Monday, 7 January 2013
Ed Balls and Workfare.
Ed Balls (leading UK Labour Party politician) advocates workfare. I’ve been arguing for that for 20 years or so. See here and here.
But workfare is tricky. Some workfare schemes work better than others.
To get the best out of workfare, it is important to understand the THEORETICAL reasons as to why it works (or doesn’t). So here goes.
The most obvious reasons why it works are thus. First, when the unemployed are given the choice of a subsidised job or benefit withdrawal, they react in one of three ways.
1. Cease claiming benefits because they have other sources of income or support and/or their claim is unjustified or fraudulent so they switch to relying on those other sources. That won’t raise GDP but it benefits taxpayers.
2. Get a regular or unsubsidised job, which raises GDP as well as benefiting taxpayers.
3. Get a subsidised job. That’s not as good as regular job, but it is hopefully better than nothing. As long as output from the job plus other benefits exceed administration costs, then GDP rises.
In addition, temporary subsidised jobs enable "learning by doing" and there is evidence that that is just as effective a form of learning as college based courses and similar. Plus those jobs maintain work habits, like getting up in the morning.
But in addition to those obvious reasons for workfare possibly bringing net benefits, there is another far less obvious reason. It’s way beyond the comprehension of 95% of those who sound off on the subject, but it’s as follows. (And even the explanation below is “dumbed down” for the sake of brevity. For a fuller explanation, see the above two “here” links.)
The declining marginal product of labour.
Assume unemployment is far too high, and assume that it is falling. As that happens, the number of unemployed on each local labour market falls, which in turn means it becomes increasingly difficult for employers to find suitable labour. And when that unsuitability reaches some level, employers resort to an increasing extent to poaching labour from each other, rather than take labour from the dole queue. And that spells inflation. (The poaching is sometimes deliberate and sometimes unconscious in that given excess demand, employers offer more generous pay, which amounts to poaching.)
However, if relatively unsuitable dole queue labour can be subsidised, that makes up for its unsuitability, which in turn means the level of unemployment at which inflation becomes a serious problem (NAIRU) falls, which in turn means government can implement a straight increase in aggregate demand, which in turn raises the number of vacancies (which workfare people can fill ). And note that that increase in demand does not COST ANYTHING in real terms. That is, if the latter workfare theory worked to perfection, there’d be no need to impose taxes to fund the scheme as Balls intends doing. That is, the scheme would be self-funding.
But the big problem there is as follows. Is it actually feasible to distinguish between “suitable” and “unsuitable” employees? And it’s important to make that distinction, otherwise employers will tend to employ suitable or productive employees with the assistance of the subsidy, when in fact the employer would be happy to employ the relevant individuals WITHOUT the subsidy. And that’s a waste.
Well there is trick that can be played on employers that so to speak “calls their bluff”, and forces employers to claim the subsidy only or primarily in respect of unproductive or unsuitable employees. And that is to strictly limit the time that a given subsidised employee stays with a given employer, or maybe even threaten to withdraw specific subsidised employees from their employer at random points in time. I don’t know what the time limit should be: I just have a hunch that it should be about two months.
The rationale of the above “bluff calling” is that it forces employers to think about which subsidised employees really need the subsidy and which don’t.
So there is a clash between Ed Balls’s workfare and mine: he says the jobs should last 6 months. I say about 2. And then there is the “Job Guarantee” (JG) as proposed by advocates of Modern Monetary Theory (MMT): there is no time limit at all there. To put it politely, heads need to be banged together.
Period of unemployment before subsidised work starts.
Ed Balls’s proposes a two year period of unemployment before subsidised jobs are available to the unemployed. In stark contrast, there was a subsidised job scheme in Switzerland where those concerned could get a subsidised job as soon as they became unemployed. I’m not sure whether that scheme is still running.
And under JG, the unemployed can get themselves a subsidised job immediately or soon after they become unemployed. (More head banging needed.)