Accurate economic data is essential for everyone from central bankers to business owners and job seekers. But what happens when the numbers we rely on are wrong? That’s the growing concern about the Bureau of Labor Statistics (BLS) and other government agencies that track employment, prices, and other economic indicators. In recent years, cracks have appeared in the system, leading many economists to question the reliability of key data reports.

A Bigger Problem Is Revealed

The BLS is the U.S. government agency responsible for collecting, analyzing, and reporting data on employment, wages, inflation, and other key economic indicators. In a recent and widely discussed revision, the BLS reported that 258,000 fewer jobs were created in May and June than initially announced. While some surprise is normal in economic forecasting, this correction was unusually large. Analysts say it could be a sign that something is fundamentally broken.

Bloomberg’s chief U.S. economist called the error a three-standard-deviation event, meaning it had a probability of only 0.2% over the past three decades. Errors of this magnitude can have real consequences, such as influencing decisions at the Federal Reserve about interest rates or misleading businesses that rely on employment trends to plan their strategies.

Survey Quality

The deeper issue is not just about one report, it's about the system itself. The quality of U.S. economic data has been declining for years, especially since the COVID-19 pandemic, according to economists. Before the pandemic, BLS survey response rates were around 64%. Now they’ve dropped to about 42%. Lower response rates make it harder to paint an accurate picture of the economy.

One reason for the drop is that BLS staff used to collect data in person or by phone. During the pandemic, much of that shifted online, where people are less likely to respond. Offices have closed, staff have been cut, and the agency now relies more heavily on estimation techniques like imputation (i.e., filling in missing data using numbers from other regions or similar products). In about a third of cases, BLS is now imputing data from different geographic areas entirely.

A Changing Economy

Beyond staffing and budget issues, the U.S. economy itself has become harder to measure. Jobs today often look very different than they did a few decades ago. Traditional employment measures worked well for large, centralized companies like traditional retail chains. But in a world of gig work, remote jobs, and rapid hiring and firing, traditional surveys may no longer capture the full picture. Shifts in communication technology have made things more difficult too. Landlines once made random sampling more reliable. Now, reaching people online often means using less personal, less accurate methods, and it’s unclear how these changes affect the responses.

One example of this is the University of Michigan’s consumer sentiment survey. When the survey switched from phone to online responses in 2023, inflation expectations reported by participants jumped sharply, even though actual inflation was falling at the time. This raised new questions about honesty and how well-informed respondents really are. Even more troubling, people’s economic expectations now seem closely tied to their political views. Some economists worry that partisan bias is creeping into data collection, not because of any flaw in the methodology, but because respondents are answering based on what they believe rather than what they observe.

A Call for Reform

As trust in official statistics declines, private alternatives are gaining attention. Companies and analysts are increasingly turning to unconventional data sources: real-time energy usage, commodity prices, transportation volumes, or even Google Maps mobility trends. Some platforms, like Truflation and Shadow Government Statistics, attempt to estimate inflation and jobs numbers independently.

More computing power and artificial intelligence have also made it easier to mine vast datasets for economic insights. This helps investors and businesses get faster, more customized forecasts but also risks creating confusion when different sources show different results.

Many economists agree that while alternative data is useful, the BLS and similar agencies still play a critical role. Public statistics are a common good as businesses, consumers, and policymakers all benefit when high-quality, unbiased data is available. But for that to continue, the BLS would need serious investment: more staff, better technology, and stronger connections to the private sector.

There’s also a growing need to publish uncertainty estimates alongside major reports. In other fields, such as climate science and polling, it’s common to acknowledge a margin of error. Doing the same in economic statistics could help rebuild trust and encourage more thoughtful interpretation.

The Cost of Getting It Wrong

Accurate data is essential for measuring inflation, employment, and GDP. In periods of economic uncertainty, reliable information becomes even more critical. Improving data collection and analysis is a complex process, but it plays a key role in supporting sound decision-making across the economy.

In the Classroom

This article can be used to discuss economics (Chapter 1: The Dynamics of Business and Economics).

Discussion Questions

  1. What factors have contributed to the decline in response rates for government economic surveys, and why does this matter?
  2. What challenges does the shift to online data collection present, and how might it affect the reliability of survey responses?
  3. Do you believe alternative data sources (like Google mobility trends or Truflation) can fully replace government statistics? Why or why not?

This article was developed with the support of Kelsey Reddick for and under the direction of O.C. Ferrell, Linda Ferrell, and Geoff Hirt.