Social Systems

A system is a partitioned environment which has varying levels of regularity associated with it.

In the simple system of rolling casino dice on a Craps table, outcome probabilities are well-defined, and it is known in advance that the probability of getting a "seven" is 1/6. You can find this probability by dividing the number of ways you can roll a "seven" by the total number of ways that dice roll.

Using just this simple logic, it is also known in advance that the probability of getting "snake eyes" (a "two") is only 1/36.

Social systems are more complex than casino games, but that does not mean that they cannot become well understood, so that we can know--in advance--what kinds of outcomes we can expect as we change up certain parameters.

After studying human behavior in interactional games, researchers have found some regularity with regard to how humans behave toward one another under different sets of rules [1]. Below is a partial list of such investigations into how humans respond to the behavior of others under different sets of rules.

Prisoner's Dilemma

Prisoner's Dilemma is a game characterized by the actual dilemma two prisoners would face if they were separately interrogated and invited to "rat out" their alleged partner-in-crime. A cop may tell Prisoner #1 that he will guarantee a lighter sentence if and only if Prisoner #1 rats out Prisoner #2.

In another room, the same deal is being given to Prisoner #2 (to rat out his partner in exchange for a lighter sentence). If both prisoners defect against each other, ratting each other out, then a moderate sentence is doled out for both.

If both prisoners "cooperate" with regard to one another (saying nothing), both end up with a light sentence. However, if one rats and the other stays "loyal" (saying nothing), the one who stayed loyal gets the harshest sentence possible, while the rat gets to go free.

When people were actually put into such a scenario by researchers, substituting various payments of money for the lengths of incarceration, researchers predicted they would rat each other out--because that maximizes the minimum payoff.

However, it was found that about 50% of people remain loyal--even if they did not have a chance to communicate with the other person they were "playing against."

For a while, researchers thought people were misguided creatures who did not know how to maximize their own benefit when interacting with other people. Then it was discovered that people do not just care about money, so that money wasn't the only reason for their decisions to cooperate with others.

It was also discovered that people think ahead and think in principles, aligning their behavior with strategies expected to work well over a long run (even with regard to money-maximization).

Public Goods Game

In the Public Goods game, people are given some initial money and can contribute some or all of it to a common pot, or pool of resources. That common pot will then grow by some amount (usually by 50% up to 300%), and then the increased amount gets dispersed evenly between all people playing the game (even those who chose to contribute nothing).

Again, researchers predicted a "Tragedy of the Commons" where most individuals would choose not to contribute but, instead, would "free ride" on the contributions of others (maximizing their personal payoff).

However, actual human beings, even in one-shot (nonrepeated) games, contribute about 50%.

The level of the multiplier can matter (50%-300% increase), so that, if the pot is doubled before redistribution (a 100% increase), contributions are 40% of the original endowment given them, but if the pot is quadrupled before redistribution (a 300% increase), contributions can rise to 80%.

It was discovered that if you have at least 5 people playing, that it predicts the behavior of even larger numbers--ie, there is little difference in outcomes when played with either 5 or 50 or 500 different people.

What can be said here?

In large groups, people do not want to contribute to common goals or projects, unless the project is expected to approximately double the aggregate amount of material well-being.

If a politician wanted millions of dollars for a bridge which served only 12 people (who lived on the other side), people would not choose to contribute--because it doesn't even approximately double the aggregate of material well-being for society.

Only certain things, with certain outcomes, are "worth it" to large groups of people.

Ultimatum Game

In Ultimatum, one person plays as the Proposer and the other as the Responder. The Proposer is given an initial endowment of money (the "pie") and told she can only keep it if she splits it with the Responder somehow such that the Responder agrees not to reject her offer.

If Responder rejects the offer, both get nothing.

Researchers predicted that any non-zero offer--even just 10% of the pie!--would not be rejected by the Responder, because that would be like throwing away free money. Based on this prediction, they also predicted that Proposers would propose just 10%--because that means they get to keep 90% for themselves.

However, most offers are from 30-50% of the pie, and offers as low as 20% of the pie are rejected about half the time. When a Responder rejects a low offer, it means that the Responder is willing to "pay a price" for enforcing fairness (because they then get nothing, instead of at least something).

When you think about the psychology of not being willing to let another person "exploit" you, it turns out that people are not just concerned with money, but with things such as being validated as someone who is not a push-over.

Scientificially, this psychic resistance against being victimized has both survival- and fitness value. People who stick up for themselves (even when it comes at a cost) do better, on average, than people who don't.

Dictator Game

In the Dictator game, one person gets all the money and can then "unilaterally" decide to give some of it away to the other party. Unlike Ultimatum, the recipient of this "offer" does not get a chance to reject the deal--so that they cannot punish low offers and enforce a moral norm of fairness.

On average, when there is no possibility that they will be punished for being selfish and greedy, people give away about 20% of the pie to the other player. When you think about it, it is like winning the lottery, and then determining to share some of your winnings.

If two people always bought lottery tickets together, and if it was the primary bond that they had with one another, and if they had no other commitments to others in their lives, and one of them then hits the lottery--then expect the winner to give about 20% of it to the other person.

Under conditions where people are farther apart or have more commitments to others, expect that less than 20% of lottery winnings will be given away to any single person.

Under conditions of totalitarian dictatorship, expect the supreme leader to be wanting to keep approximately 80% of the entire nation's economic output to himself--leaving only 20% of national income to be shared among the millions of citizen-subjects.

This means that, in the ideal case scenario (where top-down authoritarianism didn't harm productive efficiency), people will still be 5 times worse off under authoritarian rule.

Because this ideal is never achieved (top-down authoritarianism ALWAYS harms productive efficiency), people living under communism/socialism/fascism will always be more than 5 times worse off.

Moving to a free country would then allow for them to obtain a living standard that is more than 5 times higher than what they previously could "afford."

Trust Game

In the Trust game, one party is given money and told that if they entrust the other party with that money (initially give it away to the other person), then whatever was entrusted will be multiplied--so that the other party gets three times whatever was given.

The hope in giving someone else money--with both of you knowing it'll get tripled--is that the person receiving all that money will be grateful and then turn around and give a good share of it back to you (because your gift/generosity is what made them rich).

On average, people give about half of their original endowment to the other party, and the trustees turn around and give just slightly less back--so that trusting others, in this game, does not appear to "pay off" from the perspective of the original investor who entrusted someone else with the money.

However, sometimes an average doesn't tell you everything you need to know, and this is a special case of that. When the investor gives more than half of the money, then the trustee sees how much trust the investor gave, and responds by giving back even more (so that both are better off).

Gift Exchange Game

In the Gift Exchange game, an "employer" offers a wage to a "worker" and the wage contract stipulates a given level of "effort" on the part of the worker (between 1 and 10).

The employer's profit depends heavily on the actual effort provided by the worker, regardless of the wage contract--which remains in effect no matter how much effort the worker actually provides.

The worker's "profit" will be simply (wage paid) - (effort expended), so that slacking off at work means you take home more pay. Researchers predicted that both wages and effort would be low, because that minimizes risk to the employer.

However, the average wage contract stipulates a level 7 in effort, implying that the worker would be pocketing 44% of the revenue of the business (employer keeps 56%)--and most workers honor this wage contract!

However, because about 30% of workers always choose the lowest effort (effort = 1) regardless of what they agreed in the contract--the average effort spent is 4.4.

When wages represent 51-60% of business revenue, workers work the hardest (effort > 7)--and it actually pays off for the employer to offer wages that are that high (51-60% of all revenue). Note how this is a similar finding to the one above in the Trust game--where offering more than half resulted in net profits.

Application to Today's World

Wages were approximately that high in the 1950s to the early 1970s [2] but, due to lost economic freedom (a drift away from free market capitalism), US wages are no longer that high anymore--and everyone (employer, worker, investor, etc.) is now suffering because of that.

Total employee compensation is wages plus benefits and, due to an increasing share of employee compensation going to benefits as opposed to wages, total employee compensation (1953: 55.8% of GDP; 2017: 53.1% of GDP) has not fallen much compared to where it was before [3].

But this should not offer us hope--because only wages empower workers to discharge their economic obligation to help in directing the resources of society (the wage share is down by a full 8% of GDP; it used to be 20% higher).

To picture the economic harm, imagine no wages and all benefits--so that only crony relationships between the government and a dozen or two dozen "benefits providers" directed all of the resources of a society of 300 million people.

A society with 24 resource allocators will always underperform a society with 300 million resource allocators.

Resource allocation which is accomplished by a dispersed means (by millions choosing what to buy) is more efficient than resource allocation accomplished by a centralized means--where a select few benefits providers cozy up to the government in a "quid pro quo" crony relationship where each takes best care of the other (but with civil society losing out).

Drifting away from capitalism is a policy mistake. It'd be better to incorporate scientific knowledge about how human beings interact in economic situations, and then put that knowledge to work in improving our materal condition.

The unnecessary drift away from the capitalism of the 1950s is making our lives worse off each year. This is evidenced by the annually-compounded growth in the ratio of debt to income over the last few decades (we can no longer afford our own lifestyles).

Reference

[1] Henrich, J., et al. (2004). Foundations of Human Sociality. New York: Oxford University Press. Page 61-63.

[2] U.S. Bureau of Economic Analysis, Shares of gross domestic income: Compensation of employees, paid: Wage and salary accruals: Disbursements: to persons [W270RE1A156NBEA], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/W270RE1A156NBEA

[3] U.S. Bureau of Economic Analysis, Shares of gross domestic income: Compensation of employees, paid [A4002E1A156NBEA], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/A4002E1A156NBEA

Image Attribution

Page URL: https://commons.wikimedia.org/wiki/File:Mechanical_Aptitude_Illustration.png

Attribution: Davidparrottdesign [CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0)]