Slowing GDP growth is an inevitable consequence of technological improvement

A previous post discussed the 50 plus year trend in the decline in productivity growth from 3%/y to 1%/y. This post introduced a model that ascribed all productivity growth to the “goods producing” part of the economy. As the “goods producing” sectors become a smaller fraction of overall GDP, GDP growth declines. The model also shows it takes an additional declining productivity of the goods producing sector from 5% to 3.5% to replicate the observed declining growth rate.

A succeeding post showed how the non-goods service sector has become dominated by sophisticated rent seeking.

As the title of the first post indicates, I originally saw the overall decline in productivity as an aberration or a problem. After some deliberation I have come to the conclusion that declining GDP growth is not a sign of a problem, but an inevitable consequence of technological progress. 

Economists segregate GDP into services and other non-services categories, largely for accounting purposes. Services are any category that does not produce a physical output. However, many services are a part of the goods producing sector of the economy. When we buy a product we are paying for everything including sales and distribution and other business service expenses. The pure service sectors are government, healthcare, finance, education, and a few others.

Technology has made the goods producing part of the economy increasingly efficient. As time has passed fewer people are needed to make more goods, reducing the employment in the goods producing sector. This first happened in agriculture, where employment has gone from around 90% a century or so ago to less than 4% of all employment today. Goods producing has declined from 70% in 1945 to less than 35% today, and the trend continues unabated. By 2025 it will be less than 20%.

The improvement trend has slowed from 5% to 3.5% because not all parts of the goods sector are improving productivity at the same rate, so the slower improving sectors are reducing the overall rate of productivity improvement. Notable slow sectors are construction, where the small scale, regulatory impediments and cultural expectations have slowed overall productivity. The wholesale and retail sectors have suffered from low wages, reducing incentives to invest in improved productivity. Rent seeking patents and copyright have also contributed in certain industries. The overall increased consolidation of business along with lax government oversight of anti-trust and anti-competitive behavior has also contributed.

When historians examine economies before the industrial revolution, they break things into the 90% plus producers of everything, tied to the land and the 10% or so of elites in control who extract the small surplus of these low productivity agrarian economies. These old elites can be regarded as the government services sector of their time. In the modern US economy, the pure services sector runs on the surplus generated by the goods producing sector. The overall services sector has no productivity growth as measured by GDP. GDP growth only comes from the goods producing sector. Ultimately as technology continues to improve, the goods producing sector will shrink to a small percentage of the economy, much like agriculture has already shrunk and GDP growth will slow to very little. Economists have wrestled with the problem of measuring services sector productivity, as the output is impossible to quantify for most services. The answer is simply that services as a whole are a zero sum game with no measurable productivity. They compete for the surplus and “rent seekers” are the big winners. We need the equivalent of anti-trust and anti-competitive regulation for service sectors to control rent seeking. The classic remedy has been to allow rent seeking but tax high incomes. This is a blunt instrument.

Reducing GDP growth is an inevitable outcome, not a fundamental problem. Technology continues to advance, reducing the cost of goods and increasing the effectiveness of the services that are the vast majority of the economy. GDP just does not grow. The looming problem comes from behaving as if GDP growth will continue and basing borrowing on that assumption. For example, at the end of world war two the US had a large government debt. It never paid off that debt. Massive GDP growth made the debt as a percentage of GDP diminish to a small fraction of GDP over time. Today, GDP growth is slow and will get even slower over time. This means that increasing government debt will become an increasing drag on the economy and deficit spending which has become the norm based on growth expectations will have to reduce and eventually stop.

Also, its not just government debt that is a problem. Overall US debt, over two thirds of which is in households, business and the financial sector will not be able to rely on increasing GDP as an engine of debt reduction. Given that total US debt is about 350% of GDP, deleveraging is inevitable as GDP growth declines. Much of this deleveraging will likely involve debt write offs in some future financial crisis. The financial system will have to adapt to a low debt, low leverage world. 

This post and the last two started from an attempt to quantify the effect of the declining productivity of energy production on the overall US economy. While energy is a contributor to declining growth, it is clear the US economy has far bigger immediate issues. This reinforces the perspective that the biggest problem with energy is the longer term problem of CO2 emissions and the growing evidence of their influence on climate change. The long term costs of dealing with the negative consequences of climate change will ultimately lead to a dramatic and perhaps existential threat to the economy. 

By Edmund Kelly