The idea that government act as Employer of Last Resort (ELR) has been around for centuries, and numerous attempts have been made to implement the idea.
For example there were the work houses of the 17th, 18th and 19th centuries and the WPA in the United Sates in the 1930s.
Those currently keen on the idea include, 1, Warren Mosler, 2, Bill Mitchell and 3, L.Randal Wray. (Note: the three sites that these links lead to are far from being the only works on ERL produced by these authors).
Some trenchant and valid criticisms of ELR have been made by Malcolm Sawyer (1). However Sawyer confines himself to pointing to weaknesses in ELR without remedying those weaknesses. The paragraphs below do a bit of "remedial work".
Those advocating ELR normally assume that the work concerned should be public sector, and on the grounds that such work seems to be non inflationary because no addition to demand is required. Indeed, I don’t have much of a quarrel with Warren Mosler’s claim that ELR can in theory produce 100% full employment (with the emphasis on the "theory"). Though it would be near impossible to guarantee meaningful work for all in a small towns where one dominant employer went bust.
But there a couple of catches.
First, those doing this sort of work will tend to be relatively unskilled, plus they will turn over relatively quickly (if they are to be as available for the regular jobs as when unemployed). And if those doing this work have none of the usual associated factors of production ( permanent skilled supervisory labour, capital equipment, materials, etc) their output will be hopeless. (Hence the nickname the WPA acquired: “we piddle around”).
On the other hand if they DO have significant associated factors of production, this makes the scheme inflationary for much the same reasons as imposing extra demand on the private sector is inflationary. Indeed, ASSUMING WE DO A LIKE FOR LIKE COMPARISON BETWEEN PUBLIC AND PRIVATE SECTORS, there is no difference in the inflationary impact of the two sectors.
To illustrate, assuming an economy is at NAIRU, why is an expansion of the private hospital sector liable to be inflationary? Because it involves extra demand for various skills that are in short supply: doctors, nurses, architects to build more hospitals, labour to make specialised medical equipment, etc etc. Now assume a similar expansion of the public sector hospitals: EXACTLY THE SAME EFFECT !!!
So how come ELR can, at least in theory bring 100% full employment? Well it’s not because the employment is public sector: it’s because ELR (assuming for the moment no associated factors of production) involves, 1, creating jobs where those concerned are as available for normal jobs as when unemployed (i.e. aggregate labour supply is not reduced), and, 2, no additional demand for skilled labour is involved.
Now can't the private sector create jobs with the above two characteristics just as well as the public sector?
Conclusion so far: it doesn’t look as though the above preference for the public sector holds much water. Indeed there is much to be said for private sector ELR work because the private sector is better at employing unskilled labour than the public sector.
The second catch.
The second catch is this. If ELR has a small amount of associated factors of production, output will still be hopeless. But if it has lots of these factors, ELR becomes little different from an existing regular public sector employer. Indeed, and to put this another way, if the output of ELR people is to be maximised, they need to be found jobs in "institutions" or "firms" where the ratio of different factors of production is the same as obtains with normal employers. And what are these "institutions"? Well they are (to all intents and purposes) normal employers! (And NOT specially set up schemes like the work houses or WPA.) Put that another way: why make a distinction between "institutions" that employ ELR labour and normal employers?
So what “relatively unproductive” jobs are there that existing employers might create? Well, it’s not too difficult: sub minimum wage jobs. This is not to suggest that the take home pay per hour’s work of those involved should be below the legal minimum. The point is that for every employer, “employees” are a resource, which like every resource, involves diminishing returns.
That is employers are happy to pay for labour down to the legal minimum, the union wage or whatever. But most employers can perfectly well create further employment (i.e. sub minimum or sub union jobs) as long as employers themselves don’t have to finance such employment (or at least pay the full bill for same).
Do we want relatively unproductive work?
The answer is possibly yes, and for several reasons. 1, it might reduce the bill that taxpayers’ pay to support the unemployed. 2, Second, it is not clear that because employers pay little or nothing for an employee, that the latter’s output is worthless, and this is for two reasons.
First, advocates of more public sector goodies (public health care etc) have never argued that because customers do not pay for the product at the point of delivery, that therefore the output is worthless.
Second, imagine employment is a little above NAIRU. Why are consumers not prepared to pay for the output of those who would find employment were unemployment reduced to NAIRU? Well it’s clearly not because consumers regard such output as worthless: when consumers have the funds, they DO PAY for this output. The reason consumers are’nt prepared to pay for this output is that they CANNOT AFFORD IT because of inadequate aggregate demand!
As to the hoary old myth that people would be better off training than doing low grade work, evidence from round Europe doesn’t support this. That is, the evidence is that, at the margin, subsidised work produces better subsequent employment histories than training. Or to put in plain English, learning by doing is better than many of the half baked training schemes on offer.
1. Employer of Last Resort: Could It Deliver Full Employment and Price Stability?
by Malcolm Sawyer 2003 Journal of Economic Issues, Vol. 37, 2003.