The Perils of Peer Effects

By Joshua D. Angrist. 2014

616 citations recorded by the source corpus.

Individual outcomes are highly correlated with group average outcomes, a fact often interpreted as a causal peer effect. Without covariates, however, outcome-on-outcome peer effects are vacuous, either unity or, if the average is defined as a leave-out mean, determined by a generic intraclass correlation coefficient. When pre-determined peer characteristics are introduced as covariates in a model linking individual outcomes with group averages, the question of whether peer effects or social spillovers exist is econometrically identical to that of whether a 2SLS estimator using group dummies to instrument individual characteristics differs from OLS estimates of the effect of these characteristics. The interpretation of results from models that rely solely on chance variation in peer groups is therefore complicated by bias from weak instruments. With systematic variation in group composition, the weak IV issue falls away, but the resulting 2SLS estimates can be expected to exceed the corresponding OLS estimates as a result of measurement error and for other reasons unrelated to social effects. Research designs that manipulate peer characteristics in a manner unrelated to individual characteristics provide the most compelling evidence on the nature of social spillovers. As an empirical matter, designs of this sort have mostly uncovered little in the way of socially significant causal effects.

Subfields: Economics and Econometrics, Education, Sociology and Political Science

View Paper: https://www.semanticscholar.org/paper/2d845036e25ea96c19692ef3a57d698a22cd7d79

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

In Education

In Sociology and Political Science

Citations in the corpus

Ability Peer Effects in University: Evidence from a Randomized Experiment (2016, background): “These nationwide exams 2Moreover, Angrist (2014) points out that peer effect studies based on naturally occurring variation may suffer from weak instrument type bias.”
Contagion of pro- and anti-social behavior among peers and the role of social proximity (2019, background): “Researchers face substantial challenges when studying peer effects, especially in field experiments where allocation to peer groups is often non-random (Manski, 1993, 2000; Angrist, 2014).”
The effects of financial reporting and disclosure on corporate investment: A review (2019, methodology): “Electronic copy available at: https://ssrn.com/abstract=3429337 51 The empirical challenges in peer-effect studies are such that Angrist (2014) paints a bleak view of this literature.”
Young Adults Living with Their Parents and the Influence of Peers (2018, background): “Definition of emancipated peers Nominated friends who left the parental home no later than the respondent (1) (2) (3) (4) (5) (6) % peers living with parents 0.”
Leave‐Out Estimation of Variance Components (2018, background): “” A large experimental literature (Sacerdote, 2001; Graham, 2008; Chetty et al., 2011; Angrist, 2014) employs variants of Fisher’s ANOVA approach to infer the degree of variability attributable to peer or classroom effects.”
Do tax incentives for research increase firm innovation? An RD design for R&D (2016, neutral): “…dyads should improve precision which can be an issue in RD Design ap- proaches to estimating peer effects (see Dahl, Loken and Mogstad, 2014 or Angrist, 2014) As with the continuous, non-dyadic approach, we focus on the sample in which both firm i and firm j have total assets between €51m…”
Peer effects in corporate disclosure decisions (2020, neutral): “As noted in Section 2.1., however, this finding cannot be attributed to causal peer effects due to the reflection problem (Manski, 1993; Angrist, 2014).”
Behavioral Economics of Education (2015, background): “Hanushek 2006) and Sacerdote (2011), for example, argue that peer groups are as important for student outcomes as other inputs n the production function, although the validity of these results is sometimes questioned (Angrist, 2014).”
Identification and Estimation of Spillover Effects in Randomized Experiments (2017, background): “While some empirical specifications have attempted to relax parametric assumptions (Hoxby and Weingarth, 2005; Carrell, Fullerton, and West, 2009; Graham, 2008; Sacerdote, 2011, 2014), these models have only been analyzed from a linear regression perspective; as such, the identified parameters can b”
Exposure to More Female Peers Widens the Gender Gap in STEM Participation (2020, background): “26 Angrist (2014) shows that with chance variation in peer groups, measurement error can bias peer effects estimates.”
The Economics of Disclosure and Financial Reporting Regulation: Evidence and Suggestions for Future Research (2015, background): “See Manski (1993) and Angrist (2014) for the problems of estimating peer effects, which likely also apply to externalities.”
Territorial servitization: Exploring the virtuous circle connecting knowledge-intensive services and new manufacturing businesses (2017, methodology): “value of the dependent variable in the model to account for potential regional peer effects (Angrist, 2014).”
Heterogeneous Peer Effects and Rank Concerns: Theory and Evidence (2017, positive): “4 As noted by Angrist (2014), “research designs that manipulate peer characteristics in a manner unrelated to individual characteristics provide the most compelling evidence on the nature of social spillovers.” Given that only few experiments exist, I rely on a natural experiment that provides this ”