Policymakers, schools, and parents often rely on teachers’ perceptions about child development
to inform their decisions about investments in children. In this paper, I show that early childhood education instructors' perceptions about developmental delays are not only influenced by children's own development but also by the average level of development of other children in the neighbourhood. I quantify the influence of the reference group on instructors’ perceptions of developmental delays using both non-cognitive (socio-emotional) and cognitive (receptive language) objective development measures from the Longitudinal Study of Australian Children, assessed by psychologist-trained interviewers.  I show that instructors in neighbourhoods with lower average levels of non-cognitive development are less likely to perceive delays in both non-cognitive and cognitive dimensions of child development, conditional on objective development measures. This implies that they are less likely to recognize a developmental delay if such delays are more prevalent in the neighbourhood.  Further, the maternal perceptions of their children’s non-cognitive development are influenced by the information about delays that teachers convey. Teachers' and mothers' beliefs about delays predict investment in remedial services including children’s learning and behavioural therapy, and tutoring, as well as parental attitudes toward their children. Finally, I show that instructors with college degrees in education are more likely to identify children with low levels of development compared to instructors with diplomas or certificates.

I estimate the evolution of prices and quantities of human capital in occupational groups specializing in abstract, routine, and manual tasks to examine the explanations for rising U.S. wage inequality. To quantify changes in prices and quantities of occupation-specific human capital, I use a flat spot price identification method which exploits the wage changes for workers approaching retirement to infer the shifts in the prices of skills. Importantly, this method accommodates cohort quality changes over time, capturing the effects of shifts in the educational composition of workers in occupations as well as changes in the quality of education across generations.  My results show that the price increased for the abstract group relative to manual and routine groups between 1970 and 2017. Conversely, the prices of human capital in manual and routine groups declined and were highly correlated. This evidence favours the hypothesis of skill-biased technical change (SBTC), which predicts the increase in the relative price of high-skill human capital. Furthermore, even within the abstract group, postgraduate workers have disproportionately benefited from the changing relative demand for high-skill workers compared to workers with bachelor's degrees, supporting the implications of the SBTC theory. My findings do not align with the predictions of the routine-biased technical change (RBTC) hypothesis regarding the fall in the price of skill in routine relative to manual occupations.