Groups and Social Identities
POLSCI 240 / PSY 225: Political Psychology
February 24, 2025
Cooperation and conflict
- How do humans sustain cooperative relationships for mutual gain?
- Why do we see so much intergroup conflict in human societies?
- Why is “ethnicity” so important to human social organization and psychology?
Ethnicity
Symbolically marked boundaries associated with cultural traits. - Moya & Henrich (2016)
Self-ascribed membership [in a group] marked by seemingly arbitrary traits such as distinctive styles of dress or speech. - McElreath et al. (2003)
…[a group with common] points of personal reference [such as] perceptions of common descent, history, fate, and culture, which usually indicates some mix of language, physical appearance, and the ritual regulation of life, especially religion. - Hale (2004)
Sequence for this section
Simple rational models of conflict and cooperation: what do we learn? What does it take for cooperative strategies to succeed, and what do those conditions imply for human social and moral psychology?
How do group identities help people navigate the complexity of the social world? Why is ethnicity ubiquitous in human societies? Why is ethnic conflict so common?
What are the motivational implications of social identification and how does it shape people’s political thought and behavior? How can “internalizing” group membership promote both human cooperation and conflict?
Simple models of cooperation and conflict
The Prisoner’s Dilemma
- Left: a specific example - 1st value in each cell is payoff to row player
- Right: the generic version - all entries are payoffs to row player
| Cooperate | Defect | |
|---|---|---|
| Cooperate | (3, 3) | (0, 5) |
| Defect | (5, 0) | (1, 1) |
| Cooperate | Defect | |
|---|---|---|
| Cooperate | b - c | -c |
| Defect | b | 0 |
“I would, if you would, but I don’t trust you will, so I won’t”
Evolved solutions and moral psychology
Let’s ask a slightly different question
- Under what conditions can a cooperative strategy succeed in a population with many potential defectors?
- Can cooperators “survive” and prosper? If so, how?
We are asking an evolutionary question, but we can think of it either as biological or cultural evolution
- Strategies with higher payoffs “reproduce”, whether through imitation or inheritance
Payoffs
| Cooperate | Defect | |
|---|---|---|
| Cooperate | b - c | -c |
| Defect | b | 0 |
Expected payoff to Cooperate strategy: \(p(b - c) + (1 - p)(-c) = pb - c\)
- \(p\) is the probability of playing against another Cooperator
- \(1 - p\) is the probability of playing against a defector
Expected payoff to Defect strategy: \(pb\)
- Defectors always dominate when interactions are random
Positive assortment
Positive assortment is when people who share a trait interact at higher rates than expected under purely random interaction
Compare the payoffs for the two strategies again, but this time let’s allow \(p\) to vary across them:
- Cooperate: \(p_1 b - c\)
- Defect \(p_2 b\)
If \(p_1\) is much greater than \(p_2\), this means Cooperators are able to interact mostly among themselves
- Cooperate gets mostly \(b\) payoff
- Defect gets mostly 0 payoff
Stable, local interactions
One way to get a high \(p_1\) is when people only interact locally
In such cases, cooperation can build on local clusters to spread
We end up with cooperative enclaves or clusters that sustain themselves by interacting among themselves
Cooperation survives by being parochial
- Limited in scope, narrow
- Insular, inward-looking
Example
Example (\(c = \frac{1}{5}b\))
Non-local interactions
But once we allow for movement (“migration”), rather than perpetual local interactions, this falls apart
- Once defectors can invade clusters of cooperators, cooperation cannot be sustained
- Cooperators can no longer “trust” that any given interaction partner will be another cooperator
The general solution is some form of conditional cooperation
- Cooperate against other cooperators, but not with defectors
- Must be able to recognize other cooperators
Tit-for-tat
A norm of reciprocity is a strategy of cooperating if the other person does, and defecting if they do
“tit-for-tat” (TFT) is a strategy that always cooperates on the first round, and then copies what the other person did for every subsequent round
TFT is a “nice, but wary” strategy (“wary cooperative”)
- Initially cooperative, but quick to defect
An example:
\[ \begin{align} \text{C C C D D C ...} \\ \text{C C D D C C ...} \end{align} \]
Tit-for-tat
A norm of reciprocity is a strategy of cooperating if the other person does, and defecting if they do
“tit-for-tat” (TFT) is a strategy that always cooperates on the first round, and then copies what the other person did for every subsequent round
TFT is a “nice, but wary” strategy (“wary cooperative”)
- Initially cooperative, but quick to defect
TFT against always-defect:
\[ \begin{align} \text{C D D D D D ...} \\ \text{D D D D D D ...} \end{align} \]
Tit-for-tat
A norm of reciprocity is a strategy of cooperating if the other person does, and defecting if they do
“tit-for-tat” (TFT) is a strategy that always cooperates on the first round, and then copies what the other person did for every subsequent round
TFT is a “nice, but wary” strategy (“wary cooperative”)
- Initially cooperative, but quick to defect
TFT against TFT or always-cooperate:
\[ \begin{align} \text{C C C C C C ...} \\ \text{C C C C C C ...} \end{align} \]
TFT and conditional cooperation
TFT can be sustained in a population with defectors, because it is conditionally cooperative
- TFT will encounter defectors, but will quickly copy defection, neutralizing benefit to D
- And when it encounters itself (or an unconditional cooperator), it will be able to sustain that cooperation
If # of “rounds” of the game is long enough, the benefits of sustained cooperation outweigh costs of losing first rounds to defectors
- TFT can “resist” a D take-over if \(wb > c\), where \(w\) is the probability the game continues for 1 more round
- TFT can “invade” a D population if there is also positive assortment
Cooperative markers
Another mechanism for positive assortment is leveraging visible markers of cooperative dispositions
- Cooperate with people who have the marker
- Defect against those that don’t
Issue is that defectors will “disguise” as cooperators
- Need markers that are credible, hard to fake - may not always (or even often) be possible
- One obvious exception: kinship and its correlates (markers of common ancestry)
Summary
Our analysis suggests that evolved human cooperative psychology (“moral psychology”) is likely to evince a few important characteristics:
- Reciprocity: give what you get, eye for an eye
- “Wary cooperation”: initially cooperative, but quick to punish, and slow to forgive
- Parochial: initial cooperation (“blind trust”) reserved for those who are “close” in some sense
- Cooperative identity markers: concerned with “credible” signals of a shared cooperative community
Yet, the “scale” (spatial, cultural) of ethnic distinctions seems hard to square with the results from analyzing the prisoner’s dilemma - especially the need for “difficult-to-fake” signals
Summary of last class
We are interested in the requirements for sustaining cooperation, what these might tell us about human social psychology, and how they are related to ethnicity and ethnic conflict
If we model human interactions as a one-shot prisoner’s dilemma, “defection” is the equilibrium
If we allow for the possibility of continued interactions over time (and in) space, cooperation can survive, but only if cooperators can interact selectively with other cooperators
Such conditional cooperation is seen in:
- “nice, but wary” strategies like TFT
- in “spatially” isolated communities of cooperators
- in cases with visible, hard-to-fake markers of cooperative dispositions
Problems with this account?
While this analysis might help us understand some of human social and moral psychology, these pathways to cooperation seem limited in their ability to explain the “scale” of ethnicity and ethnocentrism
- Large populations of people who mostly don’t know one another
- Often spatially separated (e.g., shared ethnicity across borders)
- Often heterogeneous, with only a relatively weak degree of shared kinship or ancestry
- Strong reliance on rather arbitrary “markers” for group boundaries, yet PD analysis suggests markers need to be hard-to-fake
A different kind of game
Coordination games have a different structure from prisoner’s dilemma games (payoffs shown for row player, b > 0):
| Behavior X | Behavior Y | |
|---|---|---|
| Behavior X | b | 0 |
| Behavior Y | 0 | b |
In this game, both players just want to match each others’ behavior
A relevant example is local cultural norms, such as etiquette
- There is very real sense in which these norms are arbitrary
- But once established, they foster smooth social interaction
How might this model help us understand the emergence of ethnic group boundaries?
The role of markers
Imagine there are two groups, initially segregated (no interaction)
Group 1
| Marker A | Marker B | ||
|---|---|---|---|
| Behavior X | 60 | 20 | 80 |
| Behavior Y | 15 | 5 | 20 |
| 75 | 25 |
Group 2
| Marker A | Marker B | ||
|---|---|---|---|
| Behavior X | 10 | 10 | 20 |
| Behavior Y | 40 | 40 | 80 |
| 50 | 50 |
- Differ in which behavior is most common
- Both share same most-common “marker”
- Within groups, no association of behavior and marker
The role of markers
Imagine some migration: 10 members of Group 2 migrate to Group 1
Combined population: Group 1 + Group 2 migrants
| Marker A | Marker B | ||
|---|---|---|---|
| Behavior X | 60 + 1 | 20 + 1 | 82 |
| Behavior Y | 15 + 4 | 5 + 4 | 28 |
| 80 | 30 |
Markers associated w/ groups
- \(\text{Pr}(2 | A) = \frac{5}{80} \sim 0.06\)
- \(\text{Pr}(2 | B) = \frac{5}{30} \sim 0.17\)
Markers associated w/ behaviors
- \(\text{Pr}(X | A) = \frac{61}{80} \sim 0.76\)
- \(\text{Pr}(X | B) = \frac{21}{30} = 0.70\)
With success-based imitation and homophily
- Group 1 members will increasingly adopt Behavior X and Marker A
- Group 2 members will increasingly adopt Behavior Y and Marker B.
Extension of group differences
Once there is a difference between groups on one marker:
- Other markers correlated with the first become informative
- Over time, groups develop clusters of differentiating attributes
- Different “cultures”: language, dress, physical markings, music, etc.
And evolution here favors a psychology that uses these arbitrary markers
- The tendency toward similarity-based learning and homophily is reinforced
- Humans end up with a psychology that places much weight on “ethnicity” for navigating the social world, and this reinforces (feeds back into) ethnic differences (which reinforce ethnic psychology)
3-year-olds’ preferences for similar others
Learning from similar others
Hale: groups and uncertainty reduction
Social groups are strongly associated with various “distinctions” that matter for individual flourishing
- Intrinsically important distinctions (e.g., communication, norms)
- Distinctions imposed on people (e.g., allocation of material resources)
by identifying oneself as a group member, one effectively replaces aspects of individuality and unshared attitudes and behaviors with an “ingroup prototype” that prescribes shared beliefs, attitudes, and behaviors appropriate to that particular categorization. - (Hale 2004)
What is identity?
It is useful to treat the notion of identity as the set of points of personal reference on which people rely to navigate the social world they inhabit, to make sense of the myriad constellations of social relationships that they encounter, to discern their place in these constellations, and to understand the opportunities for action in this context. It is, in a certain way, a kind of social radar, a perceptual device through which people come to see where they stand in relation to the human environment.In the most basic sense, then, groups are defined by common relationships to points of social reference. - (Hale 2004)
What is special about ethnicity?
Hale (2004) argues ethnicity is special with respect to these important “distinctions” that people are concerned with - and this is why it is so important to human social affairs
- Cultural (e.g., religion, etiquette) and linguistic barriers to communication
- The salience of physical traits and their ease of use as cues
- Imagery of kinship and myths of common origin
- Geographic concentration and correlation with things like economic production and development
- More broadly, relation to socioeconomic class, often because of historical migration patterns and prejudice, both institutionalized and not
Ethnocentrism: from coordination to conflict
Ethnocentrism: evaluating other cultures using the norms and standards of one’s own culture; generalized ethnic prejudice
a predisposition to divide the human world into in-groups and out-groups…to us versus them…Members of in-groups are assumed to be virtuous: friendly, cooperative, trustworthy, safe, and more. Members of out-groups are assumed to be the opposite: unfriendly, uncooperative, unworthy of trust, dangerous, and more. - Kinder & Kam (2009), Us Against Them
Example: Kinder and Kam, Us Against Them
From American National Elections Study:
“Now I have some questions about different groups in our society.
I’m going to show you a seven-point scale on which the characteristics of the people in a group can be rated.
In the first statement a score of ‘1’ means that you think almost all of the people in that group tend to be ‘hard-working.’
A score of ‘7’ means that you think most people in the group are ‘lazy.’
A score of ‘4’ means that you think that most people in the group are not closer to one end or the other, and of course, you may choose any number in between.”
Example: Kinder and Kam, Us Against Them
Ethnocentrism measure:
- Calculate average (positive) rating of respondent’s out-groups on each trait
- Subtract average from in-group rating on each trait
- Average across all measured traits
Example: Kinder and Kam, Us Against Them
Ethnocentrism in contemporary US politics
Politics and positive feedback
Ethnocentric decision making
- Instrumentally useful b/c of consequential group distinctions
- e.g., inequalities of wealth, political power
↔︎
Consequential group distinctions
- Created/Reinforced by ethnocentric decision making
- e.g., institutionalized discrimination
Two views on identity and political behavior
Identities as heuristic tools
The social world is complex
Group differences correlated with interests
People use identity as a heuristic for own interests
- Who to trust/listen to, what policies to support, etc.
- Intergroup conflict can arise out of competing interests
Social identity theory
Callard on groups and decision making
The essential feature of kinship bonds is that they offer communal answers to questions…A kinship relation dictates that “this is how we live, here is how we do things, these are the things we care about, this who we are committed to protecting and providing for.” Belonging to the group gives you an answer.
Much of a person’s basic ethical “stance” is underwritten by one or another kinship relation. Group membership, in effect, stands in for the kind of justification one might have arrived at, had one been in a position to ask who or what one should care about.
Callard, Agnes. Open Socrates: The Case for a Philosophical Life (p. 82-83). W. W. Norton & Company. Kindle Edition.
Example: “linked fate”
Use the interests of one’s group as a heuristic to decide what policies and candidates to support
- From American National Elections Study: “Do you think that what happens generally to BLACK PEOPLE in this country will have something to do with what happens in your life?”
Two views on identity and political behavior
Identities as heuristic tools
The social world is complex
Group differences correlated with interests
People use (imperfectly) identity as a heuristic for own interests
- Who to trust/listen to, what policies to support, etc.
- Intergroup conflict can arise out of competing interests
Social identity theory
- All this is true, but purely “cognitive” approach misses a lot
- Intergroup conflict is “hotter”, easier to instigate, than heuristic account suggests
- Role of group membership is “deeper”, more integrated into person’s self-concept
Discrimination in minimal groups
Discrimination in minimal groups
Discrimination in minimal groups
SIT: primary theoretical claims
Social identification is an internalization of group membership
- Group becomes part of self-concept, a way to understand self
Social groups are represented cognitively by prototypes (“stereotypes”): the “typical” characteristics of group members
- People want to belong to groups w/ prototypes similar to themselves
People strive for positive distinctiveness for the in-group
- Group’s successes and failures are internalized, so want to belong to group with high status
Strategies for dealing with low status
Identifying with a low-status group has negative implications for self-esteem and people want to resolve this
- Depending on context, different strategies are more or less feasible
- Conflict: work to change the relative status of the groups
Implications for cooperation and collective action
Prisoner’s Dilemma again
What if you care about the total payoff, not (just) your own?
| Cooperate | Defect | |
|---|---|---|
| Cooperate | (3, 3) | (0, 5) |
| Defect | (5, 0) | (1, 1) |
Internalization of other group members’ payoffs can change the nature of cooperation games:
| Cooperate | Defect | |
|---|---|---|
| Cooperate | 6 | 5 |
| Defect | 5 | 2 |
Public goods game
Centrality of politics in self-concept
Asked two survey questions to a national sample of U.S. residents:
- “My political attitudes and beliefs are an important reflection of who I am”
- “In general, my political attitudes and beliefs are an important part of my self-image.”
Not only did many people agree with these statements, but agreement is increasing in political engagement (participation, interest, etc.)
- Correlation = 0.53
Examples: Partisanship as social identity
Huddy et al. (2015), “Expressive Partisanship”
Partisan identity and participation
Huddy et al. (2015), “Expressive Partisanship”
Partisan identity and emotions
Huddy et al. (2015), “Expressive Partisanship”
Examples: group consciousness
Group consciousness: a measure capturing the key components of SIT’s claims regarding the conditions for intergroup conflict
- Degree of identification with group
- Perception that the group has (illegitimate) low status in society
- Belief that social change is possible through collective action
Group consciousness and participation
Strategies for dealing with low status
Identifying with a low-status group has negative implications for self-esteem - people want to resolve this
- Depending on context, different strategies are more or less feasible
- Conflict: work to change the relative status of the groups
- Creativity: change dimensions of comparative evaluation
- Exit: identify with a different, higher-status group
One lesson from SIT is that intergroup conflict is not inevitable! People will use “easier” way to restore self-esteem, if possible
Modeling politics w/ SIT
SIT says that a full model of political behavior includes not only material self-interest, but also:
- People choosing social identities based on similarity and status
- Social identities affecting people’s outcomes (“utility”, value)
- People choosing political alternatives based on their combined interests, material and identity-related
Moreover, we need to consider the fact that everyone in the population is making these choices, and these choices affect each other’s outcomes
Moses Shayo’s model of SIT
Citizens are members of social groups (e.g., class, ethnicity, nation)
They choose to identify with one of the groups to which they belong, where identification is defined as:
- wanting to be similar to the group prototype
- wanting the group to have high relative status
There is also a policy issue being debated, on which each citizen will vote, which affects the status of individuals and, through individuals, the status of groups
Citizens choose a policy and an identity that maximizes their own utility (both material and identity-related)
Example: tax policy, identity, and group inequality
- Citizens have two group memberships, class (low or high) and nation (everyone is a member), and an income
- Citizens’ distance to in-group prototype, \(D\), is proportion w/ different income and class
- Citizens choose an identity, \(g\), either their class or the nation
- Citizens vote on flat tax rate, \(t\), and median voter’s preference is adopted
- Taxes are divided evenly among citizens giving final income, \(I\)
- Groups’ status, \(S\), determined by difference in average income between groups
Example: tax policy, identity, and group inequality
A person’s utility, \(U\), is a function of tax rate, \(t\), chosen identity, \(g\), and relative importance placed on income, distance (\(\beta\)), and status (\(\gamma\)):
\[ U(t,g) = I(t) + \beta D(g) + \gamma S(t,g) \]
- \(I(t)\) is read: “final income depends on the tax rate”
- \(D(g)\) is read: “distance to the in-group prototype depends on the group you identify with”
- \(S(t,g)\) is read: “identity group status depends on the tax rate and the group you identify with”
Person from lower class w/ lower class identity
- Their income (\(I\)) is always starting income + a tax transfer from rich
If they identify with their class:
- Their identity distance, \(D\), is minimized (shared income, class, and nation)
- Their status, \(S\), is relatively low, but increases as tax rate increases
- Will want to choose a high tax rate for both individual and identity reasons
- Will prefer class identification when status difference between classes is low (low inequality and/or high taxes) and/or national status is sufficiently low
Person from lower class w/ national identity
- Their income (\(I\)) is always starting income + a tax transfer from rich
If they identify with the nation:
- Their identity distance, \(D\), is higher (nation is mix of incomes and classes)
- Their status, \(S\), is (typically) higher, b/c status now depends (to some extent) on the incomes of the rich
- Will generally choose a lower tax rate than if identify with class
- Will prefer national identification when class status difference is high (high inequality and/or low taxes) and/or national status is high
Person from upper class w/ upper class identity
- Their income (\(I\)) is always starting income - a tax transfer to poor
If they identify with their class:
- Their identity distance, \(D\), is minimized (shared income, class, and nation)
- Their status, \(S\), is relatively high, but decreases as tax rate increases
- Will want to choose a low tax rate
- Will generally prefer class identification, unless national status is very high
Person from upper class w/ national identity
- Their income (\(I\)) is always starting income - a tax transfer to poor
If they identify with the nation:
- Their identity distance, \(D\), is higher (nation is mix of incomes and classes)
- Their status, \(S\), is (typically) lower, b/c status now depends (to some extent) on the incomes of the poor
- Will generally choose a higher tax rate than if identify with class
- Will only prefer national identification if national status is especially high
Implications
National identification reduces preferred tax rates among poor, but increases them among rich
High levels of inequality increase national identification among poor, but class identification among rich
In general, the poor are more likely to identify with the nation than the rich
In general, high inequality countries will have higher levels of national identification, but lower levels of redistribution
Two equilibria: U.S. vs Europe
National identity, income, preferences
Shayo (2009)
National identity, income, redistribution
Shayo (2009)
Status, conformity: individual & cultural differences
A person’s utility, \(U\), is a function of tax rate, \(t\), chosen identity, \(g\), and relative importance placed on income, distance (\(\beta\)), and status (\(\gamma\)):
\[ U(t,g) = I(t) + \beta D(g) + \gamma S(t,g) \]
Individual and/or cultural differences in the relative importance of these three components can shape the outcomes that obtain
- e.g., poor face a trade-off between in-group status and similarity to the prototype when choosing between class and nation
Social identity theory (SIT)