« Aprilpyssel | Main | Begreppslig utredning och försvar att använda "papret" i föregående blogganteckning »
april 04, 2005
Analys av Google Answers
Papret Why Voluntary Contributions? Google Answers (PDF) av Tobias Regner är en analys av Google Answers och förhållande mellan pris, "tipping" och reputation utifrån data mellan juli 2003 och januari 2004. Notera att en del av papret är rätt tekniskt (formler och slika saker).
Abstract:
We study the pricing and tipping behaviour of users of the online service `Google Answers'. While they set a price for the answer to their question ex ante, they can additionally give a tip to the researcher ex post.
We develop a model that is based on reciprocal theories of social preferences pioneered by Rabin (1993) and extended by Dufwenberg and Kirchsteiger (2004). The predictions of our model are empirically tested with the field data we obtained.
The reasons for leaving a tip are analysed. A significant amount of users are motivated by social preferences. We also find strong support for reputation concerns. Moreover, researchers appear to adjust their effort based on the user's previous tipping behaviour.
We conclude that an endogenous incomplete contracts design encourages people to contribute voluntarily. This is motivated by reciprocity when people are socially minded, but also generally by strategic behaviour to build up a good reputation. Efficiency is increased when contracts are left open deliberately as high effort is sustained.
Keywords: social preferences, reciprocity, moral hazard, reputation, internet
En kort summering från introduktionen:
About 25% of all answers have been tipped and the three main conclusions from our empirical analysis follow. The more questions users ask over time the more likely they are to tip. Thus, there are reputation concerns that make people tip. It seems that they aim to build up a fine reputation in order to encourage good effort of researchers for future questions. Naturally, individuals have this concern no matter what their preferences are. Moreover, even single users tip, though: around 18%. This clearly deviates from self-interest. Hence, social preferences of some kind have to be involved. Finally, tipping — motivated by social preferences or out of reputation concerns — seems to pay off. Our data confirms that researchers take the past tipping behaviour of users into account and put more effort into the answer, if the user has frequently tipped before. The higher effort increases the benefit of the user and the researcher gets fairly compensated for the extra effort.
Via nep-evo mailinglista.
Posted by hakank at april 4, 2005 05:28 EM Posted to Spelteori och ekonomi