Labour Market Monopsony

This is a continuation of an argument raised in a previous post.

Labour market monopsony doesn't exist except by government fiat. Well I'm glad that's sorted. Now we can all go home! In all seriousness, though, real economists, not mousy bloggers like me, have already put paid to this idea.

Ultimately if you've gotten over the death of objective value theory (and if you haven't, try) you are likely into labour exploitation on the margin per Joan Robinson et al. This is the idea that the gap between marginal revenues as a discrete quantity and labourers' wages is the means by which to measure exploitation.

Marx would have plumped for total income. Robinson isn't quite so ridiculous. She looks at marginal income (net income / profit) only, recognising that a business is a necessary point of financial intermediation between employees on the one hand and customers and suppliers on the other. Some non labour expenses must be paid directly by the business itself, and wages as a device save the employees having to wait on every quarterly statement before knowing whether they'll get paid.

So the only thing that's egregious to Robinson is the gap between profit and wages, that all profits should be added to wage bills as a bonus each month or quarter. This is obviously impractical in anything but a very intimate service enterprise in which everybody knows everybody else and capital costs are very low, saving the need for savings and investment from the pool of previous profits.

But what of the analysis itself, the idea that wage workers are being treated unfairly, and that the reason is the monopsony power of employers? Robinson identifies the imperfect market, characterised by imperfect competition, as the cause of the market monopsony power of employers.

Diagram showing the imperfect market according to Robinson
B= average cost (including rent)
AR= average revenue
α=marginal cos
MR=marginal revenue
OM=monopoly output
OQ=competitive output
DQ=competitive price
PM=monopoly price
C=intersection of marginal revenue and marginal cost

Mobility of labour is an important concept to engage with here. Labour mobility is the ease with which labourers can move from one narrowly-defined labour market (barista) to another (checkout) measured using the availability of other potential employers, whether in the same labour market or a different but attainable one.

The sharp-eyed reader will notice that the diagram above deals with monopoly, not monopsony, but Robinson's theory is that both arise for the same reasons. To see monopsony at work another diagram is called for, this time from Wikipedia;

Diagram showing the depressive effect of monopsony power on employment levels and wages;
w & L are wage and employment levels under monopsony
w' & L' are wage and employment levels under perfect competition

The diagram above shows that the wage rate and employment level will both be lower under monopsony than under perfect competition. Full disclosure, I don't believe in perfect competition, efficient markets, or full employment. It's antithetical to my Austrian approach to humour such ideas in a serious analysis. But humour them I must, if briefly, since Robinson's own work is intimately tied up in Marshallian and Jevonsian diagrams and mathematics respectively.

This monopsony is significant to Robinson because of its welfare implications. Welfare economics refers to that body of economic theory which deals with how well or how badly people do out of different economic arrangements including different market structures. Expressed as an equation the welfare implications of Robinson's working can be summed up as;

R' is marginal revenue
w is the wage rate
e is the rate of exploitation

The rate of exploitation, e, is therefore the gap between marginal revenue and wages. Why this constitutes any sort of undesirable exploitation is a mystery to me, but for the fact that overall fewer people are employed and are paid lower wages than in a competitive (on both sides) labour market. So we have the model explained, and its supposed implications for folks' well-being.

In a fancy study by the NBER it was found that larger firms paid employees more than small ones for basically the same work. This is the opposite of what Robinson's model predicts.

The researchers found some starting facts. For instance, after controlling for individual and store characteristics, firms with at least 1,000 employees pay 9% to 11% more than those employing 10 or fewer. Looking at individual establishments rather than firms, large stores pay 19% to 28% more than small ones. This is consistent with research in non-retail industries that finds, all else equal, big firms tend to pay more. In addition, they find that retail managers make more per hour than non-managers in manufacturing, and that there are more managers in retail than in manufacturing.

That makes Robinson's model wrong then. Well, maybe. But there's more. What about that labour mobility discussed earlier? How often is it that a particular labour market is actually reduced to a 'buyer's market' by imperfect competition or outright monopsony? Not a whole lot, according to Don Bellante.

Bellante identifies the serious theoretical deficiencies that most hamstring Robinson's (and other) monopsony theories, that they attack what is little more than a straw man in the first place (perfect competition) and that the actual conditions and phenomena they expect just don't exist in the real world;

One could argue that the model retains some applicability if a group of employers act as a cartel, but the long run sustainability of such a cartel would be subject to the same stability problem as any product market cartel. Namely, in the absence of government enforcement of the cartel (or government acquiescence to a quasi-governmental enforcer of the cartel), it is unlikely to survive the built-in inducements toward collapse.

And so Bellante concludes;

Economic analysts sometimes have a tendency to compare the real-world workings of the labor market with the abstract, idealized, friction-free model of perfect competition with its assumption of perfect, costless information and mobility. That model of perfect competition often serves as a straw man which, when knocked down, serves the purposes of those with an anti-market mentality. In the case of the new monopsony theory, it is an abstract, highly stylized model, the conclusions of which are based crucially on an improbable assumption, which is compared to the model of a perfectly competitive labor market that is found wanting. It is an even less meaningful comparison

Italics are my own.

So there we have it, no such labour market cartelisation has ever existed absent some kind of incredible government intervention like choosing for people what careers they will pursue. Time to get on with life, dudes and dudettes.

EDIT - January 15th 2017

Seems there's a lot more evidence coming out to back me up on this.

EDIT - April 30th 2017

More evidence again, this time about low-wage versus higher wage worker's commute habits versus minimum wage rates.