Saturday, 12 August 2017

Want a living wage for everyone? Raise productivity first

Jim Rose (economic adviser to the Auckland Ratepayers' Alliance) wrote in the New Zealand Herald back in June:
The Auckland Council's new living wage policy will lift the pay of its minimum waged employees by 30 per cent. Of course, no minimum wage worker will be shortlisted for these jobs in the future because these jobs will be paying $20.20 per hour. Minimum wage workers will be crowded out by better quality applicants who already earn a similar pay.
Rose's point is that increases in wages need to be underpinned by increases in productivity. To see why, consider our simple model of labour demand, as shown in the diagram below. The VMPL curve is the value of the marginal product of labour (also called the marginal revenue product of labour) - it's the extra value that one additional hour of work provides to the employer, in terms of additional revenue. Essentially it is the productivity of the marginal worker (the amount they would produce in that additional hour), multiplied by the value of that output. A profit maximising employer will be willing to hire a worker for an hour provided that the VMPL is at least as great as the wage. If the wage is higher than the VMPL, then hiring a worker for that hour would reduce profits. So, if the wage is equal to W0, then the quantity of labour employed will be where VMPL is exactly equal to W0, which in the diagram below is Q0.

Now, consider what happens if you implement a living wage that is higher than W0 (say, at W1). Let's assume that the value of output is the same regardless of which worker produces it (which seems reasonable). When the wage goes up to W1, relatively low-productivity work hours (previously producing VMPL between W0 and W1) will no longer be profitable for the employer, so it will cut back on labour hours (from Q0 to Q1). Fewer people will be employed, or those who are employed will be employed for fewer hours.

Advocates for the living wage argue that it will increase productivity, as Rose notes:
Living wage activists prefer to talk about the costs supposedly being offset by labour productivity gains - higher staff morale, fewer absences and reduced staff turnover. These are supposed to make everything right and low risk.
This is essentially an efficiency wage argument, which I have discussed before. As I noted then, it only works provided not all employers are paying the living wage, since it relies on alternative jobs for employees paying much less. However, if only a limited number of employers are using the living wage, then it could increase productivity for those living-wage-paying employers, by ensuring that the most productive employees go to work for them (and not for other employers). However, other employers would be left with less-productive workers (and would pay lower average wages as a result). So wages on average across the economy remain unchanged, because the overall productivity of the economy remains unchanged.

Rose's overall point is that we can't simply legislate higher wages, through proposals like the living wage:
Living wage activists and unions are right to point out that we live in a low-wage economy compared to Australia. The solution is not to vote ourselves a pay rise.
Increasing productivity, innovation and entrepreneurship is the only way to catch up.
If we want higher wages across the economy as a whole, we have to be more productive first.

Read more:


  1. thanks. My paper for the taxpayers union had a chapter on normative sociology of the efficiency wage arguments. The efficiency wage and fair wage arguments were neo-Keynesian theories of mass unemployment. all the rage by the end of the 1980s.

    1. Thanks Jim. I hadn't seen your report, but I've got it now and look forward to reading it.