Our confused response to market power
Fri 12 Jun 2009 by Infometrics Ltd.

The size of profits generated by banks, power companies and telecommunication firms are a frequent source of public outrage.  When those profits arise from market manipulation, this outrage is justifiable.  Yet the analogous application of market power by labour unions to raise wages is widely accepted and often applauded.   How consistent is this?

There may be a tendency to confuse total profits with rates of return on assets, and simply view large sums as obscene.   But this gut reaction has little to do with the economic theory that underpins our regulation of dominant firms capable of influencing market prices.

The economic concern is not strictly with the distribution of the gains from market transactions.   Rather it is that the exercise of market power reduces the welfare of society as a whole.

The textbook model is that firms with a large market share can raise their profits by reducing production and selling fewer goods at higher prices (extracting economic rents).   It is the dead-weight loss from this behaviour that concerns us – resources are unutilised, and goods that could profitably be produced and consumed are not.

It is this surprisingly simply model that is the basic economic justification for society’s decision to regulate voluntarily agreed upon private transactions.

The logic of this principle is largely unquestioned.   But it is difficult to prove that firms with market power are necessarily bad for society, especially when considered over a longer-time period.

Large profits may arise from being the most efficient producer (due to economies of scale).   Large firms may still innovate to protect their market position.   Regulation can be equally problematic, discouraging efficiency gains and expansion.   Then there is the practical problem that it is difficult and costly to prove the existence of excessive market power.

In summary, a simple yet relatively abstract principle underlies a large intervention into free market transactions, the net benefits of which are probably positive but may not be.   Given this, the visceral public disapproval of large profit numbers seems out of proportion to the loss suffered.

Public reaction seems even more contradictory when viewed in context with the widespread support for labour market bodies and regulatory principles that seek to increase market power and rents for workers.

Labour unions, for instance, openly act to increase the market power of their members and secure higher wages.   We can easily model unions as firms that sell labour, and intend to maximise their profit (in this case, member wages).

The textbook model then proceeds on the same basis as before.   Unions with sufficient market power can raise wages, but only to the extent that they reduce the amount of labour supplied in equilibrium.   Again, we get the result of fewer goods and services produced and enjoyed by society.

Of course, the example of labour unions is more complicated.   Although they wield market power, they tend to be located in industries with a single, dominant employer with market power of its own.     Public sector workers are the classic example of this.   When both sides have substantial market power, it is unclear whether dead-weight losses will result.

Continuing with the public sector example, however, one recent estimate is indicative that labour market power can potentially be exercised to the large advantage of workers.   Professor John Gibson estimated that wage premium for public sector workers over their private sector counterparts was as high as 21% in 2007, after controlling for individual and job characteristics.

Such estimates are subject to a number of caveats about whether unobservable characteristics explain some of the premium.  Nor can we can conclude that the premium is solely due to the market power of public sector labour unions.   However, the rapid increase in the premium to 21% from almost 0% in 2003 is strongly suggestive of public sector workers capturing increased rents rather than a change in the composition or productivity of the workers.

Let us assume then that the 21% premium is accurate, and that it reflects the market power of the public service unions.   If applied to the average-full time wage, and multiplied by 250,000 full-time public sector workers, we would come to the conclusion that the public service labour force makes a "profit" of $2.5bn annually from its market position.   That is roughly comparable in size to the annual profits of the big four banks (although the component of bank profits that could fairly be called rents would be much smaller).

The $2.5bn figure does not tell us the size of the dead-weight loss.   But, intuitively, if public sector wages were lower, more workers would be hired and more public goods created for society as a whole to enjoy.

Large firms and large unions can be modelled in exactly the same terms.   Levels of outrage (or lack thereof if market rents do not concern you) should be similar about both the profits made by bank share-holders and the wages made by teachers, nurses, and public servants.   Few readers may agree with this conclusion; but on economic principles there is no clear distinction between the two cases.

Related Articles