A minimum wage raises unemployment

The argument

Proponents of this view argue that minimum wages reduce employment in three ways, each of which operates independently of the others. The first runs through prices. Higher labour costs increase the cost of producing goods or providing services, and employers pass those costs on to consumers in the form of higher prices, which leads to fewer consumer purchases of those goods. Reduced demand then causes manufacturers and service suppliers to reduce supply, and producing less requires fewer workers, leading to layoffs or slowed job creation. On this account the job losses are not a matter of employer ill will but an ordinary consequence of demand falling when prices rise. The second runs through technology. Employers choose between human labour and machines partly on relative cost, so raising the cost of human labour relative to machines and technology shifts that calculation. Higher labour costs therefore prompt employers to accelerate automation, encouraging more of them to automate their workforces and producing the same outcome of layoffs and reduced hiring. A 2014 study on the impacts of introducing a $15 an hour minimum wage in San Francisco estimated that the measure would lead to the loss of more than 15,000 jobs in the city. The third runs through geography. Firms can relocate work as well as replace it. As domestic wages increase, companies look to offshore operations to countries with lower labour costs in order to reduce overheads, and that offshoring leads to supply chain relocation. The jobs are not automated away or priced out of existence; they simply move somewhere else, reducing the number of domestic jobs. Taken together the three mechanisms point the same way. If a minimum wage raises unemployment through prices, automation and offshoring, then the workers it is meant to help are among those most likely to lose their jobs, which is why proponents of this position hold that a minimum wage is ineffective at reducing poverty.

Context

Minimum wages increase earnings for those in employment, but they cause job losses which leads to higher unemployment.

Premises

Counter-arguments

When implemented, minimum wages do not reduce employment. A case study undertaken in 1994 comparing the fast food sectors in Pennsylvania and New Jersey revealed no link between a higher minimum wage and reduced employment. In 1992, New Jersey increased its minimum wage by $0.80 (to $5.05 per hour). Pennsylvania kept the same minimum wage ($4.25 per hour). Contrary to predictions, job growth in stores in New Jersey paying a higher minimum wage was faster than in Pennsylvania.

Rejecting the premises

[Rejecting P1] The chain is theoretically clean and empirically weak. Labour is a fraction of most retail and food-service costs, so the price effect is small — studies of fast-food chains find pass-through of a few per cent — and the premise omits the offsetting mechanisms: lower turnover and recruitment costs, higher productivity, and the additional spending of low-paid workers, who consume most of a raise. Card and Krueger's 1994 comparison of New Jersey and Pennsylvania found no employment fall after New Jersey's increase, and the large literature since clusters around small or negligible effects at moderate increases. [Rejecting P2] Automation decisions turn on the cost of the technology, the scale of the operation and interest rates at least as much as on wages — and the sectors most affected by a minimum wage, restaurants, retail, care and cleaning, are the ones least able to offshore, since the work has to be done where the customer is. The premise applies a manufacturing argument to jobs that cannot move. [Rejecting P3] The San Francisco figure is a projection from a modelling exercise, not an observed outcome, and the outcome is now available: the increase was enacted and losses on that scale did not materialise. The conclusion also does not follow even if the premises are granted — that a policy carries an employment cost does not make it ineffective against poverty, which depends on the net effect on household income across everyone affected.