Limited Information Econometrics
|
|
- Marjorie Shelton
- 5 years ago
- Views:
Transcription
1 Limited Information Econometrics Walras-Bowley Lecture NASM 2013 at USC Andrew Chesher CeMMAP & UCL June 14th 2013 AC (CeMMAP & UCL) LIE 6/14/13 1 / 32
2 Limited information econometrics Limited information econometric models leave the determination of some endogenous variables unspeci ed. Single equation or systems? Computation, e ciency, robustness - identi cation? AC (CeMMAP & UCL) LIE 6/14/13 2 / 32
3 Limited information econometrics Incomplete models may not be point identifying when there are: discrete outcomes (binary, ordered, counts, multiple discrete choice) rich speci cations of heterogeneity (random coe cients, individual e ects, frailties in duration analysis). The common feature is that unobservables are not single valued functions of observed variables. When unobservables are set valued functions of observable variables incomplete models are generically partially identifying. AC (CeMMAP & UCL) LIE 6/14/13 3 / 32
4 Limited information econometrics Reactions? ignore endogeneity, work up enthusiasm for a point identi ed parameter, beef up the restrictions and use point identifying models. AC (CeMMAP & UCL) LIE 6/14/13 4 / 32
5
6 Limited information econometrics Use the model which embodies your beliefs about the economic process being studied. If that model is partially identifying then so be it. Keep Calm and Carry On. Barriers to the use of partially identifying models? Methods for conducting inference? Recent advances. Developing usable characterizations of sharp identi ed sets? We provide a simple characterization of the identifying power of a wide class of limited information models. I explain, and illustrate, drawing on research done with my colleagues Konrad Smolinski and Adam Rosen. AC (CeMMAP & UCL) LIE 6/14/13 6 / 32
7 Example: binary outcome IV model This linear model for endogenous Y = (Y 1, Y 2 ), exogenous Z, unobserved U 1 : Y 1 = α 0 + α 1 Y 2 + U 1 U 1? Z can point identify α 1. But this binary model set identi es α 1 Y 1 = 1[α 0 + α 1 Y 2 + U 1 > 0] U 1? Z even in a parametric probit case. This IV probit model is studied in Chesher (2010, 2013). Here is a picture of the identi ed set for (α 0, α 1 ). AC (CeMMAP & UCL) LIE 6/14/13 7 / 32
8 Probit IV model: Iden4fied set for α 0 and α 1 α α 1 = α 0
9 Complete triangular model A complete recursive, triangular model is often employed: Y 1 = 1[α 0 + α 1 Y 2 + U 1 > 0] Y 2 = g(z, U 2 ) (U 1, U 2 )? Z Can be point identifying because it implies conditional exogeneity of Y 2 : U 1? Y 2 ju 2 Only useful if U 2 is a single valued, identi able (control) function of (Y 2, Z ). so, only scalar U 2 allowed, no discrete Y 2 (but see e.g. Chesher (2005), Jun, Pinkse Xu (2012), for partial identi cation) AC (CeMMAP & UCL) LIE 6/14/13 8 / 32
10 Complete triangular model This control function solution is the industry standard Y 1 = 1[α 0 + α 1 Y 2 + U 1 > 0] Y 2 = g(z, U 2 ) (U 1, U 2 )? Z see e.g. STATA s ivprobit command, Heckman (1978), Smith and Blundell (1986), Rivers and Vuong (1988), Altonji and Ichimura (2003), Blundell and Powell (2003, 2004), Chesher (2003), Vytlacil and Yildiz (2007), Florens, Heckman, Meghir, and Vytlacil (2007), Imbens and Newey (2009), Jun, Pinkse and Xu (2009), Petrin and Train (2010), Shaikh and Vytlacil (2010), Blundell, Chen and Kristenssen (2013), and many more. And a lot of applications. AC (CeMMAP & UCL) LIE 6/14/13 9 / 32
11 Plausible restrictions Many other complete point identifying models can be entertained. But can we believe the additional restrictions embodied in these models? AC (CeMMAP & UCL) LIE 6/14/13 10 / 32
12 "I can't believe that!" said Alice. "Can't you?" the queen said in a pitying tone. "Try again, draw a long breath, and shut your eyes." Alice laughed. "There's no use trying," she said. "One can't believe impossible things.
13 "I dare say you haven't had much praccce," said the queen. "When I was your age, I always did it for half an hour a day. Why, somecmes I've believed as many as six impossible things before breakfast. Lewis Carroll (1871)
14 Exploring partially identifying power of models Choosing a point-identifying completion of the incomplete model is selecting a point in the incomplete model s identi ed set. it s worth at least a look at that set... AC (CeMMAP & UCL) LIE 6/14/13 12 / 32
15 Characterizing identi ed sets Our models restrict a function h : R YZU! R that delivers sets Y level sets: Y h (u, z) fy : h(y, z, u) = 0g U level sets: U h (y, z) fu : h(y, z, u) = 0g Example: a linear model, y = (y 1, y 2 ) with z excluded y 1 = α 0 + α 1 y 2 + u has: h(y, z, u) = y 1 (α 0 + α 1 y 2 + u) Y level sets: Y h (u, z) = f(α 0 + α 1 y 2 + u, y 2 ) : y 2 2 support of Y 2 g U level sets: U h (y, z) = fy 1 α 0 α 1 y 2 g AC (CeMMAP & UCL) LIE 6/14/13 13 / 32
16 Identi ed sets in incomplete models Example: a binary model, again y = (y 1, y 2 ) with z excluded y 1 = 1[α 0 + α 1 y 2 + u > 0] has h(y, z, u) = y 1 1[α 0 + α 1 y 2 + u > 0] non-singleton U level sets 8 < (, (α 0 + α 1 y 2 )], y 1 = 0 U h (y, z) = : ( (α 0 + α 1 y 2 ), + ], y 1 = 1 AC (CeMMAP & UCL) LIE 6/14/13 14 / 32
17 Distributions of unobservables and structures Our models restrict a function h : R YZU! R that delivers sets Y level sets: Y h (u, z) fy : h(y, z, u) = 0g U level sets: U h (y, z) fu : h(y, z, u) = 0g Our models also restrict a collection of probability distributions: G U jz fg U jz =z : z 2 R Z g where (notation) for a set S R U : G U jz =z (S) = P[U 2 SjZ = z] So models de ne admissible structures (h, G U jz ). AC (CeMMAP & UCL) LIE 6/14/13 15 / 32
18 What data tell us Data informs about distributions of observable variables: F Y jz ff Y jz =z : z 2 R Z g where for a set T R Y : F Y jz =z (T ) = P[Y 2 T jz = z] The collection of structures admitted by a model that can deliver a family of distributions F Y jz is that model s identi ed set. In our papers we use concepts and methods drawn from the theory of random sets, introduced into econometrics by Beresteanu, Molchanov and Molinari (2011). AC (CeMMAP & UCL) LIE 6/14/13 16 / 32
19 Characterizing sharp identi ed sets For sets S R U de ne a set: J h (S, z) fy : U h (y, z) Sg where U h (y, z) fu : h(y, z, u) = 0g J h (S, z) contains all values of Y which structural function h says eventuate (when Z = z) if and only if U 2 S. AC (CeMMAP & UCL) LIE 6/14/13 17 / 32
20 Characterizing sharp identi ed sets For sets S R U de ne a set: J h (S, z) fy : U h (y, z) Sg U h (y, z) fu : h(y, z, u) = 0g The identi ed set delivered by F Y jz and a model M, comprises structures (h, G U jz ) admitted by M that satisfy: G U jz (Sjz) F Y jz =z (J h (S, z)), a.e. z 2 R Z for a su ciently rich collection of test sets S R U. proof uses results in Artstein (1983), Molchanov (2005), Norberg (1992). sharp identi ed set, characterized by moment inequalities. simpli es under independence U? Z. G U (S) sup z2r Z F Y jz =z (J h (S, z)) AC (CeMMAP & UCL) LIE 6/14/13 18 / 32
21 Example: Interval censored endogenous variable The model: Y 1 = g(y 2, U) U? Z 2 R Z P[ Y 2 Y 2 Y 2 ] = 1 g is strictly increasing in U and Y 2. g is normalized so that U Unif (0, 1). Observed: Y (Y 1, Y 2, Y 2 ) and Z. Manski and Tamer (2002) meets Chernozhukov and Hansen (2005). Identi ed set comprises admissible functions g such that, for all [t, t] [0, 1]: where t t sup z 2R Z F Y jz =z (J h ([t, t], z)) J h ([t, t], z) fy : g(y 2, t) y 1 g(y 2, t)g AC (CeMMAP & UCL) LIE 6/14/13 19 / 32
22 Numerical example Generate data : (probability distributions) using a Gaussian triangular model. U V Y 1 = γ 0 + γ 1 Y 2 + U Y2 = δ 1 Z + V 0 σuu? Z N, 0 σ uv σ uv σ vv Y 2 = Y 2 W 2 Y 2 = Y 2 + W 3 W i Exp(λ i ) Data generating parameter values, ( E [W i ] = 1/λ i ): γ 0 γ 1 δ 1 σ uu σ uv σ vv λ 2 λ f1, 2g f5, 10g f5, 10g Z 2f 1, 1g AC (CeMMAP & UCL) LIE 6/14/13 20 / 32
23 Numerical example The model is parametric Gaussian: Y 1 = g 0 + g 1 Y 2 + su U? Z U N(0, 1) Calculate identi ed set for (g 0, g 1, s). P[Y 2 Y 2 Y 2] = 1 Use a selection of intervals [t, t]: for m = 1/M 2 (0, 1): 2 3 [0, m] [0, 2m] [0, 3m] [0, 1] [m, 2m] [m, 3m] [m, 1] 6 4 [2m, 3m] [2m, 1] and M 2 f4, 5, 6, 7, 8, 9, 11g, involves from 9 to 133 inequalities. AC (CeMMAP & UCL) LIE 6/14/13 21 / 32
24 1.0 g g s
25 1.0 g g s
26 1.0 g g s
27 1.0 g g s
28 1.0 g g s
29 1.0 g g s
30 1.2 g s g 0 0.1
31 1.2 g s g 0 0.1
32 s g g 1
33 s g
34 The e ect of family size on female employment Angrist and Evans (1998), Angrist (2001), Angrist and Pishke (2009). Sample: 1980 US Census Public Use Micro Samples 254,654 married mothers aged with at least 2 children, oldest < 18. Binary outcome: Y 1 = 1 if worked for pay in 1979, Y 1 = 0 otherwise. Explanatory variable: Y 2 = 1 if 3 or more children, Y 2 = 0 if 2 children. Instrumental variables: Z 1 = 1 if rst two children are same-sex, 0 otherwise. Z 2 = 1 if at 2nd birth there are twins, 0 otherwise AC (CeMMAP & UCL) LIE 6/14/13 23 / 32
35 Parameters of interest I consider two models both with the structural equation: Y 1 = 1[ a by 2 < U] U? Z Focus on two parameters. The counterfactual probability: that Y 1 = 1 (work) when Y 2 = 0 (only 2 children) ρ 0 = Y 1 (0) P[ a < U] delivered by the marginal distribution of U. The Average Treatment E ect (ATE), the di erence in counterfactual probabilities: ρ 1 = Y 1 (1) Y 1 (0) P[ a b < U] P[ a < U] AC (CeMMAP & UCL) LIE 6/14/13 24 / 32
36 A complete model Heckman s (Econometrica 1978), point identifying, triangular model Y 1 = 1[ a by 2 < U] U V Y 2 = 1[ c dz < V ] 0 jz N 0 which has (Φ is the standard normal CDF) 1 r, r 1 ρ 0 = Φ(a) ρ 1 = Φ(b + a) Φ(a) ML estimates using the same-sex instrument: ˆρ 0 = (0.011) ˆρ 1 = (0.029) AC (CeMMAP & UCL) LIE 6/14/13 25 / 32
37 Incomplete model The incomplete model has no equation for Y 2, only this: Y 1 = 1[ a by 2 < U] U? Z The same-sex instrument has low predictive power for advancing beyond 2 children. P[Y 2 = 1jsame-sex] = 0.41 P[Y 2 = 1jnot same-sex] = 0.35 AC (CeMMAP & UCL) LIE 6/14/13 26 / 32
38 Incomplete model: identified set. Complete model: MLE 1.0 Same sex instrument 0.5 ρ ρ 0
39 Twins instrument The twins instrument is a much better predictor of advancing beyond 2 children (Y 2 = 1) P[Y 2 = 1jtwins] = 1.00 P[Y 2 = 1jnot twins] = 0.38 I amend Heckman s model to allow for perfect prediction. ML estimates using the twins instrument: ˆρ 0 = (0.007) ˆρ 1 = (0.017) AC (CeMMAP & UCL) LIE 6/14/13 28 / 32
40 Twins instrument The twins instrument is a much better predictor of advancing beyond 2 children (Y 2 = 1) P[Y 2 = 1jtwins] = 1.00 P[Y 2 = 1jnot twins] = 0.38 I amend Heckman s model to allow for perfect prediction. ML estimates using the twins instrument: ˆρ 0 = (0.007) ˆρ 1 = (0.017) The incomplete model point identi es ρ 0 + ρ 1 = Y 1 (1), but neither ρ 0 nor ρ 1 individually. AC (CeMMAP & UCL) LIE 6/14/13 28 / 32
41 Twins instrument The twins instrument is a much better predictor of advancing beyond 2 children (Y 2 = 1) P[Y 2 = 1jtwins] = 1.00 P[Y 2 = 1jnot twins] = 0.38 I amend Heckman s model to allow for perfect prediction. ML estimates using the twins instrument: ˆρ 0 = (0.007) ˆρ 1 = (0.017) The incomplete model point identi es ρ 0 + ρ 1 = Y 1 (1), but neither ρ 0 nor ρ 1 individually. The identi ed set for (ρ 0, ρ 1 ) is a manifold - a one-dimensional line with slope 1 and intercept equal to ρ 0 + ρ 1. AC (CeMMAP & UCL) LIE 6/14/13 28 / 32
42 Incomplete model: identified set. Complete model: MLE 1.0 Twins instrument 0.5 ρ ρ 0
43 Incomplete model: identified set. Complete model: MLE 1.0 Identified set: same sex instrument Identified set: twins instrument 0.5 ρ MLE: same sex instrument MLE: twins instrument ρ 0
44 Remarks Incomplete single equation models are generically partially identifying when unobserved variables are not single valued functions of observed variables. discrete outcomes non-scalar heterogeneity some models characterized by inequality restrictions, some models admitting multiple equilibria Most of the applied work on these sorts of problems uses point identifying complete models, but there are many ways to complete these models and data cannot distinguish one from another. Now we can characterize the sharp identi ed sets identi ed by these incomplete models. Identi ed sets are characterized by systems of moment inequalities. AC (CeMMAP & UCL) LIE 6/14/13 30 / 32
45 Remarks We have particular results for discrete outcome cases. binary and ordered outcome IV models, Chesher (2010, 2013), Chesher and Smolinski (2011, 2012), multiple discrete choice with endogenous explanatory variables. Chesher, Rosen and Smolinski (2013), Research proceeds on cases with continuous endogenous variables. How to select from an uncountable in nity of moment inequalities? Estimation and inference procedures are available - Andrews and Shi (2013), Chernozhukov, Lee and Rosen (2013). Computational challenges remain. AC (CeMMAP & UCL) LIE 6/14/13 31 / 32
46 Remarks An incomplete model s identi ed set contains all the points identi ed by its various completions. So our results provide a tool to: conduct systematic studies of fragility of inference to failure of whimsical assumptions. taking (some of) the con out of econometrics? AC (CeMMAP & UCL) LIE 6/14/13 32 / 32
What do instrumental variable models deliver with discrete dependent variables?
What do instrumental variable models deliver with discrete dependent variables? Andrew Chesher Adam Rosen The Institute for Fiscal Studies Department of Economics, UCL cemmap working paper CWP10/13 What
More informationInstrumental variable models for discrete outcomes
Instrumental variable models for discrete outcomes Andrew Chesher The Institute for Fiscal Studies Department of Economics, UCL cemmap working paper CWP30/08 Instrumental Variable Models for Discrete Outcomes
More informationInstrumental Variable Models for Discrete Outcomes. Andrew Chesher Centre for Microdata Methods and Practice and UCL. Revised November 13th 2008
Instrumental Variable Models for Discrete Outcomes Andrew Chesher Centre for Microdata Methods and Practice and UCL Revised November 13th 2008 Abstract. Single equation instrumental variable models for
More informationLectures on Identi cation 2
Lectures on Identi cation 2 Andrew Chesher CeMMAP & UCL April 16th 2008 Andrew Chesher (CeMMAP & UCL) Identi cation 2 4/16/2008 1 / 28 Topics 1 Monday April 14th. Motivation, history, de nitions, types
More informationAn Instrumental Variable Model of Multiple Discrete Choice
An Instrumental Variable Model of Multiple Discrete Choice Andrew Chesher y UCL and CeMMAP Adam M. Rosen z UCL and CeMMAP February, 20 Konrad Smolinski x UCL and CeMMAP Abstract This paper studies identi
More informationSharp identified sets for discrete variable IV models
Sharp identified sets for discrete variable IV models Andrew Chesher Konrad Smolinski The Institute for Fiscal Studies Department of Economics, UCL cemmap working paper CWP11/10 Sharp identi ed sets for
More informationGeneralized instrumental variable models, methods, and applications
Generalized instrumental variable models, methods, and applications Andrew Chesher Adam M. Rosen The Institute for Fiscal Studies Department of Economics, UCL cemmap working paper CWP43/18 Generalized
More informationCounterfactual worlds
Counterfactual worlds Andrew Chesher Adam Rosen The Institute for Fiscal Studies Department of Economics, UCL cemmap working paper CWP22/15 Counterfactual Worlds Andrew Chesher and Adam M. Rosen CeMMAP
More informationControl Functions in Nonseparable Simultaneous Equations Models 1
Control Functions in Nonseparable Simultaneous Equations Models 1 Richard Blundell 2 UCL & IFS and Rosa L. Matzkin 3 UCLA June 2013 Abstract The control function approach (Heckman and Robb (1985)) in a
More informationEndogeneity and Discrete Outcomes. Andrew Chesher Centre for Microdata Methods and Practice, UCL & IFS. Revised April 2nd 2008
Endogeneity and Discrete Outcomes Andrew Chesher Centre for Microdata Methods and Practice, UCL & IFS Revised April 2nd 2008 Abstract. This paper studies models for discrete outcomes which permit explanatory
More informationEndogeneity and Discrete Outcomes. Andrew Chesher Centre for Microdata Methods and Practice, UCL
Endogeneity and Discrete Outcomes Andrew Chesher Centre for Microdata Methods and Practice, UCL July 5th 2007 Accompanies the presentation Identi cation and Discrete Measurement CeMMAP Launch Conference,
More informationCharacterizations of identified sets delivered by structural econometric models
Characterizations of identified sets delivered by structural econometric models Andrew Chesher Adam M. Rosen The Institute for Fiscal Studies Department of Economics, UCL cemmap working paper CWP44/16
More informationNew Developments in Econometrics Lecture 16: Quantile Estimation
New Developments in Econometrics Lecture 16: Quantile Estimation Jeff Wooldridge Cemmap Lectures, UCL, June 2009 1. Review of Means, Medians, and Quantiles 2. Some Useful Asymptotic Results 3. Quantile
More informationAn instrumental variable random coefficients model for binary outcomes
An instrumental variable random coefficients model for binary outcomes Andrew Chesher Adam M. Rosen The Institute for Fiscal Studies Department of Economics, UCL cemmap working paper CWP34/12 An Instrumental
More information13 Endogeneity and Nonparametric IV
13 Endogeneity and Nonparametric IV 13.1 Nonparametric Endogeneity A nonparametric IV equation is Y i = g (X i ) + e i (1) E (e i j i ) = 0 In this model, some elements of X i are potentially endogenous,
More informationIdenti cation of Positive Treatment E ects in. Randomized Experiments with Non-Compliance
Identi cation of Positive Treatment E ects in Randomized Experiments with Non-Compliance Aleksey Tetenov y February 18, 2012 Abstract I derive sharp nonparametric lower bounds on some parameters of the
More informationA Course in Applied Econometrics Lecture 14: Control Functions and Related Methods. Jeff Wooldridge IRP Lectures, UW Madison, August 2008
A Course in Applied Econometrics Lecture 14: Control Functions and Related Methods Jeff Wooldridge IRP Lectures, UW Madison, August 2008 1. Linear-in-Parameters Models: IV versus Control Functions 2. Correlated
More informationIV Models of Ordered Choice
IV Models of Ordered Choice Andrew Chesher and Konrad Smolinski CeMMAP & UCL December 4th 2009 Abstract This paper studies single equation instrumental variable models of ordered choice in which explanatory
More informationAn instrumental variable model of multiple discrete choice
Quantitative Economics 4 (2013), 157 196 1759-7331/20130157 An instrumental variable model of multiple discrete choice Andrew Chesher Department of Economics, University College London and CeMMAP Adam
More informationUnconditional Quantile Regression with Endogenous Regressors
Unconditional Quantile Regression with Endogenous Regressors Pallab Kumar Ghosh Department of Economics Syracuse University. Email: paghosh@syr.edu Abstract This paper proposes an extension of the Fortin,
More informationPRICES VERSUS PREFERENCES: TASTE CHANGE AND TOBACCO CONSUMPTION
PRICES VERSUS PREFERENCES: TASTE CHANGE AND TOBACCO CONSUMPTION AEA SESSION: REVEALED PREFERENCE THEORY AND APPLICATIONS: RECENT DEVELOPMENTS Abigail Adams (IFS & Oxford) Martin Browning (Oxford & IFS)
More informationCENSORED DATA AND CENSORED NORMAL REGRESSION
CENSORED DATA AND CENSORED NORMAL REGRESSION Data censoring comes in many forms: binary censoring, interval censoring, and top coding are the most common. They all start with an underlying linear model
More informationThe Econometric Evaluation of Policy Design: Part I: Heterogeneity in Program Impacts, Modeling Self-Selection, and Parameters of Interest
The Econometric Evaluation of Policy Design: Part I: Heterogeneity in Program Impacts, Modeling Self-Selection, and Parameters of Interest Edward Vytlacil, Yale University Renmin University, Department
More informationAn Instrumental Variable Approach to Dynamic Models
An Instrumental Variable Approach to Dynamic Models work in progress, comments welcome Steven Berry and Giovanni Compiani Yale University September 26, 2017 1 Introduction Empirical models of dynamic decision
More informationPartial Identi cation in Monotone Binary Models: Discrete Regressors and Interval Data.
Partial Identi cation in Monotone Binary Models: Discrete Regressors and Interval Data. Thierry Magnac Eric Maurin y First version: February 004 This revision: December 006 Abstract We investigate identi
More informationMC3: Econometric Theory and Methods. Course Notes 4
University College London Department of Economics M.Sc. in Economics MC3: Econometric Theory and Methods Course Notes 4 Notes on maximum likelihood methods Andrew Chesher 25/0/2005 Course Notes 4, Andrew
More informationECONOMETRICS II (ECO 2401S) University of Toronto. Department of Economics. Spring 2013 Instructor: Victor Aguirregabiria
ECONOMETRICS II (ECO 2401S) University of Toronto. Department of Economics. Spring 2013 Instructor: Victor Aguirregabiria SOLUTION TO FINAL EXAM Friday, April 12, 2013. From 9:00-12:00 (3 hours) INSTRUCTIONS:
More informationWhat s New in Econometrics? Lecture 14 Quantile Methods
What s New in Econometrics? Lecture 14 Quantile Methods Jeff Wooldridge NBER Summer Institute, 2007 1. Reminders About Means, Medians, and Quantiles 2. Some Useful Asymptotic Results 3. Quantile Regression
More informationInformation Structure and Statistical Information in Discrete Response Models
Information Structure and Statistical Information in Discrete Response Models Shakeeb Khan a and Denis Nekipelov b This version: January 2012 Abstract Strategic interaction parameters characterize the
More informationSensitivity checks for the local average treatment effect
Sensitivity checks for the local average treatment effect Martin Huber March 13, 2014 University of St. Gallen, Dept. of Economics Abstract: The nonparametric identification of the local average treatment
More informationIdentifying Structural E ects in Nonseparable Systems Using Covariates
Identifying Structural E ects in Nonseparable Systems Using Covariates Halbert White UC San Diego Karim Chalak Boston College October 16, 2008 Abstract This paper demonstrates the extensive scope of an
More informationEcon 273B Advanced Econometrics Spring
Econ 273B Advanced Econometrics Spring 2005-6 Aprajit Mahajan email: amahajan@stanford.edu Landau 233 OH: Th 3-5 or by appt. This is a graduate level course in econometrics. The rst part of the course
More informationEstimating the Fractional Response Model with an Endogenous Count Variable
Estimating the Fractional Response Model with an Endogenous Count Variable Estimating FRM with CEEV Hoa Bao Nguyen Minh Cong Nguyen Michigan State Universtiy American University July 2009 Nguyen and Nguyen
More informationEstimation of Dynamic Nonlinear Random E ects Models with Unbalanced Panels.
Estimation of Dynamic Nonlinear Random E ects Models with Unbalanced Panels. Pedro Albarran y Raquel Carrasco z Jesus M. Carro x June 2014 Preliminary and Incomplete Abstract This paper presents and evaluates
More informationSupplementary material to: Tolerating deance? Local average treatment eects without monotonicity.
Supplementary material to: Tolerating deance? Local average treatment eects without monotonicity. Clément de Chaisemartin September 1, 2016 Abstract This paper gathers the supplementary material to de
More informationThe Identi cation Power of Equilibrium in Games: The. Supermodular Case
The Identi cation Power of Equilibrium in Games: The Supermodular Case Francesca Molinari y Cornell University Adam M. Rosen z UCL, CEMMAP, and IFS September 2007 Abstract This paper discusses how the
More informationIdentification in Nonparametric Limited Dependent Variable Models with Simultaneity and Unobserved Heterogeneity
Identification in Nonparametric Limited Dependent Variable Models with Simultaneity and Unobserved Heterogeneity Rosa L. Matzkin 1 Department of Economics University of California, Los Angeles First version:
More informationAPEC 8212: Econometric Analysis II
APEC 8212: Econometric Analysis II Instructor: Paul Glewwe Spring, 2014 Office: 337a Ruttan Hall (formerly Classroom Office Building) Phone: 612-625-0225 E-Mail: pglewwe@umn.edu Class Website: http://faculty.apec.umn.edu/pglewwe/apec8212.html
More informationThe relationship between treatment parameters within a latent variable framework
Economics Letters 66 (2000) 33 39 www.elsevier.com/ locate/ econbase The relationship between treatment parameters within a latent variable framework James J. Heckman *,1, Edward J. Vytlacil 2 Department
More informationPartial Identification of Nonseparable Models using Binary Instruments
Partial Identification of Nonseparable Models using Binary Instruments Takuya Ishihara October 13, 2017 arxiv:1707.04405v2 [stat.me] 12 Oct 2017 Abstract In this study, we eplore the partial identification
More informationEconomics 241B Estimation with Instruments
Economics 241B Estimation with Instruments Measurement Error Measurement error is de ned as the error resulting from the measurement of a variable. At some level, every variable is measured with error.
More informationNBER WORKING PAPER SERIES
NBER WORKING PAPER SERIES IDENTIFICATION OF TREATMENT EFFECTS USING CONTROL FUNCTIONS IN MODELS WITH CONTINUOUS, ENDOGENOUS TREATMENT AND HETEROGENEOUS EFFECTS Jean-Pierre Florens James J. Heckman Costas
More informationIdentification in Triangular Systems using Control Functions
Identification in Triangular Systems using Control Functions Maximilian Kasy Department of Economics, UC Berkeley Maximilian Kasy (UC Berkeley) Control Functions 1 / 19 Introduction Introduction There
More informationMicroeconometrics. C. Hsiao (2014), Analysis of Panel Data, 3rd edition. Cambridge, University Press.
Cheng Hsiao Microeconometrics Required Text: C. Hsiao (2014), Analysis of Panel Data, 3rd edition. Cambridge, University Press. A.C. Cameron and P.K. Trivedi (2005), Microeconometrics, Cambridge University
More informationNew Developments in Econometrics Lecture 11: Difference-in-Differences Estimation
New Developments in Econometrics Lecture 11: Difference-in-Differences Estimation Jeff Wooldridge Cemmap Lectures, UCL, June 2009 1. The Basic Methodology 2. How Should We View Uncertainty in DD Settings?
More informationNonadditive Models with Endogenous Regressors
Nonadditive Models with Endogenous Regressors Guido W. Imbens First Draft: July 2005 This Draft: February 2006 Abstract In the last fifteen years there has been much work on nonparametric identification
More informationIdentification of Instrumental Variable. Correlated Random Coefficients Models
Identification of Instrumental Variable Correlated Random Coefficients Models Matthew A. Masten Alexander Torgovitsky January 18, 2016 Abstract We study identification and estimation of the average partial
More informationNonseparable Unobserved Heterogeneity and Partial Identification in IV models for Count Outcomes
Nonseparable Unobserved Heterogeneity and Partial Identification in IV models for Count Outcomes Dongwoo Kim Department of Economics, University College London [Latest update: March 29, 2017] Abstract
More informationPartial Identification in Triangular Systems of Equations with Binary Dependent Variables
Partial Identification in Triangular Systems of Equations with Binary Deendent Variables Azeem M. Shaikh Deartment of Economics University of Chicago amshaikh@uchicago.edu Edward J. Vytlacil Deartment
More informationIdentifying Multiple Marginal Effects with a Single Instrument
Identifying Multiple Marginal Effects with a Single Instrument Carolina Caetano 1 Juan Carlos Escanciano 2 1 Department of Economics, University of Rochester 2 Department of Economics, Indiana University
More informationDEPARTAMENTO DE ECONOMÍA DOCUMENTO DE TRABAJO. Nonparametric Estimation of Nonadditive Hedonic Models James Heckman, Rosa Matzkin y Lars Nesheim
DEPARTAMENTO DE ECONOMÍA DOCUMENTO DE TRABAJO Nonparametric Estimation of Nonadditive Hedonic Models James Heckman, Rosa Matzkin y Lars Nesheim D.T.: N 5 Junio 2002 Vito Dumas 284, (B644BID) Victoria,
More informationInference in Ordered Response Games with Complete Information.
Inference in Ordered Response Games with Complete Information. Andres Aradillas-Lopez Adam M. Rosen The Institute for Fiscal Studies Department of Economics, UCL cemmap working paper CWP36/14 Inference
More informationNonparametric Identi cation of Regression Models Containing a Misclassi ed Dichotomous Regressor Without Instruments
Nonparametric Identi cation of Regression Models Containing a Misclassi ed Dichotomous Regressor Without Instruments Xiaohong Chen Yale University Yingyao Hu y Johns Hopkins University Arthur Lewbel z
More informationAn instrumental variable random-coefficients model for binary outcomes
Econometrics Journal (2014), volume 17, pp. S1 S19. doi: 10.1111/ectj.12018 An instrumental variable random-coefficients model for binary outcomes ANDREW CHESHER, AND ADAM M. ROSEN, Centre for Microdata
More informationNonparametric Welfare Analysis for Discrete Choice
Nonparametric Welfare Analysis for Discrete Choice Debopam Bhattacharya University of Oxford September 26, 2014. Abstract We consider empirical measurement of exact equivalent/compensating variation resulting
More informationNotes on Generalized Method of Moments Estimation
Notes on Generalized Method of Moments Estimation c Bronwyn H. Hall March 1996 (revised February 1999) 1. Introduction These notes are a non-technical introduction to the method of estimation popularized
More informationCOUNTERFACTUAL MAPPING AND INDIVIDUAL TREATMENT EFFECTS IN NONSEPARABLE MODELS WITH DISCRETE ENDOGENEITY*
COUNTERFACTUAL MAPPING AND INDIVIDUAL TREATMENT EFFECTS IN NONSEPARABLE MODELS WITH DISCRETE ENDOGENEITY* QUANG VUONG AND HAIQING XU ABSTRACT. This paper establishes nonparametric identification of individual
More informationTesting for Regime Switching: A Comment
Testing for Regime Switching: A Comment Andrew V. Carter Department of Statistics University of California, Santa Barbara Douglas G. Steigerwald Department of Economics University of California Santa Barbara
More informationSemiparametric Identification in Panel Data Discrete Response Models
Semiparametric Identification in Panel Data Discrete Response Models Eleni Aristodemou UCL March 8, 2016 Please click here for the latest version. Abstract This paper studies partial identification in
More informationWageningen Summer School in Econometrics. The Bayesian Approach in Theory and Practice
Wageningen Summer School in Econometrics The Bayesian Approach in Theory and Practice September 2008 Slides for Lecture on Qualitative and Limited Dependent Variable Models Gary Koop, University of Strathclyde
More informationECONOMETRICS II (ECO 2401) Victor Aguirregabiria. Spring 2018 TOPIC 4: INTRODUCTION TO THE EVALUATION OF TREATMENT EFFECTS
ECONOMETRICS II (ECO 2401) Victor Aguirregabiria Spring 2018 TOPIC 4: INTRODUCTION TO THE EVALUATION OF TREATMENT EFFECTS 1. Introduction and Notation 2. Randomized treatment 3. Conditional independence
More informationSIMILAR-ON-THE-BOUNDARY TESTS FOR MOMENT INEQUALITIES EXIST, BUT HAVE POOR POWER. Donald W. K. Andrews. August 2011 Revised March 2012
SIMILAR-ON-THE-BOUNDARY TESTS FOR MOMENT INEQUALITIES EXIST, BUT HAVE POOR POWER By Donald W. K. Andrews August 2011 Revised March 2012 COWLES FOUNDATION DISCUSSION PAPER NO. 1815R COWLES FOUNDATION FOR
More informationEconometrics of causal inference. Throughout, we consider the simplest case of a linear outcome equation, and homogeneous
Econometrics of causal inference Throughout, we consider the simplest case of a linear outcome equation, and homogeneous effects: y = βx + ɛ (1) where y is some outcome, x is an explanatory variable, and
More informationGranger Causality and Structural Causality in Cross-Section and Panel Data
Granger Causality and Structural Causality in Cross-Section and anel Data Xun Lu a, Liangjun Su b ; and Halbert White c a Department of Economics, Hong Kong University of Science and Technology b School
More informationECON 594: Lecture #6
ECON 594: Lecture #6 Thomas Lemieux Vancouver School of Economics, UBC May 2018 1 Limited dependent variables: introduction Up to now, we have been implicitly assuming that the dependent variable, y, was
More informationContinuous Treatments
Continuous Treatments Stefan Hoderlein Boston College Yuya Sasaki Brown University First Draft: July 15, 2009 This Draft: January 19, 2011 Abstract This paper explores the relationship between nonseparable
More informationA test of the conditional independence assumption in sample selection models
A test of the conditional independence assumption in sample selection models Martin Huber, Blaise Melly First draft: December 2006, Last changes: September 2012 Abstract: Identi cation in most sample selection
More informationProgram Evaluation with High-Dimensional Data
Program Evaluation with High-Dimensional Data Alexandre Belloni Duke Victor Chernozhukov MIT Iván Fernández-Val BU Christian Hansen Booth ESWC 215 August 17, 215 Introduction Goal is to perform inference
More informationcemmap working paper, Centre for Microdata Methods and Practice, No. CWP11/04
econstor www.econstor.eu Der Open-Access-Publikationsserver der ZBW Leibniz-Informationszentrum Wirtschaft The Open Access Publication Server of the ZBW Leibniz Information Centre for Economics Chesher,
More informationNBER WORKING PAPER SERIES NONPARAMETRIC IDENTIFICATION OF MULTINOMIAL CHOICE DEMAND MODELS WITH HETEROGENEOUS CONSUMERS
NBER WORKING PAPER SERIES NONPARAMETRIC IDENTIFICATION OF MULTINOMIAL CHOICE DEMAND MODELS WITH HETEROGENEOUS CONSUMERS Steven T. Berry Philip A. Haile Working Paper 15276 http://www.nber.org/papers/w15276
More informationInference in ordered response games with complete information
Inference in ordered response games with complete information Andres Aradillas-Lopez Adam Rosen The Institute for Fiscal Studies Department of Economics, UCL cemmap working paper CWP33/13 Inference in
More informationLecture 11 Roy model, MTE, PRTE
Lecture 11 Roy model, MTE, PRTE Economics 2123 George Washington University Instructor: Prof. Ben Williams Roy Model Motivation The standard textbook example of simultaneity is a supply and demand system
More informationRank Estimation of Partially Linear Index Models
Rank Estimation of Partially Linear Index Models Jason Abrevaya University of Texas at Austin Youngki Shin University of Western Ontario October 2008 Preliminary Do not distribute Abstract We consider
More informationConditions for the Existence of Control Functions in Nonseparable Simultaneous Equations Models 1
Conditions for the Existence of Control Functions in Nonseparable Simultaneous Equations Models 1 Richard Blundell UCL and IFS and Rosa L. Matzkin UCLA First version: March 2008 This version: October 2010
More informationComments on: Panel Data Analysis Advantages and Challenges. Manuel Arellano CEMFI, Madrid November 2006
Comments on: Panel Data Analysis Advantages and Challenges Manuel Arellano CEMFI, Madrid November 2006 This paper provides an impressive, yet compact and easily accessible review of the econometric literature
More informationFlexible Estimation of Treatment Effect Parameters
Flexible Estimation of Treatment Effect Parameters Thomas MaCurdy a and Xiaohong Chen b and Han Hong c Introduction Many empirical studies of program evaluations are complicated by the presence of both
More informationEndogenous Semiparametric Binary Choice Models with Heteroscedasticity
Endogenous Semiparametric Binary Choice Models with Heteroscedasticity Stefan Hoderlein Brown University First Draft: February 15, 2006 This Draft: September 19, 2013 Abstract In this paper we consider
More informationThe Physics of Impossible Things Benjamin Schumacher Kenyon College
The Physics of Impossible Things Benjamin Schumacher Kenyon College "I can't believe that!" said Alice. "Can't you?" the Queen said in a pitying tone. "Try again: draw a long breath, and shut your eyes."
More informationGMM-based inference in the AR(1) panel data model for parameter values where local identi cation fails
GMM-based inference in the AR() panel data model for parameter values where local identi cation fails Edith Madsen entre for Applied Microeconometrics (AM) Department of Economics, University of openhagen,
More informationIDENTIFICATION IN DIFFERENTIATED PRODUCTS MARKETS USING MARKET LEVEL DATA. Steven T. Berry and Philip A. Haile. January 2010 Updated February 2010
IDENTIFICATION IN DIFFERENTIATED PRODUCTS MARKETS USING MARKET LEVEL DATA By Steven T. Berry and Philip A. Haile January 2010 Updated February 2010 COWLES FOUNDATION DISCUSSION PAPER NO. 1744 COWLES FOUNDATION
More informationFinal Exam. Economics 835: Econometrics. Fall 2010
Final Exam Economics 835: Econometrics Fall 2010 Please answer the question I ask - no more and no less - and remember that the correct answer is often short and simple. 1 Some short questions a) For each
More informationIntroduction: structural econometrics. Jean-Marc Robin
Introduction: structural econometrics Jean-Marc Robin Abstract 1. Descriptive vs structural models 2. Correlation is not causality a. Simultaneity b. Heterogeneity c. Selectivity Descriptive models Consider
More informationA Note on the Closed-form Identi cation of Regression Models with a Mismeasured Binary Regressor
A Note on the Closed-form Identi cation of Regression Models with a Mismeasured Binary Regressor Xiaohong Chen Yale University Yingyao Hu y Johns Hopkins University Arthur Lewbel z Boston College First
More informationUniversity of Toronto Department of Economics. Testing Local Average Treatment Effect Assumptions
University of Toronto Department of Economics Working Paper 514 Testing Local Average Treatment Effect Assumptions By Ismael Mourifie and Yuanyuan Wan July 7, 214 TESTING LATE ASSUMPTIONS ISMAEL MOURIFIÉ
More informationIdentification of Regression Models with Misclassified and Endogenous Binary Regressor
Identification of Regression Models with Misclassified and Endogenous Binary Regressor A Celebration of Peter Phillips Fourty Years at Yale Conference Hiroyuki Kasahara 1 Katsumi Shimotsu 2 1 Department
More informationSIMILAR-ON-THE-BOUNDARY TESTS FOR MOMENT INEQUALITIES EXIST, BUT HAVE POOR POWER. Donald W. K. Andrews. August 2011
SIMILAR-ON-THE-BOUNDARY TESTS FOR MOMENT INEQUALITIES EXIST, BUT HAVE POOR POWER By Donald W. K. Andrews August 2011 COWLES FOUNDATION DISCUSSION PAPER NO. 1815 COWLES FOUNDATION FOR RESEARCH IN ECONOMICS
More informationESTIMATION OF NONPARAMETRIC MODELS WITH SIMULTANEITY
ESTIMATION OF NONPARAMETRIC MODELS WITH SIMULTANEITY Rosa L. Matzkin Department of Economics University of California, Los Angeles First version: May 200 This version: August 204 Abstract We introduce
More informationQUANTILE MODELS WITH ENDOGENEITY
QUANTILE MODELS WITH ENDOGENEITY V. CHERNOZHUKOV AND C. HANSEN Abstract. In this article, we review quantile models with endogeneity. We focus on models that achieve identification through the use of instrumental
More informationEmpirical Welfare Analysis for Discrete Choice: Some New Results
Empirical Welfare Analysis for Discrete Choice Some New Results Debopam Bhattacharya University of Cambridge September 1, 016. Abstract We develop methods of empirical welfare-analysis in multinomial choice
More informationEconometric Analysis of Games 1
Econometric Analysis of Games 1 HT 2017 Recap Aim: provide an introduction to incomplete models and partial identification in the context of discrete games 1. Coherence & Completeness 2. Basic Framework
More informationIdentifying Effects of Multivalued Treatments
Identifying Effects of Multivalued Treatments Sokbae Lee Bernard Salanie The Institute for Fiscal Studies Department of Economics, UCL cemmap working paper CWP72/15 Identifying Effects of Multivalued Treatments
More informationInference in Ordered Response Games with Complete Information.
Inference in Ordered Response Games with Complete Information. Andres Aradillas-Lopez Pennsylvania State University Adam M. Rosen Duke University and CeMMAP October 18, 2016 Abstract We study econometric
More informationSemi and Nonparametric Models in Econometrics
Semi and Nonparametric Models in Econometrics Part 4: partial identification Xavier d Haultfoeuille CREST-INSEE Outline Introduction First examples: missing data Second example: incomplete models Inference
More informationEmpirical Methods in Applied Microeconomics
Empirical Methods in Applied Microeconomics Jörn-Ste en Pischke LSE November 2007 1 Nonlinearity and Heterogeneity We have so far concentrated on the estimation of treatment e ects when the treatment e
More informationIDENTIFICATION IN DIFFERENTIATED PRODUCTS MARKETS USING MARKET LEVEL DATA. Steven T. Berry and Philip A. Haile. January 2010 Revised May 2012
IDENTIFICATION IN DIFFERENTIATED PRODUCTS MARKETS USING MARKET LEVEL DATA By Steven T. Berry and Philip A. Haile January 2010 Revised May 2012 COWLES FOUNDATION DISCUSSION PAPER NO. 1744R COWLES FOUNDATION
More informationECONOMETRICS FIELD EXAM Michigan State University May 9, 2008
ECONOMETRICS FIELD EXAM Michigan State University May 9, 2008 Instructions: Answer all four (4) questions. Point totals for each question are given in parenthesis; there are 00 points possible. Within
More informationThe Identification Zoo - Meanings of Identification in Econometrics: PART 2
The Identification Zoo - Meanings of Identification in Econometrics: PART 2 Arthur Lewbel Boston College heavily revised 2017 Lewbel (Boston College) Identification Zoo 2017 1 / 80 The Identification Zoo
More informationThe Identification Zoo - Meanings of Identification in Econometrics: PART 3
The Identification Zoo - Meanings of Identification in Econometrics: PART 3 Arthur Lewbel Boston College original 2015, heavily revised 2018 Lewbel (Boston College) Identification Zoo 2018 1 / 85 The Identification
More informationA Course in Applied Econometrics Lecture 4: Linear Panel Data Models, II. Jeff Wooldridge IRP Lectures, UW Madison, August 2008
A Course in Applied Econometrics Lecture 4: Linear Panel Data Models, II Jeff Wooldridge IRP Lectures, UW Madison, August 2008 5. Estimating Production Functions Using Proxy Variables 6. Pseudo Panels
More informationLecture 4: Linear panel models
Lecture 4: Linear panel models Luc Behaghel PSE February 2009 Luc Behaghel (PSE) Lecture 4 February 2009 1 / 47 Introduction Panel = repeated observations of the same individuals (e.g., rms, workers, countries)
More information