It’s simple economics, really. In economic literature it is known as factor substitution: one input (factor) of production (like labour) can, given sufficient time for adjustment and sufficient resources to effect the change, be substituted for another (like capital) to produce the same output. What motivates substitution between factors is the relative cost of the factors. For example, if the cost of labour was to rise, businesses would be incentivized to shift towards using more capital and less labour in production processes. Simple really, yet seemingly either hard to understand, easy to forget or convenient to ignore.
The concept of factor substitution is obviously very pertinent to the minimum wage debate. Should the price of labour (wages) rise by decree, without a corresponding rise in productivity, one should expect the use of labour to fall and unemployment to rise. The first jobs to go will be those held by the least productive workers, where the artificial wage hike has the biggest effect. Unsurprisingly, these argument is routinely highlighted in the debate – yet seemingly also routinely ignored.
Despite the recognised adverse effect on employment of relatively poor, low-skilled workers, minimum wages are unfailingly portraited as a boon for workers – like when the introduction of the £7.20 Living Wage in 2016 was hailed as a ‘pay rise for Britain’. It may very well be, but not if you are substituted out of your job and replaced by a machine.
In August 2017, economists Grace Lordan and David Neumark from the US National Bureau of Economic Research published a report entitled People Versus Machines: The Impact of Minimum Wages on Automatable Jobs. In the report, the authors examine ‘the effect of minimum wage increases on employment in automatable jobs – jobs in which employers may find it easier to substitute machines for people – focusing on low-skilled workers from whom such substitution may be spurred by minimum wage increases.’
The findings are not surprising. Based on data from 1980-2015, the evidence shows that a $1 increase in the minimum wage lead to an average 0.43% reduction in low-skilled, automatable jobs. In manufacturing, the reduction was 0.99%. Pertinently, the groups traditionally perceived to be the weakest, and normally found deserving of special consideration, are the first to lose out: ‘these effects […] are larger for the oldest and youngest workers, for females and for blacks’, the study found.
Automation and mechanization has disrupted industries and upset workers since the late 18th century. Famously, automation in the cotton industry led to uprisings from the so-called Luddites, who saw their jobs threatened by the power loom. However, the substituted workers soon found employment in the rapidly expanding cotton industry. In Manchester alone, the number of cotton mills rose from 2 in 1790 to 66 in 1821 as the price of cotton clothes dropped and demand exploded. Technological advances and automation is a boon for society, despite the disruptive short term effects.
However, when automation is forced upon business, not because new technology has advanced to become more economical, but because the cost of labour has been artificially raised, the picture is very different. Prices will not fall, but rise, as industry substitutes an optimal structure of production for a new, sub-optimal one forced upon them by the higher price of labour. As prices rise, demand will fall and there will be no economic expansion allowing those displaced by automation to find alternative employment, but a contraction to compound the misery. Those are the pernicious unintended consequences of a well meaning, but naive policy prescription. Destroying the job prospects for the low-skilled is no admirable political ambition. A minimum wage hike is indeed a pay rise, but only if you still have a job.