Robots and jobs (again)

As an economist, I hate bad data and over interpretation of bad data.

And although a picture may say a thousand words, those thousand words might be rubbish.

That’s the case with the latest ‘Robots are stealing our jobs’ graph doing the rounds on Twitter. (I’ve written before how, in fact, robots are not stealing jobs).

A quick glance at the chart suggests that the use of robots globally is on the rise, while US manufacturing jobs are declining. Hey presto – we have evidence robots are stealing jobs.

Well, part of the problem is that the chart is comparing a global stat with at US stat; that’s apples and oranges territory.

Also, take a closer look at the trends in the chart. US manufacturing jobs don’t fall once robots use begins to rise. Nor does the big fall in US jobs coincide with a big up swing in robots. (It does, however, come once China joined the WTO in 2001).

Lastly, the robots-stealing-jobs theory implies active causation: using more robots means less jobs. The really is no causation in this chart.

UPDATE: Take a look at the small print in the chart. Much of the ‘Global industrial robots’ data is actually extrapolation and forecasts, not actual data.