There has been much written about the impact automation might have on employment, not least on this blog.  Whilst there have been predictions that as many as 40% of jobs today will be automated within a decade, these reports don’t tend to look at the number of new jobs that might be created, or indeed the jobs that might grow as a result of larger markets.

A recent paper steps into this space to examine the relationship between automation and the minimum wage.  The authors reason that most examples of automation don’t automate the entire task but rather make previously complex tasks more accessible to lower skilled individuals.

Think of it like the recent example I shared of a new platform that automates a chunk of data science work.  The aim isn’t to reduce the need for humans to do data science analysis, but rather to make it easier, and therefore accessible to a much wider range of people than can currently do data science.

Automation and the minimum wage

So what impact does the minimum wage have?  The paper argues that because automation is typically a partial job, labor will still be required.  So, the cost is not purely in technology terms but also in (relatively) low wage workers that are now able to use it.

They argue that by deploying the minimum wage the profitability of the technology is reduced, therefore if the minimum wage is increased, it makes automation less profitable.  The hypothesis runs counter to the traditional argument that a higher minimum wage will encourage companies to automate tasks in a bid to lower their costs.

The hypothesis is tested thoroughly, and the author suggests that each $1 decrease in the minimum wage sees the average use of technology rise by 30%, with a 1% drop in the routine share of the wage bill.  They suggest that the wage bill in industries with a high routine component, the minimum wage is a very important lever in the pace of technological change.