Marx claimed that the introduction of more and more advanced machinery would lead to unemployment of workers, which decreases the ability of the society to buy things, which is a vicious cycle. However, even if some workers are going to lose jobs, more jobs may be created because we need to design and produce those machines. For example, electricity makes the producers of Kerosene lamp unemployed, but it gives rise to jobs such as telegrapher, bulb maker, etc., etc.

So to what extent can we regard Marx's vicious cycle to be true?

  • 2
    Comments deleted. Please don't try to answer the question with comments. If you would like to answer, please write a proper answer which adheres to our quality standards.
    – Philipp
    Feb 8, 2018 at 15:38
  • 1
    electricity makes the producers of Kerosene lamp unemployed, but it gives rise to jobs such as telegrapher, bulb maker - Innovation in consumption does not cause unemployment, innovation in production does. In the case, electricity, as a final good, probably creates jobs. It however replaces human manual labour, and as such is a cause of unemployment. Nov 10, 2018 at 12:47

5 Answers 5


In short, it turned out to be not nearly as bad as Marx though it would be during the 20th century, and how much it will impact the 21st century is in debate nowadays.

Note that Marx has never been the only one to think this. It is called technological unemployment, and it was also what lead Keynes to think we would work 15 hours a week by now. It could have very well been true on a massive scale, like Marx or Keynes imagined it, were it not for the development of the consumption society we live in.

More precisely, it was tackled by:

  • The appearance of new markets, the telegraph market in your example.

  • The development of planned obsolescence. It consists in making people buy more stuff by making the so called stuff break faster/out of fashion/... Thus maintaining the consumption levels. This idea was inspired by the fear technological unemployment, but if industrials effectively implemented it, it was only in order to make more money.

  • A larger drive to consumption with the explosion of advertisement (in order to create "percieved obsolescence" for instance, make people buy more stuff)

  • The explosion of the service industry. The thinkers who feared this technological unemployment mostly think in terms of material goods, although they are well aware of many jobs not being about material production. It developped much more than what they imagined.

  • New needs were met due to technological improvement (see @MSalters comment); hence new jobs appeared. Those include many health related issues, communications, ... Those new jobs can be found both in manufacturing and service industry.

Note that the idea that we need to do less work is coming back. This time it is not due to sheer industrial power, like what Marx was worried about. It is due to automation in general and artificial intelligence in particular. Again, there are answers already proposed to this, for instance the universal basic income.

  • 7
    It's a bit suggestive to state that people only want new stuff because of planned obsolescence and "percieved obsolescence"[sic]. For instance, toothpaste (1880) postdates Das Kapital (1867). In general, healthcare is not solely part of the service industry, nor is it a new market.
    – MSalters
    Feb 5, 2018 at 15:54
  • @MSalters Good point, I will mention this. Feb 5, 2018 at 16:07
  • 9
    -1 for the claim that "What made unemployment a problem was pretty much the industrial revolution, before that no one heard of mass unemployment." This either is (more or less) deliberately obfuscatory or belies a profound ignorance of actual economic fluctuations in pre-industrial societies. Unemployment and underemployment were chronic in most pre-industrial societies (especially those in Europe). Maybe it just wasn't as immediate noticeable as it would become in industrial societies, but that's mostly because most people lived in rural communities, where unemployment is much less "visible".
    – Mico
    Feb 5, 2018 at 17:06
  • 4
    I don't know a lot about economics but this answer doesn't make a lot of sense to me. First, the industrial revolution changed the nature of "employment" so much that comparing "unemployment" before and after it seems illogical. Second, the asserted reasons why we still work seem to be conjectures without much support. Most importantly, it seems to me a glaring omission that despite well over 100 years (and I suspect it goes farther back in time than that) of predicting machinery putting people out of work in general, it still hasn't happened at all. Feb 5, 2018 at 18:52
  • 1
    @ToddWilcox It is not conjectures as to why we work; it is stuff that actually make us work today and that Marx didn't predict. I understand the the question this way. Feb 5, 2018 at 20:07

In order to simplify the analysis greatly, I'll assume an economy with a single good, and that marginal productivity determines wages perfectly (ie: the market is perfectly competitive)and infinite demand for said good (infinite want is a good assumption in economics, however demand for a single good is often finite, ie: a person may switch from desiring another unit of flour when someone delivers 10 tonnes of it to their door a day, to desiring say a car. The single good in this model is assumed to be both the only one, and not suffer from the desire to not have it that the flour does in the above example). In this case we say that workers with capital can produce 100 units of some good per unit time. With no capital it is say 10 units of goods per unit of times.

We will assume that capital is in slightly greater supply then labour in this initial state, ergo the marginal product of labour is 100 units, and the marginal product of capital is 0. (Now no one would bother providing capital at 0 return, however essentially it is assumed that the wage and profit approaches but does not equal those 2 amounts) Now in addition we assume that a new method of utilising capital appears that allows the old capital that needed ten labourers to be utilised with 1. The initial equilibrium is changed such that if 1 more labourer is added only 10 units of additional production is obtained, as all of the capital is engaged with a 10th of the labour force. Because of this wages fall drastically.

So there is a way that it can cause however problems with this model are that

  • It only applies transiently, as any capital accumulation will push wages back up from 10 back to a 100 and then to 1000 over time (higher then initial wages).

  • It assumes that no labour is used in producing capital, if it were demand for labour could rise greatly, pushing up wages.

  • It applies only to cases where innovations are purely labour saving, if innovations work purely on the productivity of capital rather then being labour saving, the only effect is a rise in capital owners profits, and an eventual pushing up of the wage rate later.

To simplify what I mean by labour saving vs capital productivity increasing

Let's say in a factory a worker has to press a button at a certain time and this causes 1 unit of a good to pop out. If an innovation occurs such that the workers machine can be wired to 10 other factory lines, this would reduce the need for labour by 90%, and the firms production would be unchanged. On the other hand if an innovation occurs such that the button press spits out 10 units of good, this is an improvement in capitals productivity. The firm keeps on its workers but produces 10 times as many goods as before.

So essentially Marx's assumption will only be correct in a very limited set of circumstances. Very few of them applied historically as a lot of innovations increase capital productivity, rather then save labour, and the effect is only short term.

  • What if there isn't demand for 10 times as many goods at a price which is profitable? Jul 29, 2021 at 8:28
  • In this toy model, that would imply that production would curtail until the amount of goods was where they wanted, and the workers would work fewer hours. Jun 1, 2023 at 17:24

It wasn't just Marx that said this, Keynes also speculated that increasing automation would eventually emancipate men and women from the neccessity of labour. Both of them were writing when the automation of physical labour was just becoming a reality; lately, we are seeing a second industrial revolution where cognitive labour is being automated and people have begun to speculate where this might lead us. Though neither of Marx or Keynes forecast this possibility, this is not at all surprising, as in their time no cognitive labour whatsoever had been automated - but in essence, it's part of their forecast.

Work or labour is has two distinct aspects, it's physical and cognitive aspects. If both are capable of being automated then the possibility of a world without the need for labour arises. In certain situations this is already taken for granted. For example, we see historically that certain professions have become obsolete and likewise people have forecasted that in the present and the future that certain professions too will become obsolete and this has become unsurprising; few people, though, speculate about the end of all professions - the assumption being that professions will change and adapt - and it is specific modalities of the professions that have become obsolete. After all, we no longer have engineers who know how a radio valve works and can fix it - but we still have engineers.

Hannah Arendt distinguished labour from work; the first is a neccesity due to the fact that we live in a world and must secure food and shelter; she distinguishes this from work, which is the work a man or woman freely takes up as a mode of enquiry, of expression or of politics. And this ties to her third category of human possibility - that of action; men and women are free when they are free to act. Labour negates that possibility, whereas work helps fulfill it.

However the emancipation of humanity is not merely a theoretical question, it is also a practical, pragmatic and political question; as the world stands now, the goods of the world accrue to a man due to the labour he puts in to getting them; this was the view of Locke and foundational in his political philosophy; it's also the view of the middle classes and thus part of the ideology of the West and because of the reach of the West, for much of the world now; it's also entered into myth, I mean the mythos of the American dream; this is then the countervailing pressure that works against the possibility of emancipation; and this is no small thing, such an ideology has tremendous currency and pressure; and not merely ideology, but in its practise, for it has concrete effects in the world and is organically linked to it having grown and matured from a world where labouring was a neccesity.

Theoretically speaking, though; were we to posit a world without the neccessity of labour; the question of how we were to divide up the goods of the city, or of the nation still arises and this is and would be a difficult question; and another counter-vailing pressure.

The other important point to put forward is that neither Marx, nor Arendt and nor Keynes put any kind of time line when this emancipation might occur; given that Marx thought in economic categories, I think it is safe to say that a world without the neccesity of labour lays the foundation of a different kind of economic system; when we consider that hunter-gatherer societies were around for tens of thousands of years, that city states were around for millenia, that feudalism was around for a millenia, it's not likely then from that historical view that our current epoch of capitalism will be short; if we date it from the Industrial revolution in England then we can say it's been going for two or three centuries. But can we put a timeline on it? Well one suggestion is to see that the economic epochs have been gradually getting shorter. The last epoch - feudalism - lasted a thousand years; if we halve that to allow for this progressive shortening then we can posit a (crude) timescale for Capitalism - five hundred years. So we're roughly halfway through that period. Capitalism still has a long route ahead of it, as much as it has behind it; and much will change in that period but nevertheless the trajectory will be the same; Some people in sympathy with the ideas of Marx have banded about the term 'Late Capitalism' as though Capitalism was just about to expire some time soon; but in this analysis, it's in its maturity; and this is what the term 'Globalisation' essentially captures; Capitalism had no competitors at the present time and nor on any horizon; thus at present, there is no real and authentic horizon for change; but nevertheless, change will occur. This is one of the great lessons a study of history can tell us and it was the lesson that Marx, Engels & Keynes were trying to teach from their study of economic history.


There does exists a problem where new technologies leads to more employment, but the mechanism is different than what Marx proposed. As explained in detail by David Bowie, when new technologies arrive certain jobs become obsolete, but this also creates jobs that didn't exist before. This explains why the massive changes that have happened since the Industrial Revolution have not led to massive unemployment today.

To put things into perspective, with the best available supercomputing facilities we can just about control a drone to avoid obstacles, but such a system performs a less well than a bee. And the bee is then actually flying with its brain, while the drone's brain is far to heavy to be put on the drone.

However, despite the fact that the best machines we have are too dumb to make us redundant, now do see unemployment due to recent technological advances. So, what has changed if the brainpower of the machines isn't relevant? What is happening now is that the rate of technological innovation which induces changes in the jobs market is happening on time scales that are shorter than the time scale on which the job market can adjust itself on.

When people enter the job market around the age of 20, then there are only limited retraining opportunities during their careers. As long as the technological advances that causes their jobs to become obsolete happen slowly enough so that the retraining facilities are good enough for them until they retire, there is no problem. The new generation entering the jobs market may find that the work their parents did doesn't exist anymore, but they'll have had an education that's adapted to the new situation.

What we're now seeing in certain sectors is that entire classes of jobs become obsolete on short time scales (say a decade), making the retraining needed more like going back to school for several years. This then means that employers are going to hire the newly graduated people from school. These new employees may in turn find themselves in the same boat as the people they replaced after just a decade or even sooner. This problem will have to be addressed by changing the educational system to a system where people get lifelong access to educational facilities.

Now, we can still ask if in the future when machines will start to outsmart humans, Marx will be vindicated. I don't think so, because the more intelligent a system is, the more it will want to let other systems do all the work for them. So, we'll actually get into the opposite problem, we risk becoming the slave laborers of our super-intelligent machine masters.


Marx’s fundamental conclusion about mechanization, automation, etc., is not only that it replaces labour, but that it tends to shrink profitability rate (not necessarily marginal profit. It is this marxist law why they predict the final collapse of capitalism and (only then) the subsequent inevitable rise of socialism.

Okishio was a marxist that in the 1960s tried to prove this reduce-profitability-rate law mathematically. However, he showed that didn’t happens in his mathematical model of a marxist economy due to the the eigenvalue of his (Frobenius matrix) having to be definitely positive. Subsequently, many analytical marxists have tried to disproved Okishio’s theorem (among them, Andrew Kliman).

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .