The strange paradox of AI and Robotics and productivity growth slowing down

During my Sloan program at London Business School, I had classes with Professor Andrew Scott in a subject called the Global Corporate Agenda. The question above was discussed in a paper I handed in. Below you can find the key highlights of it.

In order to explain this paradox, I will explore three reasons: redistribution, mismeasurement and implementation lags. 

Redistribution

Core to our class was the Solow model that I will use to address this question. In this economic framework technology impact is an input, other inputs are capital and labor. The produced output is (GDP) - which is expected to grow. According to Professor Scott Productivity = Output/Employment.

Above picture from Professor Andrew Scott slides - subject Global Corporate Agenda taught in LBS - illustrates the matter.

Another important concept related to the Solow growth model is the law of diminishing return that states that output capital faces diminishing returns to each new unit added and it depreciates. Consequently, in advanced economies, capital investment is high as it is depreciation. Countries tend to achieve steady states when there is an accumulation of capital. The referred question about productivity growth slowing down is of particular interest for the advanced economies that have to invest more to balance depreciation in order not to achieve the steady-state model. In this state investment = depreciation.

If you look at economic history it becomes clear that in advanced economies, such as the United Kingdom, this situation could be aggravated. According to Mokyr in his book The Past and the Future of Innovation: “Some portion of technological change and economic growth in the near future... is likely to take the form of fixing the unpaid costs ... of past economic growth by redesigning and replacing existing systems”.

The hypothesis is that advanced economies do not grow because new capital created is being used to replace old capital and therefore these countries go back to steady-state, hence redistribution of capital. If capital stock (plant, equipment, and other assets that help with production) do not grow the economy does not grow. To illustrate this matter, I was always intrigued by the Asian Tigers economies spectacular growth. There the depreciation and investment issue was clarified in a discussion in class. In these economies a large part of the infrastructure is new therefore depreciation is less, so the investment to maintain the already existing capital stock is also smaller, which liberates money to be used in new plant, equipment, and other assets that help with production which in the end, will support the GDP growth.

Now that we covered capital, there is another variable to discuss... 

The labor variable might help us to explain the slowdown of economies. According to several authors, including our Professor, labor productivity is impacted by the degree of education (human capital) but it is also affected by diminishing returns. What does this mean exactly? It means that return declines with average educational attainment. So gains are % bigger when you invest in unskilled workers, but in economies based on knowledge (like services which is the basis of several advanced economies), you have less return per each $ you invest. See below Picture from Professor Andrew Scott slides - subject Global Corporate Agenda taught in LBS

Implementation Lags

Going back once more to economic history there is another concept called Engels Pause which refers to the period from 1800 to 1840, when British working-class wages stagnated and per-capita gross domestic product expanded rapidly during a technological upheaval, we understand that phenomenon happens some years after because people (labor) needs to learn how to use technology and then need to restructure and adapt. So that might also be the case for AI and Robotics nowadays, there would be a transition period between today and the future and that could also be seen in our era.

Another aspect to be analyzed is the premise that technology is highly correlated to TFP (total-factor productivity, also called multi-factor productivity, is a variable which accounts for effects in total output not caused by traditionally measured inputs of labor and capital) growth is controversial. Mokyr defends “some portion of innovative effort in the coming decades, ...may be directed to correcting the eventual costs. Such innovations will contribute to economic output and will thus contribute to measured economic growth. But they may not show up necessarily as TFP growth. “

Simplifying, it takes time to new technology to pay off, hence implementation lags.

In a nutshell, considering the usual economic models described above, emerging technologies are currently not mature enough to deliver the full productivity and the total investment is still low to meet depreciation needed to advanced economies (e.g in case of UK, Germany, Japan) to enable the full absorption of the growth.

And how about mismeasurement?

Considering that the usual models crafted for industrial economies present problems such as implementation lags and redistribution they might not be adequate to capture and measure growth trends productivity in digital economies. A further reflection to be done in another article in the future.

But in case, you cannot wait and you would like to learn more about the subject, I recommend reading the following article "Artificial Intelligence and the Modern Productivity Paradox: A Clash of Expectations and Statistics - Brynjolfsson, Rock and Syverson". I found it while reviewing this article and believe it is very clarifying. Click below to read more:

References used in this article:

• JOBS LOST, JOBS GAINED: WORKFORCE TRANSITIONS IN A TIME OF AUTOMATION. McKinsey, 2017
• The Past and the Future of Innovation: some lessons from Economic History - Joel Mokyr, 2017
• https://www.theatlantic.com/business/archive/2014/02/the-dawn-of-the-age-of-artificial-intelligence/283730/ - 10 The Global Corporate Agenda - Class slides – session 8
• Macroeconomics: understanding the global economy. David D Miles; Andrew Scott; Francis Breedon, 2012
• Future Politics: Living Together in a World Transformed by Tech. Susskind, Jamie, 2018
• Macroeconomics: understanding the global economy. David D Miles; Andrew Scott; Francis Breedon, 2012