Identifying the average treatment effect in ordered treatment models without unconfoundedness

Lewbel, Arthur; Yang, Thomas Tao

doi:10.1016/j.jeconom.2016.05.015

Cited by 7 publications

(5 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The reliance on these assumptions is inescapable given the econometrics literature on point identification of discrete choice models. Our approach elucidates the identifying power of a single excluded regressor in models that satisfy the SCP and, in particular, the relative ranking of alternatives encapsulated in Facts 3 and 4 (see Lewbel & Yang, 2016, for related results for average treatment effects in ordered discrete choice models). We further exploit this structure to establish identification in models with substantially richer levels of unobserved heterogeneity, by allowing for dependence between consideration and preferences.…”

Section: Discussionmentioning

confidence: 99%

Discrete choice under risk with limited consideration

Barseghyan

Molinari

Thirkettle

2020

View full text Add to dashboard Cite

This paper is concerned with learning decision makers' preferences using data on observed choices from a finite set of risky alternatives. We propose a discrete choice model with unobserved heterogeneity in consideration sets and in standard risk aversion. We obtain sufficient conditions for the model's semi-nonparametric point identification, including in cases where consideration depends on preferences and on some of the exogenous variables. Our method yields an estimator that is easy to compute and is applicable in markets with large choice sets. We illustrate its properties using a dataset on property insurance purchases.

show abstract

Section: Discussionmentioning

confidence: 99%

Discrete choice under risk with limited consideration

Barseghyan

Molinari

Thirkettle

2020

View full text Add to dashboard Cite

show abstract

“…Lewbel and Yang () considered a different non‐monotonic selection mechanism for estimating the average treatment effect. They showed that the average treatment effect is identified when a binary treatment is assigned by

D = double-struck1 false(α_{0} \leq Z + V \leq α_{1} false),

where V is an unobserved random variable; Z is a continuous variable that satisfies

E false(Y_{j} false| V, Z false) = E false(Y_{j} false| V false)

for

j = 0, 1

and

V ⫫ Z

; and

α_{0}

α_{1}

are unknown parameters.…”

Section: Relation To the Existing Literaturementioning

confidence: 99%

“…Here, changes in Z 1 conform to monotonicity; but changes in Z 2 need not. Lewbel and Yang (2016) considered a different non-monotonic selection mechanism for estimating the average treatment effect. They showed that the average treatment effect is identified when a binary treatment is assigned by…”

Section: Other Non-monotonic Modelsmentioning

confidence: 99%

See 1 more Smart Citation

Identifying Effects of Multivalued Treatments

Lee¹,

Salanié²

2018

Econometrica

View full text Add to dashboard Cite

Multivalued treatment models have typically been studied under restrictive assumptions: ordered choice, and more recently, unordered monotonicity. We show how treatment effects can be identified in a more general class of models that allows for multidimensional unobserved heterogeneity. Our results rely on two main assumptions: treatment assignment must be a measurable function of threshold‐crossing rules, and enough continuous instruments must be available. We illustrate our approach for several classes of models.

show abstract

“…Other papers have considered identification under various violations of LATE monotonicity. In the case of a binary treatment, Gautier and Hoderlein (2011), Lewbel and Yang (2016) and Gautier (2020) consider various explicit selection models, while Chaisemartin (2017) shows that a weaker notion than monotonicity can be sufficient to give a causal interpretation to LATE estimands. 2 Lee and Salanié (2018) relax monotonicity in a setting with multivalued treatment and continuous instruments, generalizing results from the local instrumental variables approach of Heckman and Vytlacil (2005).…”

Section: Introductionmentioning

confidence: 99%

A Vector Monotonicity Assumption for Multiple Instruments

Goff¹

2020

Preprint

View full text Add to dashboard Cite

When a researcher wishes to use multiple instrumental variables for a single binary treatment, the familiar LATE monotonicity assumption can become restrictive: it requires that all units share a common direction of response even when different instruments are shifted in opposing directions.What I call vector monotonicity, by contrast, simply restricts treatment status to be monotonic in each instrument separately. This is a natural assumption in many contexts, capturing the intuitive notion of "no defiers" for each instrument. I show that in a setting with a binary treatment and multiple discrete instruments, a class of causal parameters is point identified under vector monotonicity, including the average treatment effect among units that are responsive to any particular subset of the instruments. I propose a simple "2SLS-like" estimator for the family of identified treatment effect parameters. An empirical application revisits the labor market returns to college education.

show abstract

Identifying the average treatment effect in ordered treatment models without unconfoundedness

Cited by 7 publications

References 49 publications

Discrete choice under risk with limited consideration

Discrete choice under risk with limited consideration

Identifying Effects of Multivalued Treatments

A Vector Monotonicity Assumption for Multiple Instruments

Contact Info

Product

Resources

About