Public Good Game
Each subject secretly chooses how much of their initial endowment to put into a public pot. The joint value in this pot is multiplied by a factor ([latex] 1 < factor < N [/latex]) and evenly paid out across all [latex]N[/latex] subjects. All unspent endowments is kept by the respective subject. In one-shot games a non-cooperative strategy is usually applied. In infinitely repeated games, subjects eventually cooperate.
However, experimental results are not in line with predictions of rational choice theory. Even in one-shot, two person prisoners’ dilemma games half the participants cooperate. Voluntary contributions to public goods fall if the game repeats and therefore participants have been called “adaptive egoists” (Mueller, 2003) or “conditionally cooperative”(Gächter, 2006).
People seem to be contributing for “warm glow” preference (utility from contributing) (Palfrey & Prisbrey, 1997), altruistic preferences (want to increase other’s utility) or due to error and learning (testing the best approach).
Gächter found that voluntary cooperation is fragile, that there exists social interaction effects in voluntary cooperation, group composition matters (likemindedness), and that management of believes matters. Also, path dependency has been observed where the first round is the most important.
In other field experiments, students were informed how many others contributed to social funds (64% versus 46%) and were influenced by the numbers (Frey & Meier, 2004).
The cooperative environment (Ostrom, 1998)
Models of complete rationality do not work well in non-competitive situations. Verbal communication can improve trust and allow groups to reciprocate. There is also some capacity to solve second-order social dilemma that change the structure of the first-order dilemma (e.g. institute punishment). Ostrom proposed a model of bounded rationality where individuals use heuristics, norms and rules to improve outcomes of non-competitive, not frequently repeated situations.
According to Ostrom, reciprocity, reputation, and trust can overcome strong temptations of short-run self-interest. Consequently, a self-reinforceing process can be created that increases the level of cooperation and leads to higher net benefits for all, but it is very fragile.
The system is stable if it has a small size, symmetry of assets and resources, long time horizon, and a low-cost production function. As the group size increases, marginal gains from contributing falls and it is more difficult to identify and punish defectors. As the stakes increase, cooperation is reduced. Other issues that may arise are: monitoring becomes more costly, economies of scale may not apply and marginalisation behaviour within the group may be questions (is it fair punishment?).
Current Research (Lanz, Wurlod, Panzone, & Swanson, 2017)
A field experiment in supermarkets in greater London area to compare quantitative impact of three measures to reduce footprint of consumption (welfare analysis). The experiment tested conditions on four types of goods: soda, milk, spreads, and meat. Three conditions were used:
- Information label
- Pigouvian tax based on relatively higher footprint of “dirty-type” product alternative
- Neutrally framed price change.
The main findings includes effectiveness of all policy interventions is higher if substitutability is higher and the motivation crowding out due to taxation relevant only for low effort products.
Critiques were raised since it was a “one-shot” game testing “warm glow” and it required a state-solution.
Way of thinking that can lead to systematic deviations from the benchmark of rationality. The concept originates from psychology and behavioural economics. Classic biases include availability heuristics (remembered events seem more likely), confirmation bias (consume new information as complementing preconceptions), endowment effect (ownerships changes value perception), and framing effect (the presentation affects the conclusion).
The problem with the state without agency and the state as a nexus of cooperation (Acemoglu, 2009) is that the government is treated as a black box into which takes as input individual preferences and provides as output a collective choice. Arrow’s Impossibility Theorem (Arrow, 1963) shows that “the only voting method that is not flawed is a dictatorship.” Arrow stipulates that even under reasonable requirements there can’t exist a ranked voting system (i.e. social welfare function or preference aggregation rule) that transforms the set of preferences into a single global societal order for two or more participants with three or more option. Arrow specifies reasonable requirements as:
- Universality (unique and complete ranking)
- Independence of irrelevant alternatives
- Pareto efficient (unanimity)
Pairwise voting alternatives do not lead to a complete ordering. Only through violation of reasonable requirements (dictatorship, Kaldor-Hicks efficiency, non-universality) can preference aggregation be performed. Cox (Cox & McCubbins, 2000) showed that voting rules shape policy outcomes even if the voter preference is fixed.
Nudges or “Libertarian paternalism”
A nudge can be understood as a change in the choice context (that would be irrelevant to the homo oeconomicus) to intentionally steer real-world agents’ behaviour in the direction of the homo oeconomicus benchmark (see organ donor debate). Critiques of such “behavioural welfare economics” are again that no global social preference exists and that without appropriate information manipulation could occur. It also implies a benevolent paternalist state, which is debatable.
Endogenous versus exogenous institutions
The difference between whether on asses how institutions arose (endogenous) or how institutions impact policy (exogenous). This course focuses on the latter.