Advanced Sociological Theory PDF

Advanced Sociological Theory Structuur: **PART 1: Foundations, building plans, and tools** - **Week 1: The structure of sociological explanation** - **Week 2: Game theory and cooperation (I)** - **Week 3: Game theory and cooperation (II)** - **Week 4: Theories of action and model build...

Advanced Sociological Theory Structuur: **PART 1: Foundations, building plans, and tools** - **Week 1: The structure of sociological explanation** - **Week 2: Game theory and cooperation (I)** - **Week 3: Game theory and cooperation (II)** - **Week 4: Theories of action and model building** **PART 2: Applications** - **Week 5: Schelling's segregation model** - **Week 6: Collective behavior** - **Week 7: Emergence of norms** - **Week 8: Summary and wrap-up** Lecture 1 Openness of the educational system: some types of problems 1. **Descriptive problems** a. Trends over time b. Difference between countries, cities/ rural areas 2. **Explanatory problems** What are the effects of early tracking on educational opportunities for different groups? 3. **Problems of institutional design** How to reduce or mitigate unequal opportunities in educations and school segregation? 4. **Normative problems** What kind of inequality in educational outcomes is acceptable? How much can you restrict freedom of choice of parents to prevent inequalities? Contents of the course ====================== **Introduction to modern theoretical sociology** 1. "Blueprints" of sociological theory 2. The toolbox of theoretical sociology: model building 3. Applications of the toolbox:\ Classical and modern example models Blueprints for sociological theory ================================== Types of explanations - Covering-law explanations - Logical schemes - Coleman's diagram, **Covering law explanation :** the phenomenon is explained by a casual law - *If a fire starts in a crowded room, people will panic leading to many fatalities* **Statistical explanations**: the phenomenon is explained by statistical relationship: - *The smaller the door in a room the more fatalities when a fire starts* - **Mechanism explanations**: the phenomenon is explained by a means of an action-based explanation incorporating social restrictions/ opportunities and the aggregation process What does the lower arrow stand for in Coleman's diagram? : Behavioral Theory Which statement(s) is (are) wrong about the Coleman's diagram? (multiple answers possible 1. **It is a thinking framework for sociological theories** 2. **It clearly distinguishes between micro and macro level** 3. **All arrows represent causal relations** 4. **Bridge assumptions link the micro-level to the macro-level** 5. **Transformation rules link the micro-level to the macro-level** Afbeelding met tekst, schermopname, Lettertype, scherm Automatisch gegenereerde beschrijving  Macro vs Micro Level ==================== **In more abstract theory formation:** - Macro can refer to any group etc. that brings together lower-level units, e.g., - micro = individual, macro = group of two, three, or more people - micro = organization, macro = industry - In more applied contexts, these labels have often a more substantive interpretation - Macro: countries, societies; micro: individuals; and in between meso: organizations, groups etc. Importants definitions ====================== - **Theory**: a set of verbal or formal assumptions and propositions on the question at hand including causal relations that lead to testable hypotheses about the question - **Model**: formalized version of a theory in which all conditions, actors, possible actions, preferences, implications of behavior for everyone and relations between these elements are made very precise such that logical implications and hypotheses can be derived - **Syllogism**: building block of a theory or model - **Conceptual model**: an overview of concepts and arrows that summarize the hypotheses following from a theory or model Five important parts of theory building ======================================= 1 Be Precise ------------ Make things less ambiguous - Definitions, but also language in general - Formalization helps! 2 Avoid black boxes ------------------- - Thinking about plausible mechanisms often opens black boxes. 1. The stronger the cohesion among a given social group, the lower the suicide rate in this group 2. Cohesion among catholics is stronger than among protestants 3. The suicide rate is lower among catholics than among protestants 3 include micro level --------------------- - Mechanism/process at micro-level not considered - Does cohesion affect individual suicide (and how)? - Moreover, individual level is not considered (third important aspect of theory building). 4 Not neglecting the transformation rule ---------------------------------------- - Coleman criticizes sociologists (classical, including Durkheim) for disregarding micro, particularly *transformation rule*  5. Behavioral theory: as simple as possible, as complex as necessary -------------------------------------------------------------------- Five important aspects of theory formation in analytical sociology 1. Making things precise 2. Search for plausible mechanisms 3. Take the micro-level into account 4. Carefully model the micro-macro transformation 5. Search for a micro-model that is "realistic" enough, and not too complex WG 2 ==== ### Coleman boats Enforcement/ Training, education teachers Pros: Taxes Grade repetition remedial teachings? Human capital theory Signaling Theory Signaling Theory: people with 5,4 will not continue as to people with 5.5 do Inequality leads to lower gdp: Taxes **V. Love at first swipe** The advent of online dating has affected how people meet future spouses. Today, online dating through apps such as Tinder and Bumble are the most frequent way couples meet. Some researchers argue that dating apps may reduce divorce rates. 1. Dating apps provide a large pool of potential suitors that compete against each other, and a potential match can be found within a few clicks. 2. People are more likely to be better matched initially if they can choose from more potential partners and better matches lead to longer cohabitation spells. 3. The social exchange theory posits that social behavior is the result of an exchange process in which people weigh the potential benefits and risks of their social relationships. When the risks outweigh the rewards, they will terminate or abandon the relationship. Sketch the Coleman boat and find a transformation rule. *Grote poule aan mensen Social bonds are stronger* *Social exchange* *Theory* Lecture 2 Cooperation in social dilemma Durkheim's mechanical vs organic solidarity - *Mechanical:* social integration of members of a society who have common values and beliefs - *Organic:* social integration that arises out of the need of individuals for one another's services Game Theory =========== Example of a trust problem: buying a used Car Three possible outcomes: 1\. No trust = no transaction 2\. Trust is placed and honored = a peach is sold 3\. Trust is placed and abused = a lemon is sold Preferences of trustor: 2 -- 1 -- 3 - To consider buying a used car at all makes sense only if outcome "peach" is best. - Given "reasonable" public transportation, having no car is preferable to a lemon. Preferences of trustee: 3 -- 2 -- 1 - In principle (and due to the "information asymmetry" between buyer and dealer), the dealer can sell a lemon for the price of a peach and thus increase his (short-term!) profits. - Selling no car at all is worse than selling a peach Lecture 3 ### Game theory and cooperation Coopeeration in repeated encounters - In repeated games, many possible strategies, because behavior can be conditional on others behaviors - Axelrods classification of strategies - Nice , conditional, retaliatory/provocable, forgiving Idea of Axelrod (if analytical results are too hard too obtain) - Many simulated players play against each other for many rounds - Against some you do better, against others worse: Who does best in whole population? - Evolution: if good strategies get more kids than bad strategies: who survives? - Tit for tat does well in competition with many strategies The repeated prisoners dilemma T\>R\>S\>P ALL D, ALL D : equilibrium Tit for tat can be equilibrium IF: *10R\> T+ 9P* *10R-10P \>T -- P* *10 (R - P)\> T - P* *10\> (T -- P )/(R -- P)* Punishment threats to stabilize cooperation - Repeated interactions - Punishment opportunity in the public goods game - Reputational damage in networks Seminar 2 Game theory - What is a game? - Description of a social situation consisting of actors, actions and outcomes - When is a game theory useful? - Put otherwise, what kind of social situations - Social interactions with strategic interdependence - What is a strategy - Decision rule specifying a complete plan of actions for each decision npde af a game Nash equilibrem - **Pareto-optimal** = No one can improve their situation without making the other worse off. - **Pareto-suboptimal** = At least one person can be made better off without hurting the other player. +-----------------+-----------------+-----------------+-----------------+ | - **game 7** | player 2 | | | +-----------------+-----------------+-----------------+-----------------+ | | P | H | | +-----------------+-----------------+-----------------+-----------------+ | player 1 | P | -1, 1 | 1 -1 | +-----------------+-----------------+-----------------+-----------------+ | | H | 1, -1 | -1, 1 | +-----------------+-----------------+-----------------+-----------------+ NE: No nash equilibrium DS: none ------------ ---------- ------ ------ **game 1** player 2 C D player 1 C 3, 3 1, 5 D 5, 1 2, 2 ------------ ---------- ------ ------ NE = DD is a strategy played by both player DS: Both D ------------ ---------- ------ ------ **game 4** player 2 S T player 1 S 2, 1 0, 0 T 0, 0 1, 2 ------------ ---------- ------ ------ Coordination Game A: NE = SS & TT DS = none ------------ ---------- ----------- ----------- ----------- --------- **game 8** player 3 C D player 2 player 2 C D C D player 1 C 6, 6, 6\* 4, 7, 4\* 4, 4, 7\* 2, 5, 5 D 7, 4, 4\* 5, 5, 2 5, 2, 5 3, 3, 3 ------------ ---------- ----------- ----------- ----------- --------- Dominant strategy 1: D Dominant strategy 2: D Dominant strategy 3: D Nash equilibrium DDD \*\* pareto optimal ------------ ---------- ---------- -------------- --------- ------------ **game 9** player 3 A NA player 2 player 2 A NA A NA player 1 A 2, 2, 2 2, 4, 2 2, 2, 4 2, 4, 4 NA 4, 2, 2 4, 4, 2 \*\* 4, 2, 4 -4, -4, -4 ------------ ---------- ---------- -------------- --------- ------------ Nash Equilibrium (NA,NA,A) (NA,A,NA) (A,NA,NA) SEMINAR 4 ========= 4\. Consider the following situation. Two people play a game where they each hold a coin, and simultaneously choose to show heads (H) or tails (T). Player 1 wins twenty euro if the coins do not match, but loses 20 euros to player 2 if the coins match. First, write down this situation in a game table. Then, give the Nash equilibrium(s) and Pareto optimal outcome(s) of the game. ------------- ---------- --------- ---------- -- -- -- -- **game 4)** player 2 H (p) T (1-p) player 1 H (q) -20, 20 20, - 20 T (1-q) 20, -20 -20 , 20 ------------- ---------- --------- ---------- -- -- -- -- Player 1 U (heads) = -20p+ 20(1-p) = -20p +20 -- 20p = -40p + 20 U (tails) = 20p + (-20(1-p)) = 20p -- 20 + 20p = 40p- 20 -40p + 20 = 40p -20 -40p -40p = -40 -80p = -40 P = 1/2 U have to make indifferent: make the chances equal, otherwise player will counter you. Player 2 U (heads) = 20q - 20(1-q) = 20q +-20 + 20q= 40q - 20 U (tails) = -20q + 20(1-q) = -20q +20 - 20q = -40q-+20 -40q + 20 = 40q-20 -40q -40q= -40 -80q = -40 P = 1/2 All options are pareto optimal VB: If p is equel to 0.8 and q is equal to 0.3 is is a nash equilibrium Fill it in here 40q - 20 = -40q+20 -40p + 20 = 40p -20 They have to be equal to each other Repeated prisoners dilemma: ALLD: Play D in each round - unconditional - not noce TF2T: play C in round 1 and 2 In the ;ater rounds (3,4,t..) - Play D if the other player played D in the previous rounds (1,2....) Repeated PD =========== ----------------- ---------- --------------- --------------- -- -- -- -- **REPEATED PD** player 2 TF2T ALL D player 1 TF2T 10R, 10R 2S+8P, 2T+ 8P ALL D 2S+8P, 2T+ 8P 10P, 10P ----------------- ---------- --------------- --------------- -- -- -- -- **Step 2:** Determine the payoffs that correspond to the different strategy combinations (See Steps 2.1-2.3 on the next pages to fill in the payoff matrix below); **Step 3:** **3.1.** Determine the best replies in the scheme below; **3.2.** Argue what the Nash-equilibria are (as mentioned: (TF2T, TF2T) is sometimes a Nash-equilibrium, show under which conditions that is the case). S\>R\> T T\> R\> P\> S Seminar 4/5 P1 P2 cooperate Cooperate + punish Freeride -------------------- ----------- -------------------- ---------- cooperate 36,36 36,36 18, 38 Cooperate + Punish 36, 36 36,36 13, 28 freeride 38, 18 28, 13 20,20 Nash equilibrium Punishment is less effective - Defection will be more attractive Multiplier is larger - Cooperation becomes more attractive LECTURE 5 PART 2: Schelling's segregation model Model building in sociology =========================== **Model:** Explanation with simplified assumptions **Where to simplify (Coleman)?** 1\. Assumptions with respect to individuals simplified as much as possible 2\. Assumptions with respect to social conditions simplified as little as possible **Goal:** derive implications regarding effects of social conditions on collective effects **Decreasing abstraction (Lindenberg**): -\> stepwise replacement of simplifications by more realistic assumptions Reality is complex: in order to understand -\> simplify Models in the course so far --------------------------- - Non-repeated games (PD, trust game, Public Goods, etc.) as models of cooperation - Repeated games as models of cooperation in socially embedded settings - Decreasing abstraction (adding complexity) at the macrolevel - Micro-level models of human decision making - Decreasing abstraction (adding complexity) at the microlevel #### Woodclap What is the explanatory problem with residential segregation? - Why does segregation emerge so easily, regardless of preferences? Agent based modeling ==================== - **Explanandum: complex social phenomena** - **Key elements:** 1. *Bottom-up* approach (methodological individualism) 2. Agents (actors) following simple behavioral rules 3. Social interaction between agents - Goal: study how social phenomena emerge from the interactions of simple agents - Typical method: computer simulation Simulation as an "automated thought experiment" 4. Build a (simplified) model of a social process 5. Translate the model into computer code 6. Let the model run on the computer 7. Make changes to the conditions ("what if...") and see what happens = deriving hypotheses from the model Schelling ========= What do we need: 1. A town, with spaces to live 2. Inhibitants of the town a. Different types b. Preferences where to love and c. Options where to move 3. Specifications of the moving process: Preferences (1) --------------- - Actors want to avoid being in a minority situation in their neighborhood → [unhappy if more than half of the actors in the neighborhood have the opposite color, happy otherwise ] - Note 1: we do not entirely assume "racism" away; actors still care about types! - Note 2: actors are still relatively tolerant; they are fine with living in mixed neighborhoods Preferences (2): Definition of the neighborhood of a household -------------------------------------------------------------- - The neighborhood of a household consists of the household itself, the four households to the left, and the four households to the right. - Households near the end of the line have fewer neighbors. (alternative: circle) Afbeelding met tekst, schermopname, Lettertype, maatstok Automatisch gegenereerde beschrijving Moving (1) ---------- - At the beginning of each round, identify "unhappy" actors - Order of moving: unhappy actors move in turn, counting from left to right. - An unhappy actor moves to the nearest point between two other actors to the right or to the left where it is not a minority. Moving (2) ---------- - If the neighborhood of a marked piece has changed such that not more than half of the neighbors have the opposite color when the turn to move comes, the marked piece does not move after all. → Interdependence between households. - End of a round: each piece that has been marked at the beginning of the round has had a turn to move. - End of the game: each piece has a neighborhood with at least half of the neighbors of the same color. - (Or: there are no suitable places to move to) Moving (3): underlying assumptions ---------------------------------- - Moves are costly; costs increase in distance - No scarcity on the housing market - No legal or economic constraints to choosing certain neighborhoods - No discrimination on the housing market **Schelling's model versus the "common sense theory"** - The model shows that high levels of segregation can result from rather tolerant individual preferences - High levels of segregation can result from both tolerant and intolerant preferences - Schelling makes the micro-macro link explicit - The model allows for derivation of new implications  **Some of the higher level lessons from schellings model:** - Complex social phenomena can result from simple individual behavior - [Collective outcomes are not a one-to-one reflection of individual preferences ] - (Computer) simulation can be a useful tool to explore this micro-macro link **some common misconceptions** - **"Schelling's model shows that segregation is not caused by things like structural inequality"** **→** Schelling's model does not rule out such alternative explanations of segregation - **"Schelling's model shows that (racial) segregation may emerge without racist preferences" →** Actors still base their decisions on ingroup/outgroup preferences Developing the model further: The method of decreasing abstraction in action ---------------------------------------------------------------------------- - Goal: keep the model as simple as possible, but as complex as necessary (Einstein) - If we replace simplifying assumptions by more realistic (= less abstract, more complex) ones, does that change the predictions of the model? - If **No**: this simplification is not problematic, we can keep it - If **Yes**: investigate further; are the new predictions better or worse, does the model still explain the phenomenon well? Do we need more complex to get better TESTING THE MODEL 1. How you define neighborhood preferences matters for the predictions of Schelling's model, under certain conditions (randomness) 2. Empirical preferences in segregated areas are very different from what Schelling assumed! 3. However, also for these preferences, the model can still explain high segregation 4. So, the modified (more realistic model) works well, but under certain conditions Summary of the discussion of Schelling's model as an example of agent-based modeling ------------------------------------------------------------------------------------ 8. Assumptions (simple version) 9. Analysis: even mild preferences lead to high segregation 10. A first comparison with data: Schelling vs. Common Sense → Schelling both theoretically and empirically stronger 11. Method of decreasing abstraction: - a\) A two dimensional town → doesn't matter) - b\) Alternative preference functions + empirical comparison → does matter, under certain conditions! Lecture 6 Which statements are true about granovetters threshold model? - It shows that the emergence of mass collection action depends mostly on the distribution of peoples opinion - It helps to understand why collective action is hard to predict Analysis result - The size of a riot depends more on the distribution of thresholds than on the mean - Small changes on the micro level can have big effects on the macro level - Collective behavior can differ dramatically between different distrubutions, even if the means of the distributions are the same - Difference between stable and unstable equilibria Some methodological conclusions 12. Small changes at the microlevel can lead to large at the macro 13. Complex macro level effects result from very simple assumptions at the micro level 14. Some analytical tool needed to understand the behavior of the model 15. Some social processes may be *fundamentally* hard to predict, even if we *know* the mechanism and the process is *deterministic* A policy maker trying to prevent social unrest should invest in: - A swift response to emerging collective action, instead of polls or other prediction A revolutionary should invest in - Widening the distribution of sentiments Lecture 7 Conjoint norms: sanctioning is done endogenously Direct and indirect sanctioning +-----------------------------------+-----------------------------------+ | Indirect | DIRECT | +===================================+===================================+ | - Punish deviation by defection | - Punish deviation by some | | | action external to the | | - Sanctioning action part of | original dilemma | | the original dilemma | | | | - Does not require repeated | | - In repeated interaction | interaction | +-----------------------------------+-----------------------------------+ Indirect sanctioning in the RPD - Conditional cooperation TFT/ Trigger Some implications norm change In general: prople may conform to social norms if they expect to be sanctioned Expectation of sanctions may depend on - Perceived conformity - Perceived costs of sanctioning - Availability of information on sanctioning - Shadow of the future - Density of networks Thus, conformity to norms may increase/decrease depending on social conditions, even if individual normative preferences do not change Sanctioning as a volunteers dilemma (diekmann, 1985) ------- ------------------------------ ----------------------- ----- **c** All others At least one other sanctions No one else sanctions you sanction 0 0 Not sanction 1 -10 ------- ------------------------------ ----------------------- ----- Hypothesis: diffusion of responsibility: the larger the group, smaller probability But if people differ with regard to their interest in the norm: - The person with the strongest interest ( benefits -- cost) in the norm will sanction Utility matrix  **Shadow of the Future** The "shadow of the future" refers to the idea that in repeated games (where players will interact multiple times), the expectation of future encounters influences current behavior. If players know they'll interact again, they might be more likely to cooperate today, hoping to build trust and ensure cooperation in future rounds. **Example** Imagine a repeated game scenario like the Prisoner's Dilemma where two individuals decide daily whether to cooperate or defect. Knowing they'll play again tomorrow (and in the future) creates an incentive to cooperate now. If they defect today, it might lead the other player to defect in future rounds, resulting in lower payoffs long-term. This \"shadow of the future\" encourages cooperation because players want to avoid retaliation in future encounters. the discount parameter with an example? Thus, with a high δ, cooperation becomes more attractive because the cumulative value of future cooperation exceeds the one-time gain from defecting VIDEOS Week 1 ====== Understanding sociological questions Sociological questions are questions about relations between macro conditions to macro outcomes - Why does panic occur in a crowded room if the fire alarm goes off? - Why do revolutions mostly occur when economic conditions are improving? Explanations should go beyond common sense claims directly on the macro relation; they should make understandable how macro conditions affect individual behavior and how behavioral reactions of many lead back to the macro outcome A thinking framework for sociological problems - To ensure that arguments on bridge assumptions, transformation rules, and the individual level theory are made explicit - Much more than a causal scheme: arguments on bridge assumptions and transformation rules can imply feedback loops and other more complex causal relationships, because eg individual behavior of one person can change macro conditions for another persom - To explicate some of these more complex mechanisms sometimes "coleman"fleets are used to make that more explicit Week 2 ====== Advanced sociological theory: modelling social interaction. Game theory: Key concepts Theory **of goal- directed** behavior in **interdependent** decision situations **Interdependent**: The outcomes for one actor depend not only on his/ her own actions, but also on the actions of another actor Game theory studies strategic behavior Well suited for the analysis of social interaction Examples of interdependent situations: - Changing lanes in traffic jams - Joining a demonstration - Bidding in negotiations - ( games like Rock, paper , scissor) An intuitive characterization of "Goal- directed" action Actors have: - Alternative actions - Goals, i.e., evaluations of the possible outcomes of actions - Expectations ( or information) on the states of the world ( for example, expectations on certain "contingencies" or on behavior of other actors Actors choose: - The action that seems most appropriate, given their expectations, to realize their goals What is a game? - Actors - Moves/ behavior - Outcomes depending on moves of all actors - Rules of the game - Strategies ( behavioral plans) Some implications - A player chooses a best reply strategy, given hid to her anticipation of the strategy, given his or her anticipation of the strategy chosen by the other player. - If a player has a dominant strategy , they will use this strategy - The chosen strategies will be a nash equilibrium ANALYZING TWO BY TWO GAMES 0, 0 10, 2 ------- ------- 2, 10 4, 4 Yellow: NASH Pareto: 4,4, 2,10, 10, 2 Prisoners dilemma ---------- ---------- -------- -------- **c** player 2 C D player 1 C R,R\* S, T\* D T, S\* P, P ---------- ---------- -------- -------- Assumptions: - T\>R\>P\>S - Simultaneous moves - No binding agreements - Information: each player is informed on their own alternative actions and outcomes, as well as alternative actions or outcomes for the partner WEEK 3 ====== Repeated prisoners dilemma -------------------------- Extension: cooperation in repeated interactions - Original explanatory problem: Why do people cooperate in social dilemmas? (problem of order) - Theory: ( non -- repeated) prisoners's dilemmas prediction: no cooperation - Observation: people do sometimes cooperate - Revision of the theory: repeated interaction - (Other possible visions: networks social preferences) Plan for the analysis 16. Specify the game 17. Figure out the payoffs of the players 18. Figure out the equilibrium The repeated prisoners dilemma is played 10 times. After each round, each player is informed on the others players behavior. Strategies Players can only play two games - ALL D - TIT FOR TAT ( start with cooperation and play whatever the other played in the round before) ----------------- ---------- ------------- ------------- -- -- -- -- **REPEATED PD** player 2 TFT ALL D player 1 TFT 10R, 10R S+9P, T+ 9P ALL D T+9P, S+ 9P 10P, 10P ----------------- ---------- ------------- ------------- -- -- -- -- All D is always equilibrium: 10P\> S+ 9P 10R\> T + 9P 10R -- 10P \> T -- P 10 ( R -- P ) \> T -- P 10 \>( T -- P) / ( R -- P ) - 10 is stability of relation - T- P is temptation do defect - R -- P is costs of conflict Cooperation is an equilibrium if 19. The players play enough games 20. The temaptaion to defect is smal enough 21. The costs of conflict are high enough - Why is the analysis flawed - All possible strategies in the repeated game should be considered - Cooperation is never a best reply in the last game Public goods game ----------------- - Game representing cooperation problem - More than 2 players - Investment ( level of cooperation) more continuous instead of yes/ no - Version with and without punishment option Analysis of the game - If eryone contributes everything, 32 for everyone - Every point invested by yourself, leads only to a 0.4 point return for yourself. Whatever the other do, every point invested cost you more than you get in return. - Zero investment id the dominant strategy - During experiments, subjects often invest in early rounds, but this becomes less throughout the game Analysis - Starting at the end of the game, punishment is costly and does not provide any benefits, so no one is expected to punish in the one shot game - Zero investment by everyone is NASH and no one punishing after that is till NASH and the only nash that takes into account that when arriving at the punishment stage, not punishing is what a rational selfish player would do. - Still there are other nash that do neglect the order of moves - With punishment cooperation is sustained in the repeated game Week 4 ====== Including non-standard preferences in two by two games Utility - Utility of a certain outcome for a player is the total value a player assigns to a certain outcome - The payoff is the number of points )money) a player earns in a certain outcome - If players only care about their own payoff, utility and payoff is basically the same' - If players also care in some way about the other players payoff, Utility id different and to analyze a game the effective utilities need to be calculated before the game can be analyzed Altruistic preferences - Utility of a player does not only involve his or her own payoff, but the payoff of the other player ass to the playoffs - Lets assume players value the point of the other player half as much as their own points - If a player earns 10 and other 6 utility is 10+ 6x0,5 = 13 - Recalculate the whole payoff matric into the effective utility matrix - Hidden assumptions: players know each other's preferences The players value the others payoff half as much as their own payoff ---------- ---------- ---------- ---------- **game** player 2 C D player 1 C 6+3 ,6+3 0+4, 8+0 D 8+0, 0+4 4+2,4+2 ---------- ---------- ---------- ---------- ---------- ---------- ------ ------ **game** player 2 C D player 1 C 9, 9 4, 8 D 8, 4 6, 6 ---------- ---------- ------ ------ NASH: yellow Inequality aversion ( Fehr -- Schmidt preferences) - Substantively: Utility = payoff -- envy - guilt - Utility is 22. OWN outcome MINUS 23. A -- amount the focal player earns less than the other, 24. B. Amount the focal player earns more than the other - Example 25. Outcome is (10,6) 26. A = 0.5 B=0.25 27. Utilities are ( 10 -- 0 -- 0,25\*4, 6 -- 0,5\*4 -- 0 ) = 9,4 Week 5 ====== Residential segregation Examples - Race - Religion - Wealth Types of problems' 1. Descriptive problems - Trends over time - Cross section: comparing different regions 2. Explanatory problems - Ehy does residential segregation remain stable over time or change? - Why are there differences in residential segregation between regions ( or not) 3. Problems of institutional design - How to reduce or mitigare residential segregation 4. Normative problems - How much residential segregation is normatively acceptable? - What are acceptable costs of reducing or mitigating residential segregation? Trends of residential segregation over time: Sketch of an explanation Common sense- theory: Percentages of whites approving residential integration Segregation have remained stable Agent based modelling Complex social phenomena 1. Bottom up approach ( methodological individualism) 2. Agents ( actors) following simple behavioral rules 3. Social interaction between agents Goal: study how social phenomena emerge from interactions of simple agents Week 6 ====== The problem of explaining mass collective action ------------------------------------------------ Examples of puzzling macro pehnomena - Mass collective action ( demonstrations, riots, revolutions etc) - (non) -- diffusion of beliefs, innovations - Fads, hypes - Economic "bubbles" Problems when explaining social change - Social changes can happen suddenly, in an unpredictable way - Explanation by simple aggregation of individual preferences (" change of norms") often not sufficient - Social changes are often u*nintended* consequences of individual behavior The complexity of the transformational mechanism - Understanding how individual actions aggregate and lead to certain collective outcomes is the most challenging task for sociologists - For this purpose formal analytical tools are often necessary Explanation by building a model Granovetters threshold model of collective behavior Main question: - How are individual preferences aggregated to collective behavior in the absence of institutionalized norms and values. - Individual level: when does someone start to protest? - Examples of explananda: riots demonstrations, diffusion of innovations Common sense theory - The stronger people preferences ( smaller costs, larger benefits) to protest the more people will protest - Granovetters point: this is at best a very partial explanation Granovetters threshold model of collective behavior --------------------------------------------------- Assumptions of the Threshold model 1. There are N actors 2. Each actors has a binary choice : join/ Not join 3. Costs and benefit of joining depend on how many others have joined 4.... But this differs between actors: each actor has their own threshold 5. The process takes place in "rounds": in each round, all actors observe how many others have joined and then update their own choice 6. Actors are goal directed: given their goals and preferences, and their perception of their situation, they act as to maximize their utility Probability of gains *increases* and risks to get in trouble *decreases* when more people protest - IF more people protest more people are willing to protest Threshold of an actor = the proportion (%) of the population who have to join before the actor will also join - T = 0% -\> Actor will *always* join - T = 50% -\> Actor will join if and only if 50% or more other actors have joined - T = 100% -\> Actor will *never* join 28. Small changes at the microlevel can have big consequences at the macrolevel 29. Consequences do not depend so much on the mean, nut more on the distribution of the threshold 30. so can we say more on how the outcomes depend on this distribution? Mathematics CDF F(x) F(20) = 0.3 -\> 30% of observations has x=20 or less Goal of the analysis - find out how many actors will join the protest, given a certain distribution of threshold - Maximum size of the protest is reached when no more actors join , and nobody leaves -\> equilibrium Finding equilibrium Afbeelding met tekst, schermopname, Lettertype Automatisch gegenereerde beschrijvingequilibrium: *r = F(r)* In summary : - The size of a riot depends more on the distribution of thresholds than on the mean - Collective behavior can differ dramatically *between* different distributions, **even if the means of the distribution are the same** - Small changes at the microlevel can have big effects on the macrolevel Week 7 ====== Social norms - Rules that permit, prohibit, or prescribe behavior for a particular population - Some degree of consensus among this population - Negative consequences in case of violation - Not enforced by the state Norms in sociological theory - Central in much classical sociology - In structural functionalis, norms are crucial to keep society functioning as a system - "economics is all about how people make choices; sociology is all about how they don't have any choices to make" ( Duesenberry 1960) - "Soft" explanations competing with more "rigorous" approaches Why norms as explananda - Explanations based on norms are at best incomplete: where do norms come from? Why do they change? - Norms are themselves social phenomena, and thus legitimate targets of sociological investigation - Our claim: an analytical approach is fruitful Given our definition, focus on - Content -\> Voss: "Demand" for norm - Distribution - maintenance -\> Voss: "Effective enforcement" of norm  Disjoint vs conjoint social norms Disjoint: beneficiaries and targets of a norm are different groups Conjoint : beneficiaries and targets of a norm perfectly overlap NB : mixed cases are also possible Explaining norms in two steps 1. Explain the demand for a norm : under what conditions would people want a norm? What would be the content of the norm? 2. Explain the realization of the norm : under what conditions is the norm actually enforced Explaining norms - Under what conditions would actors benefit from an enforced rule? - Generally ( coleman) : if actions have externalities - More specifically ( Voss): to escape social dilemma situations Social dilemmas: - Generally: situations in which individually rational behavior leads to a collectively suboptimal situation - Game- theoretical definition: game with at least one Pareto- suboptimal equilibrium Are norms always socially beneficent? - Honor codes, revenge norms - Anti achievement norms - Gender roles - Segregation norms ( apartheid) - Etc Conventions vs social norms Conventions - Solutions to variations of coordination games - Are self-reinforcing -\> NO additional sanctions needed! - Content is often more or less arbitrary ( driving left or right) - But some conventions are more efficient than others Social norms - Demand arises if there is no pareto- optimal equilibrium - Not self-reinforcing: additional mechanisms not needed READINGS\\ Watts (2011): a sociologist\'s apology (6pp) You will believe every outcome you will here and have a reasonable explanation for it. Everyone think they have a pretty good idea of wg: why someone is poor, or designing things. It is "common sense" Preface of everything is obvious, Once you know the answer Week 3 ------ *The Evolution of Cooperation* by Robert Axelrod: **Chapter 1: The Problem of Cooperation** Axelrod begins with a central question: How does cooperation emerge in a world of self-interested individuals without central authority? He explores historical and modern examples, like the U.S. and the Soviet Union during the Cold War, where each side faced incentives to both cooperate and compete. This dilemma is modeled through the *Prisoner\'s Dilemma* game, where mutual cooperation benefits both players, but defection offers a tempting payoff if the other player cooperates. However, mutual defection leads to worse outcomes for both. Axelrod highlights how this dilemma recurs in various contexts---from international relations to everyday interactions like business deals and social favors. In the absence of centralized control, cooperation can be difficult but not impossible. Axelrod\'s goal is to determine what conditions foster cooperation among self-interested parties. **Chapter 2: The Success of TIT FOR TAT in Computer Tournaments** Axelrod organizes a computer tournament where participants submit strategies for playing the iterated Prisoner\'s Dilemma game, where players interact multiple times, allowing for patterns of cooperation or retaliation to develop. Surprisingly, the simplest strategy, *TIT FOR TAT*, wins the tournament. TIT FOR TAT starts by cooperating and then mimics the other player's previous move. Axelrod finds that this strategy's success stems from four key properties: it is ***nice*** (never the first to defect), ***provokable*** (retaliates after a defection), ***forgiving*** (returns to cooperation if the opponent does), and ***clear*** (easy for the opponent to understand). The findings suggest that cooperation based on reciprocity can emerge even among self-interested players. **Chapter 3: The Chronology of Cooperation** Axelrod further explores the conditions that allow cooperation to develop and persist. He describes three stages for the evolution of cooperation: 1. **Initial foothold**: Cooperation can begin if clusters of individuals use reciprocal strategies, even in non-cooperative environments. 2. **Spread**: Once established, reciprocal strategies can thrive among various strategies. 3. **Stability**: When cooperation is well-established, it can defend against less cooperative strategies. Through these stages, Axelrod argues that cooperation can evolve naturally, even without foresight or enforced agreements, if individuals have enough interactions and adopt reciprocal strategies. These chapters lay the groundwork for Axelrod\'s examination of cooperation, showing that mutual cooperation in the absence of central authority can be understood and fostered under certain strategic conditions. Altruistic punishment in humans Ernst Fehr, Simon Gachter --------------------------------------------------------- **Altruistic Punishment**: Participants frequently punished low contributors, even though it was costly and offered no direct personal gain. This punishment was mainly performed by those who contributed above the average and was typically directed at defectors (free-riders) **Sustained Cooperation**: The possibility of punishment led to higher contributions over time. When participants could punish defectors, cooperation levels increased, whereas in the no-punishment condition, cooperation quickly deteriorated. **Emotional Drivers**: The study suggested that negative emotions, such as anger towards defectors, motivated altruistic punishment. These emotions increased with the extent to which an individual's contribution fell short of the average group investment. **No Reciprocity or Reputation Influence**: The experimental design prevented participants from forming reputations or expecting future encounters with the same group members, eliminating reciprocity as a motive. This reinforced that the willingness to punish free-riders was altruistic. \"Cooperative Networks: Altruism, Group Solidarity, Reciprocity, and Sanctioning in Ugandan Producer Organizations\" by Delia Baldassarri ----------------------------------------------------------------------------------------------------------------------------------------- **Network Position and Cooperation**: Farmers with central roles in their network, such as those who are more connected or bridge otherwise separate individuals, display higher cooperation levels. This centrality correlates with cooperative behaviors, like selling crops collectively and participating in organizational meetings. **Altruism vs. Reciprocity and Sanctioning**: Experiments using public goods and dictator games revealed that generalized altruism and group solidarity were less predictive of cooperation than mechanisms like reciprocity, which arise from repeated interaction and trust. The threat of sanctions also encouraged cooperation, although less effectively than reciprocity. **Experimental and Real-life Correlation**: The study found that prosocial behaviors observed in experimental settings translated to real-life cooperation within the farmer groups. Farmers who exhibited cooperative tendencies in the games were more likely to participate in organizational activities and adhere to group decisions. **Mechanisms Driving Cooperation**: The study concludes that reciprocity and sanctioning, supported by network connections and communication, are primary drivers of cooperation in these organizations. Unlike generalized altruism, these mechanisms encourage individuals to prioritize collective benefits over personal gains, essential for overcoming free-rider challenges. Elinor Ostrom's "Collective Action and the Evolution of Social Norms": ---------------------------------------------------------------------- **Olson's Zero Contribution Thesis**: - Olson argued that rational individuals wouldn't contribute to public goods without coercion or incentives. - This thesis is contradicted by real-world behaviors like voting and tax compliance. 1. **Empirical Evidence**: - Fieldwork shows people often self-organize to manage resources and provide mutual protection. - Government policies can sometimes hinder private provision of public goods. - Successful self-organized groups monitor and sanction to reduce free-riding. 2. **Experimental Findings**: - People contribute to public goods more than predicted by rational egoist models. - Contributions decay over time but don't drop to zero. - Communication and belief in others' cooperation boost contributions. - People punish free riders, even at a personal cost. 1. **Clearly Defined Boundaries** 2. **Rules Aligned with Local Needs** 3. **Collective Decision-Making Rights** 4. **Effective Monitoring** 5. **Graduated Sanctions** 6. **Mechanisms for Conflict Resolution** Ostrom's evolutionary theory suggests that humans have developed a propensity for social norms like reciprocity and trust through evolutionary processes. This aligns with Axelrod's findings, as "Tit for Tat" embodies these principles by rewarding cooperation and punishing defection, thus fostering a stable environment for collective action. Ostrom's design principles emphasize the importance of monitoring and graduated sanctions (Design Principles 4 and 5). Fehr and Gächter's findings provide empirical support for these principles, showing that the willingness to punish free riders is crucial for maintaining cooperation. Ostrom's principles of participatory rule-making (Design Principle 3) and conflict resolution mechanisms (Design Principle 6) are supported by Baldassarri's findings. These principles rely on strong social networks and communication to build trust and cooperation "The Handbook of Rational Choice Social Research ------------------------------------------------ **Key Concepts of Rational Choice Theory (RCT) in Social Research** **Introduction to RCT**\ RCT posits that individuals act rationally, choosing the best options to achieve consistent goals, shaping fields like economics, crime, and politics. **Rationality and Decision-Making**\ The rational choice model is based on personal preferences, beliefs, and constraints, assuming decisions are logically consistent (preferences are complete, transitive, and independent). **Extensions and Criticisms**\ Acknowledging cognitive limitations (bounded rationality) and social preferences (like fairness and altruism) refines RCT's predictive power. **Empirical Applications** - *Experimental Economics*: Lab tests affirm RCT's predictions, particularly in market behavior. - *Behavioral Economics*: Combines psychology with RCT to address non-rational behaviors. **Case Studies**\ Experiments in markets and public goods demonstrate how RCT explains social cooperation and equilibrium. **Advantages and Disadvantages of the Rational Choice Model** - **Advantages**: The rational choice model provides a clear framework for understanding decision-making processes by assuming that individuals act consistently to maximize their utility. - **Disadvantages**: It often oversimplifies human behavior by assuming perfect rationality and ignoring other factors like emotions, social influences, and bounded rationality. **Distinction Between Rational Choice Model and Selfishness Assumption** - **Rational Choice Model**: Assumes consistency in behavior, meaning individuals make decisions that are logically consistent with their preferences and constraints. - **Selfishness Assumption**: Assumes individuals act solely to maximize their own utility, without regard for others' welfare. Gächter emphasizes that these are separate assumptions and that rational behavior does not necessarily imply selfish behavior. From Factors to Actors: Computational Sociology and Agent-Based Modeling\" by Michael W. Macy and Robert Willer --------------------------------------------------------------------------------------------------------------- Key Points: 1. **ABM Basics**: Agent-based models focus on interactions between autonomous \"agents\" (individuals, households, firms), which react to each other based on simple rules. This method allows researchers to study complex social patterns, like social norms or collective behaviors, arising from these local interactions. 2. **Shift from \"Factors\" to \"Actors\"**: Sociology traditionally modeled society through variables representing population-level factors (e.g., institutions). ABM shifts to modeling interactions among individual actors, making it possible to study how group behaviors emerge. 3. **Applications of ABM**: ABMs help understand phenomena like social norms, cooperation, cultural diffusion, and segregation. For instance, Schelling's segregation model showed how individual preferences for neighbors of similar backgrounds could lead to macro-level segregation. 4. **Emergent Social Patterns**: ABMs demonstrate how individual behaviors can produce social structures that were not centrally coordinated. Examples include fads, market crashes, or collective actions. 5. **ABM Limitations and Recommendations**: The authors highlight that ABM needs to avoid overly complex, biologically-inspired models, emphasizing that ABM should be used for theoretical experiments rather than strict predictions. In essence, ABM in sociology provides a \"bottom-up\" approach to understanding social structures, making it useful for exploring complex dynamics without assuming top-down control, such as how social cooperation might evolve or how social norms become widespread Self study: Watts refers to the common notion about social science that "it's not rocket science". What is Watt's point here? Watts challenges the idea that social science is easy or intuitive. He argues that, unlike rocket science, people assume they understand social science because it deals with familiar aspects of daily life, like poverty or relationships. However, social science explores complex, interconnected systems with unpredictable patterns, making accurate explanations difficult and often counterintuitive. 2\. Watts argues that many "big questions" about society remain unsolved. What do you consider to be "big questions" for social science? Come up with at least two examples of your own. - *What are the root causes of persistent income inequality across generations?* This question digs into complex factors like family background, education access, and social mobility. - *How does technology impact mental health and social relationships?* Examining how digital interaction changes human behavior, social norms, and mental health can reveal broader social impacts. 3\. Please provide in your own words and with not more than 20 lines of text a sketch of the structure of scientific explanations. - **Identify the phenomenon** to explain, such as a societal trend or behavior. - **Formulate a hypothesis** that proposes a cause-and-effect relationship. - **Gather evidence** through observation, experimentation, or modeling. - **Analyze results** to validate or refine the hypothesis. - **Conclusion and integration** into broader theories. This helps explain how micro-level behaviors dividual actions) link to macro-level outcomes (societal tr 4\. Provide two additional examples of P-T-E sequences and two additional examples of explanations from earlier sociology courses. Schelling's segregation model demonstrates how individual preferences for neighbors similar to themselves (T) lead to patterns of segregation (E), even without overt discrimination. In studying crime rates, Durkheim's theory suggests that weak social bonds (T) contribute to higher suicide rates (E)​ 5\. Please provide in your own words and with not more than 20 lines of text a sketch of the structure of scientific explanations. 6\. Provide examples of collective and individual phenomena related to the main problems of sociology *Collective Phenomenon:* Social norms, which emerge from the repeated actions and expectations of individuals but impact entire communities or societies. *Individual Phenomenon:* Personal identity, shaped by social experiences and norms but distinctively personal and psychological in its formation​ 7\. What is the "problem of transformation\"? Sketch an example in connection with electoral systems. The problem of transformation involves how individual choices aggregate to produce collective outcomes. In electoral systems, individual votes combine to determine government representation. However, due to electoral rules, the aggregate outcome may not reflect a straightforward majority, illustrating how individual actions translate unpredictably to collective outcomes​ 8\. Provide a sketch of the idea of unintended social consequences of individual actions in your own words and on less than 15 lines of text. *I*ndividuals often act with specific intentions that lead to unforeseen collective outcomes. For example, people moving to areas where they feel culturally comfortable can unintentionally lead to high levels of neighborhood segregation, even if they did not aim to create segregated communities​ 9\. Show the connection between the explanatory scheme for collective effects and Coleman's scheme for the reconstruction of macro-micro, and micro-macro transitions. Where do you locate social conditions, bridge assumptions, transformation rules, hypotheses on individual behavior, individual effects, and collective effects in Coleman's scheme? 10\. Try to reconstruct some examples of sociological analyses that you came across in earlier sociology courses using Coleman's scheme. 11\. Try to reconstruct the bank run described by Merton in Coleman's scheme. In a bank run, individual customers (micro-level) start withdrawing funds due to fear of insolvency. Each withdrawal heightens collective insecurity, causing even more people to act similarly until the bank collapses (macro-level outcome). 12\. Try to reconstruct the discrimination of Jews as described in Merton in Coleman's scheme. Individual prejudices and behaviors (micro-level) can, through collective reinforcement and societal structures, manifest as widespread societal discrimination against Jews. Social norms, laws, or policies that institutionalize this prejudice demonstrate how individual biases can escalate into systemic discrimination Vincent is asked whether he would like to have coffee or tea and he indicates that he prefers to have coffee. Then, the person asking indicates that Vincent could also have water, and then Vincent says that he then prefers to have tea. Which choice principle is violated here (explain)? The choice principle Vincent violates is *consistency*, introducing a new option (water) should not influence the preference between the original options (coffee and tea) if preferences are stable. By switching to tea only after water is introduced, Vincent's choice shows that his preference is not independent of irrelevant alternatives 2\. Which behavior in which game illustrated in Gächter's paper shows clearly a violation of selfishness (explain)? Gächter's experiments on altruistic punishment reveal **behaviors** that clearly violate pure selfishness. Participants often **punished low** contributors (free riders) **at a personal cost**, with no direct benefit or expected reciprocity. This action demonstrates **altruism** driven by negative emotions toward defectors rather than self-interest 3.Why is proposing more than the minimal amount in the ultimatum game not per se an indication of altruism (explain)? In the ultimatum game, proposing **more than the minimum** amount may not necessarily indicate altruism. Instead, it could reflect *strategic fairness*, where **proposers offer a fair split to avoid rejection** and secure their own payout. This response, while appearing altruistic, often serves to ensure the offer is accepted, rather than being motivated by concern for the other party's welfare 4\. (Advanced) what it the minimal value of α and β (as defined in the Fehr-Schmidt model) in the PD game with payoffs (S, P, R, T) = (0, 1, 4, 10) such that defection is no longer a dominant strategy in the game? In the Fehr-Schmidt model, inequality aversion parameters α (envy) and β (guilt) reduce the attractiveness of defection. For cooperation to be a viable strategy in the Prisoner's Dilemma game with payoffs (S,P,R,T)=(0,1,4,10), α must be high enough to make the *temptation to defect* (T) less appealing, and β must increase the discomfort from any resulting inequality. The minimum value of α\\alphaα (envy parameter) required for defection to no longer be a dominant strategy is α=3/5​ or 0.6. This result implies that if Vincent's aversion to earning less than others (envy) is strong enough, specifically at least 0.6, he would no longer choose defection as the dominant strategy in this Prisoner's Dilemma game. 5\. What are the six design principles of Ostrom? - Clearly defined boundaries - Rules aligned with local needs - Collective decision-making rights - Effective monitoring - Graduated sanctions - Mechanisms for conflict resolution 6\. Which design principle(s) can be linked to the success of Tit-for-Tat in Axelrod and how? Tit-for-Tat in Axelrod's studies aligns well with Ostrom's principles, particularly *monitoring* and *graduated sanctions*. Tit-for-Tat's simplicity and reciprocity foster cooperation by monitoring previous behavior and retaliating when necessary, which reflects a form of graduated sanctioning 7\. Which degree principle(s) can be linked to the success of altruistic punishment as in Fehr and Gächter? Altruistic punishment aligns with Ostrom's principle of *graduated sanctions*, where defectors face increasing consequences. In Gächter's study, participants' willingness to punish free-riders to maintain group cooperation exemplifies this principle by promoting accountability and discouraging selfish behavior​ Diekmann: Coordination norms (conventions)/: self-enforcing and emerge when people benefit from aligning their behavior with others. Compliance occurs because it provides metal benefit, or efficiency in achieving goals without necessarily requiring additional enforcement. For example, driving on the right side of the road