On Social Power

Peter Turchin


Join 36.9K other subscribers

There were quite a lot of questions about (and some disagreements with) my tweet about the four sources of social power. As Twitter is not a great platform for thoughtful commentary, I thought that I would expand on it in this blog. What follows is a somewhat modified version of Michael Mann’s theory explained in his Sources of Social Power. This basic framework, with a few modifications, have gained a lot of acceptance among sociologists of power, for example, Bill Domhoff.

First, what is social power (as distinct from physical power measured in watts)? Simply put, it’s ability to influence other people’s behavior. Some theorists add “against their resistance” or something like that, but I am against adding this qualifier, because it eliminates one of the most effective forms of soft power—persuasion. If you are persuaded, you do it of your own volition.

Next, then, what are the main forms of social power? My approach is very much that of a natural scientist. I like to connect society-level dynamics of power to “atomic” or “elementary particle” interactions. The atom of power is a dyadic interaction between an influencer and the influenced. There are four general ways one person can affect the behavior of another, ranging from the crudest to the most subtle ones.

Coercion: I can force somebody to do my will, either physically or by threats. Mann calls this form of power “military,” but that’s too restrictive, because coercion is wielded by, yes, the military, but also by the courts and the police, as well as by gangsters. Also, coercion, or punishment, can be not only physical injury, but take other forms, such as a prison term, or imposing a fine, or threatening to take your house away.

Economic power: I can pay you to do something (or to avoid doing something). This is less crude than coercion, but still quite direct and easily understandable. Economic power is wielded, quite obviously, by wealth-holders.

Administrative power: I can order you to do something by virtue of being your boss, a superior in an organization. Mann calls this “political power” but, in my view, politicians (especially if we think about politicians in democracies or republics) start by wielding persuasion (the last form of social power), and only once they are elected exercise administrative power. In its pure form (but see below) administrative power is exercised without coercion or material pay-offs. We (or most humans) are simply used to following routines, and one of them is to follow the direction from the boss. Administrative power is the one that has the greatest effect on behavior of people living in complex societies organized as states (which means, most of us), because it regulates our day-to-day behavior—when we get up in the morning, drive to work (following traffic rules), perform our jobs, etc.

Persuasion, or ideological power: this is the most subtle, but extremely effective (when it works) form of power. It is especially effective because it is often not appreciated enough, or even not considered as a form of power. After all you end up doing what I want because you want it, right? Persuasion is exercised by aspiring politicians, orators, religious figures, thought leaders, and opinion influencers working for the traditional media or, more recently, on social media. It is also wielded by those who have a lot of “prestige,” such as popular actors and actresses and other celebrities. Uber wealthy also acquire prestige, and can influence how other people think or behave even without paying them.

One important caveat: this is a model, which breaks down complex reality in simple terms. In real life, in most situations several kinds of power are often mixed together. The most “promiscuous” one is probably the administrative type, because when your boss orders you to do something, the specters of other kinds of power often stand in the back. You can get punished (fired) or, alternatively, rewarded by a pay increase; and persuasive, charismatic bosses are more effective than those that rely merely on carrot and stick.

Understanding the sources of social power is very important if we want to understand how our societies operate, and why they sometimes step on the path of political disintegration. The key question is who are those who concentrate social power in their hands and how they cohere in networks of power. But that’s a topic for another time.

Notify of
Most Voted
Newest Oldest
Inline Feedbacks
View all comments
Joao Wemans

Nice list. I didn’t check Twitter discussion or references but, from my experience as complex systems and collective inteligence I fell there is another form if social power: group dynamics participation. Like family, self-managed companies… or mobs even? I think is missing some form of power source coming from my involvement in the group dynamics that influences others.

Other sources of power I feel are missing but maybe are included in yours:
example – do as i do.
compassion – help me because I need.
conformity – do as i told you others are doing.

Jon Labahn, Jr.

Thank you, Sir.
As the basis of transactionalism that (calculable) lends itself to Cliodynamicism (your Specialty), the simple and ‘planar repect’ facts of interactive matters is useful and priority to items such as Sesehat Database (another of the items in your technologic); grateful, as I said for the “simplistico”!

Jan Steinman

First, what is social power (as distinct from physical power measured in watts)?

Here’s a somewhat whimsical take on this I did for Communities Magazine: An Energy Primer: How We Consume Our Ancient Sunlight

Godfree Roberts

You forgot Confucius’ fave, exemplary power: the influence exercised by good example.

It’s been the core of Chinese governance for 1500 years and the Organization Department, China’s massive HR division, wouldn’t have it any other way.

“The rule of virtue can be compared to the Pole Star which commands the homage of the multitude of stars without leaving its place.” (2:1)

It helps explain why China has the lowest serious crime, lowest incarceration, lowest recidivism, unarmed cops and the most trusted legal system on earth: because their government sets an example of unselfish service, and provides everyone with homes, food, clothes and incomes. In other words, its actions are exemplary in their beneficence, too.


Given that China is also an extremely authoritarian state with deep corruption and a myriad of human rights abuses I would be skeptical about Chinese statistical data not gathered by third party sources.

Jay S Bancroft
Dennis Smithson

“It helps explain why China has the lowest … lowest incarceration”

You mean not including millions of minority members in the custody of “camps”?

“It helps explain why China has the … most trusted legal system on earth”

Well there’s not much competition among legal systems on Mars. Oh wait, you said Earth …

I wonder if your newsletter is funded by the Chinese government or a front thereof. I know, you’ll say it isn’t …

John Martin Strate

It is possible to read an introductory text in political science without seeing a clear definition of power. Yes, it is the central concept in the discipline. So it is a bit humbling to be a political scientist. I’ve used ideas from an old intro text by Fred Willhoite. Here’s what I use. Power is the capacity to produce intended effects on the behavior of others. A person wielding power can get others to do things that they would not otherwise do. There are four sources of power: authority (traditional, charismatic, legal, i.e. administrative) coercion (physical or psychological), inducements (rewards or their withdrawal), and persuasion rational , manipulation involving deceit, lies and taking advantage of cognitive biases), indoctrination). Authority is the right to give commands with the expectation or social norm that others will comply with those commands. Inducements include a large number of valued goods and services (e.g., money, promotions, special privileges, sex, honors, food items, etc.). Give limitations of information, time, and cognitive capacity, bounded rationality would qualify as rational. So there is a lot of overlap in overlap with respect to definitions of power across the social sciences.

Tom Christoffel

Is not power something that comes with authority and is based on social relationships, beginning with the family. Parents have authority; grandparents have authority; older siblings may have authority as assigned. Outside the family, in the culture, there are various authorities which have power of the economic and administrative sorts. These are legitimate. Contests for that authority may involve coercion and/or persuasion. There can be situation of nominal power and actual power. Dad does what Mom says. The Mayor does what the ___ wants, balancing interests in the moment. You don’t want to be drafted and go to war, but you can’t embarrass your family, so you do what you must and try to survive. Understanding who has power in your world is important, but it may not be easy to figure out and there may be no one to tell you. In a democracy, the Courts have authority, but they can be corrupted over time as judges are replaced, administrative power degraded.

Peter van den Engel

This is an interesting topic.
However before we can call it a science, we need to have the parameters correct.
Otherwise we don’t know what we are talking about, or it cannot even be called a
science, or alternatively it falls under a different category of a ‘science’ under a different nominator.
First let me explain why the current title “social power” bothers me. Largely because the definition is incorrect. Although a number of well defined categories with their
implicit differences are mentioned in the content. The question is of this is science: whether the
assumed consequences (effects) fall in the category power. My first observation is: they do not.
Because they all fall under the opposite definition: they all represent weaknesses.
Hence you cannot refer to it as a power. The paper itself mentions them as influences. Obviously an influence, influences a power and therefore is not the power description itself.
But a derivate.
So the first point (falsification) is taken.

This leads to the second point: what is the relation with social? Especially when the
author himself refers to the implementation of natural physical laws. Like those of
particles. * Which by itself is incorrect or incomplete: because it should be about the wave function. Humans are observers and the observer states the particle (what it is). Heisenberg’s law.
The particle is the end, not the beginning condition. Which is usually needed before science can make any predictions.

Hence without the observer and his or hers assumption, this reality would not exist. Not be relevant. The influencer is the wave function, witch is undecided, hence the influenced decides itself by chance about what information it perceives. The function is reversed to the presented theory.
So the conclusion is the observer derives the world from what he has observed: from what the community he is in has remembered. Because one (the innocent child) functions based on information witch must exist somewhere. Hence it must always be bias.
Because the effect of evolution is change. Adaptation to itself.

Put in simple words: because it is a learning process, the assumption earlier made must
contain a mistake (or incompleteness). To prove why it is mostly based on biases (or
cultural experiences which do not cover the whole: that is, the future also keeps changing as well).
These habits mean (which is correctly described as such in the paper: human
behaviour is based on habits)… they do not contain (different) information anymore!
Which is a perfect description of a bias.

Strengthening this is the information we daily get: our reality is circular (like a clock). Which makes it perfectly predictable. While evolution tells us it is constantly changing?
Which by itself already leads to the conclusion, the assumption reality is predictable, still represents a true law of nature as well. So long as it does not change.
This leads to what is dominantly called social: the average behaviour of a creature, including the farting and the lies. Which means it tries to adapt to a reality (squirming) which is overall not understood (when the boundary leads to uncertainty, it is denied: it will always stay within its niche). It is by definition conservative. Not a force, but an evasion (from death).

Second point taken: social cannot be the title of a force. But that of a weakness (limitation).
Unless circumstances have changed and provide different information.
Which is exceptional.

* Hence the baby boomers at the time were much more adventurous (creative) than humans normally are. Because of the promess of rising wealth and the introduction of new world media (television) falsifying old paradigms or at least breaking them open.

** On a side note: this explains also why the so called recurring trend of a cultural
conflict about every 50 years, as an explanation is not incorrect. But incomplete.
The ‘revolution’ of the seventies cannot be compared to a civil war. Like what happened in the Soviet Union 50 years earlier. Because this one was largely playful (with an exception for the US, because of the mandatory draft in relation to the war in Vietnam: although MASH or “We are all going to die” of course cannot be called a serious comment on the war nor a vialant protest :-).
Besides, where did WW II come from? When it did not fit the pattern, ignore it happened (one of the reasons academics have made a laughing stock out of themselves, nobody takes serious anymore).
Don’t get me wrong I highly respect you all for your intelligence and kindness, but not being able to think more holistically (complete) is a real problem.
* Your curiosity does no liberate you from your duties. It is not for free. Please grow up!

To get back to the subject. Social represents limited or bias behaviour (niche attitude) and therefore cannot be called a natural force. Its squirming is rather about its psychology: to define why it does what is does. To deal with a force it cannot handle nor understand. Thus to learn about how to persuade him, in influencing his irrational behaviour, perhaps is an (questionable) art, But certainly not a science.
Because the effect is unpredictable with a learning species. You would have to decide when it would remain dumb under what circumstances and when on the contrary it would reach a learning stage about something and what.
Because it remains independable.
Now we are touching on the real definition, in terms of its natural physical properties. Which is about a force (not a weakness) and is not social. But impersonal.

The description is: human consciousness is by definition constantly undecided. In
natural physical terms: the quantum state is endless: 0 equals infinity. Witch is his evolutionary programming: to survive under all circumstances. Because when the world is constantly changing it is unpredictable by nature. The state allows for the highest sensitivity for any information.
THAT is the FORCE.
Although conservatism (the bias) protects him from not having tested the ultimate impossibility to survive under his own stupidity witch created the secondary circumstances.
Because the world is only how he remembers it.

In the light of the incomplete Darwin theory: a species proves its survival by its offspring
(when the street is wet obviously it has rained, explaining nothing abut why and if it rained: the stupidity to elevate statistics over theory as a proper scientific conclusion is very old).
It is a scapegoat for not having any idea about the cause, although Darwin’s branches theory was brilliant.
This discloses, the genes (or nature) always includes the two relevant opposites. That’s why there are conservatives and creatives, or heroes next to cowards in the world. If one does not win (survive) predictably the other one will. Only then they get a random offspring.

Extending this it would be very interesting against the background of the proper definition, to in hindsight find out what the mentioned traits really represent: coercion, economic power, administrative power and persuasion. As human weaknesses.
Unless I had not already overstretched the space and time limitations.

Peter van den Engel

Peter van den Engel

@ Dennis Smithson
This is called social pressure. Which is well known in group cultures. The same goes for Japan.
The outcome is only few people end up in prisons, because everybody knows crossing certain rules will end up in social exclusion (total sensorship).
It leads to either extremely predictable cultures as reliable, or extremely conservative cultures, never learning.
In the early twentiest century this was exemplary for Austria. Leading to Kafka, Hitler, but also Freud and Jung as counter reactions.
Their ‘luck’ was they were not censured.
Likely because they were not taken seriously or ignored by the elites.

Bob Nuber

I see that you’ve asked a broad question, “what are the main forms of social power?”, but have narrowed your list of Social Power methods to the “ways one person can affect the behavior of another.” Like Joao Wemans, I feel that group dynamics participation is a significant form of social power missing from the list, perhaps due to your focus on a single actor.

Much social power is wielded through fear of exclusion, sometimes explicit but more often implicit and even imagined. Conformity seems a fine term for how groups influence a person’s behavior: ‘going along to get along’, ‘being a team player’, ‘rallying around the flag’, etc.

Bruce Lepper

Vladimir Putin is often described in the west as an architect of social change in Russia over the past 10 years. Is this an example of the use of social power and if so how would you rate the importance of each of the four different forms you have mentioned?

Michael Elling

At the end of the day, all 4 relate back to people’s tolerance of risk and their level of trust or belief in that power. Risk and trust have an inverse relationship. The problem is that trust can go to 100%, but risk can never go to zero. Ultimately, it is well structured institutional frameworks (aka networks) that reduce risk; or individuals’ perceived level of risk and trust of that institution/network.

Gerrit Van Wyk

All models of power suffer from the same problem; they assume a mechanistic reality. It means power is a thing, you can classify it, and one can know that thing by studying it. If, on the other hand, reality is complex, power is not a thing, instead it is something emerging from human interactions and relationships. It means you can’t talk about power without talking about how people resist it, what power you talk about depends on a context, and power waxes and wanes. Depending on the ontology you pick, modeling power becomes a very different challenge.

John Strate

The list more closely mirrors the political science understanding of power: coercion (physical or psychological); authority or the right to give commands with the expectation that they will be obeyed (traditional, charismatic, legal-rational), inducements (rewards bestowed, withheld, or taken away), and persuasion (rational using arguments based on true facts; manipulational (arguments using selective facts, priming, framing, lies); indoctrination). Plainly, there is overlap between these sources of power in any given setting or use.) If power is defined as the ability to get others to do what the power wielder wants but not what the target would want or otherwise do it may be difficult to determine the counterfactual (what would the person have done in the absence of the exercise of power).

© Peter Turchin 2023 All rights reserved

Privacy Policy