Internet discussion about having government act as employer of last resort (ELR) has flared up in the last week amongst advocates of Modern Monetary Theory. So I thought I’d set out a brief summary of a paper by an opponent of ELR: Malcolm Sawyer (Prof of economics at Leeds University in the UK).
His paper is 14,000 words, so some people might prefer something a bit shorter: the summary below is about a tenth as long. This summary is bound to be inaccurate in some respects. Don’t expect perfection on this blog.
The headings below are the actual headings used in Sawyer’s paper. After each point, I’ve put a brief comment of my own.
Functional Finance and ELR.
i) The value of output from ELR jobs is inherently low, because given full employment, a range of “normal” or regular jobs would exist which are regarded as being more worthwhile than ELR jobs. I.e. given full employment a proportion of (or all ELR) jobs are abandoned, and the relevant labour moves to regular employment.
My answer to that is: “OK, but ELR jobs are still arguably better than nothing.”
ii) Sawyer then divides unemployment up into the usual categories, frictional, structural and demand deficient. Plus he makes the conventional point that where demand deficient unemployment is at a minimum (or if you like, at “NAIRU”), frictional and structural factors are the obstacle to further unemployment reduction. That is, employers cannot find the skill mix they want.
This means that if ELR is used to deal with unemployment when unemployment is a NAIRU, then ELR has a skill mix problem.
My answer to that: “True. That’s one reason I advocate offering the unemployed temporary subsidised places with EXISTING employers, rather than ELR.” See here.
Finance and Money.
i) Sawyer sets out one of the basic ELR claims that used to be put by advocates of Modern Monetary Theory (though they’ve gone quite on this point in recent years.) This is that the costs of ELR can be funded essentially by printing money, and then controlling inflation by the sale of government bonds. But as Sawyer rightly points out, the money printing idea leads to a never ending expansion in the amount of “money plus bonds” relative to GDP, which is unsustainable.
ii) Sawyer’s next point, to quote, is “…why restrict the form of government expenditure in this way?” In other words if employment can be expanded simply by printing money, why not print money and use such money to create extra regular or normal jobs?
My answer: “Good point. In other words, the whole idea that money printing can fund ELR is nonsense, as advocates of Modern Monetary Theory now seem to have conceded.”
Costs of ELR proposals.
i) Under this heading Sawyer points out that some estimates of the cost of ELR include just the cost of labour! He claims that the costs of materials, capital equipment and permanent skilled labour are likely to double the costs.
My answer: “Good point.”
ii) Sawyer then gives another reason for costs being underestimated, namely that ELR would actually draw people into the labour force.
My answer: “What of it? This involves employing those who have given up looking for work, and are thus not classified as “unemployed”. There is nothing wrong with employing members of this “hidden unemployed” category.”
Are the Jobs Available?
i) Sawyer claims that ELR jobs need to be jobs which “do not require much skill or which use skills which are widely available in the population (e.g. literacy, ability to drive). Second, the job leads to the production of useful output, but the output is not necessary in that the output is only forthcoming when aggregate demand is low and the ELR jobs are required. Even work on capital projects (which has often been used to provide jobs at times of high unemployment) would not fit the ELR requirements. Apart from logistical problems of speeding up or slowing down capital projects depending on the state of aggregate demand, much of the work on capital projects is skilled work for which wages are usually significantly above the minimum wage. Jobs such as those in education, health service, personal social services, and care would not be good candidates for ELR jobs. Such jobs may well provide valuable public services and could be expanded as part of mainline public expenditure. But they do not provide examples of jobs which can be undertaken at the basic wage and only undertaken when there is a low level of demand in the economy, generating requirements for ELR jobs.”
My answer: “Good point.”
ii) Sawyers says, “ELR jobs have to be provided virtually instantaneously, for if they are not then someone requiring an ELR job would be unemployed (in reality if not in name). If the capital equipment, material inputs, and supervisory labour for a job are not immediately forthcoming (or standing idly by), then this job cannot be "switched on" to meet ELR job requirements.” And having capital equipment and skilled labour “standing idly by” is a waste of resources.
My answer: “Good point. That’s one reason temporary subsidised jobs with existing employers are better than ELR: the capital equipment (and skilled labour) is already there.”
iii) Next, Sawyer says “an ELR job which did draw on material inputs to a significant degree would generate demand (for those materials) in the non-ELR sector.”
My answer: “Quite right. And that’s one flaw in the claim that ELR is non-inflationary. Or put another way, to create ELR jobs with any sort of respectable output, materials and capital equipment have to be withdrawn from the regular economy. The advocates of ELR never quantify this destruction of regular employment when computing the output of ELR jobs: they just sweep this problem under the carpet.”
iv) Next, Sawyer points to the fact that unemployment can be particularly high in particular geographical areas, or suddenly rise in such areas because of the closure of a local large employer. As he points out, while there may be an argument for having a SMALL proportion of the population doing ELR type work over the country as a whole, having a LARGE proportion doing same in high unemployment areas would tend to result in pointless types of work.
My answer: “Valid point.”
ELR, Underemployment, and Unemployment.
i) Sawyer claims that if the wage on ELR type work exceeds the value of the output on such work, then the relevant employees are making a “net claim” on the rest of the economy.
My answer: “Not a good criticism. The alternative is to have the relevant people unemployed, in which case their “net claim” is probably LARGER!”
ii) Training. In the para starting “To illustrate the significance of these figures…” Sawyer gets the point (not appreciated by many ERL enthusiasts) that there is clash between on the one hand the relatively fast turnover of the unemployed and presumably equally fast turnover of ELR employees, and on the other hand, the requirement that any half decent training on ELR schemes has to last for a considerable or specific period. That is, money spent on a training course that pupils abandon half way thru, is money that is largely wasted.
This deficiency has been substantiated by empirical studies done around Europe over the last twenty years which shows that straightforward subsidised work produces better results than training on ELR schemes.
iii) Next, Sawyer points to the fact that ELR employees have to be available at a moment’s notice for mainstream jobs, which would lead to inefficiencies on ELR projects.
My answer: “True, but that is not a desperately strong criticism since peripheral or relatively unskilled employees in mainstream employment also tend to leave at a moment’s notice.”
ELR, the NAIRU, and Inflation.
i) Where ELR is voluntary, its attractions for those partaking must be superior to the attractions of unemployment. Ergo the RELATIVE attractions of regular employment are reduced. To this extent, NAIRU under ELR (sometimes called NAIBER) will be higher than in the absence of ELR.
My answer: “Correct. I tumbled to this point decades ago as did the Swedish labour market economist, Calmfors, who entitled the effect “Calmfor’s Iron Law of Active Labour Market Policy. The only way round this problem is to introduce what might be called a “workfare” element into ELR: i.e. “do this job else your unemployment benefit gets cut”.