The Accuracy and Malleability of Parental Beliefs about Child Socio-Emotional Health

We document systematic parental under-reporting of children’s socio-emotional difficulties relative to children’s self-reports, using representative data from Luxembourg, the United Kingdom, and Australia. To study the origins of this discrepancy, we develop a simple theoretical framework showing how parent–child gaps can arise from information frictions and differences in reporting styles. We complement the model with a novel survey design that elicits both parental beliefs about children’s latent socio-emotional wellbeing and parental beliefs about children’s self-reports, allowing us to disentangle the different sources of the discrepancy. Using a new survey from Luxembourg, we estimate that approximately 70% of the observed gap is attributable to information frictions. Consistent with a Bayesian model of signal extraction, belief accuracy declines when children experience high levels of distress. The precision of second-order beliefs is negatively correlated with parental education, income, and employment, and—paradoxically—with more accurate priors about aggregate parental under-reporting, a pattern we refer to as the Capacity Paradox. As predicted by the model, a randomized information intervention shifts both first- and second-order beliefs only among parents with weak priors and generates heterogeneous effects on intended parental investments. These findings highlight the central role of second-order beliefs in understanding parental misperceptions and the potential for targeted information policies to improve parental awareness of children’s socio-emotional wellbeing.