This post stems from thoughts I’ve had recently about how power functions in society—for the purpose of organizing. I have also been thinking more about G.A. Cohen and Analytical Marxists discussions of functional explanations as it relates to historical materialism (Marx’s theory of history).  Finally, I have been inspired by Anwar Shaikh’s work on reviving the classical and business-literature concept of competition.

A considerable bit of editing was done to this post, and it was updated on December 24, 2020.

In systems with a large number of complex, varied agents and interactions, we can nonetheless get a pretty decent model of agent behavior.  We do so with a methodology that focuses on the selective pressures of their environment.  For individuals acting in the context of an institution, we might look at the selective pressures of the institution that they are a part of.

With enough uncoordinated individuals, the institution can select for those who fit in with the institution’s interests.  An institution can be said to have “interests” insofar as there are specific behaviors that the institution selects for in the way that people are hired, fired, brought in, removed, commended, and reprimanded.

Furthermore, there are also selective pressures on the interests of institutions.  Capitalist companies are, for example, under the selective pressures of their environment to act in ways that are expected to create profit.

There are valid concerns and limitations to this methodology for modeling behaviors of actors in the context of institutions, and for modeling the abstract behaviors of institutions in the context of the larger selective pressures that influence them.  Nevertheless, I believe this methodology is useful to both organizers and social scientists.

There is a clear need for modeling institutions, both from a social science perspective interested in generating knowledge, and from an organizer’s interest in understanding the best routes to effecting change.  However, with so many complex individuals and interactions, it might be difficult to see how we could move beyond statistical correlation and come up with something rigorous and explanatory.  One of the most powerful tools that I have come across for understanding and predicting institutional behaviors draws from evolutionary thinking.

Although each of us is a complicated being with our own personal lives and thought processes, in the context of institutions, people tend to become somewhat predictable.  This is because an institution’s rules around hiring, firing, promoting, demoting, etc. can be thought of as a selection mechanism for the behaviors of the actors within the institution.  There are two main mechanisms for this:

• First, the institution refuses to hire, bring in, or promote people who display behaviors that are judged to go against the institution’s interests. This means that people who get brought in or promoted to decision-making positions tend to hold beliefs and patterns of behavior which are in-line with the institution’s goals.

• Second, if someone gains power, but their behavior goes against the institution’s interests, then the institution is set up to fire them or remove them from their role.

These two main mechanisms act as a selection mechanism for the types of behaviors of people in institutions. They particularly help select for the behaviors of people with decision-making roles because, to get appointed, those people would need to have successfully passed through multiple filters.

These selection filters also tend to be maintained through time.  The decision-making positions which have the power to change the selection filters have selected for behaviors that uphold the institution’s values. As the filters tend be in the interests of the institution, the decision makers tend not to change them much—if at all.

The institutional filters are usually robust against variations in individual behavior—if an individual varies too far from the institution’s rules and norms, they are ejected from the institution, and they can no longer affect the institution from the inside.  Individual actors cannot change the selection mechanisms alone—generally, it takes collective action or overwhelming pressure from outside of the institution.  This resilience against individual actions is why institutionally-selected behaviors are sometimes considered ‘structural’.  For similar reasons, people who criticize institutions sometimes use the term ‘structural issue’ or ‘structural problems’.

In Practice...

Let’s make the discussion more concrete and look at a couple of examples.  You can skip over this if you feel like you have a grasp of the concept already.

Capitalist businesses appoint CEOs and top decision-makers through shareholder votes.  Shareholders are primarily interested in profit.  If a CEO does not make decisions to effectively increase profits for shareholders, the CEO gets voted out.  The CEO is aware of this and makes their hiring/firing decisions accordingly. [1]

The pressure toward profitability is also passed down to managers lower in the hierarchy. Even if the manager is a kind person, if they will overwhelmingly choose to follow orders from higher up—for example, if they are told to fire a portion of their team or anything else that makes the working lives of those under them worse. If they refuse, they’ll get fired themselves, and a part of their team will still get fired, but by the manager’s replacement.  The selective pressures of profitability are further passed down to employees, who signal their ability to make the company profits—so that they can get hired by the company, or so that they aren’t fired and replaced by someone who looks as if they can make profits more efficiently.   People are hired based on the expectation that they are willing and able to act in the company’s interests.  After a person makes it past the hiring filter, the threat of losing one’s position is then usually enough to make people toe the company line, and internalize the company’s values.  If this threat isn’t enough, then they can always be replaced by someone who will act in the company’s interests.

Another example. Politicians are motivated by what will make them more likely to get elected or re-elected.  Politicians have a decent amount of space for variation in their actions and hiring decisions. Ultimately, however, the behaviors present in political positions are restrained by the individual politician’s internalization of their re-election pressures, as well as by the actuality of whether or not their behaviors cause them to lose their election. [2]  With more polarization, politicians have been able to exhibit a wider range of behaviors within their party. You might also observe that politicians feel freer to speak their mind or vote their conscience when they aren’t seeking reelection. Parties and wealthy donors also filter for behaviors of politicians, through selective endorsements, funding, and advertisement campaigns.

I will include a couple more examples at the bottom of this post (about non-profits and police departments). [3] More nuance is required in all of these examples, but hopefully this section has conveyed the logic and effectiveness of this way of analyzing institutions as well as the behaviors of people within them.

Institutions Also Face Selective Pressures

The behaviors of institutional actors are selected for within institutions.  However, the interests of these institutions are themselves often shaped by even larger selection processes.

Let’s take a look again at businesses under capitalism.  Why are they profit-oriented, rather than being oriented directly toward the creation of value for their communities?

First, there are the most basic selective pressures necessary for staying in existence.   If a company cannot fund its expenses, then it will not be around for long.  It simply will not have the money for things such as repairing/maintaining capital, paying rent on land, and paying their workers.  There is a selective pressure coming from the basic requirements of what a company needs to stay active and incorporated.

Furthermore, they need to remain competitive on the market because that affects their market share.  Market share, in turn, affects how much money they make, and whether they can fund their expenses.  Competition is thus a selective pressure which can hitch a ride alongside the basic needs of an active and incorporated company.

Finally, companies which are under the pressure of capitalist competition need to do more than simply balance their budget—they need to turn a profit.  Why is this? As we’ve already described, businesses are pressured to stay competitive. To do that, they enough funding to undercut the prices of their competitors until their competitors give up, differentiate, or go out of business. [4]  In that situation, the company will not be making a profit—they are cutting their prices and are likely running at a loss.  To be able to outlast the competition, companies need to get a great amount of funding from outside of the company.  The solution?—the companies go to capitalist investors.

These investors demand partial ownership of the company as a condition for their funding.  Investors also make funding choices with the goal of long-term profitability, so that even if some ventures fail, the others that succeed are more than enough to make up for the losses.  Companies are selected for based on their expected value for the investor, and capitalist investors tend to have enough bargaining power to demand ownership and a slice of profits, with no contractual upper limit.  Therefore, the only companies that last are overwhelmingly those which have a profit motive.

The capitalist mode of funding is a highly effective selection filter that essentially only allows through profit-motivated companies.  If companies want to survive competition with other well-off competitors (and they don’t have a natural monopoly or enough savings to self-fund), then by the capitalist system of funding, they will be forced to become profit-oriented.

~

In this example, we saw that there are multiple selective pressures, and that these can be used to explain why businesses under capitalism overwhelmingly tend to have profit-motivated behaviors.  This set of behaviors—abstractly, the company’s interest in profit—are reified and objectively observable in the company’s rules and norms around hiring, firing, promoting, demoting, commending, reprimanding, etc.  The selection pressures of capitalism have been passed down the levels of the hierarchy—with a surprising degree of accuracy and precision.

There are other society-wide selection pressures that affect institutions’ interests.  My plan is to write more posts about the selection constraints on political positions, universities and education, etc.  Famously, Edward S. Herman and Noam Chomsky’s Manufacturing Consent described the selection mechanisms shaping media news companies’ behaviors.  Also famously (but less so), G.A. Cohen’s Karl Marx’s Theory of History: A Defence was one of the first to rigorously describe functional explanatory power, and he would emphasize the link to evolution in later discourse with other Analytical Marxists.

Usefulness of the Method

This is an interesting way to analyze institutions and institutional actors, but does it say enough to be useful?  There are also genuine limitations to this methodology which I will get into shortly, but do these limitations render it useless?  I think it’s useful in several ways:

If we want to know the behavioral patterns of a person as they act within an institution, or the likely behavior of an institution within a larger set of selective pressures, then this modeling tool makes a good first guess.  Additionally, by observing the strength of the selection filters, we can get an estimate of the accuracy of this first guess.  From there, we can update our beliefs as we gather new and more individualized evidence.

For organizers who want to affect institutions, it gives us a way of pinpointing people’s and institutions’ likely guiding motivations—and therefore the sorts of actions we can take which they will respond to.  Want to get concessions from a business?  Threaten or hurt their profitability.  Want to get concessions from a politician?  Threaten or hurt their electability.

It also emphasizes the structural nature of the issues that organizers want to change—this helps them focus their efforts toward effective actions, rather than putting all their hopes in making a well-argued moral appeal to individuals.  Individuals tend to be constrained by the selective pressures of their position—this is why such appeals are likely to fail.  For example, politicians often care more about not risking their re-election.  Even if they agree with you morally, they usually won’t take action unless they are confident that it will benefit their job prospects.  With this analysis, it becomes clearer that: to affect change in government, organizers usually need to build a mass movement of people.  The movement can give the politician the confidence to do what they think is moral.  And if the politician doesn’t agree that the demands are moral, and they don’t follow through, then the movement has the momentum to vote in somebody who will.

Perhaps one of my favorite parts of this model for organizers is that it allows us to have a nuanced understanding of categories: both the categories of institutional actors and the categories of institutions.  For example, the lines within economic classes become more obvious.  Our organizing can be improved by looking at the different selection filters faced by non-owners, including managers, high-income workers, lower income workers, unemployed, etc.  We can see why managers tend to side with the owners even if they don’t have any ownership stake in the company—it’s how the managers keep their job.  Within the owner class, we can make useful distinctions between small business owners and large business owners, largely by looking at the different selection filters they face.  The lines between institutions also become clearer.  For example, it can be reasonable shorthand to group capitalists and politicians together into a single ‘ruling class’.  Nonetheless, there are clear differences between how politicians stay in office and how capitalists stay in business. There are also clear overlaps where capitalists affect political elections, or politicians affect capitalist profits.  Organizers can exploit this understanding to their advantage.  All of these observable differences will help us forecast how structurally-defined groups of people will come down on each of our policy fights—are they likely to be on our side, against us, or on the fence? [5]

For economists and other social scientists, this methodology can help abstractly model behaviors and interactions.  Insofar as there exists variation among agents, a large pool to select from, and selective mechanisms, these predictions will hold true on average. And even though we aren’t always able to predict when something is going to occur, this methodology helps point us to variables that can help us determine the frequency of events. That is, we may be able to make forecasts, even if we cannot make precise predictions.  The methodology also provides social scientists a way to trace some of their assumptions about behaviors back to the selective pressures on these institutions.  Additionally, having a place where behaviors converge at a more abstract, institutional level—even if it is merely stochastic—provides a robust place to begin modeling the interactions among institutions.  Finally, this methodology has the potential to create links between micro and macro models.

Limitations to the Method

There are genuine concerns, questions, and limitations to this natural-selection-inspired methodology.  This post is long enough already that I am going to save the detailed discussion of these limitations to a dedicated post.  I will link to that here when it is out.  For now though, I will quickly list off a few main concerns, limitations, and areas for future research:

• Where does the analogy with evolution break down?

• Haven’t we heard people talk about Social Darwinism before, and didn’t that turn out badly?

• How can we even say that an institution has ‘interests’? Isn’t that just metaphysical mumbo-jumbo?

• If an institution’s interests determine the behaviors they select for, and the behaviors that they select for define an institution’s interests, then isn’t this circular reasoning? How does this hold up, particularly when we know that decisionmakers in institutions can modify their practices with regards to hiring, firing, etc.?

• How can this be useful if it is only abstract? Where is the usefulness of these abstracted (stochastic) models of behaviors?

• How much are actors’ behaviors selected by the actuality of hiring and firing? How much are they dependent on having agents internalize and base their decisions around the motivations of getting hired and avoiding being fired? That is, are there patterns in how much leeway people have within institutions, and does this throw doubt on the value of the overall analysis?

• Perhaps there are agents who have different motivations than those of the institution, but they make it through the selection filters because they just so happen to act in the ways that are being selected for. Doesn’t this mean that the model lacks explanatory power about behaviors, and especially lacks predictive power about future behaviors?

• Institutions can influence each other and influence the selective pressures that affect them. For example, monopolies seem to remove some selective pressures of competition, and sometimes, politicians change the constitution to keep themselves in power and avoid the selective pressures of elections. How does this methodology handle reflexive influences like these?

• Even though selection pressures might result in a statistically tight distribution around a set of behaviors, is this useful for making predictions about the interactions of institutions? Institutions—even after the selection process—will still have enough unpredictable variation to create an unpredictable, chaotic system.

In Conclusion...

Ok, wrapping up.  In systems with a large number of complex, varied agents and interactions, we can nonetheless get a pretty decent model of agent behavior.  We do so with a methodology that focuses on the selective pressures of their environment.  For individuals acting in the context of an institution, we might look at the selective pressures of the institution that they are a part of.

With enough uncoordinated individuals, the institution can select for those who fit in with the institution’s interests.  An institution can be said to have “interests” insofar as there are specific behaviors that the institution selects for in the way that people are hired, fired, brought in, removed, commended, and reprimanded.

Furthermore, there are also selective pressures on the interests of institutions.  Capitalist companies are, for example, under the selective pressures of their environment to act in ways that are expected to create profit.

There are valid concerns and limitations to this methodology for modeling behaviors of actors in the context of institutions, and for modeling the abstract behaviors of institutions in the context of the larger selective pressures that influence them.  Nevertheless, I believe this methodology is useful to both organizers and social scientists.

Footnotes & Extra Bits

[1]

Even Milton Friedman recognizes the selective pressures on CEOs and capitalist businesses.  In a famous article in which he justifies the greed of corporate profit-seeking, ‘The Social Responsibility Of Business Is to Increase Its Profits’:

And, whether he wants to or not, can he get away with spending his stockholders, customers’ or employes’ money? Will not the stockholders fire him? (Either the present ones or those who take over when his actions in the name of social responsibility have reduced the corporation's profits and the price of its stock.) His customers and his employes can desert him for other producers and employers less scrupulous in exercising their social responsibilities.”

[2]

Having space for behavioral variation allows politicians to express their personal ethics in their decisions. When structural behavioral constraints are placed on this variation, the selective pressures tend to result in a ‘satisficing’ decision-making process. If you think that people are only thinking about their re-election and have no other interests, it seems like people are merely making adequate decisions, rather than the perfectly rational ones that maximize their chances. Of course, it’s not the case that politicians are only thinking about re-election—I believe that so-called satisficing behaviors are emergent out of the combination of motivational variations and institutional selection constraints.

[3]

Non-Profits tend to shape their organizational structure to represent the interests of their wealthy donors—for example, by becoming top-down hierarchies in their decision making so that it’s easier to give updates to their donors and easier for donors to influence the decisions of the non-profit.  (I also talked about this in my previous post.)

Police Departments often promote camaraderie among officers—it’s highly expected that they have each other’s backs in any situation.  However, this tends to extend to times when police murder civilians—any ‘good’ police officers who speak out about police misconduct will usually retire or be shunned and disrespected out of the organization by their coworkers.  In the worst scenarios, some of the police will physically threaten outspoken do-gooders, while the department refuses to hold them accountable.

[4]

The sort of competition described here is essentially Anwar Shaikh’s concept of ‘real competition’. It is inspired by classical economists like Smith, Ricardo, and Marx. Shaikh contrasts it to concepts of ‘perfect competition’ and ‘imperfect competition’, though I think those models can still be useful to a degree.

[5]

This idea comes from a 1978 poster created by a Marxist Feminist collective, Press Gang Publishers, which I saw in Erik Olin Wright’s book Understanding Class.  The poster says: “Class consciousness is knowing which side of the fence you are on.  Class analysis is figuring out who is there with you.” (p. 165)  To which Wright adds in lectures given on his work, class analysis is also about knowing who is still on the fence.