The Bureau of Labor Statistics released its Employment Situation Report for February on March 10, showing a healthy 235,000 gain in payroll employment. Asked what President Trump thought about the numbers, White House press secretary Sean Spicer said, “I talked to the president prior to this, and he said to quote him very clearly,” Spicer said. “They may have been phony in the past, but it’s very real now.”
Many of the reporters present laughed. I cringed.
Why the jobs number matters
Over the years on public television’s Nightly Business Report, I filed countless “numbers” pieces. The monthly employment reports were most closely watched. For better or worse these reports often had an immediate financial market moving impact, making them lead stories for a market driven broadcast.
I cringed because I believe attempts to undermine the credibility of these reports do a great disservice. This is not a partisan matter. I’ve heard politicians from both parties take issue with the reports. But candidate Trump took the criticism to a new level. He once claimed the unemployment rate was really 42 percent at a time the government report pegged it at 5 percent. After assuming office, during his February address to Congress, Trump said, “We must honestly acknowledge the circumstances we inherited…. Ninety-four million Americans are out of the labor force.”
To the extent that people believe these statements, reporters share responsibility. Clarity in reporting is essential so the public understands how these reports are generated. Precision in reporting is essential so they understand what they mean.
Why it’s a lagging indicator
Recruiting, training and acclimating new employees is an expensive process. For that reason, managers are slow to hire and fire. That makes the employment number what economists call a “lagging indicator”. It can take six, nine, even 12 months before changes in demand lead to changes in employment. Few economists would credit the Trump administration for the February hiring. (To be fair, there are also “leading indicators” and the stock market is an important one. Traders are looking ahead and psychology changes with every news item. Clearly the market rally following Trump’s victory is due in part to a belief on Wall Street that his administration will be good for the stock market.)
Understanding the report is hard work. You need to learn how to read the detail. You must understand the nuance and be prepared to fight your instincts and probably your editor so your story reflects the limits of the data, even if that means the loss of a great click-bait headline. If you think it is difficult to make an economic report interesting in print or on the web, try to do it on television.
Why it’s tamperproof
You can always find the latest employment report on the BLS website. The methodology used is excruciatingly transparent. The report is the product of a large team of career professionals, economists, statisticians, computer scientists and field workers. These government workers do their jobs quietly and diligently regardless of the political party of their managers, who often change with each administration. Fudging the numbers would be a career-ending and probably illegal act.
It would also be impossible, requiring a conspiracy involving many people being closely watched by observers.
It would also be detectable. Even if you could tweak the numbers for a month or two, in time you’d get caught. While the BLS survey is large and tries to be all encompassing, by its nature when you drill down to a specific sector or geographical location you lose precision. Reporters need to consider this if they use the numbers to generate stories about a region or industry.
Because of this there is a constant stream of private corporate and academic research projects designed with a narrow focus to study specific markets. These would quickly detect problems with the BLS numbers.
Why it’s not precise
It is not enough for reporters to challenge politicians when they try to discredit the reports. Reporters must also be diligent in making sure their audience understands what the reports mean.
Remember first that this is a survey. The BLS can’t question everybody in the country each month so it questions a sample. That means there is imprecision in the number reported. Special factors can skew a report. Bad weather curtails construction and transportation. A strike at a major manufacturing plant can impact the monthly report. So can a fire putting a refinery out of service.
“Seasonal adjustments” are regularly made to account for predicable factors. Schools layoff teachers in June and hire them back in September. Stores and delivery companies hire extra workers in November the let them go after the holiday shopping season. Statistical methods are employed to smooth out these factors but they are reviewed and revised from time to time.
I’ll have some specific suggestions on reporting the employment report in part two.
• For better or worse, the monthly jobs number is avidly consumed and has an immediate impact on the stock market. As a lagging indicator, it reflects economic events that occurred months in the past.
• The BLS Employment Situation Report is a monthly survey of domestic employment produced by career employees using a completely transparent methodology. There is no evidence that the reports are biased and it seems highly unlikely that they could be influenced for political or any other purpose.
• Reporters need to study the methodology carefully and be diligent in writing to accurately reflect the report contents including the limitations inherent in the data. Keep in mind that the BLS is a survey, with implicit imprecision.