Saturday, May 6, 2017

Automation update

As concerns about automation and its impact on the labour market increase, Times points to a new paper by Daron Acemoglu and Pascual Restrepo which is apparently the first to quantify large, direct, negative effects using real-world data of industrial robot usage in 19 industries in the US on labour markets in the 1990-2007 period. They write,
On average, the arrival of one new industrial robot in a local labor market coincides with an employment drop of 5.6 workers... Using a model in which robots compete against human labor in the production of different tasks, we show that robots may reduce employment and wages... we estimate large and robust negative effects of robots on employment and wages across commuting zones... The impact of robots is distinct from the impact of imports from China and Mexico, the decline of routine jobs, offshoring, other types of IT capital, and the total capital stock (in fact, exposure to robots is only weakly correlated with these other variables). According to our estimates, one more robot per thousand workers reduces the employment to population ratio by about 0.18-0.34 percentage points and wages by 0.25-0.5 percent.

Free Exchange puts this in perspective, 
Messrs Acemoglu and Restrepo reckon that one additional industrial robot per thousand workers reduces wages across the economy by 0.5%... between 1990 and 2007, each industrial robot added per thousand workers reduced employment in America by nearly six workers.
In other words, the negative displacement effect of automation appears to trump the positive productivity effect. But the last word on this is far from said.

Update 1 (14.05.2017)

From the Economist on the automation challenge (by way of e-commerce) in retailing in the US which employs 15.9 million people or one in nine US jobs,
At its current pace, by July 2018 retailing will have shed three times as many jobs as Amazon is due to create... Yet the skills required for retailing’s new jobs differ from those needed for old ones. Burning Glass found that 78% of e-commerce postings want applicants with a university degree, compared with just 12% in traditional retailing. Even warehouse positions demand more training: 53% of jobs in automated warehouses also require a degree. Couriers need less training to ferry goods to customers’ doors. Their ranks have grown to 655,000 workers last year. But that is a tiny sliver of the total retailing workforce. Retailing workers might switch to the companies that are taking over empty stores, including restaurants, beauty salons and health clinics. But it is as improbable that such firms will replace all of America’s boarded-up shops as that they will offer jobs to every former shop worker, particularly those without training...

Two-thirds of books, music and films are now purchased online, as well as over two-fifths of office supplies and toys, according to Cowen and Company, a financial-services firm... For every percentage-point increase in their share of e-commerce sales, a retailer’s margins shrink by about half a point, according to estimates by Morgan Stanley, a bank.
To put this in perspective, across the world there are 192 million retailing jobs which are apparently threatened by automation.

1 comment:

Unknown said...

The sense in which this paper by Acemoglu+ becomes interesting is that they are beginning to acknowledge the problem. That is the interesting bit. I would recommend an excellent article in Quartz and also Tyler Cowen's article in Bloomberg as to why facile comparisons to the Industrial Revolution are misplaced. I had blogged on them here: