Framed field experiments with heterogeneous frame connotation

We study label framing effects in linear public goods games. By accounting for heterogeneous frame connotation, we can identify individual framing effects. We test for such effects in a field experiment on irrigation management in India. Using membership of the water users association as a proxy for frame connotation, we find a differential impact on contributions in the game. Members contribute relatively more under the irrigation frame than non-members as compared to an alternative, neutral, frame. We conclude that experimental behaviour is sensitive to framing at the individual level but that such individual effects may cancel out on average, which explains previous studies that find mixed or only weak effects of framing.