Where Did the Productivity Go? Five Theories.


After my last post on how you can blame Silicon Valley for Donald Trump, I got a lot of questions as to why productivity is stagnating. Stagnating productivity leads to people being angry with their economic well being and turning to easy sounding solutions spouted by Mr Trump. Silicon Valley is the self proclaimed world capital of innovation, but as of yet none of the Bay Area break throughs is accelerating the sluggish productivity growth.

But why? How is it possible that giving everybody in the world access to all the information in the world doesn’t show up in economic statistics? Here are five theories.

Good things will come to those who wait

In the nineties a similar discussion was going on. All these desktop computers are great and allow us to do so many things, why don’t they show up as improved productivity? And then they did. It turns out that just having new tools doesn’t help very much. You need to figure out how to use them best and this takes time. But sometimes people just don’t and you need a new generation.

Something similar happened when factories switched from steam engines to electrical engines. Initially, against expectations electric didn’t do much better than steam. A typical steam engine factory is organized around the big steam engine and as long as you just replace that steam engine with an electrical engine, nothing much changes. Only when you realize that electrical engines can be smaller and can be placed anywhere do you get your breakthrough.

It’s possible that with mobile we’re in the same situation. Just as desktop computers in the Nineties only made a difference once we started treating them as more than electrical typewriters with screens, the iPhone will only revolutionize the world economy when it finds its role in business beyond a laptop in your pocket.

It doesn’t add up though. If the computer innovations of the seventies and eighties only showed up in the statistics from the nineties, why aren’t we seeing a productivity boost from the innovations from the Nineties and the Noughts now?

China is stealing our productivity

A second theory is that we’re taking a local perspective where a global one is required like the Simpsons paradox. Everybody knows that inequality is rising and not just in the US, also in say, China and India. Yet inequality on a global scale is falling. As we are moving from a country by country economy to a world economy, we’re moving to a situation where we have globally rich and globally poor people, rather than rich and poor in specific countries. Inequality is up in almost all countries, but globally speaking it is down.

During such a transition it would make sense to focus innovation and investment on China, India and Africa where huge strides can be made, rather than trying to eek out a few percentages in the West. We’ll return to our regular scheduled program of productivity increases after we’ve made the world economy whole.

The problem with this theory is that productivity growth in China, India and Africa recently hasn’t been impressive either. It is solid, but nothing like the spectacular rates one would expect if the productivity growth had migrated there.

Technology giveth and technology taketh

A third option is that technology does increase productivity, but that for every hour we work with increased effectiveness, there are 20 minutes completely lost to browsing Facebook. The 9 to 5 work day is dead; we check our work email at home, but we also check our home email at work. Slack let’s us communicate more efficiently with our colleagues and allows us to distract them from work anywhere using any device at any time. Maybe these forces end up canceling each other out.

This could even explain some of the increases in inequality. People with interesting jobs make more money already and take advantage of the new technology to improve their performance. People with boring jobs that don’t make so much will tend to use the distracting bit of technology. The lawyer listens to a legal podcast during their commute, the parking lot attendant kills time playing agar.io (though some presidents are also tempted).

It would be unlikely though. The two chaotic trends of technology increasing productivity and distraction would have to balance each other out every step of the way to create the smooth curve of slow productivity growth that we are seeing.

There’s more to measure than money

Productivity is defined as GDP generated per hour worked. Pundits love to reduce complex discussions to one number and as far as that goes, GDP is a winner. It is also dated.

The Soviets with their plan economy concentrated heavily on industrial production as a measure of progress, especially coal and steel. It was what drove the Industrial Revolution and what made the Great Powers of the 19th century Great. And they were successful: by 1980, the Soviet Union produced 40% more steel than the US.

By 1980 it had also become clear that economically the Soviets were irreparably behind. I’m not saying Socialism would have worked had they focussed on public consumption rather than industrial production, just that when a new economic epoch starts, we need a new way of measuring. Could that be the case now?

GDP is the value of all final goods and services produced. But how do we measure value when those goods and services are free? Having access to Wikipedia, Google Maps and Whatsapp seems very valuable, yet it doesn’t count in GDP. And what if newer versions of a product are dramatically better yet cost the same? The production of an iPhone SE adds as much to GDP as the original iPhone did, but it’s undoubtedly way more valuable. Or what if a product just becomes better overnight? Tesla made it possible for their cars to unpark themselves with a software update.

A cynic knows the price of everything, but the value of nothing; it seems the same goes for economists. If we add up the actual value, rather than the prices, productivity might look a lot better.

GDP is an outdated way of measuring economic progress. But can it really explain the observed gap? The computing power of my phone would have cost a million dollars thirty years ago, but that doesn’t make me a millionaire. Unless we can materially improve the standard of living, all the cool stuff we can do now doesn’t really help.

Technology is keeping us dumb

The last theory is interesting because if true, it would be relatively easy to do something: technology is keeping us dumb.

In the early nineties, WordPerfect for DOS was the leading word processor (and remember this is when technology did increase productivity). It didn’t bother with a mouse, what-you-see-is-what-you-get or friendly icons. It had 48 functions reachable by pressing one of the F-keys plus optionally the Ctrl and/or Shift key. It was not user-friendly by any means, but arguably once you mastered it, you were more productive than you can ever be in Google Docs.

Buying a flight ticket took a trip to a travel agency. When it was your turn, an operator would type in long strings of commands into a terminal letting you know where your money would let you fly as the result were spewed out. Again, not user-friendly. But effective. An expert still outperforms Kayak on those old green CRTs.

It used to be the case that you would tell Google what to search for. You’d type in the keywords likely to appear in the documents you were looking for. These days Google searches for what it thinks you are looking for. Most of the time this leads to a superior experience. But you can no longer be good at googling.

There are many more examples. London cabbies are experts — Uber drivers just go where their device tells them to go to. Really knowing how to use the German Railways Kursbuch was hard, but it made you an expert — now we all just use Google Maps. There will be no more good drivers once the cars drive themselves.

Technology strives to create a world where anybody can do anything without having to study. Nobody RTFMs anymore. Largely this is a good thing of course, we want complex stuff to be accessible by not just the experts.

But what if the price we pay for user friendliness is an inability to become great? Not only would it explain the technology productivity paradox, it would also offer a way out.

The tools that software engineers themselves use tend to be different. They rely on keyboard shortcuts or even on just typing in commands in a black window with small letters. There’s nothing user friendly about it, but you can always learn more arcane command and always improve your performance.

If only all our technology offered a similar path to continuous improvement, maybe we could get back to progress.

Leave a Reply