Robots Have Not Yet Displaced Many US Workers Compared to the Effects of Globalization
When I was a young geologist, working in the oil fields during the great oil boom of the 1970s to 1980s, I was paid pretty good money for what I did. But I was a professional with an advanced degree, I had excellent training, and I had a good track record. Nevertheless, I would now have to make 2.36 times as much in today’s dollars to break even with that former level of pay, back in 1984 when I was only 32. I was able to eventually do that, but not without great upheaval in both my career path (two separate ones actually) and my life, and along the way, the road has had many bumps in it. In many families, the inflation gap was covered for a long time by virtue of the fact that there were two wage-earners in the family.
Now, as I enter a sort of quasi-retirement this year, and as I look back on our economy, it seems to me that the long struggle of many in the middle class to maintain their standard of living has finally been lost, or is about to be. Unfortunately, many in the middle class have not actually seen their wages more than double since 1984, as they needed to do just to break even. In fact, real median wages are only about 6% higher now than they were in 1984 (Chart 1) in spite of real GDP per capita rising enormously. Workers contributed their fair share to the ever-increasing productivity that helps drive economic growth (Chart 2), but they did not appear to get their fair share of the gains. There are multiple factors for why this has happened, which we will discuss in what follows.
Chart 1: Real Median Wages Almost Flat Since the 1980s
Chart 2: Real Median Family Income Has Lagged In Spite of Ever Increasing Productivity
Anyway, over time this lagging real wage growth has slowly taken its toll on the middle class in America, and so for a variety of reasons, the ranks of the middle class have been declining for decades (Chart 3). In response to this economic pressure, over these same decades, millions of families have taken various measures to try to maintain their piece of the American Dream. The first of these steps taken appears to have been (in part) the transition of women to the work force (Chart 4), which had significant political and social overtones, but was also practical, in the sense that the work many women did was needed, and the wages that most women earned were also needed (and proved helpful to maintaining their family’s standard of living).
For a while this kept the wolf from the door, but the growth of women as a percentage of the work force appears to have reached its peak a few years ago. Indeed, almost simultaneously, real median household income peaked in 2000 and has been falling ever since (Chart 5). Also nearly simultaneously, millions of workers have seen their retirement benefits lost or reduced (Chart 6). Adding insult to injury, the pathway to the American Dream has been filled with additional obstacles, such as the soaring costs of healthcare and education, two essentials to the middle class lifestyle that in part define the nature of the present dilemma (Charts 7, 8).
Chart 3: Percentage of Americans Who Are Middle Class Over Time
Chart 4: Percentage of Women in the Workforce Has More Than Doubled Since 1950
Chart 5: Real Median Household Income Peaked in 2000
Chart 6: Workers Have Simultaneously Lost Retirement Benefits
Chart 7: Soaring Medical Costs Since Medicare Passed in 1965
Chart 8: Soaring Education Costs Since 1973
It is probably not a coincidence that as some of the major costs for middle class families soared and middle class income stagnated, eventually most middle class families felt that they had to come up with another measure (besides women working) to maintain their standard of living. This was for many the rapid growth of consumer credit (credit cards, student loans, automobile loans, etc.) after 1994 (cf. Charts 9, 10). But mortgage debt also grew rapidly after 2000 (Chart 11), courtesy of the Fed’s low rates, political do-gooders loosening government lending standards, and the crass greed of the Wall Street investment product machine. Mortgage debt reached dangerous levels as the housing/credit bubble peaked in 2008, and contributed to the carnage of the Global Financial Crisis (“GFC”).
Chart 9: Soaring Consumer Credit Has Partially Covered the Gap
Chart 10: Soaring Student Loan Debt Since 2003
Chart 11: Household Debt Now at $12.78 Trillion, and It’s Above the 2008 Peak
Amazingly, total household debt has just this summer passed the level reached at the 2008 peak (Maria LaMagna, 2017). However, adjusted for inflation, we are probably still about 15% below the prior peak. Furthermore, relative to disposable income, the situation seems much better than it was in 2008 (Chart 12), at least according to some statistical measures (note however that disposable income is generally only available in quantity to the upper middle class and the wealthy). There is increasing evidence that the lower middle class is not doing so well, due to the rising asymmetry of rewards in our economy (Charles Hugh-Smith, 2017).
Either way, recent increases in delinquencies suggest that financial pressure is again increasing for millions of middle class families (Chart 13) as we near the end of the current credit cycle. So not only was the over-leveraged middle class hit hard in the “GFC,” as evidenced by the huge upsurge in delinquencies on most kinds of credit as well as mortgages in 2007-2010, but also now the downward spiral appears to have begun once again. The biggest problem appears to be unpaid student loan debt this time, although auto loans and credit card debt are also seeing spikes in late payments and defaults again.
Chart 12: Household Debt Relative to Disposable Income Appears Much Better Than 2008
Chart 13: Rolling Disaster of Household Credit Problems
It appears that the middle class is surviving on a hand-to-mouth basis (in many cases, paycheck-to-paycheck), supporting their life-style with a semi-permanent and growing pile of debt, with little ability to pay it off. The average mortgage in America is now $106,132 (Student Loan Hero blog, 2017), but about one-quarter of households are still underwater on their mortgages. The average household credit card debt in 2017 has been estimated at $6,662 per family; about one-third of new car loans involve owners with negative equity from the previous car (average loss of $4,500; Philip Reed and Ronald Montoya, 2015); and the average new car loan is now $30,032. A startling 42.9 million people have overdue medical debt in the US right now. The average 2016 college graduate owes $37,172 and will likely spend a decade or more paying it off. Many of these people haven’t even gotten jobs in their chosen fields. From all sides, the middle class is being pummeled by debt, job uncertainty, soaring medical and education costs, and low or stagnant real wages.
None of what I’ve discussed so far is real news, and much has already been written about these problems. Nevertheless, it seems to me that we are at a major turning point, and the world as we knew it may be gone. You would think our media sources, the political system, and various think tanks, spurred on by rising public anger, would be seeking solutions. But an astonishing amount of the national debate on all of this has degenerated into the polarizing kind of class warfare that the media and both major political parties always revert to when big issues come up. Let us stipulate here that this kind of fault finding is going to shed more heat than light on the problem, and we are not going to be self-indulgent and add to that pointlessly irritating literature. There are a number of factors that have contributed to this dilemma, and even a cursory examination of the data suggests that it is not going to be easy to identify a simple causation or to discover straightforward solutions.
The problem is very complex, in my opinion. The main factors appear to be (in no particular order): 1) Worker displacement due to globalization; 2) technologically driven worker displacement; 3) extremely weak government and corporate support for retraining of the displaced; 4) inequality driven by the fact that the returns to labor have decreased, while the returns to capital have increased over time; 5) an increasing distributional divergence between the well educated and the poorly skilled (due to asymmetry of rewards); 6) increasing divergence between the career prospects for various educational disciplines (college majors or certificates, again due to asymmetry of rewards); 7) real GDP growth per worker has been less than real GDP growth per capita (due to women entering the workforce); and 8) declining population growth and productivity growth, which together directly affect GDP growth.
Taking the last factor first, there has been no consensus on the causes of, or solution for, the long-term decline in productivity growth. Population growth is declining for economic and social reasons that are complex; certainly a reversal of this trend is unlikely in the short term. While the impact of women entering the workforce has probably peaked and their impact on wages has likely stabilized, the fact that working women are on average more educated than the average male worker gives women a long lasting relative advantage with respect to wage growth, even if they still in some cases get paid less than men. A college education, per se, provides no particular guarantee of employment, although those who are employed will make more than other, less-educated workers. The asymmetry of rewards in our technologically-driven economy has tended in recent years to create “winner-take-all” corporations, economic sectors, and career paths. Hence the chosen few at any given moment make huge money, but everyone else lags markedly, due to educational and sectoral distribution differences.
The returns to labor have been decreasing for decades, regardless of the macroeconomic environment (Chart 14). So “financialization” of the economy probably contributed to the trend, but so apparently did globalization and inflation. So laying this problem at the door of one of the political parties is pointless and counterproductive, because no administration has had any effect on the trend. Perturbations in the trend, like the internet and housing booms, are one-off events driven by either technological advances or cheap money (credit bubbles), and their effects have not lasted.
Corporate profits have been elevated above trend for a long time, yet little has happened with regard to the real wages of workers (Chart 15). “Financialization” is part of the explanation, but so is globalization, apparently because it generally reduces the returns to labor. The robotics and Artificial Intelligence (“AI”) revolution has already had a worldwide impact on productivity (Chart 16), but the evidence for robots taking manufacturing jobs appears surprisingly weak so far (Chart 17), according to the Harvard Business Review (Mark Muro & Scott Andes, 2015). This seems counterintuitive, but it’s worth noting.
Chart 14: Declining Wages as Percentage of GDP Regardless of Macroeconomic Environment
Chart 15: US Corporate Profits Have Soared, But Labor Income Has Lagged
Chart 16: US Not Even the Leader on Industrial Robotics Applications
Chart 17: Weak Correlation Between Manufacturing Job Loss and Robotics
This leaves us with two remaining major causative factors, globalization and inflation. Both make plenty of sense. Manufacturing jobs have definitely been lost to globalization pressures (Chart 18). This long downward trend in manufacturing employment has primarily been caused by the huge increase in the global labor pool over the last 40 years. More and cheaper labor naturally created competition that has lowered the relative cost of goods for all Americans, but at the cost of many US manufacturing jobs. This kind of economic phenomenon is not something that could have easily been changed, due to its mega-scale economic drivers.
The recent election results in part reflected huge middle class frustration with this trend. President Trump’s proposed solutions appear to have helped get him elected, but I don’t see how he can reverse this trend with American labor costs so high relative to the costs of capital (Chart 19). The fact that salary and benefits (total compensation) have continued to climb steadily throughout the last 25 years (Chart 20) suggests that workers are costing more, but the cost is not wages, it’s benefits. From the corporate point of view, it hardly matters, because rates for both labor and benefits have been cheaper elsewhere. Trump’s own companies, as well as many others, have taken note of this. The ability of corporations to cut these expenses so they can better compete is variable, depending on how labor-intensive or capital-intensive their industry is, the size of the company, its relative balance sheet strength, its flexibility, and its management team’s foresight (Zero Hedge blog, 2012).
Chart 18: US Manufacturing Jobs Have Been Outsourced
Chart 19: US Labor Costs High Relative to Cost of Capital
Chart 20: US Salary and Benefits Have Increased Steadily
The impact of inflation on a cumulative basis has been enormous since 1976, as I mentioned before (Chart 21). But inflation is still a major factor in the decline of the middle class in more recent years, in spite of declining relative levels of PCE (GDP-based) inflation since 2000 (Charts 22, 23). Indeed, Fed Chair Yellen thinks that lower inflation has held back wage growth (Yellen, 2015). That may explain why the Fed is so hell-bent on its seemingly arbitrary pursuit of a 2% inflation rate. However, I can’t really see how higher inflation rates are going to trump low international labor costs and very low cost of capital when it comes to helping the middle class out of their dilemma. Furthermore, there is little evidence that the Fed has been successful in boosting inflation.
Chart 21: Cumulative Effects of CPI Inflation Since 1976 Are Enormous
Chart 22: US Wage Growth Has Declined With Inflation
Chart 23: Cumulative Inflation Up Another 40% Since 2000
Summarizing, then, it would appear that the major drivers of middle class decline have been: 1) Actual worker displacement due to globalization; 2) the steady decline of purchasing power due to inflation; 3) inequality resulting from the economic impact of the returns to labor declining as the returns to capital have climbed; and 4) distributional divergences as asymmetric rewards (“winner take all” trends) have grown, driven by educational and occupational differences between different groups. There may be some worker displacement due to automation and other technological advances, but it does not yet appear to be a major driving force in spite of popular wisdom to the contrary.
However, the failure of government and corporate entities to provide adequate support and retraining (relative to the rest of the world) for displaced workers regardless of cause (Chart 24) has made a bad situation worse. At the very least, we should be boosting these programs (assuming they actually accomplish something) quite substantially. If they don’t work, we should redesign them. If President Trump and his liberal opponents really want to do something useful for the middle class, this would be a good place to start. Adjustments to trade agreements may also help a bit, assuming they don’t end up triggering a global trade war. Nothing proposed so far would do that, but this should be something we watch carefully.
Chart 24: Extremely Weak US Support for Displaced Workers
The Fed, as I have said, has not been able to protect workers from inflation. However, if President Trump reforms the Fed either via new appointments or legislative reforms, perhaps the divergent impacts of Fed policy on capital vs. labor will improve a bit. The only thing that will really change the returns to labor would appear to be a serious labor shortage, if theory is right. However, in spite of demographic pressures and a significant dearth of qualified personnel that has already lasted for years, there is still very little upward wage pressure (Chart 25).
Chart 25: Wages Have Grown Recently, But Not Much
Another place where President Trump and his liberal opponents could help American workers is through the improvement of educational opportunities for those members of the middle class most in need of it, hopefully without increasing the debt burden on these families. While this seems like a pretty tall order, one way around this might be to create a national service program. This could be used to get cheap temporary labor for the government, give college graduates a chance to get some work experience, and allow these same people to work off their obligations on student loans through a stint of government service. I think when you look at the potentially wasted lives that would instead be better spent, and the future government welfare benefits that would instead be saved, rather than spent on those who are not doing so well economically, it might actually be cost effective. Anyway, it might prove to be a useful way to get more people moving towards the American Dream again.
I cannot recall a time when the government (and I mean both parties) was more ineffectual, or more incompetent, nor when the electorate was more polarized and angry. In the absence of significant Congressional action on these issues, I would expect the 2018 Congressional election to reflect even more voter frustration, and the 2020 presidential election to be marked by some of the most violent protests in recent memory, similar to what we saw in 1968. The issues I’ve discussed are amenable to improvement at least, even if they can’t be fully resolved. We owe it to ourselves to make a genuine effort, rather than a token one, to do something useful here.
If labor costs do actually start to rise faster than CPI, it will make sense to tilt portfolios away from companies in industries or sectors with high labor costs. I would avoid sectors like retail (XRT), entertainment (PEJ), and parts of the transportation sector (IYT). I like the idea of investing in companies in capital-intensive industries with relatively low labor costs at this point in the cycle. Examples might include energy (XLE), steel (SLX), telecoms (IYZ), and financial services (IYG).
Disclosure: I am/we are long XLE.
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.
Additional disclosure: Disclaimer: This article is intended to provide information to interested parties. As I have no knowledge of individual investor circumstances, goals, and/or portfolio concentration or diversification, readers are expected to complete their own due diligence before purchasing any stocks or other securities mentioned or recommended. This post is illustrative and educational and is not a specific recommendation or an offer of products or services. Past performance is not an indicator of future performance.