A Few Thoughts about the US Labor Market

new business

In a speech to the joint session of Congress that was widely recognized as “presidential,” US President Trump said twice that there are 94 mln Americans out of the labor market.  It is not a lie or fake news or even an alternative fact, but it is so misleading that it is incredulous.

Here is the relevant context.  The US total population is about 325 mln people.  Of that universe, there are roughly 205 mln working age people (16-65).  Of the working age people, 152 mln have jobs, and almost eight mln are looking for work.  These 159 mln people are considered in the labor market.  That leaves 46 mln (205-159) working age people not working and not looking for work.  They are not in the labor market.

Who are these 46 mln people?  They do not only include discouraged people who have given up looking for a job.  They also are disabled people, students going to school, those who have home responsibilities (taking care of children or elderly or disabled) and those that have retired early (i.e. before 65).

Still, this leaves us far from Trump’s 94 mln figure.  Who else are included?  Most of the remaining 49 mln people are retired American over the age of 65.  There are 48 mln Americans older than 65, of while almost seven mln are working.  That leaves 41 mln Americans who are not in the labor market and are part of the 94 mln that Trump touted.

Between working age people not working or looking for a job and retired people not working, we can account for 87 mln of Trump’s 94 mln figure.  Who are the other seven million people?  Rick Newman,  a journalist at Yahoo Finance, suggests the seven million are incarcerated or otherwise institutionalized.  The rate of incarceration in the US continues to rise.  It stood at 2.25 mln in 2014, for example.  And after servicing time, employment opportunities are still difficult to secure.

The 94 mln people out of work figure that Trump cited includes every American who is not employed.   That means those in school and jail, those that are disabled, and those that are retired.  It is such an inclusive measure that it is nearly meaningless.

The US President has emphasized the desire for more manufacturing jobs.  He suggests that the US trade policy and unfair practices by other countries are significant factors.  Such rhetoric, however, does not stand up to rigorous scrutiny.  There are two numbers that capture what has happened the may be overlooked.  Since 1979, US manufacturing employment has fallen from 19.4 mln to about 12.3 mln at the end of last year.  In this period, US manufacturing output has doubled.  In 1979, manufacturing output per employee was about $41.6k in 2009 dollars.  In 2015, it reached about $155.2k.

This development is explored at length in my new book, Political Economy of Tomorrow Up through WWI, to produce more goods, more inputs (labor, capital, and raw materials) were needed.  Sometime around then, a transition was made.  Output increased as inputs fell.  This is what productivity means.

It is a familiar development. The incredible increases in agricultural productivity as science was applied to the production of food, freed people up to work in factories. Science has been applied to the production of goods.  Factories no longer need as many people.  Most countries, even those that are accused of stealing US manufacturing jobs are themselves losing manufacturing jobs, including China, for example. The Center for Business and Economic Research at Ball State estimates that 85% of the loss of manufacturing jobs is due to technological advances and automation not trade.

The new industries that have risen are not as labor intensive the old smokestack industries.  Total employment in the computer and electronics sectors, for example, employed 1.87 mln Americans in 2007 and by last August the figure has fallen to a little more than a million.  The five largest US companies in that space (Apple, Alphabet, Microsoft, Facebook, and Oracle) have a market cap of almost $2 trillion, which is something on the magnitude of 80% higher than the five largest in 2000, but they have fifth fewer employees.

Consider two examples covered in the popular press.  Instagram was bought by Facebook in 2012 for $1 bln.  It had 13 employees.  WhatsAp was bought by Facebook in 2012 for $19 bln.  It had 55 employees.   The point is that the new industries and sectors are not nearly as labor intensive as the older industries.  The new technological advances are both labor saving and capital saving.

This is not to argue that unfair trade practices do not hurt the US.  They do.  There is a conflict resolution mechanism integrated into most trade agreements, including the World Trade Organization.  These mechanisms are important as there seems to always be some friction even between countries like the US and Canada which have so much in common.  Resolving conflicts helps contain disagreements and prevents a ruinous escalation that could carry into other areas.  If the US carries through with its threats not to heed WTO rulings, it will encourage other countries to defect from the international free-trade regime and increase the risks of beggar-thy-neighbor policies.

The US labor market continues to gradually improve.  The weekly jobless claims are at new cyclical lows, and due to qualification changes, the figures are not directly comparable to the earlier cycles, but the direction is clearly constructive.  Continued improvement is likely, and it will be more likely as wages increase.

We often explain what given the size of the US workforce, why hours worked is a key metric that the popular media tends to largely ignore.  However, in the current environment, we suggest that average hourly earnings (and other measures of wage pressure) are an important metric that investors will want to track closely.  The median forecast in the Bloomberg survey is for average hourly earnings to rise by 0.3% in February to 2.8% year-over-year.  If there is a surprise, we suspect it to be on the upside.  The comparison to last February is favorable as hourly earnings were flat in tow months in 2016, February and November.  The median forecast on the headline non-farm payrolls is expected to be around 190k.  In 2016, non-farm payrolls rose by an average of 187k.