An Economy of Labor and Statistics (Literally)

In this labor economy, we rely on the Bureau of Labor Statistics (BLS) — but it has a terrible name.

Not only is it a bureau, implying that it’s part of the governrnment and full of bureaucrats, but it has the words “labor” and “statistics” included in the name. These two words illicit images of hard work and college classes that no one enjoyed. And the acronym “BLS” leaves everything to be desired.

For a better acronym, the name could be changed to the Bureau of Reconciling Employment (BORE), but it wouldn’t change the facts on the ground — namely, that the title of the office and the work they perform are very different.

This must be true because the main numbers that the BLS publishes every month, the unemployment figures, are not so much based in number-crunching math as they are in fiction.

Luckily for us, when plucking these numbers out of the air, the BLS always errs on the side of growth, which means it must be a place full of optimists, where the sun always shines!

An Economy Blinded by the Sun

The Jobs Report for May 2014 showed 217,000 net jobs added to the U.S. economy. That was in line with expectations and sounds pretty good.

The bulk of the new positions were added in service areas like retail trades, wholesale, transportation, health care assistance, and leisure and hospitality.

While most analysts would rather see more jobs created in the goods-producing sector of the economy, a job is still a job. But that’s not exactly true, because the BLS has two ways of counting jobs. The first is by surveying companies, and the second is by guessing.

Each month, the BLS sends out surveys to businesses asking about their payroll. The intent is to get an understanding of how employment has changed during the preceding month.

However, the BLS doesn’t survey every business. That would be a massive undertaking. Instead, it samples a small portion of existing companies.

The choice of what companies to include isn’t completely random. Given that over 27% of private employees work for companies with more than 1,000 employees, the BLS skews its sampling to the larger firms. Never mind that those companies represent only 0.2% of all U.S. firms.

This makes sense if you’re trying to simply count workers, because it allows you to capture the most workers using the fewest survey results. But that’s not the point of the monthly jobs report. Instead, the BLS is trying to capture the change in employment. While layoffs often happen at big companies, the process of adding workers is skewed to the smallest firms. And that’s where the fight starts.

While the BLS surveys more than 60% of the companies with over 1,000 employees, the agency surveys fewer than 3% of the companies with less than 50 employees, even though more than 30% of employees work for such firms.

The issue, of course, is that there are so many small companies. Trying to survey the bulk of them would be inefficient.

So the BLS makes up for the small sample size by adjusting the weight it gives each small company in its survey, but even this doesn’t completely do the trick. With such a small sample size of small firms, the BLS understands that it can miss important hiring data.

To compensate, the BLS uses historical data to guesstimate how many new employees were hired by small firms, how many new companies were formed, and how many went out of business.

This total adjustment is called the Birth/Death Adjustment, and is added to the Non-Farm Payroll numbers each month.

Keep in mind that these are not jobs the BLS has counted, or knows exist, they are simply guessed at.

If the BLS estimates that more people went to work at small firms or started new companies than were let go, then the Birth/Death Adjustment is positive. But theoretically, the number could go either way.

While there have been a few months here and there where the BLS showed a negative Birth/Death Adjustment, in the overwhelming number of months, the adjustment is positive.

The latest employment report is no different, and is a great example of all the BLS’s guesswork.

During the month, 217,000 net new jobs were created, which sounds great! But a quick look at the Birth/Death Adjustment raises questions. In this guessed-at category, 207,000 jobs were created. This means the BLS actually counted only 10,000 net new jobs, but estimated that uncounted new positions at small businesses added the rest.

However, the guesstimates don’t simply carry into the future. Every January, the BLS matches up its work with that of the Census Bureau, taking a hard count of how many people are working in the population.

If the BLS has undercounted gains in employment during the previous year, it will add more workers to the January Non-Seasonally Adjusted numbers. If it over counted employees during the previous year, it will subtract workers from the January Non-Seasonally Adjusted numbers.

In January of 2014, the BLS adjusted down the numbers by 307,000.

This negative adjustment, which is used to unwind overly-optimistic employment numbers from the previous year, is the norm. From 2010 through 2013, the adjustments were -427,000, -339,000, -367,000 and -314,000.

Notice a trend? Every single year the BLS had to unwind some of its over-the-top optimism about the number of jobs created.

Of course, these after-the-fact corrections that happen each January don’t make it into the headline number, even though the original, erroneous estimates are plugged right into the headline each month.

It would stand to reason that if your entire job was estimating employment in the U.S. each month, you’d want to be as accurate as possible, with very little correction necessary each January.

As such, it would make sense that the error (because there will always be some, of course) would fall on both sides of the zero line, requiring a positive correction in some years and a negative correction in others.

But not at the BLS. In that organization it appears that the goal is to do what you want with estimates, as long as the effect is positive. If it has to be backed out in a future month, no worries! Most people don’t pay attention to those changes down the road. I guess in that sense, the BLS could change its name to the Font of Useless Numbers, or FUN.

At the end of the day, the problem is that people use employment numbers to gauge the health of the nation, both in a snapshot and as a measure of trend.

If we’re constantly fed rosy numbers that must be adjusted lower each year, then we’re consistently being misled about the health of the nation.

For the millions of unemployed and underemployed, I’m sure none of this comes as a surprise.

For investors and business owners, reports like this from the BLS should be viewed with a healthy dose of skepticism. You wouldn’t want to bet your portfolio or your business on guessed-at numbers.

Another report that should be viewed with skepticism is the Consumer Price Index from the BLS. The data from this index report, on the other hand, is broken down in a way that provides realistic information that informs, not misleads. We’re after real numbers, not guesstimated ones.



Follow me on Twitter @RJHSDent

Rodney Johnson

Rodney’s investment focus tends to be geared towards trends that have great disruptive potential but are only beginning to catch on to main-stream adapters. Trends that are likely to experience tipping points in the next 5 years. His work with Harry Dent – studying how people spend their money as they go through predictable stages of life and how that spending drives our economy – helps he and his subscribers to invest successfully in any market.