Looking Beyond the Data Debate

Editors’ Note: MarketMinder is politically agnostic. We assess developments for their economic and market implications only.

Three days after its release, the official jobs report continues stealing headlines—not for what the data show about the US economy, but for the fallout. Namely, President Trump’s decision to fire the Bureau of Labor Statistics’ (BLS) head. We won’t get into that, nor any of the associated politics. But we have seen some talk about potential data quality issues in the Employment Situation Report affecting markets. With this, we can help.

There are generally two big narratives about potentially flawed BLS data now. One posits that data have been politically manipulated and, hopefully with new leadership, they won’t be. The other posits that data haven’t been politically manipulated but, unfortunately with new leadership, they could be.

Whether you fall into one of these camps—or whether you are in the more neutral camp that sees declining survey response rates as responsible for large BLS data revisions and doubts that issue will change under new leadership—there is an antidote. Simply, the BLS isn’t the only employment data game in town. It hasn’t gotten much ink lately, but payroll processor ADP releases its own jobs report.

Like the BLS’s Establishment Survey, ADP reports nonfarm payrolls—specifically, private sector nonfarm payrolls. But where the BLS relies on surveys—surveys with increasingly low initial response rates—ADP uses its own records to give a monthly snapshot of payroll changes.

As Exhibit 1 shows, ADP and BLS payroll data don’t always match. After all, ADP doesn’t do every private business’s payroll. And there are other variances, as you would expect when data collection methods and seasonal adjustment factors differ. But generally, they show the same trends. Sometimes, ADP’s estimate of private-sector jobs growth is a little rosier than the BLS’s. Sometimes, the BLS’s is a little sunnier. In the highly debated most recent three months, ADP’s data were gloomier than the BLS’s revised data in May and June but a little happier in July. We are just stating facts here—not offering comment.

Exhibit 1: ADP and BLS Private Nonfarm Payroll Growth

Source: FactSet and St. Louis Fed, as of 8/4/2025. Monthly change in private sector nonfarm payrolls (ADP and BLS), January 2022 – July 2025.

Relying on any one source or reading for economic data is an error, no matter what that source or reading is. No reading is perfect. One benefit of living in the information age is that we have a wealth of sources and data. Yes, all the official government reports, with all the attendant baggage that comes with them wherever they are produced. But also a smorgasbord of data from private outlets. That includes the business surveys known as Purchasing Managers’ Indexes from the Institute for Supply Management and S&P Global (formerly IHS Markit, if you were more familiar with that branding). Home sales and housing starts from the National Association of Realtors. Consumer spending readings from credit card processors. Business activity and sentiment surveys from a wealth of trade groups. Money supply data from the Center for Financial Stability. In concert, all sketch the very messy picture that is the global economy and its trends.

In doing so, they help you see what markets have already priced and assess how things are going relative to expectations. No data, government or private sector-produced, predicts stocks. Stocks are forward-looking, pricing in likely developments about 3 – 30 months out. All economic indicators are backward-looking. The timeliest may be borderline coincident. But all can reflect only what stocks have already lived through and priced in.

Therefore, if you are wondering how you can navigate the market when economic data are questionable, that isn’t really the point. Stocks always move first. Think back to the monster recession that accompanied the global financial crisis. That bear market began in October 2007, when US economic data still looked pretty good. Data still looked ok throughout much of 2008, some readings up and some down. For most of the year, it wasn’t clear we were even in a recession. We didn’t know definitively that recession began in Q1 2008 until GDP had been revised several times. But not having clear, ironclad data didn’t prevent markets from pricing in an economic downturn.

Similarly, when the next bull market began in March 2009, all the data were really, really, really bad. The US economy was still in recession and would stay that way through June. The job market, always a late-lagging indicator, wouldn’t start recovering until early 2010. But stocks began rising in advance of all that. They didn’t wait for data to show things were ok. The simple prospect that things would be ok was enough.

So yes, every economic indicator is imperfect. All that imperfect aggregates into a broad trend that you can make sense of, which is a great thing! But that trend will always be backward-looking—especially jobs data, which are among the most backward-looking of all. The market is the best gauge going in weighing the economy’s path ahead. Trust it.