Robots and Job losses

Currently there is a lot of discussion in the Dutch press about the threats that Robots cause for employment.  Minister Asscher (social democrat PvdA) started the discussion by bluntly stating that Robots steal jobs. He is worried about the dropping prices for robotic solutions and the competition this gives to people. Point is that if we want a growing prosperity (and who doesn't want that?) we need to grow output. 

Of course due to robots some jobs will cease to exist. In the past the job losses were mainly in dirty, manual, very detailed or repetitious jobs.  In the coming years we will also see job losses that require a mental skill but can be captured in a logarithm, thus in artificial intelligence.  Every person should consider whether his career ambition will clash with robotics, because if so, he/she better chose a different job.  

Let's look at a simplified calculation example:  if the EU has a GDP of 12,5 trillion Euro and 500 million people and we ignore population growth; and if we suppose that half the population works, at 40 hours a week, 50 weeks per year, this means we have an average productivity of 50.000 Euro per worker per year or 25 Euro per worked hour.  If we want to grow our prosperity, we have to work more hours at 25 €/h (which we should not want, and this is limited by the # hours in a day) or this figure of 25 euro/hour needs to increase. A desired economic growth of 2% means that next year the added value of the average output should be 25,50 €/h and the year thereafter 26,00 €/h.  

We can raise the added value output per hour 1) by creating jobs with higher added value in services (by entrepreneurship, stimulated by more schooling and better legislation) or 2) by producing more valuable products (so investing in innovation) or 3) by increasing productivity, i.e. producing the same products with less labor time. Increasing the productivity can be done 3a) by working harder (which has a limit) or 3b) by working more efficient (lean, good IT systems, less waste etc. etc, this also has a ceiling), or 3c) by using robots.  Robots allow a worker to produce the same goods with less hours. This has no ceiling, as far as I can think of.

So robotics is needed because the other tools available for increasing added value (innovation, working more hours, working harder) all have limitations in size and they need a lot of time. Creating high added value service jobs in large enough quantities will take even more time and also in services, robotics will be needed to help raise output per hour. More importantly: those other tools do not help raising the value added in sectors that already exist, which sectors will constitute most of the economy for the foreseable future.  

So the conclusion is that without robotics there will be much less economic growth.

The Stocks & Flows of the Homo Economicus

Economists and sociologists debate whether the homo economicus exists and make rational decisions based on profit maximization and utility considerations or not. Critics of the neoclassical economic theory argue that a lot of people seem to behave differently, that markets are not always efficient, that power can play a big role, information is not available and that people have group-behavior. (e.g. debate in FD on 12 and 24 april 2014). Nico Lemmens argues that group-behavior and consensus seeking, love and emotion can still be rational from an ecological perspective. Sure.

What (neo-)classical economists and their opponents overlook is the fact that there are stocks and flows involved here. The debate described above is a debate with a linear world view. A happens thus B happens thus C happens and all in the same proportion. Even when emotions are to be included it is still linear: more love is more marriage.  This is not realistic and therefore it does not fit the empirical observations.

In Stock and Flow terms (hence “Flostock”) not only a quantity of matter is a stock, but also opinions and feelings are stocks. Any movement is a flow and linear behavior like selling or travelling is a flow.  But there is also other behavior and that is much less straightforward and non-linear. That is behavior in relation to the speed of response to a pulse. A person or a group of persons can adapt their opinion on a subject fast or slow. This behavior is rational, and rather stable over time, and for large groups it is therefore predictable, but it is non-linear.

Example: suppose a group of the population has a strong preference for supplier A (say Apple) and a second group has a strong preference for supplier B (say Huawei). Both preferences are stocks, with advertisements, coffee-machine conversations, user friendliness, price comparisons and other experiences as inflow. After a couple of years, the preference has become quite strong, a big stock. You could call this stock customer loyalty. This stock has a memory, of course, in fact it is a memory, and it generates a rather constant flow of buying products of supplier A respectively B. Now one of the two suppliers introduces a new toy, or one of the suppliers is found to be secretly spying on its consumers. This new experience and knowledge is an inflow into the preference stock and slowly changes its value. The speed at which it changes is a behavioral parameter and could be referred to as customer inertness. So customer inertness is the speed at which customer loyalty adapts to an inflow. Obviously if the stock is large and the inflow is small, the change in preference will be small. So a strong brand can survive a small PR mistake.  When the preference stock changes in value, the group will change its buying flow.  

By the way, there is a second non-linear determinant of the buying flow and that is the amount of products already owned by the group. This is also a stock and let’s call that the pool of toys. The pool of toys will make the buying behavior flow non-linear e.g. when saturation is almost reached. The speed at which the group wants to reach saturation is also a behavioral parameter that we could call customer eagerness. Customer eagerness is the speed at which customers want to follow a trend. High eagerness will create hypes, panic run-aways, bank runs and overshoots. Medium eagerness creates fashion waves. Low but positive eagerness creates stable growth.