RESULTAT

FRAGE A: the results show that people behave pro-socially, conditional on others. The more others cooperate, the more one is inclined to do so as well = Gruppe "High" hat öfters gespendet (höhere Beträge konnten sie nicht spenden, da Spendenbeträge fix)
Durch exp Bed. "High" wurde likelihood des spendens erhöht. durch exp bed "low" wurde die likelihood des spendens nicht reduziert (im vgl zu kontroll).

---ACHTUNG EINSCHRÄENKUNG: individuals have heterogeneous thresholds as to when theyare willing to cooperate,given the behavior of others. Some are willing to cooperate if only a small minority do so asEell,
While others cooperate only Ehen a clear mabority do so. In the field eHperimental setting, onlypeople Ehose threshold is betEeen 46 and 64% react to the treatments. Students who have athreshold below 46 contribute to the funds independent of the treatments and students who have thresholds above 64 also do not care about the induced beliefs. This leaves a subgroup ofthe Ehole population who respond to the pro-social behavior of others.

Ob jmnd spendet oder nicht könnte auch stark vom habitus (wie oft in letzten Jahren gespendet) abhängen.

---(((Wer lässt sich beeinflussen?
0=nie in letzten Jahren gespendet; 1 = immer; 0.5 = in 50% der Fälle
Personen von 0-0.5 lassen sich am stärksten beeinflussen, Personen bei 0 und 1 am wenigsten.
(ähnlich aus workspace: social comparisons in the working sphere, suggesting that,due to peer pressure, an induced high productivity norm increases the productivity of the leastproductive subbect, but a low productivity norm does not have much influence on the mostproductive subbects Falk and Ichino, 2003.)))

FRAGE B: A change in the perceived cooperation rate of others by ten percentage points, evaluated at the mean expectation, raises the probability of contributing by more than siHpercentage points

Interne Norm / Expectations of behaviour of others

---Expectations about the behavior of others should positively correlate with one's behavior, as found in various studies (Selten and Ockenfels, 1998] Croson, 1998, Dawes et al., 1977), Seite 10 oben)
--> vllt. auch fragen, was Personen denken wie viele Leute in % spenden an Hospiz / an wohltätige Organisationen, falls oben genannte Studien keine Feldstudien.

Kausalität Erwartung --> Verhalten
LABORFischbacher, Schächter and Fehr (2001) solve the causality problem by using the strategy method. Subbects in their laboratory public good game have to decide how much to giveto a public account, given the contributions of others. The study concludes that roughly 50 percent of the people increase their contribution if the others do so as well.
Similar results are found by Falk, Fischbacher and Schächter :2002, who get their subbects to play two separate public good games simultaneously. The authors find two social interaction effectsI firstly, people give more to the group with high cooperation rates, and secondly, the contribution within one group depends positively on others contributions.
A number of other studies in economics do not test the effects of social comparison explicitly, but the results of public good experiments show that individual contribution varies with the mean contribution of the group Keser and van Winden, 2000] Offerman et al., 199U] Sutter and Weck-Hannemann, 2003@

FELD Andreoni and Scholz :1998, who find that ones own donation depends on the donations of ones reference group. Their results show that, if the contribution of those in ones social reference group increases by an average of 10% ,then the expected rise in ones own contribution rises by about 2% to 3%.
ökonomische studie, nur Korrelation, sehr mathematisch --> zum Glück nicht genau, was wir machen https://onlinelibrary.wiley.com/doi/epdf/10.1111/j.1465-7295.1998.tb01723.x

What's the difference between external and internal norm?

An external norm is a norm that is presented/that is outside.

An internal norm is a norm that reflects the expactation of the individual. Internal norm might be understood as 1) what a person would give (her personal norm) or 2) what a person expects the external norm to be. In the Zürcher Studie, this expactation is mostly described as expectation about the behaviour of others
--> if that's the term mostly used in literature, then we should go with that :)

Result: Info über die Anzahl an Spendern in Population hat einfluss auf spendewahrsheinlichkeit von individuum

WEITERFUERENDE STUDIEN:

Analysen vorhandener Daten (correlational): Observational research uses data from donation surveys or actual giving collected from tax returns; for reviews, see Andreoni (2006), Powell and Steinberg (2006), Kolm and Ythier (2006).

LAbor-Experimente find a positive relationship between others’ contributions and one's own (Bardsley, 2000; Croson, 2007; Fischbacher et al., 2001; Keser and van Winden, 2000; Weimann, 1994). However, these studies have important limitations. First, the contributions of others are typically measured rather than manipulated, thus the experiments do not provide a tight test of the theories

Feld-Expeiment (Levitt and List, 2007). This paper addresses both these concerns by manipulating the social information donors receive in a natural (field) setting.

Feld-Experimente - nicht genau unser Thema:
List and Lucking‐Reiley (2002) (seed money), Eckel and Grossman (2005) (rebates and matching), Falk (2005) (gift exchange)

---Gründe für Entscheidung ob und wieviel geben:
Psychological research (Ajzen, 1991; Brooks, 2004) has shown that decisions about whether to act and about how much to act, although positively correlated, may be caused by different psychological motivations. Similarly, Andreoni (2006) suggested that altruism tells people what causes to give to but that warm‐glow tells people how much to give.

---Gründe, wieso social information behaviour beeinflusst (perception change)
In other work, we provide evidence for conformity to social norms as an explanation; we found that social information changes people's perceptions of both what others give to the non‐profit organisation and what the appropriate contribution is. These changed perceptions correlate at the individual level with self‐reported contribution behaviour; individuals whose perceptions change more, give more than individuals whose perceptions change less (Croson et al. 2008).

Importance of social information in economic studies/situations
Others have suggested the importance of norms in actual (Akerlof, 1982) and experimental (Fehr et al., 1998) labour markets, whether to work or live on welfare (Lindbeck et al., 1999), saving and consumption (Lindbeck, 1997) and on profit seeking entitlements (Kahneman et al., 1986).

DIFFERENCES TO OUR STUDY

• self-selecting sample of people involved in study
• via telephone (and not live at home)
• social info = ("invented") donation of last person (and not general 90-90th percentile of distribution of donations in past)
• radio station and not Hospiz

BEGRÜNDUNG HOHES PERCENTIL:

Previous work in psychology and goal‐setting suggest that behavioural goals need to be achievable and inspiring in order to change behaviour effectively (Locke and Latham, 1990).

In our setting (Radiostation Spendenmarathon 3mal im Jahr für Radiostation, 300 dollar als social reference war am effektivsten (vs 180, 75 und control)), this internal optimum is achieved at the 90th–95th percentile of contributions (in Shang and Croson forthcoming, the 80th percentile of contributions was too low to show a significant effect, while here the 99th percentile is too high to be effective).

auch movie lens study zeigt auf- und abwärtsanpassungen (2010)

KRITIK

• Personen könnten mehr gespendet haben, WEIL ANDERE PERSONEN ZUGEHOERT HABEN!!!! --> Grösse Dorf/Stadt und Sendereichweite nachforschen (in kleinem Dorf wäre dieser Effekt wohl stärker, da jeder jeden kennt...)