Understanding Instrumental Variables in Models with Essential Heterogeneity

By James J. Heckman, Sergio S. Urzúa, E. Vytlacil. Review of Economics and Statistics, 2006

887 citations recorded by the source corpus.

This paper examines the properties of instrumental variables (IV) applied to models with essential heterogeneity, that is, models where responses to interventions are heterogeneous and agents adopt treatments (participate in programs) with at least partial knowledge of their idiosyncratic response. We analyze two-outcome and multiple-outcome models, including ordered and unordered choice models. We allow for transition-specific and general instruments. We generalize previous analyses by developing weights for treatment effects for general instruments. We develop a simple test for the presence of essential heterogeneity. We note the asymmetry of the model of essential heterogeneity: outcomes of choices are heterogeneous in a general way; choices are not. When both choices and outcomes are permitted to be symmetrically heterogeneous, the method of IV breaks down for estimating treatment parameters.

Subfields: Economics and Econometrics, Gender Studies, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/dd825561b6da0135e39a891aac926d8ce4ac26bf

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 Gender Studies

In Statistics and Probability

Citations in the corpus

Econometric Methods for Program Evaluation (2018, positive): “…variables (and other program evaluation) settings was introduced by Björklund and Moffitt (1987), and later devel- oped and popularized by Heckman and Vytlacil (1999, 2005) and Heckman et al. (2006), and more recently studied by Kline and Walters (2016) and Brinch et al. (2017), among others.”
Heterogeneous Preferences and the Efficacy of Public School Choice (2008, positive): “…both heterogeneous short-run effects of attending a first choice school on academic outcomes (Heckman (1997), Heckman, Smith, and Clements (1997), Heckman, Urzua, and Vytlacil (2006)) and disparate demand-side pressure for academic improvement across low-achieving and high-achieving schools…”
The Identification and Economic Content of Ordered Choice Models with Stochastic Thresholds (2007, positive): “This specification is commonly invoked in labor economics (see, e.g., Keane and Wolpin, 1997, or Heckman et al., 2006a).”
Controlling for endogeneity with instrumental variables in strategic management research (2008, positive): “all entities induced to change their treatment status in response to changes in the value of an instrument do so in the same ‘direction’), although the strength of the response may vary among people (Heckman et al., 2006).”
Social Media, News Consumption, and Polarization: Evidence from a Field Experiment (2020, positive): “Reweight Based on Compliance P is estimated using two IV estimators, and thus its causal interpretation relies on the assumption that there is no essential heterogeneity (Heckman et al., 2006).”
Estimating distributions of potential outcomes using local instrumental variables with an application to changes in college enrollment and wage inequality (2009, positive): “Heckman, Urzua, and Vytlacil (2006) further extended the method of local instrumental variables and Aakvik, Heckman, and Vytlacil (2005) used factor structures for the analysis of the latent variable framework of Heckman and Vytlacil (1999, 2001, 2005).”
Estimating Marginal Treatment Effects in Heterogeneous Populations (2008, neutral): “…of program participation (F) on y varies with the level of participation because g is a function of F, thus inducing an inherent nonlinearity of y in F, a feature of heterogeneous treatment effects models noted by Heckman and Vytlacil (2005, p.691) and also discussed in Heckman et al. (2006).”
Identification of Social Interactions (2011, positive): “One reason involves what Heckman, Urzua, and Vytlacil (2006) have dubbed “essential heterogeneity”: if the effect of increased male graduation on a girl’s payoff differential between graduating and dropping out is heterogeneous, then it is unclear how to interpret the Cipollone and Rosolia…”
The Growing Phenomenon of Private Tutoring (2008, neutral): “…homogeneity assumption is violated, the estimated benefits of tutoring will not apply to subgroups and may be biased as well (see Heckman, Lochner, and Todd 2006 and Heckman, Urzua, and Vytlacil 2006 for discussions of the heterogeneity in returns to education in instrumental variables models).”