Scott Alexander writes,
Median wages tracked productivity until 1973, then stopped. Productivity kept growing, but wages remained stagnant.
This is called “wage decoupling”. Sometimes people talk about wages decoupling from GDP, or from GDP per capita, but it all works out pretty much the same way. Increasing growth no longer produces increasing wages for ordinary workers.
Is this true? If so, why?
He makes a valiant effort to summarize and assess the economic literature. But this is where orthodox economics is hopeless.
Productivity by definition is output divided by the amount of labor input. Let me make three points:
1. You can’t measure the numerator very well.
2. You can’t measure the denominator very well.
3. The U.S. is not just one big GDP factory. Both the numerator and the denominator are affected by shifts in the composition of the economy, even if actual productivity and wages were not changing at all.
The numerator is output. How many people work in businesses with measurable output? Scott Alexander doesn’t. I never have. Most of my readers never have. There are entire industries, like health care, education, and finance, where we do not have any idea how to measure output. And even within an industry that has quantifiable output, we still have the issue that, as Garett Jones pointed out many years ago, most workers are not engaged in actual production; they are building organizational capabilities. Even if the factory managers can count widget production, they cannot measure the productivity of the tax accountants or of the team developing a new marketing initiative.
The denominator is labor input. But most of labor input consists of human capital. To measure labor input, you need to be able to measure quality, not just quantity. What is the incremental value of X years of schooling and Y years of experience? We do not have a reliable way to do that. One approach uses wage rates as an indicator of quality, but that amounts to assuming that productivity and wage rates are tightly coupled, but that amounts to assuming away the question that Alexander is raising.
We are not in a GDP factory. As the share of GDP devoted to health care and education goes up and the share devoted to manufacturing goes down, we are giving more weight to a sector where real output and the quality of labor input are extremely difficult to measure.
I think that for economists to say anything useful about productivity and wages, they should try to study individual units of individual firms. My guess is if you were to undertake such a study, you would be overwhelmed by doubts about the precision of your measurements and the difficulty of obtaining a decent signal-to-noise ratio. It’s perverse that you would instead look at the aggregate statistics cranked out by the Commerce Department and the Labor Department and pretend that it’s 100 percent signal.
I believe you, but remember that the ways society makes stuff is the real economy and all the stuff about flows of money is just an allocation mechanism. If economists don’t understand the economy at even a very basic, aggregated level, that’s a pretty serious problem.
Quite agree. I call it “money go round”.
A small portion of what is paid for a good or service goes into its manufacture and the raw materials, the rest into compliance costs. The problem arises when taxes are levied on the whole sum and this is regarded as “productivity” or “value added”. Governments insist on compliance which have costs, and then benefit from GST or VAT thereon.
A man builds a wall that someone wants, and that wall represents value.
A lawyer produces a document is something the client has to produce on demand as otherwise he is punished.
The real value being generated is the organisation of substances dug from the ground into goods that people want. The rest is money go round.
A “Money-Go-Round”.
I like it – well said!
Exactly!
1950s: The GDP factory churns out 2 GDP units with 1 unit of labor and 1 unit of capital as inputs. Labor owners get 50% (1 GDP unit) and capital owners get 50% (1 GDP unit).
2010s: The capital intensity of the GDP factory is greatly increased. The GDP factory churns out 5 GDP units with 1 unit of labor and 4 units of capital (intellectual, physical, etc) as inputs. Labor owners get 20% (1 GDP unit) and capital owners get 80% (4 GDP units).
There is nothing inherently unfair about the labor share of GDP declining from 50% to 20%. In a socialist society, where workers own all the capital supposedly, the worker share is 100% in both cases. In the quasi-socialist society advocated by Bernie et al, the labor share is some arbitrary number between 20% and 100%, achieved through redistribution.
Drop worker from the ratio, measure relative efficiency of individual sectors, then measure standard of living, then match those two. The intermediate term should be dollars, if money is working. Total sector efficiency can be scaled to match total worker hours, but it only works if most sectors keep a bettable savings to loan balance, they are cash flow accounters.
Thus, it is a currency variance issue, currency fails capture a representative sample of the sectors. In this model the passage of enterprise cash through a savings t loan machine forces all sectors to be slightly squeezed by money, the cost of banking. In the squeeze, interest charges should reflect pricing efficiency and neutralize currency bias. Good banking makes GDP factory model work, to a known bound.
Has anybody, to your knowledge, gone through that exercise, and if so, where can we read their work?
A line from his work:
My conclusions from this section, such as they are, go:
1. Arcane technical points about inflation might explain between 33% and 66% of the apparent wage stagnation/decoupling.
2. “Explain” may not mean the same as “explain away”, and it’s not completely clear how these points relate to anything we care about
#2 especially is relevant, especially if you read the whole section.
Timothy Taylor had a nice piece the other day about gross national happiness in Bhutan in which he quotes Solow:
“My own favorite comment on the connection from GDP to social welfare is from a 1986 essay by Robert Solow (“James Meade at Eighty,” Economic Journal, December 1986, pp. 986-988), where he wrote: “If you have to be obsessed by something, maximizing real National Income is not a bad choice.” At least to me, the clear implication is that it’s perhaps better not to be obsessed by one number, and instead to cultivate a broader and multidimensional perspective. If you want to refer to that mix of statistics as Gross National Happiness, no harm is done. But yes, if you need to pick one number out of all the rest (and again, you don’t!), real per capita GDP isn’t a bad choice. To put it another way, a high or rising GDP certainly doesn’t assure a high level of social welfare, but it makes it easier to accomplish those goals than a low and falling GDP.”
Similarly, if you have to obsess about productivity, maybe GDP per hour of labor, multifactor productivity would probably be the one index to focus on, especially in the US.
But thankfully the OECD’s excellent Compendium of Productivity Indicators 2018 (available online free) provides a fine survey of the many alternative indicators (many even disaggregated by industry!) available to help one understand the various factors at play in producing the decoupling of wages and productivity. It does a particulary fine job of pointing to inherent limitations and situations in which cross-country comparisons are ill-advised. The compendium’s explanations of decoupling, as well as the influence of different measures of inflation, are clearly and crisply laid out with graphics illustrating various OECD countries’ performance.
The multifactor productivity indicator is an efficiency measure and that would seem to be the appropriate focus. As the OECD points out, income from self-employment is imputed in labor compensation. With the longstanding and sustained growth in self employment in the US, and especially in countries with substantial informal labor markets, assuming the self-employed earn average compensation seems dubious.
Understanding trends in multifactor productivity takes a lot more effort. Personally, the tax code’s channeling of talent into non-productive tax-exempt industries, the rise of education institutions as employment gatekeepers, the increased politicization of the court system, the legal guild’s “common law” tax on productive economic activity, and the profound competitive disadvantage imposed by regulatory excess and bloated government employment, would all appear to be significant contributing factors to the US’s declining multifactor productivity and thereby to lower wages.
Looking at the legal environment, one reads daily about multi-billion dollar fines aganst productive businesses being handed down all the time. To survive, firms need to have full deep pockets and the money that is in those pockets is not being used to pay shareholders, invest in capital, or to pay employees.
The money the tax code encourages firms to buy good will with through “charitable giving” has also been increasing steadily and it too is money that is not going to shareholders, capital, or labor.
And one has to wonder how significant an influence on decoupling has been the rise of ego firms, that is firms like Amazon and Google, which will never pay a dividend, and exist solely as monuments to their owner’s egos. The vast sums spent by these firms on elaborate tropical rainforest office spaces and other extravagences, are expensed for tax purposes but contribute nothing to productivity. Especially in firms like these, replacement of the current corporate profits tax with a VAT would no doubt have salutory influences on national accounting of multifactor productivity.
GDP factory perhaps not, but there are ways to highlight recognizable output variance between the standard pre-1973 tradable sector relative dominance, and what came afterward. Consider where traditional scale (the progress trajectory that relies on less labour over time to achieve progress, rather than increased overall participation) sometimes manages to find productive applications in non tradable sector product, for instance. It does so where it can overcome the time/place constraints on traditional scale, without disrupting what people seek in final product.
Plus, firms and other organizations in non tradable sectors presently take similar approaches, re achieving output they hope to be capable of meeting ongoing revenue requirements. In finance, much of this is has been achieved by moving final product away from direct connections to place scarcity, via derivatives and the like. Education relies extensively on broadcast measures rather than person to person instruction and of course what they can capture of the money-go-round, to reduce their own costs burdens resulting from time based product scarcity. Whereas healthcare has relied on extensive rent or wealth capture to meet its own extensive overhead costs due to place based scarcities which it can’t erase unless it completely changes its form.
There was a policy memo about this in the Aspen Institute employment report last month, by Michael R Strain and titled “The Link Between Wages and Productivity Is Strong”. He covers many of the difficulties you mention, and also quotes Standsbury & Summers 2017 paper:
He also provides an updates productivity/wage graph, concluding:
Median wages tracked productivity until 1973,
Simple alternative theory: Labor supply with Baby Boomers starting in mid-60s grew at higher rates than wage levels could afford. (So 1974 – 1982 we accepted high inflation. So in 1974 after the Oil Embargo Black Swan, the US hit the middle income trap.)
So what would happen 40 years later: Labor supply growth is stagnant and how many employers are whining about not finding good workers. (Note the primary reason why unemployment was contained in The Great Recession was the Boomer retiring and unemployment claims are lower than anytime since early 1970s.)
Our grumpy 4% economy is contained from higher growth by lower labor supply due to lower fertility rates and pressure to contain Immigration.
I would call this the Japan Inc trap.
In industries where productivity can be measured, such as in manufacturing, there has been a decoupling of wages and output.
I sure hope the national policy becomes one that is focused like a laser on improving real, take-home wages.
Otherwise you will be toasting AOC in the White House soon enough.
Pretty much by definition, the much touted Service Economy will have lower productivity growth. Provision of personalized service is what is prized and that has physical human limits. Try to say up the productivity of a doctor measured by patients seen, the perceived quality of the service and, where possible, the payment for the service declines.
Also, the use of “wages” is a bad measure. Assuming “wages” quietly includes all compensation, the mix of cash, i.e., paycheck, to non-cash compensation, such as healthcare, has changed dramatically over the last 50 years. Then you throw in the cost of the labor, which encompasses total compensation, as well as payments government that are driven by having employees or hours worked, then you see a real shift in the mix with more labor cost going to the government(s) as taxes and fees than the employee as compensation and especially cash wages.
When all the hands grab at the amount of money for employee labor, cash wages is the one that gets what is left over instead of first pick.