Technology displacing workers

[Franknburger]

So far, what we see, in most developed countries, is shortage of low-skilled labor. To the best of my knowledge, they haven't invented robots that are both good at making beds, cutting grass, helping the elderly - and that are cheaper than human labor.

I had prepared the following diagram for this thread, but it fits here as well:

Current German data. "Cutting grass" is within the "Real Estate & Facility Management" group, and in fact one of the strongest employment generators in there, together with security.

At least at the moment, it is not the low-skilled workers that have problems in the German labour market. Pressure is mostly on lower-level white-collar workers, especially in banking & insurance, that see their jobs replaced by customers doing it themselves via the internet. Retail may follow soon (in fact, a good part of the employment gain in trade shown in the diagram has already been eaten up by the collapse of two major German retail chains). However, internet-based retail will create quite a lot of (equally poorly paid) new jobs in transport & storage. 

[Franknburger]

In some parts of the world even basic technological advances that are commonplace are shunned due to pressure from workers who don't want to lose their simple jobs. For example, companies have workers unload entire trucks by passing goods from one person to the next instead of simply buying a forklift. I'm not saying that Americans should do this, merely driving home that technological innovation has, is, and will play a huge role in unemployment. It's not at the root, of course. The root is always money. When it becomes cheaper to automate a process, technology displaces workers. But full employment could be acheived if many technologies disappeared overnight. Most people would be underemployed, sure, but employed nevertheless.

Let me illustrate the point with a concrete example: When I did a study on the Ethiopian construction sector a few years ago, I was astonished they did not use wheelbarrows to transport sand or cement on-site. Instead, such loads were carried on handbarrows held by two people.

Asking the manager about it, he came up with a simple calculation: An unskilled worker gets around 1 USD/day. Wheelbarrows would increase labour productivity by some 50% (you only need one instead of two workers, however, you can't access all locations by wheelbarrow, so you  may have to do some extra loading/unloading). As such, investing in a simple, 30 USD wheelbarrow would pay off in around three months (20 workdays/month). However, cheap wheelbarrows would not last much longer than three months, if they were not already stolen before. As such, there was no economic benefit from using them. Asides, he preferred to create jobs, local unemployment was already high enough.

Bottom line: Technology only destroys jobs, if these jobs are comparatively expensive, i.e. adequately qualified labour is relatively scarce, and/or if it improves process/output quality (unlike wheelbarrows, which are inferior to handbarrows when it comes to accessing specific points on a construction site). 

[Franknburger]

http://m.technologyreview.com/view/519241/report-suggests-nearly-half-of-us-jobs-are-vulnerable-to-computerization/

Researchers have concluded that as many as 45% of current jobs are vulnerable to automation in the next two decades. Obviously that does not equate to 45% unemployment, but it does equate to an enormous burden on the creativity of society to sustain an economy where tens of millions of people have less disposable income. There is no plan from either side to combat this problem. Of course, it's impossible to accurate gauge the future. 45% seems particularly high, for starters. To combat or grow in such an environment, the article suggests outlawing automaton in some industries (for the former) and shifting our educational focus towards creative and highly social skills (for the latter).

While I am not able to point out at specific weaknesses of the cited study for lack of information on details, I think the 45% figure is a gross overestimate. I have done an estimate for Germany, based on the latest available detailed employment breakdown by sector (end of 2012), and grouping sectors according to their likelihood of jobs being replaced by automation:

Low automation risk (49.4% of total employment)
- Professional services, consulting, IT (actually, they should continue to show job creation)
- Equipment & devices, including their installation and maintenance: Automated performance monitoring should considerably reduce maintenance. Otherwise low automation potential in manufacturing. As they should be the ones to build the automated devices, overall employment in this sector is more likely to increase then decrease.
- Health & care (for Germany, a lack of up to 1 million professionals is predicted until 2050)
- Education & training
- Construction
- Facility management (security, cleaning, gardening etc.)
- Agriculture & Food (already heavily automated, except for quality control, which requires specific sensoric skills)
- Quarries, glass, ceramics
- Forestry, wood processing, furniture
- Waste disposal & recycling
- Personal services (hairdressing, cosmetics etc.)

Medium automation risk (36.4 % of total employment)
- Trade & repair (incl. vehicles): Low risk for the repair and online trade parts, quite a risk for wholesale, substantial risk for stationary retail. Unfortunately, available sub-sectorial breakdowns do not always follow the above lines, so I had to aggregate figures and put the total into "medium risk".
- Hospitality: Substantial automation risk for fast-food. Otherwise (restaurants, entertainment, hotels, etc.) low automation risk.
- Metallurgy, metal processing, vehicle building, chemicals & pharmaceuticals, plastics, textile & clothing.
- Public administration, interest representation, churches, etc.

High automation risk (14.2%):
- Transport & storage (not so sure about the storage part, in fact, especially when it comes to commissioning for delivery in internet-based retail).
- Banking and insurance
- Time-work: These people are the first to go when jobs are automated. OTOH, they are typically hired to cover demand/production peaks, so maybe the category rather belongs into "medium risk".
- Publishing & printing
- Utilities (automated metering & performance control).

All in all, I think for Germany we are talking at maximum about some 10-15% of jobs being endangered. This has to be compared to some 3% new job creation in care (that is where the fast food workers will go), plus substantial job creation in IT, machine-building etc.