Before reading this post, you should listen to Andrew McAfee, principal research scientist at the Centre for Digital Business, at the MIT Sloan School of Management.

Robots and algorithms are getting good at jobs like building cars, writing articles, translating — jobs that once required a human. So what will we humans do for work? Andrew McAfee walks through recent labour data to say: We ain’t seen nothing yet. But then he steps back to look at big history, and comes up with a surprising and even thrilling view of what comes next.

I’m not interested in science fiction and even less so in intelligent machines.  I assume the existence of intelligent machines to revisit economic history. 

Economic history in this analysis is highly stylised. Historians will be unsatisfied and possibly shocked; economists, used to stylise facts, will feel rather at ease.

Until the dawn of the industrial revolution (1750-1800), economic activities can be stylised by a production function with two factors: labour and land. Labour, draught animals, wind and hydraulics powered economic activities.

In the pre-industrial age, only an increase in the endowment of labour could increase total output of the economy, hence the role of slaves and war captives. (1) Nevertheless, total population was constrained by the amount of food that the agricultural sector was able to supply. The pre-industrial model, also called the Malthusian model, “had a negative feedback loop whereby, in the absence of change in technology or in the availability of land, the size of the population was self-equilibrating. (2)

In the pre-industrial there was no labour market. Occupation was determined by birth and by highly regulated guilds. We can refer to a world of ruled occupations, (3) with neither geographical nor social mobility.

The industrial age can be stylised by a production function that has capital (machines) rather than land as the main production factor associated with labour. Through investment, machines are forming the capital stock. The output per hour worked or productivity is no longer constant. To quote Alan Blinder, productivity is the name of the game. In fact, at the beginning, machines were creating fewer jobs than they were destroying. It would take a century ending around 1850-1875 to bring the labour market to its equilibrium, the point where the accumulated capital stock reached a level where some scarcity appeared on the labour market, forcing the equalisation of wages and productivity.

The change in the production function led to a dramatic institutional change: the establishment of a free labour market.

If landowners needed a permanent and stationary workforce, manufactures needed to freely hire and fire the labour force according to changing market conditions. The dawn of the industrial age is associated with the hideous army of proletarians but it is also the dawn of civil freedoms and education for the masses. “The capitalists who were striving for an educated labour force had interest in policies that promoted the education of the masses, whereas the landowners whose interest lied in the reduction of the potential mobility of the rural labour force, favoured policies that deprived the masses from education.(4)

The appearance of a free labour market is the “Great Transformation” (5) engendered by the industrial revolution that allowed an always wider geographical and social mobility. The real change that is brought about by the 18th century “is not the necessity to work, but the freedom to work”.(6)

Scientific research and technological development became an integral part of the economic sphere, fostering continuous innovation in goods and processes. The level of technology becomes endogenous and increases over time.

Are ICT leading to another “Great Transformation”, overtaking the industrial age? The computerisation of human activities could be stylised by a production function including a new production factor: intelligent machines. In the same way that the industrial revolution disconnected the production of goods from the limits of human physical force, we would be at the dawn of a technological era where the superior functions of the human being are commoditised, reproduced at wish, at ever-lower cost (computing power of semiconductors).

Robin Hanson (7) examines the economic consequences of intelligent machines. To allow for the possibility of intelligent machine, labour encompasses now human labour and intelligent machines (and they are perfect substitutes).

The model leads to the following conclusions. When computers are very expensive, they do few jobs. Human labour is very important then, and so human labour wages rise as the economy grows. Economic growth is mainly determined by improving general technology magnified by increases in ordinary forms of capital (productivity).

When computers do most jobs, human labour is relatively unimportant, and whether human wages rise or fall depends on whether owners of capital place a strong special value on services that only humans can provide. Actually, human wages might fall below human subsistence levels.  However, per-capita income for humans can increase if humans retain a constant fraction of capital, perhaps including the wage of intelligent machines.

The formal result that wages might fall below subsistence levels is best illustrated by the Roman Empire, at least if we consider a slave as an intelligent machine.

From an institutional point of view, the slave belongs to the owner and is similar to capital.(8) Rome’s imperial expansion increased dramatically the inflow of slaves and farming was increasingly done on large estates using slaves. Free peasants no longer could compete.  According to Keith Hopkins, Between 80 and 8 BC, in two generations, roughly half the free adult males in Italy left their farms. Many went to Rome, forming a proletariat of free landless citizens, depending on handouts of free food supply.(9)

The political solution was Panem et circenses!

Conclusion

With technological progress and increasing captal-labour ratio, labour productivity and wages increase

The distribution of income between wages and profits depends on the relative price of labour and capital and the ease wih which they can be substituted

Assuming perfect subtitution and considering that technological progress embodied in intelligent machines will make them cheaper over time, the share in total icome of human labour will decline

Could intelligent machines lead to another “Great Transformation”, overtaking the industrial age? Yes, in so far that the labour market herited from the industrial revolution would have to be redesigned.

1. Nils-Petter Lagerlöf (2002), The Roads To and From Serfdom. “For slavery to dominate, agricultural technology must be sufficiently advanced, so that a surplus can be generated to feed both slaves and non- working guards. Moreover, since guards are unproductive, slavery also requires a sufficiently large population to be profitable. One indication of this is the seemingly parallel emergence of larger-scale warfare and slavery and the common use of war captives as slaves in early civilizations.
2. Oded Galor and David Weil (1999), From Malthusian stagnation to Modern Growth.
3. Robert Castels (1995), Les métamorphoses de la question sociale.
4. Oded Galor, Omer Moav and Dietrich Vollrath (2006): Land inequality and the emergence of human capital promoting institutions.
5. Karl Polanyi (1944), The Great Transformation.
6. Robert Castels (1995), Les métamorphoses de la question sociale.
7. Robin Hanson (1998), Economic Growth Given Machine Intelligence.
8. Aristotle (350 BCE),: “The master is only the master of the slave; he does not belong to him, whereas the slave is not only the slave of his master, but wholly belongs to him. Hence we see what is the nature and office of a slave; he who is by nature not his own but another’s man, is by nature a slave; and he may be said to be another’s man who, being a human being, is also a possession. And a possession may be defined as an instrument of action, separable from the possessor.” Politics, Book 1, Part IV.
9. Keith Hopkins, On the Political Economy of the Roman Empire.