Income disclosure and consumer judgment in a multilevel marketing experiment

Consumer protection concerns have been raised related to income misrepresentation in multilevel marketing (MLM) recruitment. Though not required by law, some MLM firms choose to voluntarily disclose income information about previous participants. Through replication and extension of the experiment created in Bosley, Greenman, and Snyder (2020), we investigate the impact of these disclosures on consumer interest and earnings expectations. We test the external validity of their findings with subjects from Mechanical Turk and explore issues regarding treatment heterogeneity and better-than-average bias. Supporting prior findings, we present evidence that income disclosures, on average, do not significantly affect subjects' interest in the MLM opportunity, but they do decrease earnings estimates for most while increasing earnings estimates for a few. These effects depend on a subject's numeracy skills and whether they see their earnings potential as better-than-average. We also find that asking about earnings estimates first tends to decrease interest.