Bidding for nothing? The pitfalls of overly neutral framing

ABSTRACT

Neutral framing is a standard tool of experimental economics. However, overly neutral instructions, which lack any contextual clues, can lead to strange behaviour. In a contextless second-price auction for a meaningless good, a majority of subjects enter positive bids – likely a case of cognitive experimenter demand effect. Subjects may interpret the lack of context as being tasked with bidding in the experiment. Adding another auction that has a context drastically reduces the positive bids in the meaningless auction.

KEYWORDS:

• Context
• neutral framing
• experimenter demand effect
• experiment
• second-price auction

JEL CLASSIFICATION:

• C90
• D44

Acknowledgments

We thank Christoph Engel, Christoph Vanberg and Israel Waichmann for valuable comments. Financial support by the DFG through SFB 504 is gratefully acknowledged.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ For a discussion of neutral framing, see Abbink and Hennig-Schmidt (2006).

² The more general effect of subjects changing their behaviour when observed by experimenters has long been studied by experimental psychologists under the name ‘Hawthorne effect’ (Adair 1984).

³ The act of destroying was not further specified. Most subjects opted for either crossing out the envelope with pens, or ripping the envelope into pieces.

⁴ A full description of the instructions and test questions can be found in the online appendix: http://www.uni-heidelberg.de/md/awi/professuren/with2/Duersch-Mueller-BfN-appendix.zip and in Duersch and Müller (2015).

⁵ In one-half of 1A the envelopes were empty (as in NC in treatment 2A) and the other half the envelopes were filled with paper money (as in C in treatment 2A). There was no significant difference in bids (MWU test, 2-sided, , obs. = 33), therefore we pool the data.

⁶ In psychology, context effects are studied in the design of surveys. They refer to question order, phrasing of questions and the impact that earlier questions have answers to subsequent ones (Schwarz and Sudman 1992). We use the word context in a broader, more general meaning and refer with it to the whole experimental environment given in the instructions.

⁷ The distributional choice is similar to the one in a dictator game (Kahneman, Knetsch, and Thaler 1986) or ultimatum game (Güth, Schmittberger, and Schwarze 1982) with a restricted choice domain.

⁸ Note that the auction always has a winner; therefore, the allocator’s pay-off is reduced in all cases. The good sold in auction C is not punishment, but punishing personally. When run on its own, this auction produces positive but low bids, such that the bidding is comparable to the results in 1A (see Duersch and Müller 2015). However, we are not interested in the results of auction C here. Instead, we want to compare the auction without context, NC, in two situations: once on its own, and once while being contrasted with an auction with some form of strong context.

⁹ The average bid of €0.03 in auction NC of treatment 2A might be an indication of the real value subjects place on destroying an envelope.

¹⁰ The low average bid in auction NC of treatment 2A also makes competing explanations for bids in treatment 1A, such as ‘desire to win’ (van den Bos et al. 2008) or misunderstanding the second price auction, less likely.

Additional information

Funding

Financial support by the DFG through SFB 504 is gratefully acknowledged.