Understanding the Impact of robots on employment — Part 2

By Siddharth Singh, 15th May, 2015

In an earlier post, we looked at a paper co-authored by Jeff Sachs on the impact of automation on jobs (“…increase in robotic productivity will temporarily raise output, but…can lower wages and consumption in the long run…”) We also saw that “routine jobs” are being lost to automation. An accompanying example was a Chinese air-conditioning factory where industrial robots are out to swallow 6,000 of the existing 30,000 jobs.

Automation has historically consumed blue collar jobs, which are “routine” or repetitive in nature. That trend is far from over. For instance, warehouses are being increasingly automated, with startups working on creating new technology towards the same goal. What’s peculiar is that in China, several provincial government is promoting greater automation, which will come at the cost of blue collar jobs. Adam Minter writes:

Late last month, the government of Guangdong Province, the heart of China’s manufacturing behemoth, announced a three-year program to subsidize the purchase of robots at nearly 2,000 of the province’s — and thus, the world’s — largest manufacturers. Guangzhou, the provincial capital, aims to have 80 percent of its factories automated by 2020.

Then there are non-routine blue collar jobs which are a little harder to automate. For instance, gardening involves a variety of small tasks in a complicated, natural environment. That has not stopped the automation of the farm. Here’s a description of a farm managed by robots:

Harnessing high-powered computing, color sensors and small metal baskets attached to the robotic arms, the machine gently plucked ripe strawberries from below deep-green leaves, while mostly ignoring unripe fruit nearby.

Such tasks have long required the trained discernment and backbreaking effort of tens of thousands of relatively low-paid workers. But technological advances are making it possible for robots to handle the job, just as a shrinking supply of available fruit pickers has made the technology more financially attractive.

Such automation is not simply an outcome of improving technology. The “non availability” of labour due to (in my view, faulty) immigration policies in the United States is the trigger in the case of the automated farm. Ilan Brat writes:

An abundant supply of workers, particularly from Mexico, willing to plant, pull weeds and harvest ripe crops for relatively low pay also had suppressed the need for mechanization. But the number of unauthorized immigrants in the U.S. workforce has been declining since its peak in 2007, according to the Pew Research Center, in part because of increased job opportunities in Mexico, as well as tighter U.S. border patrols.

Regressive politics* and technological advancements have in the past consumed routine jobs, but an emerging trend is the impact of automation on white collar jobs. The Wall Street Journal reports:

Robots are taking over corporate finance departments, performing work that often required whole teams of people. Big companies (…) are among those using software to automate many corporate bookkeeping and accounting tasks.

Businesses use these programs to save time and staffing costs. Since 2004, the median number of full-time employees in the finance department at big companies has declined 40% to about 71 people for every $1 billion of revenue, down from 119, according to Hackett Group, a consulting firm.

(…)

Software has helped Verizon, which had $127.1 billion in 2014 revenue, cut the manual entries its workers punch into Excel spreadsheets annually by a quarter — to 10,500 from 14,000. It aims to cut another 1,400 manual entries by the end of this year for an overall reduction of 35%.

This replacement of white collar employees by computer programmes isn’t unique to bookkeeping. Even journalistic tasks are being undertaken by computer programmes now: Associated Press publishes over a 1000 articles every month generated by artificial intelligence. There is no human byline; just a sentence at the end that says, “This story was generated by Automated Insights.”

In fact, artificial intelligence has become so good at generating written content, New York Times created a quiz which tests your ability to figure out human writing from computer generated one. You can take the quiz by clicking this link.

It remains to be seen to what extent can automation crowd out human labour. But at least two things can be said about this trend: one, the scale and scope of automation today is unprecedented, and two, such automation will ultimately give birth to a new kind of labour politics.

The author can be followed on Twitter @siddharth3

* More on that, later.