S. data collection system is the most advanced of its kind, and its strength lies in constant improvement, not in pretending it is infallible. In an economy this size, a revision of the scale seen over the two month period of this past May and June should not derail a solid economic analysis. Especially when the broader trend already showed the economy slowing.

Throwing out the data because it isn’t perfect is like refusing to use a map because it doesn’t show you where there is road construction or where the potholes are located. The map will get you from A to B, but you still need some skill while navigating.

Revisions and Politics

Here’s where things get messy: politics. During the Biden presidency, downward revisions to jobs data were regularly labeled as “fake” numbers. Critics argued they were manipulated for political gain. The accusation was that the BLS released stronger numbers at first to boost sentiment, then released the revised lower number later to help Biden’s job approval which in turn was bad for Trump. Now, during the Trump administration, the exact same pattern of stronger initial jobs data which is later revised downward is happening again, and yet that same pattern is also being cast as a political hit job against Donald Trump while he is President too. Somehow the same scenario hurt Trump when he was out of office and hurts him now that he is in office.

It’s the same damn thing! So, calling it politically biased is ludicrous. If the exact same revisions happened under Biden and were said to “help” Biden, but now the same pattern is happening under Trump and somehow that’s also “bad” for Trump, then the argument collapses under its own weight. Both can’t be true. This attempted argument defies all logic.

The truth is less dramatic. This is simply the process of refining estimates in a $30.5 trillion economy with over 30 million businesses and about 135 million employees. It only appears political if you don’t follow how the system works. For those who do, revisions are expected and a necessary part of the system.

Is the data perfect? No, and it never will be. But it doesn’t have to be. The U.S. produces the most comprehensive economic data in the world, continuously improved and benchmarked to ensure reliability. Dismissing it as “fake” misunderstands its purpose and how to use it properly.

If the recent labor revisions were a surprise to or blew up a person’s economic model(s), then their models were seriously insufficient to begin with.

What About the New Head of the BLS?

President Trump has nominated E.J. Antoni to lead the U.S. Bureau of Labor Statistics (BLS) after firing Erika McEntarfer in August 2025. Her dismissal came right after a weak jobs report, with Trump accusing (without evidence) the agency of manipulating data.

Erika McEntarfer’s Record

McEntarfer was confirmed in January 2024 with overwhelming bipartisan support (86–8), including votes from Republicans like JD Vance (now Vice President) and Marco Rubio (now Secretary of State). She built a 20-year career at agencies such as the Census Bureau and Treasury, serving under both Republican and Democratic administrations.

Colleagues consistently described her as objective, technically rigorous, and known for an “impeccable reputation” in avoiding politics. Even William Beach, a Trump-appointed BLS commissioner, called her firing “baseless.” Former commissioners from both parties, through the “Friends of the BLS,” endorsed her qualifications and stressed her fit with the agency’s nonpartisan tradition. In short, no one questioned her impartiality until Trump did.

Enter E.J. Antoni

Antoni’s résumé looks very different. With a Ph.D. from Northern Illinois University, he taught at community colleges before joining the conservative Heritage Foundation as chief economist. He has been a frequent presence in right-wing media, a contributor to Project 2025, and his nomination was pushed by Steve Bannon and other MAGA figures.

Unlike past commissioners, Antoni has little experience running a large statistical agency or producing government data. Most of his predecessors came from deep backgrounds in economics, statistics, or academic research. Compared to that standard, Antoni’s record is thin.

What Economists Are Saying

Skepticism about Antoni has come from across the political spectrum, including conservatives:

• Stan Veuger (American Enterprise Institute, conservative think tank): labeled him “utterly unqualified and as partisan as it gets.”

• Justin Wolfers (University of Michigan): argued Antoni lacks credentials even for a junior BLS economist role.

• Alan Cole (Tax Foundation, a right-leaning organization): flagged “stunning” errors in Antoni’s work, including basic mistakes that could not be replicated.

• Jason Furman (Harvard, Obama-era economist who typically refrains from questioning the credentials of political appointees): called Antoni “completely unqualified” and “an extreme partisan.”

The concern is clear: Antoni’s appointment risks politicizing an agency whose authority rests entirely on credibility and impartiality. While the Heritage Foundation praises him as a “truth-teller,” even many conservative economists remain wary.

Antoni’s Plan for the Jobs Report

Antoni has also stirred controversy with his proposals to reshape BLS data releases:

• Suspending monthly reports. He initially suggested halting them in favor of quarterly releases, arguing that survey-based monthly figures are unreliable because they are often revised.

• Reviewing BLS methodology. He wants to rework statistical models to reduce the size of revisions.

• Backtracking after backlash. Following sharp criticism from economists and markets, Antoni walked back his suspension idea. Monthly reports will continue, but he still intends to push for methodological changes and greater emphasis on revisions.

The Debate Over Revisions

It’s true that monthly jobs numbers are revised, sometimes significantly. Antoni points to this as evidence of “flawed” data. Economists counter that revisions are a normal part of the process. Early estimates reflect the best information available at the time, and as more complete data comes in, the picture is refined. Revisions don’t mean the system is broken, they mean it’s working as designed.

Why Economists Are Concerned

Critics warn that Antoni’s approach could destabilize markets and erode trust in U.S. data:

• Markets depend on timeliness. The monthly jobs report is the single most important gauge of labor market health. Without it, the Fed, Wall Street, and businesses would be flying blind.

• Credibility matters. If the U.S. appears to delay or suppress economic data, global confidence in American statistics could collapse.

• Policy risks. Imperfect but timely data is better than waiting months for “perfect” numbers that arrive too late to inform decisions.

The Two Big Issues That Need Fixing

There are two key problems the BLS needs to solve to make our jobs data more reliable:

1. Update the “Birth-Death” Model
The Bureau of Labor Statistics (BLS) relies on something called the “birth-death” model. Since its survey can’t directly capture every new business opening or closure, this model uses historical patterns to estimate how many jobs are added when new businesses start and how many are lost when businesses shut down.

That approach usually works reasonably well. But in the post-pandemic years, we saw an unusual surge in new business filings. Many of these were smaller, more fragile ventures (side hustles or short-lived startups) much more prone to failure. Because the model assumes survival rates closer to historical norms, it has tended to overstate how many jobs are being created. In other words, the model isn’t fully adjusted to this new reality.

2. Reverse Falling Response Rates
Business participation in BLS surveys has collapsed, from about 64% a decade ago to just 42% today. Fewer responses mean heavier reliance on assumptions, which weakens accuracy.

Fixing these two issues would go a long way toward strengthening the system. Throwing out decades of progress because some claim “the whole system is broken” is not smart. It’s economic extremism. And when people abandon facts for extremes, history shows the outcome is never freedom or prosperity.

The Bottom Line

The United States currently has the most advanced system of economic data collection and reporting in the history of the world. This constant measuring and transparent release of information has been a major reason why the U.S. continues to serve as the global economic engine. Thanks to this system, anyone can track the government’s handling of the economy in near real-time. Something no other generation has ever enjoyed. And best of all, it’s both free and accessible.

But accessibility doesn’t mean simplicity. Properly interpreting economic data requires years of observation, practice, trial and error, and a framework for modeling how the economy works. The danger lies in dismissing the data, or “throwing the baby out with the bathwater,” just because amateurs without genuine curiosity or training misunderstand it. When people are easily misled by numbers they don’t fully grasp, it risks eroding trust in a system that is central to America’s strength. After all, one of the reasons the U.S. attracts so much global capital is because investors know they can rely on robust, regular, and transparent economic data. Undermine that, and you undermine one of our greatest advantages.

Antoni insists that less frequent but more “reliable” reports will restore trust. Most economists believe the opposite, stating that sidelining monthly data would undercut the credibility of the BLS itself. The fight here isn’t just about statistics. It’s about whether the United States will maintain its tradition of timely, transparent, and trusted data, or allow politics to weaken one of the world’s most respected statistical agencies.

Antoni’s leadership, should he be confirmed, will ultimately be judged not by rhetoric, but by whether the numbers he generates through new collection and modeling systems continue to align with historical patterns and past credit cycles. Because in the end, the truth is always in the data.

I wish him luck. But based on his track record and statements so far, expectations should remain tempered.

Click the Leave a comment button if you have any questions or comments, or need something clarified. Don’t be shy. The main point here is to improve constantly. Questions and comments help us both and tells me what you are interested in learning/hearing more about.

If you enjoyed this post or found it useful, do me a favor and hit the like (heart button all the way back to the top of the post and to the left) and share it with others.

Learn. Improve. Pass on.