What Has Happemed to Manufacturing in the U.S.?

 I have written this in reaction to an article from The New Yorker, (https://www.newyorker.com/news/us-journal/will-trumps-broken-promises-to-working-class-voters-cost-him-the-election) which I encourage you to read.  And my question is, as the title of this post says, What has happened to manufacturing in the U.?  There are two aspects to this.  The first--and, as I see it, more important--is what has happened to manufacturing output.  The chart below begins by looking at manufacturing output in the U.S. from 1972 through 2020.

Manufacturing output rose steadily from 1972 until the 2008 recession.  When the recovery got underway in 2010, manufacturing output rose again, at roughly the same rate as in the 1972-2008 period.  The pandemic of 2020 occasioned a sharp decline (falling by 11% between December 2019 and April 2020), with a rebound between April 2020 and September.  At that point, manufacturing output was about halfway back to t\o its end of 2019 level.  But the decline in manufacturing output does not, overall, to be a long-term trend for the economy.  The declines in manufacturing outpput that have occurred are related to economy-wide events, like recessions and the pandemic.

Manufacturing employment, on the other hand, has behaved differently.  

Manufacturing employment was roughly unchanged between 1980 and 2005, at around 170,000.  But even before the 2008 recession, manufacturing employment had begun  to decline,  Between 2005 and the end of the 2008 recession, about 50,000 jobs were lost (about 25,000 before the recession hit, and an additional 25,000 in the recession)/.  The recession clearly hit manufacturing harder than the other sectors of the economy.  While manufacturing output fell by about 11%, though, employment obviously fell a lot more.  On the other hand, employment recovered fairly well, rising by about 12% (and 12,000 jobs) between 2008 and 2020.  All of this, however, tells us one very significant thing:  Manufacturing employment as a percentage of total employment was generally declining.  When manufacturing employment was remaining stable, overall employment was rising.  When recessions hit, manufacturing was hit harder. 

In the early 1970s, manufacturing accounted for about 25% of all jobs,  The decline from 1972 to 2010 was basically a constant percentage decline; by 2010, manufacturing accounted for only about 9% of all jobs--on average, about 0.4 percentage points a year for 38 years.  What's interesting, though, is that the decline (in manufacturing jobs as a percentage of all jobs) seems to have stopped.  For the last decade, manufacturing has accounted for about 8% to 9% of all jobs in the U,S.  And, between 1960 and the present, manufacturing has accounted for roughly 12% of total US output, with very little variation (https://www.brookings.edu/wp-content/uploads/2016/06/us-manufacturing-past-and-potential-future-baily-bosworth.pdf).  

The relatively constant ratio of manufacturing output to total output, coupled with a decline in manufacturing's share of total employment has an implication.  That is that the productivity (output per worker, output per hour) in manufacturing has been consistently rising.  This should mean consistently rising (inflation-adjusted) earnings for workers in manufacturing.  And, since 2010, compensation in manufacturing has increased by about 25%.

 

Is the US manufacturing Sector Collapsing

Everyone seems to look at the decline in manufacturing employment.  And, yes, it is decling.  But... 

Here's another set of facts:  The Fed has a monthly index of real manufacturing output (from 1972 on) and data on personal consumption expenditures monthly as well.  I used the personal consumption deflator data to calculate real PCE, and then constructed a PCE index.  If you graph real manufacturing output and real consumption expenditures, they move almost in perfect sync.  (The difference is that manufacturing output fell much more sharply than did consumption during the 2008 recession, and, although it has increased from its trough at about the same rate as PCE, it never made up the lost ground.)

Basically, then consumption and manufacturing output have moved together.  The decline in manufacturing employment is a result of rising productivity--higher output per worker-hour--in manufacturing... 

The Failure of Net Investment (as a % of GDP) to Return to its Historic Levels

This began life as a comment on a blog post.  I have decided to post it here...

When I think about investment spending, I tend to focus most on net investment--total investment minus depreciation--as a % of GDP.  Private sector investment is not0riously volatile, falling dramatically during recessions.  So, for example, net investment (as a % of GDP--from here on out, all net investment figures will be as a % of GDP)fell from 7.5% around 1965 to just less than 4% in the 1970 mini-recession, rebounded to 7% by 1973, fell to about 2.2% in the 1974-75 recession, rebounded to 7% by 1980, fell to a little more than 2% in the 1980-82 recession, rebounded to about 6.5% in that recovery, and, from the mid-1980s to about 1992, fell to abuot 1.5%.
During the Clinton years, investment again recovered, to about 6%.  In the recession beginning in 2000, it fell to about 2%.  The recovery was fairly weak--to only about 3.5% by 2007.

Then--the real collapse.  Net investment turned negative for the first time since the late 1940s-- minus 2%.  The recovery was again weak, to just over 4% by 2009.  Since 2009, net investment has not exceeded 3.5% of GDP in any calendar quarter, and is now about 2.5%, which was the bottom-of-a-recession level in the 1960s and 1970s.  
This remarkably low level of investment comes despite the tax cuts, despite everything.  Right now, and, really, since the 2000 recession, the best done in terms of net investment is , essentially, no better than the worst from 1960 to about 2005.  

Incentives and behavior

I read an interesting blog post (well, interesting to someone who spent much of his life teaching and doing research about labor economics). and you. too, can read it here: http://conversableeconomist.blogspot.com/2020/04/are-atternative-work-arrangements.html)

What it did, in part, was to make me think about my third summer job, in the summer of 1966.  I think I got the job because my brother Dave had worked there in 1965 (and was working there in 1966, as well), and he had done a good enough job that they hired me.

We were retail route salesmen...for Banquet Dairy...we were milkmen.  And we did not get paid an hourly wage or a weekly salary, we got paid on the basis of how much we *collected* from out customers.  As I recall, the commission for most of what we sold was 17% (of course, we also had to collect it).  Now, if you're doing this permanently, you obviously have to be concerned both with how much you collected *and* how much you sold.

My routes (Monday-Thursday; Tuesday-Friday; Wednesday-Saturday) had larger than normal unpaid balances; in fact, those routes were open because the guy whose routes they had been sold a lot, but didn't collect very much.  So I figured out fairly quickly that my job #1 was to get people to pay up.  And, yes, I also sold a fair amount of product.  But I spent a lot of time knocking on doors, even of people who were not currently buying anything--and might not have been the people who ran up the unpaid balances, and talking them into paying at least something. [1]

I might add that none of my routes were in a really good part of Indy; they were all in the "near north" neighborhoods.  Basically, for those of you who know Indy, north of Tenth Street, south of 30th Street; Park Ave. east to College.  It's a somewhat better neighborhood now, and probably half of the places I delivered to have been torn down and some replaced--a lot of vacant land these days.  Gentrification there is proceeding *very* slowly.

So I collected and collected and collected, and by the end of the summer, the unpaid balances were somewhat less than half what they'd been at the beginning of summer.  I collected a lot more than I sold.  Earned more that summer than I did in any of my other summer jobs, but I did not try to go back for the summer of 1967...

I had, by that time, learned something about how incentives can shape behavior...and that, sometimes, the usual incentives (in this case, for whoever followed me, and who looked on this job as more-or-less permanent) do not work particularly.  (I felt sorry for whoever got those routes in September 1966; they had a fairly small unpaid balance, which meant they had to sell more than a normal amount, at least for a while--there was no cushion.)

[1] I wondered then, and for some time after, why someone who had not bought that unpaid-for milk would pay me anyway.  I suspect that some of them had moved there with unpaid balances elsewhere, and some of them were just intimidated...not that I thought of myself as intimidating...

Employment Growth,Population Growth, and the Recent Recession/Recovery

Lately, there has been a lot of discussion about whether the current rate of job creation is greater than usual or less than usual.  And the discussion has been framed in terms io the number of jobs that have been added.  This is almost certainly going to provide a biased look at job creation; what is likely to be a better indicator is the percentage increase in employment.  In doing this, it will also be important to look at the change in the number of people in the relevant age group, for which I will use the civilian non-institutional population age 16 and over. [1]


The annual growth in people with jobs is shown in Figure 1; the population (age 16 and over) is shown in Figure 2.  If employment is growing more rapidly than population, the employment-population ratio (EPR) is rising.  If employment grows more slowly than population, the EPR falls.  The latter will occur in recessions; the former is more likely to occur during recoveries.

Figure 3 is the (annualized) change in employment minus the (annualized) change in population.

Figure 3 may well be the most important of these charts.  A positive number here tells us that employment is growing faster than population, or, in an economist’s jargon, the employment-population ratio is rising.  When employment is rising more slowly than population, the economy might be experiencing a recession (more likely) or the economy might be experiencing a population boom that is not being matched by employment growth (less likely).  All of the instances of negative numbers in Figure 3 are recessions.

What we want to do here is look at the far right side of Chart 3.  It clearly shows the seriousness of the recession that began around the beginning of 2008 and the recovery that took hold in mid-to-late 2010.  What it does not show is any noticeable acceleration of employment growth (relative to population growth) after 2016.  (Figure 4 zooms in on the 2000-2020 period.)  What does appear to be the case is that employment growth has fluctuated between 0.5 percentage points and 1.0 percentage points faster than the growth in population.  Relative to other recoveries from recessions, this recovery has been both fairly prolonged—going on 10 years—and fairly weak.  The recovery from the 1974-75 recession saw employment growing between 1.5 and 2 percentage points faster than population.  The recoveries from the early 1980 recession and the early 1990s recession both saw employment growing as much as 1.5 percentage points faster than population.

So the most recent (2007-2009) recession was relatively severe and has had a relatively less robust recession, compared with recessions past.  It has, on the other hand, lasted a relatively long time. 


[1] I’ll be measuring, to be precise, the (not seasonally-adjusted) number of people with jobs, according to the Current Population Survey, and calculating the percentage change from the period 12 months earlier.  So, for example, the percentage increase in the number of people with jobs in January 2020 is 1.31% greater than the number of people with jobs in January 2019.  I will also be using the BLS data on population.  With the population data, there three anomalies, one centering on the 1950 Census population (which shows a period in which the US population allegedly fell) and the other two on the 1990 and 2000 Censuses (showing an unusually large increases in population).

I wish Phil Ochs Were Still With Us

I wish Phil Ochs were still with us...In the early 1970s, he re-wrote "Here's to the State of Mississippi" into "Here's to the State of Richard Nixon." We could use "Here's to the State of Donald Trump."

https://www.youtube.com/watch?v=KrrOY0vwuPE
and
https://www.youtube.com/watch?v=AeVzjIEXB7c

Greatest Economy Ever? Not So Much

President Trump touts his record on the economy as being “the best economy ever,” with a special emphasis on job growth.  And, to be fair, job growth since he was inaugurated has been reasonably strong and reasonably consistent.  Beginning with January 2017, and measuring employment growth over each successive 12-month period (to smooth out month-to-month blips), employment growth has been steadily around 1.5% to 2% for every 12-month period.  


But…In the period beginning in January 2012 through the end of 2016, employment growth was almost always over 2% for every 12-month period through the end of 2016.  Overall employment growth from the beginning of 2012 through the end of 2016 averaged 2.1% across all 12-month periods.  In the subsequent period (through October 2019), employment growth has averages 1.6% per 12-months.  And, in fact, the raw number of new jobs added per month was greater from 2012 through the end of 2016 than it has been since (202,000 per month from the beginning of 2012 through the end of 2016, compared to 195,000 since.


Then there’s the issue of wages.  Inflation adjusted wage growth on an annualized basis fas 1.02% from 2012 through 2016.  Since the beginning of 2017, it’s been 1.03%...or, essentially, the same rate of growth in inflation-adjusted wages before and after…And that’s markedly better than over the full length of the time series (since 1979)—over the past 40 years, the average weekly wages for full-time workers (adjusted for inflation) has increased at 0.2% per year.  But the increase in wages has been anything but smooth—real wages fell by about 10% between 1979 (to about $300 per week), recovered almost all of that loss by 1988, only to see wages fall back to about $300 weekly by 1992.  Again, wages recovered to about $340 per week by2002.  This was followed by a period of stagnation; wages resumed their growth in 2014 and have increased fairly steadily since then. 


But, again, the total increase since 1979 is only 7.5%.  Which is, frankly, awful.  Real total output of the U.S. economy—per person—has grown at an average annual rate of 1.9% since 1947.  If real weekly wages had grown at that rate since 1979, real weekly wages would have (roughly) doubled since 1979…if wages had increased at half the rate of increase of real GDP, the increase would have been 50%.


So…greatest economy ever?  Not so much.
(All data from the Bureau of Labor Statistics web portal or from the Federal Reserve’s web portal.)