Zero marginal productivity (ZMP) workers are those who would not increase production if employed. They are those whose labor would not be purchased at any price, and may even be refused even if the buyers were paid to purchase the labor (as when the cost to employ them were non-zero).
Picture a modern car factory. In many cases, adding a new worker will not increase the output of the factory, because a handfull of employees can operate the factory, and new machines can be added without the need for additional workers to operate them.
An economy is a little more complicated than a factory: the factory benefits from outsourcing the human input, so that engineers across the country design the machines and, once designed, they can be plugged in and increase output without needing new human input. So couldn’t workers displaced from the factory become engineers? Not always. In many cases, adding a new employee will decrease the effectiveness of other employees (at least over some time period). Training a new employee is costly, and distracts from other work (or from training other employees). This may well be the case for engineering new machines that make cars, because the cost to train an employee to a point where they can contribute to such a project is very high. If a new employee in any industry will decrease the effectiveness of the others by more than they will add in new productivity, they will be a ZMP worker.
There is debate about whether ZMP workers exist in the modern economy. Certainly for most of human history they did not. But as technology advances, the possibility that some workers will be ZMP increases. At some points, some humans will just get in the way of projects already under way. This seems certain if automation continues. And it presents a big problem for policy. What do we do with such workers? Do we just watch unemployment increase? Do we create make-work jobs to keep them busy? Do we provide sufficient welfare to keep them alive? I don’t seem most modern policy approaches as particularly well poised to deal with this problem, but it’s one we would be wise to consider before it presents itself.
Not sure if this is serious, so I’ll respond to it as if it is.
Lawyers are certainly going to be hit hard by the advance of technology in the coming years. Much of what many lawyers do can already be done better by a machine, but the disconnect between people in law and people in tech seems to make deploying that technology much slower than in other industries (other plausible explanations for the slow roll-out is that 1) lawyers are largely managed by other lawyers, who resent the idea that a machine can replace everything they’ve done in their career, and 2) lawyers as advocates have expertise in convincing others, and they can make a persuasive (if not sound) case that machines can’t actually replace them).
But not all lawyers are ZMP. For one thing, lawyers do a lot of things, so it’s hard to lump them all together. For another, a lot of the things that are disparaged as parasitic and unproductive are valuable (which is not to say that there aren’t a lot of parasitic and unproductive branches of legal practice).
But the concept of a ZMP worker doesn’t apply to industries, but to individuals. It’s unlikely that most people working as lawyers will be ZMP workers, at least in the near future. The legal system in the US selects for a constellation of attributes that are valuable in other fields: intelligence, reasoning, discipline, responsibility, communication, research, social skills. Not all lawyers have all these; not all lawyers are above average in any one. But the average lawyer will be above average in most, and so will probably be able to find some form of labor in which they can add productivity, even if the legal industry is largely wiped out by automation (which will happen in the near future).
Even if useless in said fashion, perhaps ZMP workers have societal and other values. For example, a useless worker may have a family who don’t grow up to be useless, they also buy products, and a factory is useless if it doesn’t build what they want.
To function et al, future society may have to employ an entirely ‘useless’ workforce.
“Coming years”
“near future”
Lawyers contribute nothing to production now.
I did.
A lawyer contributes absolutely nothing to the production of a factory. If anything, the existence of lawyers hinders production of automobiles by bogging down the process with inefficient laws. (Laws ‘produced’ by still more lawyers.)
Indeed. We produce everything for us, and the margin between the proportion of products and that of the user, is a deficit where the user cannot buy the products. Ultimately a work force which can afford all products as they wish, manifests the ultimate degree of benefit.
I think they do, just not as workers. The concern is that modern political systems don’t seem to value individuals who can’t be employed. I would argue that they should.
Phyllo, whether an individual contributes zero marginal productivity in law or in coal mining, they contribute zero marginal productivity.
Indeed. They may be thinking of ways out of that, inventing products or contributing ideas, but what’s more important is that they belong to society, have families etc, and the way they are treated is a detriment to society in the greater sense. Then there is cash jobs in construction, and the same people who complain about the unemployed are more than happy to get work done cheap [in what i’ve seen anyhow].
I think the main thing is that products are built for people, and are not self serving only. The amount of people visiting food banks and/or homeless, even around here [oxfordshire] is astonishing for such a rich place.
Maybe it’s the way that you frame these things. You seem to be so concerned with efficiency, technology, information and markets.
In Japan, you don’t kick an employee out of a job when the economy takes a downturn. You find a way to keep him around until things turn around. You recognize the social ties which bind humans together.
America is such a disposable economy. Use things and people and throw them away. Wasteland.
Everybody can be employed. It just takes a bit of effort to figure out how. Society is created for humans, not for markets or machines.
I don’t see how any of what you said responds to the issue I’m raising here. I’m not saying everyone can’t be employed, I’m saying that some people won’t add to overall productivity through their employment.
Nor am I saying that such workers have no value as human beings because they have no value as workers. I firmly believe that over the next century or so virtually all humans will become ZMP workers as automation begins to contribute more and more to net productivity. It is nonsensical to suppose that the very persons who define value in the first place could collectively be found to be of zero value, so, far from denying human value, I think the existence and inevitability of ZMP workers proves its existence independent from employment or productivity.
But the way society currently operates will not work. And I think there’s good reason to think that focusing on employment as the solution will fail (if we really get to a point where workers are net-negative producers, it will tend to reduce net social well-being to try to shoehorn everyone into employment). Better would be to start thinking about how a post-scarcity economy can be run, and how we can transition to one.
As a young punk I lived in a subculture where many didn’t work, my hope is that humans will be fine without work. I think people will be into creative pursuits, which if it were one’s job [like writers etc today], we’d be fine with it. So much of it is in attitude? If made to feel worthless, an increasing future unemployed sector probably wont be positive about it, which I can only imagine will be negative.
This post looks a lot like you stopped reading my post after the first sentence. Let me splice your list of assumptions with quotes to show why it comes across that way:
Fortunately for both of us, they aren’t assumptions I’ve made and they aren’t necessary to my point.
This isn’t always true. Makework is demoralizing. People like feeling like they’re contributing, and they can see through jobs that are invented for the sole person of giving someone a job. Even actually-productive jobs can be dehumanizing.
And setting that aside, this response fails to balance competing interests. I’ve presented a problem: some workers (maybe eventually all workers, as James would have it) don’t add to productivity, and may decrease net productivity, when employed. One solution is to say, ‘we should still employ everyone, because the psychic harm of unemployment is more costly than the loss of productivity.’ But that makes the assumption that there will never be a margin where the addition of human labor won’t impose costs on society great enough to outweigh the cost of the psychic harm of unemployment (not to mention the assumption that nothing but employment can provide a sense of accomplishment, responsibility, mutual obligations, etc. etc.).
Even taking as a given what I think is a not well-supported assumption that unemployment causes significant individual harms that can only be alleviated by employment, it is still possible, even likely, that a time will come when despite those harms, it is still worse for society to try to arrange for all (or even most) humans to be employed.
“Child labor is the future of factory production. Adult workers will be unnecessary and they will not be able to find work. We need to get ready for the modern child-labor economy.”
We don’t need child labor. There are good reasons not to have child labor.
Can you make that analogy more explicit? Is child labor wrong only because it competes with adult labor? Or is automation wrong because it’s exploiting vulnerable robots and depriving them or a more innocent youth?
I don’t see how child labor is a relevant comparison to automation.
They certainly do, but probably more often in the boardrooms than on the factory floors.
I don’t agree. Automation of manual processes has been evolving for at least 200 years yet unemployment rates have not progressively risen. The biggest transition was during the industrial revolution. What that showed is that the end result is that a) productivity rises and people start buying more stuff b) jobs start transitioning to higher skill jobs (making, programming and overseeing the machines responsible for automation process)
The most important policy decisions that need to be made are in terms of education and training - we have to make sure that the workforce is able to cope with with the higher level jobs. By and large, the people struggling in first world nations nowadays are those without the skills and training to cope with higher level jobs. UX designers, for example, are massively in demand, as are instructional designers, in the UK right now - there isn’t enough people to fill these jobs. This goes for engineers and many other trades too.
Only on the factory floor. They need to buy the equipment from a high tech manufacturer. They need to use software packages to track and control production, to monitor their ever more complex financial operations, and for all sorts of other things, and someone needs to build and maintain those.They need digital marketers, tele-salespeople, and I.T. departments to mange their I.T infrastructure.
If you have any good evidence that ZMP workers are on the rise and that this is causing unemployment, fine, but otherwise its a narrative I just don’t buy in to.
In both cases, child-labor and automation, someone is assuming that it is inevitable or desirable or unstoppable or necessary. In fact, we choose the society that we construct.
Do we need or want machines to make products for us? Do we want to be passive consumers?
Do we want humans to make products for other humans? Do we want humans to be active and fully involved in the entire cycle of production?