Ricardo on horses, soldiers, and the rise of machines

The idiots

Rereading David Ricardo’s “On machinery”, I was led back to that chapter in the Principles (On the principles of political economy and taxation, 1817) thinking about Samuelson’s claim that there is something like absolute disadvantage, which is the lack of any comparative advantage.

I thought the idea of comparative advantages implied that there is a place for everybody in the labour market, though that does not necessarily secure employment (one may be voluntarily or involuntarily unemployed or inactive, though employable). But if somebody has no ability in their set of abilities that they are comparatively good at, then they might be out of a job and would be out of the labour market altogether.

They would be somebody out of place like Prince Myshkin in Dostoevsky’s novel “The idiot” – an idiot.

How can that be?

Assume that individual abilities, the aggregate of which is the social set of abilities, are all directly or indirectly useful in the production of goods and services (leisurely activities, and consumption more generally, being indirectly productive with welfare being the end and production the means). Aggregation counts on each and every ability.

People specialize in one of the abilities, called functionings in the jargon (or endowments in Arrow-Debreu), in which they have the comparative advantage.¹ Other abilities that the individuals have are either made inoperative or are used to support the one the person specialises in.

Thus, though every individual ability counts in the social aggregate, one per individual counts in the social set of abilities.

One can assume, with Plato, that individual abilities are distributed by nature, one for each person, or alternatively we could follow Aristotle and assume that there is a social selection process which allocates specific individual abilities into social functionings.

E.g. institutions that attribute social roles and statuses in anthropology and sociology; or the labour market’s self-selection of tasks in labour economics; or the rules that guide the choices and actions in decision and game theories; or indeed the evolutionary rules of selection in changing circumstances.

Selecting individual and social abilities according to comparative advantages is one such mechanism. One can think of the division of labour with each and every person’s role being determined by their comparative advantages.

In the first case, Plato’s case, absolute advantages (what one does best) are the same as comparative advantages (what one does best compared to what other people do best) while in the second case, Aristotle’s case, comparative advantages out of diverse abilities become absolute advantages after specialisation.

One could also follow Descartes and assume that while people differ in physical abilities, res extensa, in cognitive abilities, res cogitans, people are the same. Or, physical abilities exhibit decreasing returns to scale while cognitive abilities deliver increasing returns to scale.

So, differences in physical abilities across individuals are Platonist, they are distributed by nature; differences in cognitive, learning, or more broadly moral abilities are Aristotelian, they are social by nature.

When it comes to res extensa abilities, machines could complement people, who may be more or less deficient in some ability, so that there is the equality of abilities across individuals, but they can also be used as substitutes for specific abilities of people; while in the case of res cogitans, to the extent that they have to be developed through learning, it is the circumstances, including the cultural, religious, and political ones, which account for the differences in people’s abilities.

One can add the requirement that social set of abilities is efficient in Plato’s sense of proper allocation of individual specialisations according to their absolute advantages, which is Pareto optimal.

Alternatively, the selection of individual abilities into social functionings in Aristotle’s sense can be seen as the proper allocation of people’s capabilities, as defined by Sen, i.e. the specific sum total of individual abilities, perhaps enhanced by machines, and moral choices which, the allocation, is at least Pareto optimal, and also just.

In any case, after specialisation, comparative advantages are absolute advantages too (as the other abilities may suffer for lack of practice and thus lead to hysteresis). Put differently, the individual set of abilities shrinks to one in which the person specialises, while the social set of abilities is diverse, so:

Individuals specialise, society diversifies.

That raises the issue of social stability and thus of individual security. Or, put differently, individual specialisation and social diversification raise the issue of proper allocation of risks of the impermanence of social demand for individual specialisations within potentially shifting social diversity, where the balance between social and individual (or private) insurance of risks must be struck.

In Plato’s case there are risks of misallocations of individual abilities to social functionings in part due to commerce, foreign in particular, due precisely to the influence of comparative advantages; and to military conflicts, but also due to internal Machiavellian, Thrasymachus’ type of political passions: tyrannical, democratic and populist ones, even in a closed state, in an autarchy.

So, though comprehensive protectionism is necessary to contain commercial opportunities, it is not enough to contain the political passions.

While in Aristotle’s case there are additional risks arising from wrong moral choices and corrupt political set-ups. In the latter case, moderation and democracy, rather than Platonist specialisation and technocracy is the way to good life and government.

E.g. defying comparative advantages, individuals may choose to specialise in their absolute advantages or in other abilities in which they have the comparative disadvantage, e.g. they choose the wrong vocation. That is an inefficient allocation of abilities and can prove to be unjust and morally corrupt.

The inefficiency, and indeed unhappiness, can also be the outcome because the selection or the aggregation mechanism, the allocation mechanism in general, is inadequate or is being tempered with, as we know not just from Dostoevsky but also from modern game theory and from the efficient allocation mechanisms theory, and of course from the theory of general competitive equilibrium, all of which are basically Platonist.

The social set of abilities, or functionings, is a proper subset of the set of individual abilities, the two being coextensive after specialisation in equilibrium. Different functionings complement each other, while the same are substitutes, they compete for employment.

So, the issue is: are there functionings which are not complements?

In other words, are there individual abilities (functionings) which fall outside of the social set of abilities (functionings).

Put aside the dependents and in particular: the old, who have mostly lost their employable abilities, and the young who are yet to develop their abilities and enter the labour market. The inclusion of dependents leads to issues with the maintenance of social stability and thus with Samuelson’s social contract across overlapping generations and consequently with secular demographic changes, which include the issues of risk-taking and risk-sharing across generations and populations, i.e. the stability of the moral and social web between people living today and the future generations, e.g. within families and through public finances.

From the point of view of the future, we might all be idiots. Thus, history might just be “a tale told by an idiot”.

Appendix – Abilities: Individual and social, Venn diagrams

The blue circle in the Venn Diagrams is the social set of abilities or functionings. The white circle is the individual set of abilities, functionings, or capabilities. The former is the aggregate of the latter.

With comparative advantages aggregating all specialised individual abilities into the social set of abilities, Diagram 1 should be the equilibrium outcome. It is also the equilibrium outcome with self-sufficient, diversifying, individuals within a self-sufficient society (or state or republic). Thus, it covers both Plato’s case and Aristotle’s.

With gradual technological progress, i.e. with gradual introduction of machines, Diagram 2 should be one phase in the process of adjustment to the equilibrium – which could happen through the expansion of the social set of abilities or through the shrinking of the set of individual abilities.

Diagram 3 is the Malthusian adjustment to a technological revolution. It covers the case of comparative advantages across countries due to foreign trade and investment, which make for the interdependence of social and individual abilities across countries.

Diagram 4 shows the labour augmenting productivity technological revolution with declining population. Thus, these are inter-generational comparative advantages with increasing technological sophistication.

There are two adjustment mechanisms to the equilibrium of Diagram 1. Either the set of individual activities adjusts to the social set of activities or the social set of activities adjusts to accommodate the diversity of the individual set of activities, i.e. to accommodate individual capabilities. Both developments can be influenced by policies. Ricardo considers trade and fiscal policies.

To the extent that the adjustment works through cross-countries comparative advantages, the increasing use of machines requires the emergence of some kind of an enduring cosmopolitanism. That ensures, with appropriate policies, that the equilibrium of the Diagram 1 is established or re-established after technology changes.

Enter horses and machines

Now, Ricardo introduces the example of a horse being substituted for a worker. Workers displaced or outcompeted by horses would be driven out of their employment, but not out of the labour market. They had lost their jobs, but not their abilities and comparative advantages. If the landlords or capitalists, the owners of the horses and of the products, decide to invest the extra profits which the employment of horses brings, rather than spend them on luxuries (on Sraffa’s non-basics), there will be jobs on offer for the temporarily unemployed workers.²

Horses are substitutes for people, and the same is true of the substitution of people with machines.

What people do, horses and machines do at least as productively. Horses substitute people out of employment, but not out of the labour market. The subsistence wage will have to decline at least temporarily to bring displaced people back into employment.

Similarly, the profits and the rents may stay the same or increase after the introduction of machines, while either there will be less employment or wages will decline after an inconsiderable period of time (Ricardo counts in years, thus after a year’s time). Because resources will be shifted from the production of subsistence or wage goods to production of machines (e.g. as in Marx’s two-sectoral model).

The mechanisms of adjustment to employment of horses and machines may indeed be different, however. Horses and people may compete at least in part for the same subsistence goods, while machines do not. Thus, employing horses will compete people out of their jobs if horses’ subsistence costs less, if the use of horses instead of people cuts costs, while machines drive out production of subsistence goods, compared to production of machines, and thus either lower wages permanently or reduce employment of people and perhaps drive them out of the labour market altogether.

Which is why horses are substitutes for people and may drive them out of employment but not out of the labour market, while machines may also make some people’s specialisations redundant, i.e. comparatively disadvantaged, driving the displaced people out of the labour market altogether.

In the end, the effects of the use of horses and machines in production on employment and wages, though via different routes, end up at the same outcome – temporarily lower employment with wages temporarily below the previous level or even below the level of subsistence.

Except for the potential effect of the introduction of machinery on the people’s comparative advantages or rather absolute disadvantages. With potentially opposite demographic consequences of the increased employment of horses and machines – the latter having the negative effect.

The machinery problem

Ricardo’s machinery problem is that of the possibility that the net domestic income (profits plus rents) may at least stay the same (which includes the case of zero profits) while the gross domestic income (net income plus wages) stays at most the same– when machinery is preferred to labour. So, the demand for employment may be deficient (Say’s and Walras’s Laws being satisfied however).³

Thus, if investment in machinery cuts into employment, capitalists will not employ the laid off workers as long as their profits are as high as they were before they substituted machinery for labour.⁴

In general, the machinery problem being just an example of Ricardo’s main insight, as brought to light most clearly by Sraffa, that the allocation of abilities and resources is not independent from the distribution of rewards, reflected in this case in the dependence of the gross income on the net income, or put more specifically, reflected in the dependence of wages on profits.

However, in a growing economy, the growth being perhaps due to the technological progress, i.e. due to growing investments in machinery, though the workers are hurt by the greater use of machines, this is temporary, and should not have, and that is important, Malthusian demographic consequences – the population need not adjust, need not shrink to the lower level of employment in order to adjust to the lower supply of means of subsistence.

In Ricardo, as wages are at subsistence level in equilibrium, adjustment to lower supply of wage goods can only work within the business cycle, where in good times wages are above their equilibrium level and in the bad times below. In case of larger swings, Malthus takes over with population adjusting, though not necessarily misadjusting, to the availability of wage or subsistence goods. The dynamics of employment, however, is based on the assumption that labour complements machines:

People can produce without machines; machines cannot produce without people.

So even though temporarily there will be unemployed people, they will get employment over time. Though, for the return to full employment, growth of population may have gradually to slow down or even turn negative – depending on the way and the speed that the constant, Von Neumann or golden path, of machines to labour ratio is being approached. The introduction of robots modifies this assumption somewhat, with labour complementing machines at one additional degree of separation thus putting additional pressure on population growth.

That is so as long as the laid off workers continue to keep their comparative advantages, even if those have to change. Their wages may have to decline with the growing use of machinery, especially if there is the threat that investments might move abroad, but the labour market will adjust to full employment and the wages will grow back up over time.

The introduction of machinery and technological progress in general may present a temporary or transitory, within the business cycle, employment or rather unemployment problem with wages providing for the equilibrating mechanism in the neoclassical manner, though around the subsistence level which is where the wages settle in equilibrium according to Ricardo.

Thus, though wages are flexible, they have much less way to go on the downside, as in Keynes.

Exit soldiers

At the very end of what is indeed an essay rather than a chapter in the treatise, when discussing the foreign trade and investment aspect of technological progress, which is what the growing use of machinery is, Ricardo does not bring in directly the possible changes in comparative advantages. But the context is about changes in productivity and of the exchange rates, which clearly have implications for employment and wages, but also for shifts in comparative advantages.

Before coming to that, he discusses the case of wars and of specialisations they bring about with implications for growth of population or indeed its decline.

When a war ends, there are all those soldiers whose special skills of soldiering are not needed any more. That leads not just to the decline of wages when they enter the labour market but may lead to the decline of population as the soldiers may not have the skills which are needed in the time of peace – they might have specialised them away.

If peace endures, one needs to adjust this possibility of an unemployable specialisation with the consideration of the possible expansion of various criminal activities or of the spread of military dictatorships which may increase the demand for the skills of demobilised soldiers – which Ricardo however does not consider directly (he does perhaps indirectly, within the fiscal context, more on that later), but was very much at the centre of political thought from Xenophon and Plato onwards.

E.g. some kind of praetorianism is a distinct possibility.

So, in the example of demobilised soldiers, what happens is that the set of individual abilities, e.g. soldiering, falls in part outside of the social set of abilities which complement each other when peace breaks out.

Military skills and soldiering as the specialisation have no complement within the social set of abilities in time of peace.

Thus, with complete specialisation, which ensures efficient use of resources, where employment is guided by comparative advantages, if the social set of functionings shifts, or rather shrinks in this case, there will be people who will find themselves not only without employment, but outside of the labour market altogether. Except if they turn to revolutionary politics, take over and run the state, thus changing the whole system of production and distribution.

One remembers that both for Plato and Aristotle good government is rarer than a bad one – it is almost utopian.

Enduring peace is one such shift in the social set of functionings. There will be no specialisation for soldiers, e.g. unlike in Plato’s state, and no need to grow the population, by birth or conquest, to have enough soldiers to at least maintain the balance of powers in the manner of Hobbes or Hegel.

The same might be the effect of the rise of machines, e.g. of the introduction of robots. Ricardo, still not considering the problem within the context of comparative advantages, argues that only the shock of sudden and widespread introduction of new technologies may lead to people with redundant specialisations being driven out of the labour market and indeed out of the social set of abilities altogether, so not just the level of employment, but that of population would have to adjust, in the Malthusian way (which is to say population will adjust if there is not enough food or other necessary goods at the subsistence level for all).

Warring nations might need growing populations, while peaceful ones may not.

Plato thought that the political geography might help, as some kind of natural fortification, so population can stay at the constant, golden land-to-labour ratio. But international security threats are still needed for all those natural-born soldiers to have employable abilities in the well-functioning state (to use Rawls’ terminology).

Except of course if machines take over the art of war and warriors take over the art of politics. These are themes Ricardo does not go into directly, but can be addressed indirectly as he looks at the fiscal consequences of wars and their aftermaths. Soldiers are paid from taxes, and they can continue to draw on fiscal resources after peace settles using political influence exerted in one way or another.

Exit people

Ricardo addresses the question of cross-border investments which may grow if the introduction of innovations, e.g. of machines, not to mention robots, is retarded in one nation but not in the others. Capital movements across borders are almost inevitable in a world with trading nations. So, Ricardo does not need to introduce it as an additional assumption.

In terms of innovations and investments, his world is global. Investment will go where machines are, if machines cannot be employed at home.

That will keep the workers in the domestic labour market, but as unemployed. This is worse than being displaced at home, because of the assumption that machines need people to operate. But, if those are people in other countries, there will be no jobs and indeed no need for workers at home.

They have the abilities, but there is no demand for them.

Machines would have not only substituted people away, but people who had retained their functionings would have lost the needed complements to stay in the labour market, they would have lost their comparative advantages.

People would have to change their specialisations, perhaps reigniting the use of their sheer labour power while moving across industries and territories.

Over time wages would have to adjust downwards to support employment and the exchange rate would have to depreciate to support exports; or population growth will have to stop as the downward adjustment of wages cannot be as deep as needed, if they are to sustain all the people given that they are close to the subsistence level in equilibrium.

So, Malthusian albeit gradual demographic adjustment will have to take over.⁵

Ricardo does not consider emigration explicitly. Though this is to be expected if machines, albeit via foreign investments, are pushing out people. In addition, emigration would be the recourse if the social set of abilities of one’s nation does not include the particular individual ability.

People without any comparative advantages in their own countries might move or emigrate in order not to drop out altogether from the labour market.

E.g. to become a mercenary if one is a soldier in a peaceful country, or if one works with robots in a country, e.g. a backward or a developing one, where the use of robots is not profitable. Or perhaps people will migrate out of countries where there are relatively cheap horses or machines.

All that mobility is in accordance with the extended Stolpar-Samuelson theory – people moving from where they are abundant to where they are scarce while land is abundant.

So various types of emigration and indeed of colonisation are possible. Ricardo considers some of the potential consequences of colonisation for trade (in the chapter On Colonial Trade), but not because of the cross-border movement of people but as a way to benefit from protectionism. In this context, acquiring colonies might support domestic employment with protectionist policies which might also slow down the introduction of machinery, both at home and especially in the colonies.

He does, considering the introduction of machinery, compare the availability of food in America and in England, labour being scarce and land abundant in the former while the opposite in the latter, and argues that machines are more competitive with labour in the latter than in the former country.

That would be conducive to emigration from England and immigration into America.

That example also supports or rather complements his discussion of the opposite cross-border investments from places with high food prices and a low level, let us say, of industrialisation, due to protectionism designed to deter investments in machinery, which would drive investments out of the latter place and drive wages down while supporting outward migration to new worlds with affordable prices of food or subsistence goods in general.

This is apart from the influence of demographic changes due to growing use of machines on migration flows – from more technologically advanced countries to new worlds and from backward countries to the more advanced world.

Or it could drive people to migrate out of existence altogether.

His Malthusian mechanism of population growth and decline is indeed another way to bring in migrations, though not cross-border ones directly. People migrate in and out of employment and out of the labour market and indeed out of the population altogether.

Substituting the exit from existence, i.e. secular demographic decline with regional and cross-border mobility is straightforward.

Rise of machines

Ricardo considers volumes of produced goods and services, of commodities, not values. Wages are subsistence goods and services. So, the fewer people employed, the less subsistence goods needed for their wages. Or, rather, more goods being used to produce machines, less subsistence goods and thus less people employed. While profits and rents, the net income, stay at least the same, there are more machines and less subsistence goods, so wages decline leading either to a lower subsistence level or to less employment.

Secular rise of machines need not lead to persistent unemployment or to permanently lower wages, though the share of wages in the gross income may indeed decline.

However, like the outbreak of the enduring peace, technological revolution which changes dramatically the distribution of comparative advantages will have strong immediate and irreversible long-term effects. At least on the current generation of people.

Assume that a technological revolution brings in robots that change completely the way things are done, e.g. mechanically.

To see what that means, take Plato’s state. It has three classes: legislators (philosophers), producers, and soldiers.  Everybody is fully specialised in one or the other role. Permanent peace breaks out, so there is no need for soldiers and security services in general. And assume that production of food, agriculture, but also food processing, is entirely mechanised, so neither people nor horses are needed to produce food. And assume that profits are distributed to households as in the general equilibrium theory.

Only engineers and technocrats or philosophers, who are also the promotors of the technological advancement, remain in demand by the process of production; otherwise the whole mechanism of production would have no welfare purpose. Even they need to specialise because of the rise of artificial intelligence.

People specialising in physical tasks, in soldiering or in toiling the fields, would lose their comparative advantages and given Plato’s natural distribution of abilities, would simply go extinct as there will be nothing for them to do. Except if, as already mentioned, they find a way to run the world, which the power enhancing technology would actually make easier.

Technological revolution, i.e. the shock of massive use of new machines, e.g. robots or artificial intelligence, might trigger a wholesale loss of employment or spur mass mobility or exit of people and indeed lead to an even more profound shake up of the labour market with a large number of people losing their comparative advantages and being driven out of the labour market in the dramatically changed economy.

Technological revolution does not just change comparative advantages, requiring adjustments in the specialisation of almost everybody, but in addition shifts the social set of functionings far away from the feasible specialisations of large number of people, who end up with no comparative advantages in the newly mechanised economy.

They have to move from country to city or from industry to services. And then out of services to science and learning industry.

That would produce Malthusian effects on the level of population, on the rate of growth of population, and indeed on the stability of the economy, society, and of course of politics akin to mass demobilisation.

Ricardo thought that piece-meal technological progress will not have negative demographic consequences. But that is only as long as the set of individual abilities is coextensive with the social set of abilities. If some of the individual abilities drop out of the social set when they are substituted by machines, e.g. by robots, then the demographic decline may come slowly, but it will still happen.

One should expect that with people becoming comparatively disadvantaged due to the increasing use of machines, there will be a decrease in the growth of population with changing set of individual specialisations which get included into the social set of functionings.

And with the technological revolution, there will be a Malthusian dramatic type of demographic adjustment.

Diversify to adjust

Ricardo considers labour market adjustments with people moving out from industry to (domestic) services. That, though to a limited extent, does require that people are not only endowed with but also preserve diverse abilities.

The difference between the employment of horses and the use of machines comes out here. In agriculture, horses specialise, while people diversify. This is so in Aristotle’s economics, which is about husbandry. The use of diverse abilities is needed to attend to production on land, with horses, tools, and all that. Horses substitute for sheer labour power but complement other abilities of people.

Horses do not increase the division of labour as it were. Machines, however, may require adjustment in the comparative advantages.

Industrialisation of agriculture as well as the rise of industry and the increased introduction of machines support specialisation. But they also increase the demand for services e.g. for domestics. Eventually, the industrialisation of services will change that too.

Thus, people always face some risk not only of losing their current jobs, but of being driven out of the labour market altogether.

Diversification is one way to insure against changing demand for specialisations due to technological changes. And the fact that society diversifies over people’s abilities is not enough of an insurance and can indeed be an additional source of risks due to being based on individual specialisation.

Thus, people need to diversify over their abilities or functionings, which is what one means by capabilities. Individuals need to be able to move from one specialisation to another – they need to have the necessary functionings, the opportunity to use them in parallel or consecutively, and be free or have the rights, they have to be able to diversify.

People need to hold on to their capabilities, to store them, as it were, as human capital, as they specialise in one of their abilities, or functionings. So, they can have feasible opportunities to adjust their specialisation to changing social demand for functionings.

The idea of capabilities is Aristotelian in the sense that it suggests self-sufficiency – of individuals, economies, and states.⁶

Assume that everybody is endowed with a set of abilities which come down to what they are capable of in changing environments. With specialisation, they need to invest in Arrow-Debreu securities in abilities – which amounts to maintaining diverse abilities in order to be capable to dealing with changing comparative advantages. This is what the appropriate demographic adjustment should provide at least in part.

While diversifying one’s abilities is inefficient in terms of comparative advantages with certainty, it is efficient if the risk of increased introduction of machines and possible displacement from the labour market is considered.

Mobility is one way to diversify, as Ricardo suggests, e.g. from country to city, or from agriculture to industry and then to services, not necessarily in that order. Demographic change is another, as Ricardo also suggests.

In general, some combination of individual and inter-generational social insurance of risks through diversification will be required.

There are obviously limits to diversification of physical or inborn abilities, but there are no inherent limits to the scope of diversification of cognitive abilities, though the scale of diverse knowledge is dependent on circumstances, on the history of science and technology, and on the system of education. The latter also being dependent on the individual and social circumstances of all kinds, e.g. on those ascribed to people by birth, geography, social practices, or by culture, and especially by religious beliefs.

Which means that both individual capabilities and the social set of abilities or functionings will need to be managed – by moral and political means. In general, individuals will have to invest in the diversification of their abilities over consecutive generations, which is the moral issue, or the public will have to invest in the adjustment of the individual abilities, e.g. through retraining, or policies need to be introduced which interfere either with the distribution between net and gross income or with the working of the comparative advantages.

Enter citizens

Ricardo envisages the transition from industry, which is increasingly dominated by machines, to services provided by domestics. This coincides with urbanisation, which is also supported by industrialisation and the rise of machines. And with those, come not just classes, but also citizens.

To see the difference that citizens bring to the economic and political set up, comparison with slaves and aliens might prove helpful.

Slaves, Aristotle argued, clearly criticising Plato, his teacher, specialise in what they are by their nature best at. They need citizens, who are political beings, to guide and direct them. In that they are like horses or machines.

Citizens, and even aliens, the latter including Aristotle himself, need to, and aliens indeed must, diversify, though in different ways. If nothing else, citizens have to divide their time between pursuing their private interests and the public interest, even if in the case of the latter that meant turning to soldiering when needed. Aliens take up activities for which they have comparative advantages, which might differ from those they might have originally specialised in. Slaves specialise in what they are best at, as everybody does according to Plato, though Plato did not approve of slavery.

Slaves for Aristotle are like machines, or machines are slaves – specialised single-purpose devices that allow citizens to engage in diverse productive, political or leisurely activities. So, in an Aristotelian set up, slaves are used for their absolute advantages, immigrants for their comparative advantages, while citizens have the opportunity to exercise their capabilities, i.e. chose and change their specializations.

Slaves can be substituted with machines, immigrants can be sent back, while citizens can rely on politics to exercise their capabilities.

To see this, take Aristotle’s definition of justice: to treat the equal equally, and the unequal unequally (citizens diversify, slaves specialise). Then compare it to Marx’s definition of justice (which is in fact Ricardo’s specialisation with subsistence, and equivalent to the neoclassical welfare economics): from each according to their abilities, to each according to their needs. Plato’s definition of justice could be put thus: treat the unequal equally.

The needs, the Marxist welfare aspect of justice, do not appear in Plato. This is because Marx (though his meaning is not altogether clear) had in mind specialisation in capabilities, i.e. the opportunity to make use of diverse abilities, as individual productive contributions, and income as the compensation that supports people’s capabilities and thus includes the distributed net income from Arrow-Debreu securities, as in general equilibrium theory.

Plato, however, had in mind the contributions of specialised abilities or functionings, with the command of land or tools needed for their exercise with the provision of an adequate subsistence income. Philosophers or legislators for instance, do not need much in the way of either wealth or income.

One can think of specialised workers, domestics, and the rentiers and capitalists in Ricardo as Plato’s specialised members of the state or as slaves, aliens, and citizens in Aristotle. One can see the three being combined in a liberal and democratic state that emerged with the industrial revolution, i.e. with the rise of machines.

Capitalism in one country

Ricardo identifies at least the transitory negative consequences of the increased use of machines and considers protectionism to slow down industrialisation or technological advance and the rise of the machines and finds that it is self-defeating due to cross-border movements of investments. Though protectionism does support increased diversification of economic activities and thus of the social set of abilities by interfering with the working of comparative advantages through foreign trade.

In the chapter On Colonial Trade, Ricardo argues that tariffs on the rest of the world, on the other countries besides the colonial empire, enables the colonial power to produce and sell to its colony goods that the rest of the world has comparative advantages in. That enables the colonial power to impede specialisation at home and also to make the colonies structurally dependent, in Hirschman’s sense, on the empire. That slows down or distorts the technological advancement in the colonial power and in the colony respectively.

So, capitalism in one country or within an empire or within regions of a country leads to a slowdown of technological advancement and to regional divergence.

That is a self-defeating policy, if protectionism is limited to trade and there are no restrictions to cross-border investments.

Negative effects on employment from investments abroad could be mitigated by taxation though. Ricardo gives the example of taxation to finance wars. As taxes fall on the net income (rents plus profits), they can be used to finance not just the army, but whatever other employment the government may want to support. That could see the increase of wages at the expense of profits and rents.

So, trade and fiscal policies could manage to increase the social set of abilities and thus to either keep the given set of individual activities and specialisations or even increase the latter given the set of individual abilities. That, however, would come at the expense of the increased use of machines or of technological advancement.

The use of fiscal policy supports individual and social diversification in violation of cross-border comparative advantages and might require democratisation in order to use taxation powers to support employment and wages which are enduringly above the subsistence level. That would be an Aristotelian economic policy.

The aim being to enable the individuals to realise their capabilities, those that they can freely choose or have the opportunity to choose, irrespective of whether they have comparative advantages in the capabilities they choose.

One could see this policy mix – some amount of closedness to international trade with public investment and employment – which allows for greater social diversification and preserves the individual abilities and employment as a kind of Aristotelian Keynesianism.

The refinement of taste

Similar to Plato, Ricardo did not consider the changing demand for goods with the view to increasing the individual welfare, at least for those living off of their labour. Though in Ricardo’s case specialisation comes through the social or economic mechanism of comparative advantages, which regulates social demand for individual functionings, and not because of the natural distribution of specialised abilities as in Plato.

However, Ricardo was impressed by Smith’s observation that there are diminishing returns to the satisfaction of physical needs, but not to the refinement of taste. Refined taste is that which can appreciate diverse goods and services. The aesthetics of refined taste is one of diversity and of increasing appreciation.

So, reproduction of res extensa determines the needed amount of subsistence goods, while res cogitans can very well exhibit increasing returns and the appetite for knowledge and experience may never be satiated.

Increasingly, individual welfare will depend on the diversity or the refinement of taste rather than on subsistence goods.

Ever finer and more diverse taste will demand diverse individual abilities to satisfy them. So, to the extent that machines substitute people in all kinds of activities, and indeed make some of them absolutely disadvantaged, there will be growing demand for diverse individual abilities in providing services which can satisfy the growing refinement of taste or, to put it differently, for improved welfare.

Perpetual peace

Take the world where there is Athens and Sparta, and then there is Persia. And take the Ricardian world with cross-border trade and investment. Demographics are in part determined by the Malthusian mechanism in the international set up of the balance of power. Countries maintain armies and go to war periodically in part to support larger populations. They also get involved in trade wars in order to promote technological advancement or to protect from it.

Assume that Alexander the Great follows Aristotle’s advice and globalises the world, and let it be run by the Kantian cosmopolitan confederation. Soldiers will have to find another occupation or drop out. And the same may happen to other specialisations. Efficiency will require specialisation and thus potentially machines may push people out of the global set of functionings and capabilities.

The cosmopolitan, liberal, democratic – meaning benevolent – government may tax and spend Keynes- or Haavelmo-style to coordinate the growth of net and gross global income and thus of all the incomes of the members of the confederation, but the population growth would have to moderate and potentially turn negative in order for the wages and generally incomes which are not profits and rents (i.e. wages and transfers) to grow.

So, the Plato-Malthus conclusion of demographic stagnation or of the decline of population is the consequence of the rise of machines, of the Aristotle-Sen diverse capabilities, and of the Kantian globalisation and perpetual peace.

Thus, if Kantian perpetual peace breaks out, population growth would slow down or even reverse due to specialization in production, the cosmopolitan insurance from risks of the rise of machines, and because of the increased productivity of science and knowledge which would keep consumption at the subsistence level, which of course will depend on the refinement of tastes.

Thus gradually, Descartes and the machines will take over.

Peščanik.net, 10.10.2019.

1. Comparative advantage is the ability that one is the best at, given the abilities of others. Absolute advantage is the ability one is best at. So, people specialise not in what they are best at, but in what they are best at given other people’s specialisations. In general, most people will be employed to do what they are comparatively best at and not what they are absolutely best at. Thus, most people, workers in particular, could do better, i.e. they are capable of doing better, than the specialisations they are assigned to do by the workings of the comparative advantages. Thus, capabilities are different from functionings in the sense that people would choose to do what they are best at rather than what they are comparatively best at.
2. As Sraffa argued, I think correctly, issues of the returns to scale – constant, increasing or decreasing – are not relevant for Ricardo’s economics. So, assuming constant or even increasing returns to scale is harmless if helpful. Indeed, given the complementarity of activities guided by comparative advantages, increasing returns to scale are to be expected.
3. Say’s Law is an accounting identity of aggregate supply equalling aggregate demand. Walras’ Law is an assumption or a theorem of aggregate zero net excess demand and thus of the equality of aggregate supply and demand.
4. Neoclassical economics unlike Ricardo assumes that the marginal return to capital declines with the growth of capital and labour ratio. That leads to the problems that were at the heart of the so-called Cambridge controversy on theory of capital.
5. There is the so-called transfer problem here, which I discuss in a separate essay.
6. Sen takes capabilities to be positive freedoms. Those are abilities (or functionings) which individuals are not only able to exercise but are actually capable of exercising. So, not only are they free to develop their physical and cognitive abilities but they can choose those in which they want to specialise and can indeed exercise them. These capabilities remain available even as the set of social abilities changes. Given the risks involved in individual choices and social changes, both individual and social diversification that insures against those risks will be needed. In that sense, both individuals and societies are self-sufficient.