The Real Unemployment Story

by: Jim Mosquera


What does the unemployment figure, U3, tell us?

Are people eager to work?

The government figure that can't be fudged.

The unemployment report was recently released and with it comes the usual scorn challenging the veracity of the figure. If you work for the current Presidential administration, you are crowing about the results. If you are unemployed, you probably question why you still are. If you are an employer with an unfilled position, you may tend to agree with the unemployment rate.

The official unemployment rate we all hear about is called U3. The U5 measure includes discouraged and partially attached workers, while the U6 includes those who work part-time purely for economic reasons. As you might surmise, as you move from U3 to U6, the figure gets larger. The peak, non-seasonally adjusted U3 came in January of 2010, some six months after the official end of the recession. This is instructive since U3 peaks tend to come well after a recession's end and bottoms tend to arrive before the start of a recession. For instance, after the early 2000s recession ended, the U3 peak did not arrive for over a year.

The published U3 rate is subject to seasonal adjustments like the one the Bureau of Labor Statistics (BLS) used in January to account for retailer slowdown after the holidays. The most recent seasonal adjustment amounted to 2.165 million jobs with a reported monthly gain of 151,000 positions. The BLS can also make adjustments based on their perceptions of the job market. My intent is not to cast aspersions on BLS methodologies. I would like to illustrate by way of several charts and numbers another way of looking at the unemployment picture.

The first illustration is the unemployment rate or U3. The figure below (recessionary periods shaded), containing 55 years of data, shows how U3 tends to bottom (low unemployment rate) immediately before a declared recession. As I noted earlier, peak unemployment rates tend to come after, sometimes well after, the official end of a recession. Consider a low U3 rate to be a precursor to a recession. The question then becomes, how much lower can the current unemployment rate go?

The next chart is a combination of the earlier figure with the civilian employment to population ratio. The BLS defines the civilian population as those 16 and older residing in the 50 states and D.C. who are not inmates of institutions and not on active duty in the Armed Forces. We would expect U3 to run opposite to this ratio. This is to say that when the unemployment level goes down the ratio goes up and vice versa. The chart, which has 55 years of data, reflects this relationship. Since about 2000 though, there has been a steady erosion of the employment ratio. If you read my books, you will understand why the year 2000 is important for economic cycle measurement. For the two official recessions after 2000, the employment ratio's rebound after the recession is weaker than in the previous years of data. Moreover, the current bounce is weaker than the one immediately preceding it. In absolute terms, the ratio is much lower than it was in 2000. When considering the earlier question about how much lower can the U3 rate go, this ratio should be of concern.

Part of the following illustration may be familiar to readers. In this chart, I combine the U3 chart with the civilian labor force participation rate for the period after the official end of the last recession. What is troubling about this chart is that despite a significant improvement in the unemployment rate, the civilian participation rate continues to decline. In my last book, I addressed the segmentation in this figure by noting that those who were closer to retirement (55+) actually increased their participation compared to the younger sect. There could be a whole discussion on the decreasing labor participation rate but suffice to say, until that figure improves, it will be difficult for the U3 rate to get much lower.

If you believe, as I do, that services are ancillary to production, the next couple of charts should cast a shadow on our employment and wage picture. The first is an illustration of the number of people employed in manufacturing. Since about 1980, the number of people employed in manufacturing has fallen and if one considers this as a ratio, the fall would be steeper since the labor force has increased in raw numbers. Also troubling, is the post-recession manufacturing employment after 1990. After the end of the early 2000s recession, manufacturing employment kept falling!

Has manufacturing fared better since the end of the last recession? While there has been a slight improvement in manufacturing employment, the opinion of purchasing managers has not reflected this. The ISM Manufacturing Employment Index surveys purchasing managers about their views on manufacturing jobs. Figures below 50 reflect a degree of pessimism. This index is far from perfect since it treats all firms, regardless of size, equally. It does, however, provide an indication of trend. Despite the unemployment rate getting lower, the opinion in the manufacturing sector is far from rosy. This does not bode well for increases in employment in this sector.

So if manufacturing employment growth has been tepid and prevailing opinion about its growth bearish, does that mean the growth has been in services? The next chart, with the ISM Index for non-manufacturing employment, shows the inverse relationship between the overall unemployment rate (U3) and the perception of services employment. When the index rises, unemployment falls and vice versa. The trouble with the rise in non-manufacturing employment is the wages associated with these positions. Dent Research indicated that 60% of the jobs added in January fell below the median wage and the great majority of those fell between $12 and $16 per hour. So while there is optimism about job growth in non-manufacturing areas, we should expect wages from those positions to pale in comparison.

There is a very definitive measure that describes labor activity and is not subject to opinions or numerical adjustments like seasonality or other discretionary manipulation. The U.S. Department of the Treasury publishes a Daily Treasury Statement (DTS) that shows deposits and withdrawals. Within that statement is an entry called "Withheld Income and Employment Taxes". This statement accurately documents the labor input of the economy in real time. Any worker who has taxes withheld will show up on this entry. On February 8th alone, the Treasury Department collected $16 billion in withheld taxes.

The DTS aggregates data according to the federal government's fiscal year which ends on the last business day in September. The next chart shows the entirety of collected withholdings for fiscal years back to 2007. Notice how withholdings bottomed in 2010, a year after the recession officially ended. So while the current sub 5% unemployment rate may be questionable given its narrow definition and subjectivity, the government is undeniably collecting more tax up front. This speaks to a lower unemployment rate.

So while the government is collecting more, what do trends in the DTS suggest? The next chart shows the growth in withholdings beginning in 2008. The left bar reflects the entire fiscal year over year growth and the right bar indicates the growth of the first four months compared to the same period in the previous fiscal year. I captured the first four months of the fiscal year since that period encompasses the very important holiday season when payrolls naturally swell. The figure clearly depicts the recession and the subsequent recovery. The trend in this data, however, is less comforting. Since peaking in 2012-2013, withholdings growth has slowed. Although 2016 data won't be complete until the end of September, we do have the first four months of the current fiscal year available. If we drew this bar into the figure it would be half as high as 2015. Clearly we are experiencing a slowdown in employment growth.

While the U3 figure gets challenged, I believe it can be helpful in identifying employment trends. The trend in unemployment, as it is measured by the BLS, reflects an improved jobs picture since 2007. That said, there are many warning signs. The wage quality associated with this improvement has deteriorated. The labor force participation rate and the employment to population ratio present a significant headwind towards any further improvement. The growth in tax withholding is undeniably slowing. If you believe the stock market shakeup last summer and the first part of this year is a precursor to economic weakness, then the recent unemployment figure published by the BLS will likely mark a low before heading up during the next recession.

As with any published statistic, it is always instructive to look at the components surrounding it and make your judgment accordingly. The unemployment figure is adequate for a general discussion, but as always, the devil's in the details.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.