Working Paper: CEPR ID: DP14681
Authors: Corrado Giulietti; Michael Vlassopoulos; Yves Zenou
Abstract: We provide first evidence that peer depression in adolescence affects own depression in adulthood. We use data from Add Health and an identification strategy that relies on within-school and across-cohort idiosyncratic variation in the share of own-gender peers who are depressed. We find a significant peer effect for females but not for males. An increase of one standard deviation of the share of own-gender peers (schoolmates) who are depressed increases the probability of depression in adulthood by 2.6 percentage points for females (or 11.5% of mean depression). We also find that the peer effect is already present in the short term when girls are still in school and provide evidence for why it persists over time. Further analysis reveals that individuals from families with a lower socioeconomic background are more susceptible to peer influence, thereby suggesting that family can function as a buffer. Our findings underscore the importance of peer relationships in adolescence with regard to the development of long-lasting depression in women.
Keywords: Causal; Peer Effects; Depression; Gender; Adolescence
JEL Codes: I12; Z13
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| peer depression in adolescence (C92) | own depression in adulthood (I12) |
| increase in the share of same-gender depressed peers (C92) | own depression in adulthood (I12) |
| peer depression (C92) | own depression shortly after (within one year) (I12) |
| peer depression (C92) | reduced likelihood of college attendance (D29) |
| peer depression (C92) | reduced likelihood of employment (J68) |
| peer depression (C92) | reduced income (E25) |
| lower socioeconomic backgrounds (I24) | more susceptibility to peer influence (C92) |