My editorial

I continue reviewing the literature pertaining to behavioural economics, and, even more broadly, to the behavioural approach in social sciences. I started in my last update in French (see Parfois j’ai du pot et parfois pas tout à fait ), with Herbert Simon and Thomas Bayes. Reviewing the work of those two thinkers helped me to somehow nail down the concept of behavioural learning. Now, I turn towards someone who, most probably, had never considered himself as an economist, let alone a behavioural economist: Charles Darwin. I take on one of his works, maybe less known than his evolutionary writings: ‘The Expression of The Emotions in Man and Animals’, published in London, in 1873, with John Murray. I picked up this one for two reasons. Firstly, it is a good example of early theoretical approaches to the empirical observation of behaviour and to experimental methods in this respect. Secondly, there are those three inside me: the happy bulldog, the curious ape and the austere monk. Those three, taken together, make out of me the intellectual equivalent of those baboons that people observe in the outskirts of big cities in Africa: I keep rummaging in those piles of things that people throw away, as apparently useless, and when I find something interesting, I am just fascinated with it.

Charles Darwin approached emotions in quite a behaviourist way, i.e. by observing their external expression rather than the verbal description of the emotional content. In the introduction to his work (pages 1 – 15), he states something really interesting: if we can express emotions with our facial muscles (e.g. smile or frown), those muscles have a function of communicating our emotions, i.e. we developed these muscles in order to communicate with other people. If a dog can express its emotions by the movements (or by the rigid immobility) of its tail, it means that the muscles responsible for these movements have a social function, i.e. they evolved in order to express the canine emotions to other dogs. That, in turn, means that we need to have the right muscles, trained the right way, in order to communicate correctly with other people (dogs?). Strolling further down this avenue, I call by a casual remark that a muscle – any muscle – is a resource: I use those precious gram-molecules of neurotransmitter to give it orders, I use those gram-molecules of glycogen to give it power for carrying out the orders I give etc. A muscle is costly. Still, it gives work in exchange. Basically, my muscles are like a workforce employed to move me around. My facial muscles are a specific workforce, focused on making my emotions intelligible to other people. Partial conclusion: we develop, generation after generation, specific and individually appropriated resources, which, in turn, serve to create and maintain social cohesion, possibly contribute to our position in the social hierarchy etc.

When you talk (right, when I talk) about resources appropriated mostly for socializing, ‘The Theory of the Leisure Class’ by Thorstein Veblen [1] is really hard to ignore. This another social thinker, which we do not necessarily associate with behavioural analysis, and yet there is a lot of behaviourism in what he wrote. Studying the middle class of the Western society, and especially the peculiar form it took during the second half of the 19^th century, Thorstein Veblen stated something quite interesting: there are instances of societies, where most of the available resources are directed on creating internal cohesion, and that cohesion is being built as a subtle interplay of overlapping, informal hierarchies. People acquire valuable things solely in order to fit into the desired bracket of the social structure. Veblen used to derive the source of property rights, as social institution, precisely from that phenomenon. When we acquire things in order to show something to someone, possibly to many someones, and those other someones do the same, we suddenly have a lot of things to show to each other, and it becomes important to delineate the claims that people have on each particular thing.

Good. Now I am trying to connect those four behavioural theories by Thomas Bayes, Herbert Simon, Charles Darwin and Thorstein Veblen, with my own sketch of behavioural approach (see Any given piece of my behaviour (yours too, by the way) ), and with my progressively emerging business concept for investing in smart cities (see Smart cities, or rummaging in the waste heap of culture ). Smart cities make a peculiar social environment, where we are surrounded by technologies, whose main function is to learn faster than we do. Living in smart cities makes a fundamental change, as compared to other environments. We, humans, are used to figure out things about things around us. In a smart city, the smart technologies incorporated in virtually any piece of manufactured goods around us figure out things about us. When you design smart cities, you are the active subject of innovation. Still, when you live in a smart city, you become the object of innovation.

In business terms, the behavioural analysis I am developing serves me to formulate testable hypotheses as for how will people behave when transitioning to and into the peculiar environment of smart cities. In a next step, I want to design an experimental environment to test those hypotheses, and this is, as I am figuring it right now, my most immediate operational goal in that business plan for smart cities. Yes, now I can phrase out clearly that the business I want to start consists in gathering investors with capital around the development of such an experimental environment.

Anyway, starting with the late reverend Thomas Bayes, may he be resting in peace despite my recurrent attempts to talk him into my science, as people plunge into the environment of a smart city, they learn. They learn by repeated experiments, and the sequence of experiments leads them to a given sequence of successes and failures. Charles Darwin whispers in my ear that in the process of learning, people develop resources functionally oriented on communicating and building social cohesion. Thorstein Veblen adds that with any luck, most of the social effort is going to be oriented on developing whole markets of resources valuable in developing and maintaining social cohesion, and social hierarchy. Herbert Simon claims that if we want to ascribe any kind of economic rationality to human behaviour, we should assume there is some broad and not quite explored range of possible ways to do things, and we should assume that our decision making consists in exploring that universe of opportunities and picking, imperfectly, just some of those ways as our personal strategies. I claim that in smart cities, the key factors shaping the social structures in place are: super-fast technological change and quick development of new energy sources. These two factors are likely to create very high monetization of local economic systems in smart cities (a cartload of monetary mass for each unit of real output, as compared to the world outside), and very steep social hierarchies, as well as a new geography of settlement. In a smart city, actively participating in the local technological change builds one’s position in the social hierarchy. Active participation in technological change requires the capacity to mobilize quickly significant amounts of capital, thus it requires liquid capital, monetary or similar in liquidity.  On the behavioural level, I assume that all those changes are likely to get depth and sort of sink in on the condition that people develop the corresponding patterns of recurrent, ritualized behaviour accompanied by socially recognized rules of conduct.

That was fast. I think I need to go back to Thomas Bayes, and go through the whole sequence once again, more sort of step by step. Sequence is the key concept here. I make a weak assumption that in any given experiment the probability P_S of me succeeding (p) is exactly the same as the probability P_F of experiencing a failure (q). Thus, P_S = P_F = 0,5. I consider a sequence of 10 attempts, where my score is fifty-fifty, i.e. 5 successes and 5 failures, or p = q = 5, and n = p + q = 10. This is the crest separating negative behavioural learning from the positive one, as with 5 successes and 5 failures I am, basically, equally likely to discourage myself, or, conversely, to develop valuable new skills. According to the classical Bayesian development, there are (10⁵) / 5! = 833,3333 etc. ways of having 5 successes out of 10 attempts, and the probability of each of those 833,3333 ways actually taking place is equal, with those odds being fifty-fifty, to 0,5⁵*0,5⁵ = 0,000976563. Thus, in terms of actual probability, that crest between the positive behavioural learning and the negative one is P = 0,813802083 wide. As probabilities come, P = 0,813802083 is quite a lot, it leaves only 1 – P = 0,093098958 of room available for other types of occurrences.

The point of this little mathematical mindf*ck is that in the absence of any strong assumptions as for our ways to learn things, the most probable scenario is the one which gives us equal odds to give up or to form new, valuable skills. Still, as I have a closer look at that P = 0,813802083, I sort of figure out there are some of those 833,3333 ways, which are much more likely to develop valuable learning than others. If I experience a sequence like ‘q – q – q – q – q – p – p – p – p – p’, thus five failures followed by five successes, I am highly likely to drop off, or to cling desperately to the outcomes of my first success, and to develop very narrow a range of skills.

Similarly, should I experience ‘p – p – p – p – p – q – q – q – q – q ’, I will have that ‘punch-in-the-face’ syndrome, when all of a sudden, everything I had previously learnt, in a sequence of successes, gets mauled by the sequence of failures. I drop off, discouraged, or I develop a narrow span of skills with a lot of emotional drive to perpetrate them. Those two sequences, namely ‘5 failures and then 5 successes’ as well as ‘5 successes followed by 5 failures’, are likely to eliminate some people from the process of collective learning (those who drop off), and to produce, in those who manage to stay in the game, the type of behaviour we can observe in outstanding, highly driven individuals with very sharp views and little flexibility.

If, on the other hand, I experience something like ‘p – q – p – q – p – q – p – q – p – q’ or ‘q – p – q – p – q – p – q – p – q – p’, this is sort of smooth learning, when I quickly reinterpret each success in the context of the subsequent failure and vice versa. There is little adversity and lots of positive learning. The odds of dropping off are low, and the probability of developing a broad, flexible range of skills is high. This is the kind of learning that makes a skilful business person, an efficient salesman etc. Both intuition, and maths suggest that, other factors held constant, the chances of going through something like a like ‘p – q – p – q – p – q – p – q – p – q’ sequence are generally (i.e. without additional assumptions) higher than experiencing something in the lines of ‘q – q – q – q – q – p – p – p – p – p’. Any group of people, who learn something by trial and error, are likely to develop a central path of social change based on the development of social flexibility, and two fringes, one made of the temporary losers, and another one consisting of highly driven, outstanding role models. All that learning can lead to deep social change if it translates into behavioural rituals and rules of conduct, and thus we have like three streams of collective learning: rules and rituals for quick adaptation in the central path (e.g. rules and rituals for changing one’s job or for making a sale), those for developing an outstanding position in the social hierarchy (rules for becoming a mayor, for example), and, finally, those for dropping off any aspirations (rules for getting jailed etc.).

Now, I go down the Darwinian path and I assume that each type of learning leads to developing, in each person, a set of resources functionally oriented on communication and social cohesion, i.e. the equivalent of facial muscles. In smart cities, if they evolve the way I think they will evolve, inhabitants will be learning and developing personal resources for getting money, which, in turn, will allow them participating in the local, super-fast technological change. The issue of personal resources makes me bounce back to that initial idea of mine, namely to gather capital for developing a centre for experimental research regarding the technologies of smart cities. If, as my personal hypothesis states it, smart cities of the future will host steeper social hierarchies than we presently have in an average city, personal resources functionally connected to social skills are likely to be even more valuable. Hence, an experimental environment could expose people to various paths of learning with new, smart technologies. For example, I invite a group of participants to follow a path of smooth learning, sort of ‘p – q – p – q – p – q – p – q – p – q’, and then, suddenly, I expose them to learning the hard way, like ‘q – q – q – q – q – p – p – p – p – p’. How many of them will drop off? How exactly will they be dropping off? Same for the tough and successful ones: how exactly will they be becoming tough and successful? How will the exposure to tough learning narrow down, or, conversely, broaden their personal horizons?

And so I ventured myself into the realm of what people think they can do, and I see a wise man sitting by a tree. I assume he is wise, ‘cause if you see a lonely man sitting by a tree and you assume he is stupid, logically you should take another path of enlightenment, just in case the stupid man has a gun. Anyway, I assume that the man I see is wise, I introduce myself, and he does the same, and so I learn his name’s Gigerenzer, and he eagerly introduces me into the Brunswikian theory of confidence [2]. It states two important things, regarding my line of research. Firstly, and contrarily to what Adam Smith used to claim, people who know they have to make an important decision will make very fine probabilistic distinctions, as in that reasoning I have just developed a few paragraphs ago. In other words, the Brunswikian theory of confidence claims strongly, with experimental results to support the claim, that overconfidence in one’s knowledge about the situation at hand is an exception rather than the rule. Secondly, the knowledge we use to make decisions is a combination of long-term memory about the environment, on the one hand, and short-term judgment regarding the structure of the task at hand. Logically, what can introduce a lot of bias in our decisions is quick change in the environment, too quick for us to internalize it into our long-term memory. Here comes that interesting thing about smart cities: they are supposed to make really a fluid environment, with technologies figuring things out about us even faster than we do about them. What kind of cognitive bias can it induce in the decisions made by people living and doing business in smart cities?

I am consistently delivering good, almost new science to my readers, and love doing it, and I am working on crowdfunding this activity of mine. You can consider going to my Patreon page and become my patron. If you decide so, I will be grateful for suggesting me two things that Patreon suggests me to suggest you. Firstly, what kind of reward would you expect in exchange of supporting me? Secondly, what kind of phases would you like to see in the development of my research, and of the corresponding educational tools?

[1] Veblen, T. (1899). The theory of the leisure class: An economic study in the evolution of institutions. Macmillan.

[2] Gigerenzer, G., Hoffrage, U., & Kleinbölting, H. (1991). Probabilistic mental models: a Brunswikian theory of confidence. Psychological review, 98(4), 506