Employment Policy

I’ve been thinking quite a lot about the question of employment.
It’s an unfamiliar line of thought for me — I’ve always held to the libertarian line on unemployment: it’s a result of an obstructed market, let the market clear and unemployment will not exist by definition. Subsidise unemployment through the welfare system and you’ll get more of it.
That isn’t wrong — within the libertarian framework it’s completely true. But I’ve left the framework behind. Political power will be gained and held by people who believe that gaining and holding power are always a first-order consideration. I hope for a government whose hold on power is so solid that it does not depend on interfering in the market for labour, but that is not relevant to any present government or any feasible near-future one. Welfare is here to stay (even if based on private charity rather than the state, it would still have market-distorting effects), and unemployment will therefore always need to be addressed.
In the meantime, we’re faced with questions like the one a commenter asked back when I was kicking around AI possibilities:

Don’t you think we’re already way past the point where diminishing   returns of replacing human activity with automated activity kicked   in? Most people are just not that smart, they can’t all be designers   and scientists (or can’t be made smart quickly, it is not important   which is true as practical results are similar in both cases), and   it appears to me that we, the societies of the developed countries,   don’t know how to employ these people.

Yes, in a free market these people would have jobs, even if at a wage below what is generally seen as the poverty line. But, if I’m going to be more serious than simply advocating that, I need to face the question: is the combination of politically necessary price control in the labour market with technological advance making high unemployment unavoidable?
I don’t think it has to. There are other causes to high unemployment, some of which can be treated rapidly, and some which are more long-term projects.
High taxation is one of the biggest. Doing work yourself instead of employing someone is by far the greatest area of tax avoidance. I’ve spelled out the arithmetic before: I can do an hour’s work for myself, or I can do more of my normal job and use the income to pay someone to do it. The former is tax free, the latter, at minimum, involves 40% tax on my extra income, then 20% tax on what I pay out. I have to earn £1.68 to put a pound in someone else’s pocket, without taking into account NI or VAT, which may or may not come into it as well, pushing the effective tax on the extra activity up towards 100% of the real cost. That applies to jobs around the house (leading to the dreaded DIY); it also applies, less obviously but probably more significantly, to any good or service I buy where the supplier could, by employing extra labour, provide a better or more complete service. It applies to having the supermarket deliver my goods rather than make me carry them, or even to them having someone work on a checkout rather than wave me to a self-service.
Lower taxes would directly lead to lower unemployment. Also, lower taxes on economic activity would directly lead to lower unemployment — I’ve never written about the question, but I’m pretty sure the Land Value Tax crowd are basically right. Land Value Tax is still a tax, and is still bad, but it’s less economically destructive than the taxes we currently have.
That’s one area then for attacking unemployment: reduce tax, and shift what’s left from income, sales & profit taxes to land.
Next?  Well, the education system. It’s not that it’s failing to teach people “what they need to get jobs”. Rather, the purposeless and ineffective attempts to control unacademic children are actively teaching them not to work. Being forced to do schoolwork is a fairly crappy training for doing real work, but today the bottom stratum aren’t even getting that training. The result is they’re unemployable, not for lack of skill so much as lack of socialisation. It may be only a few percent, but the risk to the employer of getting one of them, and the costs if you do, push a large swathe of the lower classes out of employability.
A demonstration of what I mean came to my mind a couple of weeks ago: a hundred years ago, some huge proportion of the population worked in domestic service. I’ve been meaning to look it up: Tim said yesterday it was 25%.
That dropped sharply from the First World War to about zero by the 1960s, in large part due to the high demand for unskilled labour from mass manufacturing industry. Now that demand has subsided (for good, and inevitably — as also pointed out yesterday by Tim, busy chap). What is the reason why we can’t have domestic service back?  We’re always hearing about how ridiculously stinking rich the the rich are getting, so it can’t possibly be that they can’t afford what the moderately rich of a century ago were happy to pay for. The answer is all too obvious — the equivalent today of the people who were domestic servants a hundred years ago are people that no sane rich person would allow into his house under any circumstance. The late twentieth-century education system prepares normal people for the easily-supervised assembly line jobs that no longer exist, but not for any role requiring any degree of trust or self-discipline. (Having said that, the mass of civilised but somewhat dim people doing largely pointless make-work in the bureaucracy would possibly be capable of roles as butlers or housekeepers supervising the helots… worth thinking about).
The education system doesn’t need to be improved, it just needs to be in large part abolished. Actually doing useful work, for the family or for someone else, is not only a better preparation for being a useful adult than our schools are, it’s probably a good deal more personally satisfying and rewarding as well. The norm should be for people to be in full-time employment by the age of 16, and 13 or 14 is probably a good idea in a lot of cases. The wealthy can do what the hell they want as long as they pay for it themselves, and a sane education system not lumbered with uneducable teenagers should be able to grab anyone from any background with the right talent into a more academic channel, as was routine in this country up until the introduction of comprehensive education in the 60s-70s.
This is a tricky change to introduce, not least because if you already have high unemployment, throwing the bulk of the 16-21 age group into the job market is going to make things worse in the short term. But in the longer term, I think it would improve the situation. Unemployment is not simply the result of lack of work available, but due to the unfitness of a chunk of the population for what work could be available, due to artificially created and prolonged adolescence. The problems raised by Robert Epstein and Lenore Skenazy are relevant here.
Other options?  Well, there’s the immigration question. Again, the libertarian reasoning is entirely correct: if an immigrant is making a living, that means he is producing more value than he is consuming; in aggregate he is making all of us better off. But if immigrants are overwhelmingly competing with poor people in providing services to rich people, on top of that aggregate benefit there is a transfer effect from poor to rich. If it is politically necessary to compensate the poor for this transfer, and if the mechanisms for doing so are unavoidably clumsy and inefficient, then the aggregate benefit can be entirely eaten up. I’m not convinced that that is practically the case, but all the steps in the argument are plausible, and so I am not convinced that it isn’t.
The same form of argument could be made for other forms of protectionism: after all, foreigners compete with natives whether they actually come here or not. But I draw the line at that with moderate confidence. There are so many different ways in which overseas trade affects the domestic economy, all of them beneficial in the aggregate, and while some of them may harm some particular interest or other, the wide distribution of harms means that for almost everyone, the net effect of free trade is positive, and for the aggregate, the effect is so enormously positive that it should not be rejected.
And the same even more strongly for technological development — it is so hugely beneficial that restraining it in order to protect a politically influential constituency from competition is always a bad policy. The reason for paying attention to unemployment and being realistic about the necessity of reducing it is to make it politically easier to hold onto the huge benefits of technology and trade, which it is disastrous to give up.
(The effect of current tax and industrial policies is mainly to encourage more investment in capital goods rather than employing low-skill labour. There is no need for that. But neither is there a need for opposite policies).