Discrete Choice Modeling William Greene Stern School of Business New York University.

Slides:



Advertisements
Similar presentations
7. Models for Count Data, Inflation Models. Models for Count Data.
Advertisements

Econometrics I Professor William Greene Stern School of Business
Discrete Choice Modeling William Greene Stern School of Business New York University Lab Sessions.
Empirical Methods for Microeconomic Applications University of Lugano, Switzerland May 27-31, 2013 William Greene Department of Economics Stern School.
3. Binary Choice – Inference. Hypothesis Testing in Binary Choice Models.
12. Random Parameters Logit Models. Random Parameters Model Allow model parameters as well as constants to be random Allow multiple observations with.
[Part 1] 1/15 Discrete Choice Modeling Econometric Methodology Discrete Choice Modeling William Greene Stern School of Business New York University 0Introduction.
Discrete Choice Modeling William Greene Stern School of Business New York University Lab Sessions.
Empirical Methods for Microeconomic Applications William Greene Department of Economics Stern School of Business.
Econometrics I Professor William Greene Stern School of Business
Part 12: Random Parameters [ 1/46] Econometric Analysis of Panel Data William Greene Department of Economics Stern School of Business.
Discrete Choice Modeling William Greene Stern School of Business New York University Lab Sessions.
1/62: Topic 2.3 – Panel Data Binary Choice Models Microeconometric Modeling William Greene Stern School of Business New York University New York NY USA.
Part 23: Simulation Based Estimation 23-1/26 Econometrics I Professor William Greene Stern School of Business Department of Economics.
8. Heterogeneity: Latent Class Models. Latent Classes A population contains a mixture of individuals of different types (classes) Common form of the.
Discrete Choice Modeling William Greene Stern School of Business New York University Lab Sessions.
Discrete Choice Modeling William Greene Stern School of Business New York University.
2. Binary Choice Estimation. Modeling Binary Choice.
Econometric Methodology. The Sample and Measurement Population Measurement Theory Characteristics Behavior Patterns Choices.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Empirical Methods for Microeconomic Applications William Greene Department of Economics Stern School of Business.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Empirical Methods for Microeconomic Applications University of Lugano, Switzerland May 27-31, 2013 William Greene Department of Economics Stern School.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Empirical Methods for Microeconomic Applications University of Lugano, Switzerland May 27-31, 2013 William Greene Department of Economics Stern School.
[Part 4] 1/43 Discrete Choice Modeling Bivariate & Multivariate Probit Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
[Topic 9-Latent Class Models] 1/66 9. Heterogeneity: Latent Class Models.
1/62: Topic 2.3 – Panel Data Binary Choice Models Microeconometric Modeling William Greene Stern School of Business New York University New York NY USA.
1/68: Topic 4.2 – Latent Class Models Microeconometric Modeling William Greene Stern School of Business New York University New York NY USA William Greene.
Discrete Choice Modeling William Greene Stern School of Business New York University.
1/53: Topic 3.1 – Models for Ordered Choices Microeconometric Modeling William Greene Stern School of Business New York University New York NY USA William.
[Part 2] 1/86 Discrete Choice Modeling Binary Choice Models Discrete Choice Modeling William Greene Stern School of Business New York University 0Introduction.
6. Ordered Choice Models. Ordered Choices Ordered Discrete Outcomes E.g.: Taste test, credit rating, course grade, preference scale Underlying random.
[Part 5] 1/43 Discrete Choice Modeling Ordered Choice Models Discrete Choice Modeling William Greene Stern School of Business New York University 0Introduction.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Applied Econometrics William Greene Department of Economics Stern School of Business.
1/26: Topic 2.2 – Nonlinear Panel Data Models Microeconometric Modeling William Greene Stern School of Business New York University New York NY USA William.
5. Extensions of Binary Choice Models
Microeconometric Modeling
Microeconometric Modeling
William Greene Stern School of Business New York University
William Greene Stern School of Business New York University
Discrete Choice Modeling
Discrete Choice Modeling
Discrete Choice Modeling
Microeconometric Modeling
Microeconometric Modeling
Microeconometric Modeling
Discrete Choice Modeling
Microeconometric Modeling
Microeconometric Modeling
Microeconometric Modeling
Microeconometric Modeling
Microeconometric Modeling
Econometric Analysis of Panel Data
Discrete Choice Modeling
Microeconometric Modeling
Microeconometric Modeling
Econometrics I Professor William Greene Stern School of Business
Microeconometric Modeling
Empirical Methods for Microeconomic Applications University of Lugano, Switzerland May 27-31, 2019 William Greene Department of Economics Stern School.
Empirical Methods for Microeconomic Applications University of Lugano, Switzerland May 27-31, 2019 William Greene Department of Economics Stern School.
Econometrics I Professor William Greene Stern School of Business
Empirical Methods for Microeconomic Applications
Presentation transcript:

Discrete Choice Modeling William Greene Stern School of Business New York University

Part 8 Models for Count Data

Application: Major Derogatory Reports AmEx Credit Card Holders N = 13,777 Number of major derogatory reports in 1 year Issues: Nonrandom selection Excess zeros

Histogram for MAJORDRG NOBS= 1310, Too low: 0, Too high: 0 Bin Lower limit Upper limit Frequency Cumulative Frequency ======================================================================== (.8038) 1053(.8038) (.1038) 1189(.9076) (.0382) 1239(.9458) (.0183) 1263(.9641) (.0130) 1280(.9771) (.0076) 1290(.9847) (.0038) 1295(.9885) (.0046) 1301(.9931) (.0000) 1301(.9931) (.0015) 1303(.9947) (.0008) 1304(.9954) (.0031) 1308(.9985) (.0008) 1309(.9992) (.0000) 1309(.9992) (.0008) 1310(1.0000) Histogram for Credit Data

Doctor Visits

Basic Model for Counts of Events  E.g., Visits to site, number of purchases, number of doctor visits  Regression approach Quantitative outcome measured Discrete variable, model probabilities  Poisson probabilities – “loglinear model”

Poisson Model for Doctor Visits Poisson Regression Dependent variable DOCVIS Log likelihood function Restricted log likelihood Chi squared [ 6 d.f.] Significance level McFadden Pseudo R-squared Estimation based on N = 27326, K = 7 Information Criteria: Normalization=1/N Normalized Unnormalized AIC Chi- squared = RsqP=.0818 G - squared = RsqD=.0601 Overdispersion tests: g=mu(i) : Overdispersion tests: g=mu(i)^2: Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X Constant|.77267*** AGE|.01763*** EDUC| *** FEMALE|.29287*** MARRIED| HHNINC| *** HHKIDS| ***

Partial Effects Partial derivatives of expected val. with respect to the vector of characteristics. Effects are averaged over individuals. Observations used for means are All Obs. Conditional Mean at Sample Point Scale Factor for Marginal Effects Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X AGE|.05613*** EDUC| *** FEMALE|.93237*** MARRIED| HHNINC| *** HHKIDS| ***

Poisson Model Specification Issues  Equi-Dispersion: Var[y i |x i ] = E[y i |x i ].  Overdispersion: If i = exp[  ’x i + ε i ], E[y i |x i ] = γexp[  ’x i ] Var[y i ] > E[y i ] (overdispersed) ε i ~ log-Gamma  Negative binomial model ε i ~ Normal[0,  2 ]  Normal-mixture model ε i is viewed as unobserved heterogeneity (“frailty”).  Normal model may be more natural.  Estimation is a bit more complicated.

Poisson Model for Doctor Visits Poisson Regression Dependent variable DOCVIS Log likelihood function Restricted log likelihood Chi squared [ 6 d.f.] Significance level McFadden Pseudo R-squared Estimation based on N = 27326, K = 7 Information Criteria: Normalization=1/N Normalized Unnormalized AIC Chi- squared = RsqP=.0818 G - squared = RsqD=.0601 Overdispersion tests: g=mu(i) : Overdispersion tests: g=mu(i)^2: Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X Constant|.77267*** AGE|.01763*** EDUC| *** FEMALE|.29287*** MARRIED| HHNINC| *** HHKIDS| ***

Alternative Covariance Matrices Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X | Standard – Negative Inverse of Second Derivatives Constant|.77267*** AGE|.01763*** EDUC| *** FEMALE|.29287*** MARRIED| HHNINC| *** HHKIDS| *** | Robust – Sandwich Constant|.77267*** AGE|.01763*** EDUC| *** FEMALE|.29287*** MARRIED| HHNINC| *** HHKIDS| *** | Cluster Correction Constant|.77267*** AGE|.01763*** EDUC| *** FEMALE|.29287*** MARRIED| HHNINC| *** HHKIDS| ***

Overdispersion  In the Poisson model, Var[y|x]=E[y|x]  Equidispersion is a strong assumption  Negbin II: Var[y|x]=E[y|x] +  2 E[y|x] 2  How does overdispersion arise: NonPoissonness Omitted Heterogeneity

Negative Binomial Regression

Negative Binomial Specification  Prob(Y i =j|x i ) has greater mass to the right and left of the mean  Conditional mean function is the same as the Poisson: E[y i |x i ] = λ i =Exp(  ’x i ), so marginal effects have the same form.  Variance is Var[y i |x i ] = λ i (1 + α λ i ), α is the overdispersion parameter; α = 0 reverts to the Poisson.  Poisson is consistent when NegBin is appropriate. Therefore, this is a case for the ROBUST covariance matrix estimator. (Neglected heterogeneity that is uncorrelated with x i.)

Testing for Overdispersion  Regression based test: Regress (y-mean) 2 on mean  Neyman – Pearson tests in NegBin regression | Overdispersion tests: g=mu(i) : | | Overdispersion tests: g=mu(i)^2: | Dispersion parameter for count data model Alpha

NegBin Model for Doctor Visits Negative Binomial Regression Dependent variable DOCVIS Log likelihood function NegBin LogL Restricted log likelihood Poisson LogL Chi squared [ 1 d.f.] Reject Poisson model Significance level McFadden Pseudo R-squared Estimation based on N = 27326, K = 8 Information Criteria: Normalization=1/N Normalized Unnormalized AIC NegBin form 2; Psi(i) = theta Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X Constant|.80825*** AGE|.01806*** EDUC| *** FEMALE|.32596*** MARRIED| HHNINC| *** HHKIDS| *** |Dispersion parameter for count data model Alpha| ***

Partial Effects Should Be the Same Scale Factor for Marginal Effects POISSON Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X AGE|.05613*** EDUC| *** FEMALE|.93237*** MARRIED| HHNINC| *** HHKIDS| *** Scale Factor for Marginal Effects NEGATIVE BINOMIAL AGE|.05767*** EDUC| *** FEMALE| *** MARRIED| HHNINC| *** HHKIDS| ***

Model Formulations for Negative Binomial E[y i |x i ]=λ i

NegBin-1 Model Negative Binomial Regression Dependent variable DOCVIS Log likelihood function Restricted log likelihood NegBin form 1; Psi(i) = theta*exp[bx(i)] Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X Constant|.62584*** AGE|.01428*** EDUC| *** FEMALE|.33028*** MARRIED|.04324** HHNINC| *** HHKIDS| *** |Dispersion parameter for count data model Alpha| ***

NegBin-P Model Negative Binomial (P) Model Dependent variable DOCVIS Log likelihood function Restricted log likelihood Chi squared [ 1 d.f.] Variable| Coefficient Standard Error b/St.Er Constant|.60840*** AGE|.01710*** EDUC| *** FEMALE|.36386*** MARRIED|.03670* HHNINC| *** HHKIDS| *** |Dispersion parameter for count data model Alpha| *** |Negative Binomial. General form, NegBin P P| *** NB-2 NB-1 Poisson

Partial Effects for Different Models Scale Factor for Marginal Effects POISSON Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X AGE|.05613*** EDUC| *** FEMALE|.93237*** MARRIED| HHNINC| *** HHKIDS| *** Scale Factor for Marginal Effects NEGATIVE BINOMIAL - 2 AGE|.05767*** EDUC| *** FEMALE| *** MARRIED| HHNINC| *** HHKIDS| *** Scale Factor for Marginal Effects NEGATIVE BINOMIAL - P AGE|.05143*** EDUC| *** FEMALE| *** MARRIED|.11038* HHNINC| *** HHKIDS| ***

Zero Inflation – ZIP Models  Two regimes: (Recreation site visits) Zero (with probability 1). (Never visit site) Poisson with Pr(0) = exp[-  ’x i ]. (Number of visits, including zero visits this season.)  Unconditional: Pr[0] = P(regime 0) + P(regime 1)*Pr[0|regime 1] Pr[j | j >0] = P(regime 1)*Pr[j|regime 1]  “Two inflation” – Number of children  These are “latent class models”

Two Forms of Zero Inflation Models

Notes on Zero Inflation Models  Poisson is not nested in ZIP. tau = 0 in ZIP(tau) or γ = 0 in ZIP does not produce Poisson; it produces ZIP with P(regime 0) = ½. Standard tests are not appropriate Use Vuong statistic. ZIP model almost always wins.  Zero Inflation models extend to NB models – ZINB(tau) and ZINB are standard models Creates two sources of overdispersion Generally difficult to estimate

ZIP(τ) Model Zero Altered Poisson Regression Model Logistic distribution used for splitting model. ZAP term in probability is F[tau x ln LAMBDA] Comparison of estimated models Pr[0|means] Number of zeros Log-likelihood Poisson Act.= Prd.= Z.I.Poisson Act.= Prd.= Note, the ZIP log-likelihood is not directly comparable. ZIP model with nonzero Q does not encompass the others. Vuong statistic for testing ZIP vs. unaltered model is Distributed as standard normal. A value greater than favors the zero altered Z.I.Poisson model. A value less than rejects the ZIP model Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X |Poisson/NB/Gamma regression model Constant| *** AGE|.01140*** EDUC| *** FEMALE|.13129*** MARRIED| *** HHNINC| *** HHKIDS| *** |Zero inflation model Tau| ***

ZIP Model Zero Altered Poisson Regression Model Logistic distribution used for splitting model. ZAP term in probability is F[tau x Z(i) ] Comparison of estimated models Pr[0|means] Number of zeros Log-likelihood Poisson Act.= Prd.= Z.I.Poisson Act.= Prd.= Vuong statistic for testing ZIP vs. unaltered model is Distributed as standard normal. A value greater than favors the zero altered Z.I.Poisson model. A value less than rejects the ZIP model Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X |Poisson/NB/Gamma regression model Constant| *** AGE|.01100*** EDUC| *** FEMALE|.10943*** MARRIED| *** HHNINC| *** HHKIDS| *** |Zero inflation model Constant| *** FEMALE| *** EDUC|.04114***

Partial Effects for Different Models Scale Factor for Marginal Effects POISSON Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X AGE|.05613*** EDUC| *** FEMALE|.93237*** MARRIED| HHNINC| *** HHKIDS| *** Scale Factor for Marginal Effects NEGATIVE BINOMIAL - 2 AGE|.05767*** EDUC| *** FEMALE| *** MARRIED| HHNINC| *** HHKIDS| *** Scale Factor for Marginal Effects ZERO INFLATED POISSON AGE|.03427*** EDUC| *** FEMALE|.97958*** MARRIED| *** HHNINC| *** HHKIDS| ***

The Vuong Statistic for Nonnested Models

A Hurdle Model  Two part model: Model 1: Probability model for more than zero occurrences Model 2: Model for number of occurrences given that the number is greater than zero.  Applications common in health economics Usage of health care facilities Use of drugs, alcohol, etc.

Hurdle Model

Hurdle Model for Doctor Visits Poisson hurdle model for counts Dependent variable DOCVIS Log likelihood function Restricted log likelihood Chi squared [ 1 d.f.] Significance level McFadden Pseudo R-squared Estimation based on N = 27326, K = 10 LOGIT hurdle equation Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X |Parameters of count model equation Constant| *** AGE|.01088*** EDUC| *** FEMALE|.10244*** MARRIED| *** HHNINC| *** HHKIDS| *** |Parameters of binary hurdle equation Constant|.77475*** FEMALE|.59389*** EDUC| ***

Partial Effects Partial derivatives of expected val. with respect to the vector of characteristics. Effects are averaged over individuals. Observations used for means are All Obs. Conditional Mean at Sample Point.0109 Scale Factor for Marginal Effects Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Mean of X |Effects in Count Model Equation Constant| AGE| EDUC| FEMALE| MARRIED| HHNINC| HHKIDS| |Effects in Binary Hurdle Equation Constant|.86178*** FEMALE|.66060*** EDUC| *** |Combined effect is the sum of the two parts Constant| * EDUC| *** FEMALE|.96915***

Panel Data Models Heterogeneity; λ it = exp(β’x it + c i ) Fixed Effects  Poisson: Standard, no incidental parameters issue  NB  Hausman, Hall, Griliches (1984) put FE in variance, not the mean  Use “brute force” to get a conventional FE model Random Effects  Poisson  Log-gamma heterogeneity becomes an NB model  Contemporary treatments are using normal heterogeneity with simulation or quadrature based estimators  NB with random effects is equivalent to two “effects” one time varying one time invariant. The model is probably overspecified Random Parameters: Mixed models, latent class models, hiererchical – all extended to Poisson and NB

A Peculiarity of the FENB Model  ‘True’ FE model has λ i =exp(α i +x it ’β). Cannot be fit if there are time invariant variables.  Hausman, Hall and Griliches (Econometrica, 1984) has α i appearing in θ. Produces different results Implies that the FEM can contain time invariant variables.

Also: Allison and Waterman (2002), Greene, Econometric Analysis (2011)

| Panel Model with Group Effects | | Log likelihood function | Hausman et al. version. | Unbalanced panel has 7293 individuals. | FENB turns into a logit | Neg.Binomial Regression -- Fixed Effects | model |Variable| Coefficient | Standard Error |b/St.Er.|P[|Z|>z]| Mean of X| HHNINC | EDUC | HSAT | | FIXED EFFECTS NegBin Model | | Log likelihood function | ‘True’ FE model. Estimated | Bypassed 1153 groups with inestimable a(i). | by ‘brute force.’ | Negative binomial regression model | |Variable| Coefficient | Standard Error |b/St.Er.|P[|Z|>z]| Mean of X| Index function for probability HHNINC | EDUC | HSAT | Overdispersion parameter Alpha |