Article
Whom do we trust?
A field experiment addressing ethnic discrimination in a multicultural society
-
1We find evidence of discrimination in market transactions against buyers with culturally distant ethnic backgrounds (7% more real sellers prefer fictitious buyers with Catalan, Spanish or Latin-American names over fictitious buyers with Arab or Chinese names). Discrimination occurs only if the offer is made at the listed price.
-
2We do not find evidence of discrimination in market transactions against ethnic minority sellers: real buyers do not make more bids for products sold by fictitious sellers with Spanish or Catalan names than when the fictitious sellers have Arab, Chinese, or Latin-American names.
-
3The presence of discrimination against buyers and the absence of discrimination against sellers of culturally distant ethnic backgrounds could be due to experimental effects, but also to forms of statistical discrimination based on stereotypes about minorities’ behaviour as buyers (e.g., being seen as less trustful) and sellers (seen as equally trustworthy).

Differences between fictitious buyers distinguished by their native or foreign background in the percentage of positive responses received from real sellers (left), and in the number of bids made by real buyers for fictitious sellers’ products (right). We find evidence of discrimination against ethnic minorities only on the buying side, not on the sellers’ side. The average number of positive enquiries received by fictitious sellers from real buyers was similar regardless of whether their names denoted native or minority background (about 2 messages each).