Evidence of Employment

Friday, the US Bureau of Labor Statistics (BLS) released February’s Employment Situation report, aka, the unemployment report. In the month, the unemployment rate held steady at 8.3%, matching analysts’ estimates. Private payrolls, however, beat estimates—rising 233,000. Government payrolls continued to fall, this time by -6,000. So there you have it. The labor market continued improving last month, at least according to government data.

Yet many remain skeptical regarding the improvements—noting government employment data have statistical issues. (Some even take it far, far further.) Without doubt, we certainly agree taking government data with a grain of salt is wise. Consider, for example, the unemployment rate can move (up or down) even if there are absolutely zero people hired or fired in a given month. But it isn’t as though BLS data are without value.

BLS unemployment data have a very long history, contain many different statistical points and are widely watched by analysts and investors. In that sense, how these points relate to expectations can be important. Also, the data provide one potential way to assess health of the labor market—a way to see trends. And trends indicated by government data can easily be tested using non-government-sourced data.

Payroll services firm ADP has published a series of payroll data since 2001. Now, that’s perhaps too short to draw many major macroeconomic conclusions on its own (few economic cycles are contained in that timeframe). But it could add color to one’s assessment of government labor market data.

Exhibit 1: ADP Private Payrolls vs. BLS Private Payrolls (Monthly Change, in Thousands)

Sources: ADP, US Bureau of Labor Statistics, Federal Reserve Bank of St. Louis.

As shown, the two tend to track each other closely—and both show relatively similar-sized job gains in recent months.

But one needn’t stop with ADP. The Institute of Supply Management (ISM) conducts monthly surveys regarding employers’ expectations—and one component is hiring. Now, these indexes don’t necessarily show how many jobs were added. But they do show the prevalence of expected hiring by private businesses—painting a picture of expected job supply. And recently, ISM’s reports on both manufacturing and services employment have shown readings above 50—in expected hiring territory.

Exhibit 2: ISM Manufacturing and Non-Manufacturing (Services) Employment Indexes

Sources: Institute of Supply Management, Federal Reserve Bank of St. Louis.

Employers’ hiring expectations may or may not come to fruition. But considering the increased expectations the past few years—and more recently, in America’s large services sector—an increased pace of hiring isn’t very surprising. Moreover, a survey of CFOs reported last week indicates hiring is expected to continue ahead. And it’s quite possible job seekers are beginning to see this—employment website Monster.com’s index of online job postings has shown increased “help wanted” ads for some time.

Exhibit 3: Monster Online Job Posting Index (Year-Over-Year Percent Change)

Source: Monster.com.

But of course, hiring is only one side of the equation—layoffs matter, too. The BLS’s weekly initial claims for unemployment—a frequently cited metric of newly unemployed workers—have fallen significantly in recent months.

Exhibit 4: US Weekly Initial and Continuing Unemployment Claims

Source: US Bureau of Labor Statistics.

The slowing pace of layoffs is echoed by private-sector outplacement firm Challenger, Gray and Christmas’ (CGC) planned layoff data. Both BLS and CGC data show noteworthy improvement.

Exhibit 5: Challenger, Gray and Christmas Quarterly Planned Layoffs

Source: Challenger, Gray and Christmas.

There are no perfect measures to statistically break down unemployment to the person. All—and we mean all—have their flaws. But taking multiple and varied reports in context, it’s relatively easy to see recent employment gains aren’t at all likely to be solely the function of government data manipulation. And that such hiring gains are occurring now—after three years of bull market and ten quarters of US economic growth—is to be expected. So yes, be skeptical of government data. But it’s equally dangerous to merely be skeptical for the sake of being a contrarian. The specific figures matter much less than the trend—and that trend seems clearly pointed toward a healing labor market.