Stasis, not decline

In 1589, William Lee was sitting in a house in a small village outside of Nottingham, watching his wife knit. Billy was upset because, apparently, he wanted to have Biblical relations with his wife, while she preferred to knit. And so, in a fit of revenge, he invented the stocking frame, a mechanical knitting machine. Well, that’s the story anyway.

Billy’s stocking frame first could knit wool with 8 needles per inch; soon after he improved it so that it could knit silk with 20 needles per inch. He demonstrated the device for Queen Elizabeth and hoped to be granted a patent. She refused, saying, “Thou aimest high, Master Lee. Consider thou what the invention could do to my poor subjects. It would assuredly bring to them ruin by depriving them of employment, thus making them beggars.” King James later also refused to grant Billy a patent; he took his knitting machine to France and died broke. The textile revolution engulfed England a century later.

The fear of technology replacing humans has been with us for centuries — it’s embedded into our view of technology, even when it’s false. So, a series of observations about productivity and efficiency gains, technology, and macro-economics:

1. Productivity gain is all that matters. Our economy is based on ‘growth,’ which is economic growth beyond population growth. So you can get GDP growth by growing your economy … or by keeping your economy the same and shrinking your population, or some combination. Generally, growth is the result of productivity gains, not nominal population gains (2000 years ago productivity and population correlated but don’t anymore in the west; decoupling began about 1000 years ago).

2. We’re in stagnation. Economic growth in the West has been largely stagnant since about 1970. Most ‘growth’ has been the result of jiggering with CPI, inflation, etc.[1]. Our real growth as a percent of nominal growth has magically inflated massively. But it’s a matter of consensus among economists that growth has slowed significantly since about 1970.

Productivity growth before 1970 average 2.82% annually. Since 1970, it’s averaged 1.62%. Since 2006, growth has declined further, averaging 1.3%. Depending on how you finagle inflation, that’s basically no growth. In the same period, output per hour fell from 1.8 to .9 and hours of work from 1.4 to .4. Growth has declined regularly, at any unemployment rate, for the last twenty years.

Going nowhere fast

2. Most of this decline has been in tech. The last 30 years has been the most technologically stagnant of the last 200 years. We’ve exhausted our capabilities for major improvements (7nm chips is a good example). Over the last 200 years, each generation experienced social revisions. We moved from wood to steel, introduced machines, instant communication with the telegraph and telephone and radio, made plumbing and indoor heating/cooling and electricity widespread. Travel time massively decreased with the Boeing 707 that began flying in 1958. It has a cruising speed of about 600mph, which is the same as today.[2]

Median household income, 1975–2014, in 2014 dollars. So much tech. So little to show for it.

In the past 200 years, we’ve invented the concept of the job, of schools and education, of cities, of the ability to transmit data and people and everything in between rather quickly.[3] With the rise of textile mills and mass production of clothing, standard clothing sizes are invented (1860s). With rail, we could now transport perishables quickly, and cities devoted to this are created (Chicago, 1870s). With the steamship, we could likewise transmit perishables quickly, and we get new foods (bananas, 1880s). We’ve broken down barriers between night/day and time/space, and each generation has experienced a new kind of civilization — up until about the 1970s.

The problem is global…and the variance is all debt.

And yet, despite apparent continued improvements, the evidence suggests otherwise. The first killer app was VisiCalc. People actually bought the hardware to use that software, and VisiCalc essentially made the desktop PC business. When VisiCalc was released, there were two million bookkeepers in the U.S., and everyone fretted like Queen Elizabeth that bookkeepers would be put out of business. Today, there are one million bookkeepers, so while there was a 50% decline, there’s still a significant number of bookkeepers. More importantly, there are 1.5 million additional financial analysts today than there were in 1980. So VisiCalc did decrease the number of bookkeepers but greatly increased to total number of jobs related to using such software and data. Bottom line: we actually have more people working with that data today than before (with little to show for it in form of improved productivity). Think of how much we hear of the e-commerce revolution. It’s a relentless pounding of hype. But since 2008, the total number of people working in retail in the U.S. has increased by 200,000. Increased. E-commerce hasn’t so much as replaced retail jobs as augmented them, and the resulting picture is muddled at best or decreased efficiencies at worst. From a macro-economic perspective, the great scale/efficiency revolution in retail was delivered by Wal-Mart and was done by the 1990s; Amazon actually hasn’t had much of an effect. Construction hasn’t changed for generations. Education has become remarkably less efficient. From a productivity/efficiency growth perspective, we’ve been largely spinning our wheels since 1970.

ATMs are another good example. When they were developed in the 1960s, it was assumed they would eliminate the need for tellers. ATMs began to spread in the 1970s and were ubiquitous by the 1980s. And yet, the number of bank tellers did not decline — — at all.[4]

The digital revolution let me save to buy mom’s old Toyota.

4. Exhaustion Point v. Population Growth. If you’ve reached exhaustion point, your choice for continued growth then is decreased population, which also makes sense given the decline of certain types of jobs.[5] You cannot retrain anyone to do anything. Lots of people (maybe 15%) will just be jobless in the future. (Tech advancement today is often more a political problem than a technical one.)

Declining efficiency. Declining returns.

Schools are one good example of exhaustion — there’s no evidence that we can improve the outcomes of Detroit or Chicago schools. None, despite 100 billion+ spent on those two districts over the last two generations. Vary the inputs all you want … same outputs. Macro education data has been stagnant since 1960 (actually declined in some areas).

5. Requirement of Growth. The necessity of growth is not capitalist but consumerist; the two are not necessarily tethered. The consumption model requires growth regardless; capital model can operate purely under productivity gains (with or without wealth gains). Lots of examples of that.

6. State Pensions. The benefits model that is used today was developed under the pre-1970s economic model with its growth assumptions. Much of the data is there but we keep lying to ourselves about how much growth has slowed down (with the assistance of fraudulent economists, which is most of them) and benefits get increasingly out of whack. Look at how many benefits models assumed >7% annual growth … then look at how often that actually happened. Yeah, it’s not the 1950s anymore. Misaligned benefits models are one of many timebombs that can massively devalue currency (because the solution will be bailouts, QE, printing more money, etc). Wanton money devaluation was the end of Spain, Rome, China and any number of other empires.

People often misconstrue what the “Dark Ages” were. It wasn’t a period of decline as much as a period of transfer and stasis. First, when Rome “fell,” its best and brightest didn’t evaporate. They moved — increasingly to Byzantium and often to North Africa and Spain. The regions South, East and West of Rome increased in dynamism while Rome suffered stasis. And that’s the second point: the dark ages weren’t a period wherein humans reverted back to attempting to reinvent the wheel and domesticate dogs. Rather, it was a period wherein the core inventions of the Greeks, after centuries of being applied and improved by Romans, had reached a point wherein further improvements were marginal and negligible and generally not worth attempting. In other words, the Dark Ages were a period of stasis.

To be clear, there were great improvements during the Dark Ages, even in Europe. The great cathedrals that started to rise in the 10th century weren’t conjured from nothing. The institution on which modern Europe would build, the Catholic Church, was establishing itself as the center of the European project during the Dark Ages. And in many ways, Europe struggled with rediscovering the Greeks while bypassing the slave-heavy Rome — new machines would be required to supplant the lack of human labor. Of course, the Black Plague lit a fire under that impulse. So statis doesn’t mean “nothing happened” or a reversion to barbarism but rather refers to a systemic decline in the rate of improvement (or productivity growth).

So the Dark Ages were, more than anything, a period of Greco-Roman stasis (in core-Europe). Are we in a shiny new dark ages now?

Next: Economic exhaustion begins with cultural and institutional exhaustion; the need to rediscover ‘work’ and replace ‘jobs.’ Further explorations of the issues discussed here:

[1] The Boskin Commission of 1996 is one such example. The GDP number that’s typically reported is nominal GDP (total growth) minus inflation, which leaves us with the amount the economy has actually grown. For example, if you have 4% nominal growth and 2.5% inflation, then your GDP growth is 1.5%. Economists (and the governments that employ them) can engage in shenanigans with that inflation number. For example, if you have the same 4% nominal growth but only 1% inflation, then you have GDP growth of 3%. The only number that changed is what economists decided what the inflation was … and if you can decrease inflation, you increase GDP growth. The Boskin Commission magically decreased inflation by 1.1%, which magically increase GDP by the same amount. Remove Boskin Commission adjustments, and GDP is remarkably stagnant, particularly since about 2005. Worth keeping in mind that most/all countries finagle these numbers in similar ways, some much more egregiously than the U.S. (hi China).

Countries routinely (and massively) manipulate their GDP numbers, often by manipulating their inflation numbers. Here’s a June 2019 article on how India likely overestimated its GDP growth number by over 50% for the years 2011–2017 (India claimed gangbusters GDP growth. A former government adviser says it was a mirage).

[2] There is an argument for improvement around cost declines; it takes the same time today to go from NYC to Chicago (plane or train) as it did two generations ago, but the costs have declined and thus access has increased. Interestingly, this is much of the story of improvement since the 1970s; the technology isn’t new (though perhaps improved), but access has increased. It’s worth noting that overall this ‘improvement’ doesn’t really show up in the macroeconomic data. It’s also worth noting how untrue it is in many areas — even adjusted for inflation, the costs of many things — phones, college, construction — has increased. The average selling price of a car today is about $35,000, which is about $5,400 in 1970. But the average price of a car in 1970 was $3,500. So people today are paying >50% for a car than in 1970. And, of course, people are paying much more for phones and technology in general. The ‘increased access’ argument seems more about increased access to financing than to declining costs (in tech or anything else). Doubt it? Check out the explosion of personal debt since 1970. Further, the increased access to technology has made no significant contribution in increased productivity, despite what Microsoft and Google tell you.

[3] Over two centuries ago, there was ‘work,’ not ‘jobs.’ Jobs were largely invented because of the factory (a ‘job’ is the human input in the factory model). This definition was then applied to education, etc.

[4] It could be argued that while the population grew over this period, holding bank tellers constant constitutes growth. That’s a fair point but it’s also my point: are we really in a period wherein ‘growth’ means simply compensating for population increases? That’s it? We have a few more people so we grow a few more turnips? That’s a Dark Ages economic model.

[5] Certain jobs will disappear; their replacement jobs, if any, do not have fungible skills … which means you can’t simply assign people to such new jobs. For example: the job building the self-checkout scanner cannot necessarily be filled by the now-jobless cashier.

About Nathan Allen

Formerly of Xio Research, an A.I. appliance company. Previously a strategy and development leader at IBM. His views do not necessarily reflect anyone’s, including his own. (What.) Nathan’s academic training is in intellectual history; his next book, Weapon of Choice, examines the creation of American identity and modern Western power. Don’t get too excited, Weapon of Choice isn’t about wars but rather more about the seeming ex nihilo development of individual agency … which doesn’t really seem sexy until you consider that individual agency covers everything from voting rights to the cash in your wallet to the reason mass communication even makes sense….