Estimating the Technology of Cognitive and Noncognitive Skill Formation

Exploring the role of parental engagement in non-cognitive skill development over the lifecourse

“…Our empirical analysis departs from the perspective that LOC observed at any age a is the result of a cumulative dynamic process that depends on past inputs, some fixed mental capacity, shocks and education opportunities (Todd and Wolpin, 2003;Cunha and Heckman, 2008;Cunha et al, 2010;Fiorini and Keane, 2014). The LOC production function of individual i at age a is:…”

“…We apply the human capability production framework to model the determinants of middle age control beliefs as a function of past control beliefs, parental behaviors, and socioeconomic opportunities that shape a child's life (Todd and Wolpin, 2003;Cunha and Heckman, 2008;Cunha et al, 2010;Fiorini and Keane, 2014). Importantly, we utilize a teacher-reported measure of parental interest in education to avoid some of the bias associated with parent self-reported measures (e.g., Reynolds, 1992).…”

“…Non-cognitive abilities are generally understood as the product of a complex combination of social learning processes and parental investments in child development, as well as genetic factors, education, and other aspects of the environment in which a child is raised. Early stocks of non-cognitive (and cognitive) skills feed into the production of later ones, such that compromised skill formation early on can hinder skill development processes down the track (e.g., Cunha et al, 2010).…”

School Closures and Effective In-Person Learning during COVID-19: When, Where and for Whom

“…Next we consider EIPL by local affluence, education and family structure, which are prominent inputs for models of human capital accumulation (e.g. Cunha et al, 2010), as well as student race. We proxy local affluence by average household income, education by the share of households with a college degree or higher, and family structure by the share double-headed households with children, all measured at the zip-code level of the school and based on 2016-2019 estimates from the American Community Survey (ACS).…”

Structural models for policy-making: Coping with parametric uncertainty

“…Instead, we conduct an as-if analysis where the point estimates serve as a stand-in for the true parameters. We study the implications of our models at the point estimates (Eisenhauer et al, 2015;Adda et al, 2017;Blundell et al, 2016;Eckstein et al, 2019) and rank competing policy proposals based on the point predictions (Cunha et al, 2010;Todd and Wolpin, 2006;Blundell and Shephard, 2012). In their handbook article, Keane et al (2011) state that they are unaware of any work that has ever reported the distribution of policy simulations that account for parametric uncertainty.…”