Economic Productivity Needs to be Studied in Physical Terms

I have been ruminating on the lessons of the book The Rise and Fall of American Growth after finishing it a few weeks ago. The final parts focused on future growth projections, and I somehow expected to have stronger feelings about this part than what I did. Overall, it was a fantastic piece of work, but there’s one item that I strongly feel it failed to support in a manner worthy of the rest of it — it’s core thesis.

It’s claim is that productivity growth through the 20th century is gone, and the main reason for this is because those years saw enormous improvements in quality of life that can not be replicated moving forward because those improvements can only be made once. There were also efforts to make comparisons between the 20th century advancements and potentially similar advancements that we now have on the horizon, with the latter falling far short in magnitude.

Where the Analysis Cuts Short

Macroeconomics is a study of accounting for various factors, mostly as components of national income. The author followed those division to a very meticulous detail when accounting for the changes in economic product. Beyond that, there were a lot of details to teasing out the hourly productivity from the overall economic product, which involves correcting for hours worked and labor force participation rate over several historical shifts like women entering the workforce.

There was an airtight case for how the directly quantifiable factors will change going in the future. Demographics is a dead giveaway.

But what the book did not attempt to qualify was any real justification about future growth of pure productivity growth from technological and political shifts. Some cornerstone driving forces really can be predicted and pointed out, but it’s much less clear how much these will affect productivity.

Once we really think about the question, it’s pretty clear to me that we don’t have any good empirical picture of what drove past economic productivity changes. Productivity is a measured value, or an assumption that goes into a larger pot to produce a growth number. But it’s not something that anyone seems to do any good prediction or derivation of from something more akin to first-principles.

Relevance of Labor Productivity for our Future

These are no small potatoes either. Given the economic headwinds we fact, we should want to obtain substantial productivity growth for the future to keep up with our expectations. This is exactly what our politicians should want, and there’s a question on the table of whether we’re willing to do what it takes to get that growth. Or, for that matter, is this even something we can affect? I think the answer is “yes”.

I have a hypothetical dream project of my own about the economics of moving into space. I read a lot of economics books, and those seem to largely start from the data and work backwards. For the task I have in mind, we have no choice but to start from the first-principles and work ahead.

Within the book The Rise and Fall of American Growth, I was actually surprised to see so little mention of the impacts of economies of scale in its most pure form. The very idea of a factory is a product of economies of scale, and with it, the modern notion of economic specialization. On the other hand, the implicit effects of economies of scale as a driving factor do get significant mentions. It is mentioned that anti-trade policies have a meaningful risk in the coming decades to eat away at our productivity growth — providing another headwind.

We should not content ourselves to stop the discussion at this point when it comes to productivity growth components. We really can enumerate the abstract economic components that productivity is made of:

  • Economies of scale (improved productivity through specialization)
  • Direct efficiency improvement through sustaining innovation
  • Replacement of certain roles with new, more productive, technologies, what I would call disruptive innovation
  • Expansion of economic inputs other than labor

This might be an arbitrary list, and at this point, it may still be futile to try to fully decompose this topic. That doesn’t mean I don’t want to try.

The hard work comes when you try to apply your theory to real things. Corn fields have increased yield over time, which is a directly measurable form of productivity increase. Ok, so, let’s break down this specific example. In this particular scenario, the “other economic inputs” are fixed when it comes to land, although we must allow some provision for increased fertilizer use and such. Seed, pesticide, and fertilizer use probably has comprised a large fraction of the productivity growth seen here. Some amount of that might comprise of direct sustaining innovation, and other components might come from more disruptive innovation. However, for an established industry like corn, I favor assuming the former more-so than the latter.

IMO, the rubber will really meet the road once we start talking about space colonization with these new kinds of economic glasses on. Space fundamentally entails the input of new economic inputs, and this is something that we maybe otherwise be absolutely starved for. The “filling up” of planet Earth is a bad thing seen from an macroeconomics perspective along several ways of looking at the problem. When there is no more globalization to be had, then there is no more economies of scale. Robots themselves and AI may help to constitute a new form of specialization beyond what the human population alone may suggest is possible. I see this as a particular appealing concept when you consider the impacts of microgravity manufacturing and settlement development.

Emerging Efforts

The kind of productivity accounting I’m talking about is hard. New more highly data-driven economics may play a very big role in this, and a project with has always been a big inspiration to me is the Atlas of Economic Complexity.

This is but one attempt to put real quantifiers on the abstract notion of economies of scale (specialization), involving grouping related and mutually self-supporting industries together. This kind of grouping is tremendously fascinating to me, and a particular extension of the logic I have in mind are seed factories:

Somewhat like a light version of Von Newman probes, the idea is that particular clusters of economic activities can become self-supporting in a way that enables extraordinary levels of productivity growth. In a sense, self-replicating technologies play a part, but reasonably we only need to imagine a form of partial self-replication. But this isn’t realistic in the current economic and physical circumstances that we live in. Expanding the other economic inputs (such as what we have available through the rest of the solar system) seems to be a necessary requisite in order for a new kind of economic takeoff to take place.

These are just some new interesting directions that I see as providing a foundation for a first-principles accounting of economic productivity. I think The Rise and Fall of American growth is basically leaving the narrative at the foot of this intellectual frontier.

Obligatory analytical writing, online participation account for Medium. Engineering, software, books, space, constant daydreaming.