James Morley

Let me begin with a riddle: how is it possible for income inequality to increase within countries and across countries, but to decrease for the world as a whole?

The answer to this riddle is “long-run economic growth”.

Specifically, income per capita in China and India has grown rapidly during the past half century. So even though inequality has increased within China and India, this growth has led the living standards for hundreds of millions of people to converge towards those in Europe, North America and Australia.

This convergence is nothing to fear. Economic growth is not a zero-sum-game. But what drives this growth?

Variation in standards of living across countries is clearly associated with different amounts of physical capital such as public infrastructure. So should we simply invest more in more roads and bridges to increase our standard of living?

The former Soviet Union tried this approach and failed.

The problem is that physical capital only explains about one-third of the variation in income per capita across countries. The other two-thirds are “explained” by a more nebulous concept that economists refer to as total factor productivity, or TFP for short.

I have to put quotes around “explained” because we can only measure TFP as the residual component of income per capita not explained by capital.

The point is that massive investment in infrastructure would only ever get even the poorest country one-third of the way to catching up with rich countries. Worse yet, capital accumulation is subject to diminishing returns for all countries.

Yes, income per capita would increase with the amount of machinery and equipment per worker, but nowhere near proportionately. Rather than an easy path to prosperity, capital accumulation quickly becomes a lot like squeezing blood from a stone.

The prediction of diminishing returns to capital was the main insight of the Solow-Swan “neoclassical” theory of economic growth that undergraduate students in macroeconomics have been struggling with for decades – and one that politicians looking for growth elixirs in the form of public or private infrastructure have ignored at their peril.

But it is possible that some of the assumptions and predictions of neoclassical growth theory are wrong. And here’s a big example why.

China’s single-child policy may easily have been the largest social policy experiment in history, with many negative social consequences. But did the policy in part explain China’s rapid economic growth?

The Solow-Swan neoclassical growth theory, which predicts a lower rate of population growth will boost income per capita, would say yes.

Yet interestingly, empirical estimates of the effects of lower population growth on China’s economic growth are relatively small in magnitude, even when assuming neoclassical growth theory is correct about the existence of such effects.

In some sense, small effects of the single-child policy should not come as much of a surprise, because according to neoclassical growth theory, a decrease in population growth only generates a transitory increase in economic growth.

Specifically, the generally accepted theory says, long-run growth depends only on “exogenous” technological change — that is, it is assumed to be unaffected by population growth or capital accumulation.

Economist Paul Romer has developed a theory of economic growth with “endogenous” technological change — that is, it can depend on population growth and capital accumulation.

His endogenous growth theory ties the development of new ideas to the number of people working in the knowledge sector (think of this as effort devoted to R&D). These new ideas make everyone else producing regular goods and services more productive – that is, ideas increase TFP.

There are many variants of endogenous growth theory, but a robust prediction is that an increase in population or an increase in the share of people working in the knowledge sector will increase economic growth.

This theory is quite radical for two reasons.

First, the prediction of higher economic growth for a larger population suggests that neoclassical growth theory, not to mention even more pessimistic economic theories of population going back to Thomas Malthus, got things completely wrong.

Evidently, China’s single-child policy was a mistake, not just for social reasons, but also for economic reasons. According to endogenous growth theory, China and the rest of the world could have had more growth because China would have produced more new ideas with an even larger population.

Second, because ideas are what economists label as “non-rival” (meaning that my use of an idea, like a recipe or a mathematical formula, doesn’t prevent your use of it), there will only be an economic incentive for more people to work in the knowledge sector if there are intellectual property rights such as patents and copyright.

Thus, it is necessary to restrict competition in the knowledge sector in order to stimulate growth, even though this leads to other distortions and disparities in the economy.

As with neoclassical growth theory, it is difficult to point to a particular policy that was implemented because a policymaker sat down and read an academic article on endogenous growth theory.

Romer has recently despaired publicly that disputes among academics working on growth theory have hindered its influence on the real world.

Yet Romer can take heart that politicians are now widely using the concepts highlighted by endogenous growth theory, such as “human capital” and “intellectual property rights”, in debates on how to stimulate economic growth.

James Morley is a professor of economics and an associate dean (research) at UNSW Business School. A version of this post appeared on The Conversation.