Health Econometrics

Using Stata

Partha Deb
Hunter College, CUNY and NBER

Edward C. Norton
University of Michigan and NBER

Willard G. Manning
University of Chicago

A Stata Press Publication

StataCorp LLC
College Station, Texas
 ®
Copyright c 2017 StataCorp LLC
All rights reserved. First edition 2017

Published by Stata Press, 4905 Lakeway Drive, College Station, Texas 77845
Typeset in LATEX 2ε
Printed in the United States of America
10 9 8 7 6 5 4 3 2 1

Print ISBN-10: 1-59718-228-1

Print ISBN-13: 978-1-59718-228-7
ePub ISBN-10: 1-59718-229-X
ePub ISBN-13: 978-1-59718-229-4
Mobi ISBN-10: 1-59718-230-3
Mobi ISBN-13: 978-1-59718-230-0

Library of Congress Control Number: 2016960172

No part of this book may be reproduced, stored in a retrieval system, or transcribed, in any
form or by any means—electronic, mechanical, photocopy, recording, or otherwise—without
the prior written permission of StataCorp LLC.
Stata, , Stata Press, Mata, , and NetCourse are registered trademarks of
StataCorp LLC.
Stata and Stata Press are registered trademarks with the World Intellectual Property Organi-
zation of the United Nations.
NetCourseNow is a trademark of StataCorp LLC.
LATEX 2ε is a trademark of the American Mathematical Society.
 Dedication to
Willard G. Manning, Jr. (1946–2014)

Will Manning joined the RAND Corporation in 1975, a few years after completing
his PhD at Stanford. He quickly became involved in the RAND Health Insurance Ex-
periment. Will was the lead author of the article that reported the main insurance
results in the 1987 American Economic Review, one of the most cited and influential
articles in health economics. He also published many seminal articles about demand for
alcohol and cigarettes, the taxes of sin, and mental healthcare. In 2010, the American
Society of Health Economics awarded the Victor R. Fuchs Award to Will for his lifetime
contributions to the field of health economics.
But perhaps his strongest influence was on empirical methods central to applied
health economics research. With others at RAND, he advocated moving away from
tobit and sample-selection models to deal with distributions of dependent variables
that had a large mass at zero. The two-part model, in all of its forms, is now the
dominant model for healthcare expenditures and use. He also understood the power and
limitations of taking logarithms of skewed distributions. And woe to the author who did
not deal adequately with heteroskedasticity upon retransformation. Will continued to
push the field of health econometrics through the end of his career. He helped develop
new methods and advocated the work of others who found better ways of modeling
healthcare expenditures and use. His influence on applied health economics is deep and
lasting (Konetzka 2015; Mullahy 2015).
Will had three other characteristics that we grew to appreciate, if not emulate,
over the years. He was absolutely meticulous about research—data, methods, and
attribution. Precision is not merely an abstract statistical concept but an essential part
of all steps in a research project.
Will was extraordinarily generous with his time. We know of many junior economists
who were amazed and profoundly grateful that Will took the time to give detailed
feedback on their article or presentation—and to explain why their standard errors
were all wrong.
Finally, Will was hilariously funny. We know this is a rare trait in an economist, but
a little levity makes the daily give and take in search of truth that much more enjoyable.
We dedicate this book to our friend and colleague, Will Manning.

Partha Deb and Edward C. Norton

Contents
List of tables xiii
List of figures xv
Preface xvii
Notation and typography xix
1 Introduction 1
1.1 Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Themes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Health econometric myths . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 Stata friendly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.5 A useful way forward . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2 Framework 7
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Potential outcomes and treatment effects . . . . . . . . . . . . . . . . 9
2.3 Estimating ATEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3.1 A laboratory experiment . . . . . . . . . . . . . . . . . . . . 11
2.3.2 Randomization . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3.3 Covariate adjustment . . . . . . . . . . . . . . . . . . . . . . 12
2.4 Regression estimates of treatment effects . . . . . . . . . . . . . . . . 13
2.4.1 Linear regression . . . . . . . . . . . . . . . . . . . . . . . . 13
2.4.2 Nonlinear regression . . . . . . . . . . . . . . . . . . . . . . 15
2.5 Incremental and marginal effects . . . . . . . . . . . . . . . . . . . . 16
2.6 Model selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.6.1 In-sample model selection . . . . . . . . . . . . . . . . . . . 19
2.6.2 Cross-validation . . . . . . . . . . . . . . . . . . . . . . . . . 20
viii Contents

2.7 Other issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3 MEPS data 23
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2 Overview of all variables . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Expenditure and use variables . . . . . . . . . . . . . . . . . . . . . . 26
3.4 Explanatory variables . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.5 Sample dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.6 Stata resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4 The linear regression model: Specification and checks 33
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.2 The linear regression model . . . . . . . . . . . . . . . . . . . . . . . 34
4.3 Marginal, incremental, and treatment effects . . . . . . . . . . . . . . 35
4.3.1 Marginal and incremental effects . . . . . . . . . . . . . . . 36
4.3.2 Graphical representation of marginal and incremental effects 38
4.3.3 Treatment effects . . . . . . . . . . . . . . . . . . . . . . . . 41
4.4 Consequences of misspecification . . . . . . . . . . . . . . . . . . . . 47
4.4.1 Example: A quadratic specification . . . . . . . . . . . . . . 47
4.4.2 Example: An exponential specification . . . . . . . . . . . . 50
4.5 Visual checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.5.1 Artificial-data example of visual checks . . . . . . . . . . . . 52
4.5.2 MEPS example of visual checks . . . . . . . . . . . . . . . . 55
4.6 Statistical tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.6.1 Pregibon’s link test . . . . . . . . . . . . . . . . . . . . . . . 58
4.6.2 Ramsey’s RESET test . . . . . . . . . . . . . . . . . . . . . 59
4.6.3 Modified Hosmer–Lemeshow test . . . . . . . . . . . . . . . 59
4.6.4 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.6.5 Model selection using AIC and BIC . . . . . . . . . . . . . . 66
4.7 Stata resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5 Generalized linear models 71
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Contents ix

5.2 GLM framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

5.2.1 GLM assumptions . . . . . . . . . . . . . . . . . . . . . . . . 73
5.2.2 Parameter estimation . . . . . . . . . . . . . . . . . . . . . . 75
5.3 GLM examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.4 GLM predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.5 GLM example with interaction term . . . . . . . . . . . . . . . . . . 78
5.6 Marginal and incremental effects . . . . . . . . . . . . . . . . . . . . 81
5.7 Example of marginal and incremental effects . . . . . . . . . . . . . . 82
5.8 Choice of link function and distribution family . . . . . . . . . . . . 85
5.8.1 AIC and BIC . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.8.2 Test for the link function . . . . . . . . . . . . . . . . . . . . 87
5.8.3 Modified Park test for the distribution family . . . . . . . . 89
5.8.4 Extended GLM . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.9 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.10 Stata resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
6 Log and Box–Cox models 93
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
6.2 Log models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
6.2.1 Log model estimation and interpretation . . . . . . . . . . . 94
6.3 Retransformation from ln(y) to raw scale . . . . . . . . . . . . . . . 96
6.3.1 Error retransformation and model predictions . . . . . . . . 96
6.3.2 Marginal and incremental effects . . . . . . . . . . . . . . . 99
6.4 Comparison of log models to GLM . . . . . . . . . . . . . . . . . . . 99
6.5 Box–Cox models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
6.5.1 Box–Cox example . . . . . . . . . . . . . . . . . . . . . . . . 101
6.6 Stata resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
7 Models for continuous outcomes with mass at zero 105
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
7.2 Two-part models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
7.2.1 Expected values and marginal and incremental effects . . . . 108
x Contents

7.3 Generalized tobit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

7.3.1 Full-information maximum likelihood and limited-infor-
mation
maximum likelihood . . . . . . . . . . . . . . . . . . . . . . 110
7.4 Comparison of two-part and generalized tobit models . . . . . . . . . 111
7.4.1 Examples that show similarity of marginal effects . . . . . . 112
7.5 Interpretation and marginal effects . . . . . . . . . . . . . . . . . . . 116
7.5.1 Two-part model example . . . . . . . . . . . . . . . . . . . . 116
7.5.2 Two-part model marginal effects . . . . . . . . . . . . . . . 118
7.5.3 Two-part model marginal effects example . . . . . . . . . . 120
7.5.4 Generalized tobit interpretation . . . . . . . . . . . . . . . . 122
7.5.5 Generalized tobit example . . . . . . . . . . . . . . . . . . . 122
7.6 Single-index models that accommodate zeros . . . . . . . . . . . . . 128
7.6.1 The tobit model . . . . . . . . . . . . . . . . . . . . . . . . . 128
7.6.2 Why tobit is used sparingly . . . . . . . . . . . . . . . . . . 130
7.6.3 One-part models . . . . . . . . . . . . . . . . . . . . . . . . 131
7.7 Statistical tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
7.8 Stata resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
8 Count models 133
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
8.2 Poisson regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.2.1 Poisson MLE . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.2.2 Robustness of the Poisson regression . . . . . . . . . . . . . 138
8.2.3 Interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . 139
8.2.4 Is Poisson too restrictive? . . . . . . . . . . . . . . . . . . . 142
8.3 Negative binomial models . . . . . . . . . . . . . . . . . . . . . . . . 144
8.3.1 Examples of negative binomial models . . . . . . . . . . . . 146
8.4 Hurdle and zero-inflated count models . . . . . . . . . . . . . . . . . 149
8.4.1 Hurdle count models . . . . . . . . . . . . . . . . . . . . . . 149
8.4.2 Zero-inflated models . . . . . . . . . . . . . . . . . . . . . . 155
Contents xi

8.5 Truncation and censoring . . . . . . . . . . . . . . . . . . . . . . . . 158

8.5.1 Truncation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
8.5.2 Censoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
8.6 Model comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
8.6.1 Model selection . . . . . . . . . . . . . . . . . . . . . . . . . 160
8.6.2 Cross-validation . . . . . . . . . . . . . . . . . . . . . . . . . 161
8.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
8.8 Stata resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
9 Models for heterogeneous effects 165
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
9.2 Quantile regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
9.2.1 MEPS examples . . . . . . . . . . . . . . . . . . . . . . . . . 167
9.2.2 Extensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
9.3 Finite mixture models . . . . . . . . . . . . . . . . . . . . . . . . . . 173
9.3.1 MEPS example of healthcare expenditures . . . . . . . . . . 176
9.3.2 MEPS example of healthcare use . . . . . . . . . . . . . . . 186
9.4 Nonparametric regression . . . . . . . . . . . . . . . . . . . . . . . . 192
9.4.1 MEPS examples . . . . . . . . . . . . . . . . . . . . . . . . . 193
9.5 Conditional density estimator . . . . . . . . . . . . . . . . . . . . . . 198
9.6 Stata resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
10 Endogeneity 201
10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
10.2 Endogeneity in linear models . . . . . . . . . . . . . . . . . . . . . . 202
10.2.1 OLS is inconsistent . . . . . . . . . . . . . . . . . . . . . . . 202
10.2.2 2SLS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
10.2.3 Specification tests . . . . . . . . . . . . . . . . . . . . . . . . 206
10.2.4 2SRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
10.2.5 Modeling endogeneity with ERM . . . . . . . . . . . . . . . 209
10.3 Endogeneity with a binary endogenous variable . . . . . . . . . . . . 210
10.3.1 Additional considerations . . . . . . . . . . . . . . . . . . . 216
xii Contents

10.4 GMM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216

10.5 Stata resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
11 Design effects 221
11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
11.2 Features of sampling designs . . . . . . . . . . . . . . . . . . . . . . . 222
11.2.1 Weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
11.2.2 Clusters and stratification . . . . . . . . . . . . . . . . . . . 223
11.2.3 Weights and clustering in natural experiments . . . . . . . 225
11.3 Methods for point estimation and inference . . . . . . . . . . . . . . 225
11.3.1 Point estimation . . . . . . . . . . . . . . . . . . . . . . . . 226
11.3.2 Standard errors . . . . . . . . . . . . . . . . . . . . . . . . . 226
11.4 Empirical examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
11.4.1 Survey design setup . . . . . . . . . . . . . . . . . . . . . . . 228
11.4.2 Weighted sample means . . . . . . . . . . . . . . . . . . . . 229
11.4.3 Weighted least-squares regression . . . . . . . . . . . . . . . 231
11.4.4 Weighted Poisson count model . . . . . . . . . . . . . . . . . 233
11.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
11.6 Stata resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
References 237
Author index 247
Subject index 251
Preface
This book grew out of our experience giving presentations about applied health econo-
metrics at the International Health Economics Association and the American Society
of Health Economists biennial conferences. In those preconference seminars, we tried to
expose graduate students and early career academics to topics not generally covered in
traditional econometrics courses but nonetheless are salient to most applied research on
healthcare expenditures and use. Participants began to encourage us to turn our slides
into a book.
In this book, we aim to provide a clear understanding of the most commonly used
(and abused) econometric models for healthcare expenditure and use and of approaches
to choose the most appropriate model. If you want intuition, meaningful examples,
inspiration to improve your best practice, and enough math for rigor but not enough
to cause rigor mortis, then keep reading. If you want a general econometrics textbook,
then put down this book and go buy a general econometrics textbook. Get ready to try
new methods and statistical tests in Stata as you read. Be prepared to think.
Despite years of training and practice in applied econometrics, we still learned a
tremendous amount while working on this book from reading recent literature, compar-
ing and testing models in Stata, and debating with each other. We particularly learned
from our coauthor Will Manning, who unfortunately died in 2014 before seeing our
collective effort come to fruition. Will was a fountain of knowledge. We think that his
overarching approach to econometrics of repeated testing to find the best model for the
particular research question and dataset is the best guide. The journey matters, not
just the final parameter estimate.
In closing, we want to thank some of the many people who have helped us complete
this book. David Drukker, editor and econometrician, had numerous suggestions, large
and small, that dramatically improved the book. We are grateful to Stephanie White,
Adam Crawley, and David Culwell at StataCorp for help with LATEX, editorial assis-
tance, and production of the book. We thank Betsy Querna Cliff, Morris Hamilton,
Jun Li, and Eden Volkov for reading early drafts and providing critical feedback. We
thank the many conference participants who were the early guinea pigs for our efforts
at clarity and instruction and especially those who gave us the initial motivation to
undertake this book. Our wives, Erika Bach and Carolyn Norton, provided support and
encouragement, especially during periods of low marginal productivity. Erika Manning
cared for Will during his illness and tolerated lengthy phone calls at odd hours, and
Will’s bad puns at all hours.
Partha Deb and Edward C. Norton
1 Introduction
Health and healthcare are central to society and economic activity. This observation
extends beyond the large fraction of gross national product devoted to formal healthcare
to the fact that health and healthcare affect each other and numerous other decisions.
Health affects people’s ability to engage in work and leisure, their probability of mar-
riage, probability of living to a ripe old age, and how much they spend on healthcare.
Healthcare affects health mostly for the better, although side effects and medical errors
can have drastic consequences. The desire for better health motivates decisions about
smoking, drinking, diet, and exercise over a lifetime. Therefore, it is important to un-
derstand the underlying causes of health and how health affects people’s lives, including
examining the determinants of healthcare expenditures and use.
Economic theory and policy motivate research questions on healthcare expenditures
and use. The healthcare sector is one of the best areas to test economic theories of in-
surance, the behavior of nonprofit firms, principal-agent relationships, and the influence
of peers. Moreover, governments are intimately involved with healthcare. For example,
in the United States, the federal and state governments administer Medicare and Med-
icaid health insurance, the Veterans Administration provides healthcare to veterans,
and the government regulates tobacco, alcohol, and prescription drugs. Understanding
what drives healthcare expenditures and use is essential for policy. National domestic
political debates often center around policies that aim to enhance public health, im-
prove quality of healthcare, and ensure affordable access to health insurance. Health
economists have much to offer by studying these issues.
The last few decades have also seen a proliferation of sophisticated statistical meth-
ods. Researchers now have many alternatives to ordinary least squares (OLS) to analyze
data with dependent variables that are binary, count, or skewed. Researchers can adjust
estimates to control for complex survey design and heteroskedasticity. There are classes
of models [for example, generalized linear models (GLM)] and statistical methods (for
example, maximum likelihood estimation and generalized method of moments) beyond
least squares that provide powerful and unified approaches to fitting many complex
models. Advances in computing power mean that researchers can estimate technically
complex statistical models faster than ever. Stata (and other statistical software) allows
researchers to use these models quickly and easily.
Like the people behind the statistics, data come in all shapes, sizes, and ages. Re-
searchers collect population health and census data, episode-level claims data, survey
data on households and on providers, and, more recently, individual biometric data—
including genetic information. Datasets are often merged to generate richer information
over time. The variety of data is dizzying.

1
 2 Chapter 1 Introduction

The importance of the research and policy questions requires that we use the econo-
metric models with care and that we think deeply about the correct interpretation.
Faster computers do not obviate the need for thought.
In this book, we lay out the main statistical approaches and econometric models
used to analyze healthcare expenditure and use data. We explain how to estimate and
interpret the models using easy-to-follow examples. We include numerous references to
the main theoretical and applied literature. We also discuss the strengths and weak-
nesses of the models we present. Knowing the limitations of models is as important as
knowing when to appropriately use them. Most importantly, we demonstrate rigorous
model testing methods. By following our approach, researchers can rigorously address
research questions in health economics using a way that is tailored to their data.

1.1 Outline
This book is divided into three groups of chapters. The early chapters provide the
background necessary to understand the rest of the book. Many empirical research
questions aim to estimate treatment effects. Consequently, chapter 2 introduces the
potential outcomes framework, which is useful for estimating and interpreting treatment
effects. It also relates treatment effects to marginal and incremental effects in both
linear and nonlinear models. Chapter 3 introduces the Medical Expenditure Panel
Survey dataset, which is used throughout this book for illustrative examples. Chapter 4
illustrates how to estimate the average treatment effect, the treatment effect on the
treated, and marginal and incremental effects for linear regression models. Chapter 4
also shows that misspecifications in OLS models can lead to inconsistent average effects.
It also includes graphical and statistical tests for model specification to help decide
between competing statistical models.
The core chapters describe the most prominent set of models used for healthcare
expenditures and use, including those that explicitly deal with skewness, heteroskedas-
ticity, log transformations, zeros, and count data. Chapter 5 presents GLMs as an al-
ternative to OLS for modeling positive continuous outcomes. Generalized linear models
are especially useful for skewed dependent variables and for heteroskedastic error terms.
Although we argue that GLM provides a powerful set of models for health expenditure,
we also lay out the popular log transformation model in chapter 6. Transforming a de-
pendent variable by taking its natural logarithm is a widely used way to model skewed
outcomes. Chapter 6 describes several versions that differ in their assumptions about
heteroskedasticity and the distribution of the error term (normal or nonnormal). We
show that interpretation can be complex, even though estimation is simple. Chapter 7
adds observations with outcomes equal to zero. Most health expenditure data have a
substantial mass at zero, which makes models that explicitly account for zeros appeal-
ing. Here we describe and compare two-part and selection models. We explain the
underlying assumptions behind the often misunderstood two-part model, and show how
two-part models are superficially similar, yet strikingly different from selection models
in fundamental ways. Chapter 8 moves away from continuous dependent variables to
 1.2 Themes 3

count models. These models are essential for outcomes that are nonnegative integer
valued, including counts of office visits, number of cigarettes smoked, and prescription
drug use.
The book then shifts to more advanced topics. Chapter 9 presents four flexible
approaches to modeling treatment-effect heterogeneity. Quantile regression allows re-
sponse heterogeneity by level of the dependent variable. We describe basic quantile
regressions and how to use those models to obtain quantile treatment effects. Next,
we describe finite mixture models. These models allow us to draw the sample from
a finite number of subpopulations with different relationships between outcomes and
predictors in each subpopulation. Thus, finite mixture models can uncover patterns in
the data caused by heterogeneous types. Third, we describe local-linear regression, a
nonparametric regression method. Nonparametric regression techniques make few as-
sumptions about the functional form of the relationship between the outcome and the
covariates and allow for very general relationships. Finally, conditional density estima-
tion is another flexible alternative to linear models for dependent variables with unusual
distributions. The last two chapters discuss issues that cut across all models. Chap-
ter 10 introduces controlling for endogeneity or selection-on-unobservables of covariates
of policy interest to the researcher. Chapter 11 discusses design effects. Many datasets
have information collected with complex survey designs. Analyses of such data should
account for stratified sampling, primary sampling units, and clustered data.
This book does not attempt to provide a comprehensive treatment of economet-
rics. For that, we refer readers to other sources (for example, Cameron and Trivedi
[2005; 2010], Greene [2012], Wooldridge [2010; 2016]). Instead, we focus on healthcare
econometric models that emphasize three core statistical issues of skewness, zeros, and
heterogeneous response. We focus on providing intuition, a basic mathematical frame-
work, and user-friendly Stata applications. We provide citations to the literature for
original proofs and important applications. Much promising theoretical and applied
work continues to appear in the literature each year. Jones (2010) describe some of the
recent research as well as a wide range of econometric approaches.

1.2 Themes
Although we present numerous alternative models and ways to check and choose between
those models, it should be no surprise that we do not determine a single best model for
all situations or a good second-best model for all cases. Instead, researchers must find
the model that is most appropriate for their research question and data. We recommend
comprehensive model checking, but model checking is not a simple check list. It requires
thought.
We aim to provide the tools to find the best model to consistently estimate the
answer to the research question. This answer will often be a function of E(y|x), such as
the average treatment effect, or the marginal effect of a covariate on the outcome. We
are also concerned about the precision of those estimates, measured by the variance of
the estimators.
 4 Chapter 1 Introduction

Of all possible statistical models, we focus on those that address three key issues
that often appear in health expenditure and use data: skewness, a large mass at zero,
and heterogeneous response. Health expenditure data are often wildly right skewed.
Transforming the dependent variable to generate a dependent variable with a more
symmetric distribution may improve the statistical properties of the model fit but may
make it harder to interpret. The distributions of many interesting health outcomes—
such as total annual healthcare expenditures, hospital visits in the calendar year, and
smoking in the last 30 days—typically have a substantial fraction of zeros, which can
pose difficulties for standard statistical models. Consequently, health economists have
developed models to deal with such outcomes, allowing for a rich understanding of
how variables affect whether the outcome is positive (extensive margin) and the mag-
nitude of the outcome (intensive margin). While a single, summary marginal effect is
sometimes of interest, we often expect heterogeneous treatment effects across different
subpopulations. Modeling the heterogeneity explicitly can reveal new insights.
In summary, our aim is to improve best practices among health economists and
health services researchers.

1.3 Health econometric myths

Despite the tremendous recent advances in econometrics, we have noticed a number of
misconceptions in the published literature. We hope the following myths will disappear
in future generations:

1. Model selection by citation is safe. The lemming approach to econometrics is to

follow blindly what others have done. But each research question, each dataset,
and each model requires individual attention. We advocate artisanal handcrafted
research, not mass-produced cookie-cutter research (see chapter 2).
2. Trim outliers. Outliers are so annoying. They are highly influential, do not fit
nicely onto graphs, and are just, well, different. Why not trim them from the
data? The reason is that each outlier represents a real person or episode. As
much as a hospital administrator would like to assume away an ultraexpensive
patient, the patient exists and is an important feature of many datasets. Embrace
the diversity; start by exploring outliers in the Medical Expenditure Panel Survey
data described in chapter 3.
3. OLS is fine. OLS regression has many virtues. It is easy to estimate and interpret.
Under a set of well-known assumptions—including that the model as specified is
correct— OLS is the best linear unbiased estimator, except when the assumptions
fail, which is often. We demonstrate the limitations of OLS in chapter 4.
4. All GLM models should have a log link with a gamma distribution. Several early
influential articles using GLM models in health economics happened to analyze
data for which the log link with a gamma distribution was the appropriate choice.
Different link and distributional families may be better (see chapter 5) for other
data.
 1.4 Stata friendly 5

5. Log all skewed dependent variables. Health economists have developed a com-
pulsive, almost Pavlovian, instinct to log any and all skewed dependent variables.
While the log transformation makes estimation on the log scale simple, it makes
interpretation and prediction on the raw scale surprisingly difficult (see chapter 6).
6. Use selection models for data with a large mass at zero. When the data have sub-
stantial mass at zero, some researchers reach for the two-part model, while others
reach for selection models. Their choices often lead to considerable argument
over which is better. We advocate the two-part model for researchers interested
in actual outcomes (including the zeros), and we advocate selection models for
researchers interested in latent outcomes (assuming that the zeros are missing
values). We set the record straight in chapter 7.
7. All count models are Poisson. Ever wonder why some researchers reflexively use
Poisson, and others use the negative binomial? We explain the tradeoff between
inference about the conditional mean function and conditional frequencies while
providing intuition and pretty pictures (see chapter 8).
8. Modeling heterogeneity is not worth the effort. Tired of assuming monolithic
treatment effects? Want to spice up your research life? We introduce four ways to
model treatment-effect heterogeneity that can enrich any analysis (see chapter 9).
9. Correlation is causation. Actually, virtually all researchers know that statement
is false. However, knowing it and correctly adjusting for endogeneity are two
different things. We discuss ways to better establish causality by controlling for
endogeneity, which is pervasive in applied social science research (see chapter 10).
10. Complex survey design is just for summary statistics. Most large surveys use
stratification and cluster sampling to better represent important subsamples and
to use resources efficiently. Model estimation, not just summary statistics, should
control for sample weights, clustering, and stratification (see chapter 11).

1.4 Stata friendly

We assume the reader has a basic understanding of Stata. To learn more, read the Stata
manuals and online help, or consult introductions to Stata by Long and Freese (2014)
(see especially chapters 2–4) and Cameron and Trivedi (2010) (see chapters 1 and 2 for
overview). Stata is easy to learn, easy to use, and has a powerful set of tools. Many
of them are built-in, but others are provided by dedicated users who share their code
via Stata packages. Once the reader has a grasp of the basics, our book will be fully
accessible.
Merely reading about econometrics is not the best way to learn. Readers must
actively analyze data themselves. Therefore, we provide user-friendly Stata code, so
interested readers cannot only reproduce all the examples in the book but also modify
the code to analyze their own data. The data and Stata code in this book are publicly
available. We have designed this not only to be user-friendly but also to be interactive.
Dig in!
 6 Chapter 1 Introduction

1.5 A useful way forward

Finally, we agree with the observation by Box and Draper (1987) that “all models are
wrong, but some are useful”. Our intent is to provide methods to choose models that
are useful for the research question of interest.
4 The linear regression model:
Specification and checks

4.1 Introduction
The linear regression model is undoubtedly the workhorse of empirical research. Re-
searchers use it ubiquitously for continuous outcomes and often for count and binary
outcomes. With relatively few assumptions—namely, the relationship between the out-
come and the regressors is correctly specified, and the error term has an expected value
of zero conditional on the values of the regressors—ordinary least-squares (OLS) esti-
mates of the parameters of the model are unbiased and consistent. In other words, given
those two assumptions, OLS delivers estimates that are correct on average. In addition,
if the distribution of the errors have constant variance across the sample observations
and are uncorrelated across sample observations, then OLS produces estimates that have
the smallest variance among all linear unbiased estimators.
The formal statement of these properties is the Gauss–Markov theorem. Many text-
books formally discuss the assumptions and proof, including Wooldridge (2010) and
Cameron and Trivedi (2005). The Gauss–Markov theorem has two main implications.
First, OLS estimates have the desirable property of being unbiased under relatively weak
conditions. Second, there is no linear estimator with better properties than OLS. These
desirable features mean, in many cases, we can use the linear regression model to esti-
mate causal treatment effects and marginal and incremental effects of other covariates,
as we outlined in chapter 2. In this chapter, we show how these can be implemented in
Stata and discuss the interpretation of various effects.
The Gauss–Markov theorem applies to OLS models only when the assumptions are
met. If a regressor is endogenous, for example, the conditional expectation of the
error term is not zero. If it was, it would violate one of the Gauss–Markov theorem’s
assumptions, and the OLS estimates would be inconsistent. With observational data,
researchers should always be aware of the possibility of endogenous regressors. We
address these issues in chapter 10.
The other main assumption is that the model specification is correct. Estimation
of a linear model without serious consideration of the model specification can lead to
substantially misleading answers. One of the most important features of any model
is the relationship of the covariates to the dependent variable. Correct specification
of the relationship is a key assumption of the theorem. In practice, while researchers
cannot claim to know the true model, they should strive to specify good models. A

33
 34 Chapter 4 The linear regression model: Specification and checks

good model includes all the necessary variables—including higher-order polynomials

and interactions terms—but no more. A good model includes variables with the correct
functional relationship between the covariates and outcome. Choosing the correct model
specification requires making choices. There is tension between simplicity and attention
to detail, and there is tension between misspecification and overfitting. We address
these issues in this chapter.
In this chapter, we show with two examples how easy it is to estimate inconsistent
marginal effects when the fit model is misspecified. Marginal effects are surprisingly
sensitive to model misspecification. If we include a variable in the model, but the
relationship between it and the dependent variable is not correct, the estimated marginal
effects of that variable are sensitive to the distribution of that covariate and to whether
the marginal effects are conditioned on a specific value of that covariate.
Some readers may wonder why we obsess about model specification. A commonly
held belief is that the estimate, βbk , of the average marginal effect (AME), ∂E(yi |xi )/∂xk ,
of covariate, xk , is consistent even if it is estimated using a misspecified model. We show
that this can easily be false. In addition, we believe that the focus on average effects
is too narrow a view, because policy interest is often about the response to a covariate
for a specific value of that covariate. For example, we may care only about the effect of
a weight-loss drug on those with an unusually high body mass index, rather than the
entire population. In the case of health insurance, we might be worried about the effect
of raising the coinsurance rates or deductibles in the least generous health insurance
plans rather than for all health insurance plans. In such situations, the marginal effect
for the subsample of interest may be inconsistent, even if the average of marginal effects
for the full sample are not.
Because we never know the correct model specification (theory rarely provides guid-
ance for model specification), it is important to know how to make informed choices.
To this end, the final sections in this chapter describe visual and statistical methods to
test model specification.

4.2 The linear regression model

It is useful to begin with a precise, mathematical formulation of the linear regression
model, in which
yi = x′i β + ui
where yi is the outcome for the ith observation (i = 1, . . . , N ), x′i is a row vector of k
covariates including a constant, β is a column vector of k coefficients to be estimated
including the intercept, and u is the error term. A linear specification can include
nonlinear terms in xi but is always linear in β. Specifications that are nonlinear in β
generally cannot be transformed into a linear specification.
As we showed in chapter 2, if the model is linear in variables, then the estimates of
treatment, marginal, and incremental effects are all simply regression coefficients. Non-
linear terms, which are interactions between covariates or polynomial terms of covari-
 4.3 Marginal, incremental, and treatment effects 35

ates, are functions of parameters and covariate values. We must estimate and interpret
them more carefully.

4.3 Marginal, incremental, and treatment effects

We begin with a fairly simple OLS regression model to predict total annual expenditures
at the individual level for those who spend at least some money, using the 2004 Medical
Expenditure Panel Survey (MEPS) data (see chapter 3). Our goal is to interpret the
results using the framework of potential outcomes and marginal effects (see chapter 2).
To be clear, the model we fit may not be appropriate for a serious research exercise: it
drops all observations with zero expenditures, and its specification of covariates is rudi-
mentary. Additionally, we do not consider any possible models besides OLS, especially
ones that may be better suited to deal with the severe skewness in the distribution of
this outcome, and we do not control for design effects or possible endogeneity. In short,
we ignore all interesting features of this typical health outcome variable, knowing that
we will return to each of these issues throughout the book. The focus of this section is
to provide a framework for interpreting regression results.
In this regression model, we estimate the effect of age (age), gender (female is
a binary indicator for being female), and any health limitations (anylim is a binary
indicator of whether the person has health limitations) on total healthcare expenditures
for persons with any expenditures (exptot > 0), using the MEPS data (see chapter 3).

. *** MEPS data

. use http://www.stata-press.com/data/heus/heus_mepssample
(Sample of MEPS 2004 data)
. *** Restrict to subsample with positive total expenditures
. drop if exp_tot <= 0
(3,440 observations deleted)

We include an interaction term between age and gender, allowing the effect of age
to differ between men and women. It is essential to use the notation with ## between
c.age and i.female so Stata understands that those variables are interacted. The
prefix c. indicates that the variable age is continuous; the prefix i. indicates that the
variable gender is binary. We estimate robust standard errors.
The results appear to show that healthcare expenditures increase with age and are
higher for women. However, the interaction term is negative and statistically significant,
indicating that we must put more effort into fully understanding the relationship be-
tween these demographics and total expenditures. Unsurprisingly, expenditures are far
higher for those with at least one limitation. All coefficients are statistically significant
at p < 0.05.
 36 Chapter 4 The linear regression model: Specification and checks

. *** Regression to predict total expenditures

. regress exp_tot c.age##i.female i.anylim, vce(robust)
Linear regression Number of obs = 15,946
F(4, 15941) = 221.41
Prob > F = 0.0000
R-squared = 0.0773
Root MSE = 10187

Robust
exp_tot Coef. Std. Err. t P>|t| [95% Conf. Interval]

age 94.70939 8.664488 10.93 0.000 77.72601 111.6928

female
Female 1593.099 459.1437 3.47 0.001 693.1256 2493.073

female#c.age
Female -22.56145 10.64236 -2.12 0.034 -43.42168 -1.701215

anylim
Activity .. 4456.928 237.8453 18.74 0.000 3990.724 4923.132
_cons -1772.051 380.9481 -4.65 0.000 -2518.752 -1025.349

4.3.1 Marginal and incremental effects

We first interpret the results for age and gender in more detail. Following chapter 2, we
interpret regression results for the continuous variable age as a marginal effect (deriva-
tive) and for the dichotomous variable female as an incremental effect (difference). One
way to interpret the effects (not necessarily the most informative way for this example,
as we will see) is to compute the average marginal and incremental effects using the
Stata command margins, dydx(). Because of the interaction term between age and
gender, the average marginal and incremental effects will not equal any of the estimated
coefficients in the model.
Women spend more than men by an average of $523, averaged across all ages in the
sample. The AME of age is $81, meaning that on average (allowing for the interaction
with female) for this sample, an increase in age of 1 year corresponds with an increase
in total expenditures of $81. We note that this interpretation, in a model with covariates
entered nonlinearly, is only true because the model is affine in age. If one is interested in
knowing what happens if each person in the sample becomes one year older, it would be
better to do that computation directly. We show how this can be done using margins
in section 5.7.
4.3.1 Marginal and incremental effects 37

. *** Average marginal effects for age and female

. margins, dydx(age female)
Average marginal effects Number of obs = 15,946
Model VCE : Robust
Expression : Linear prediction, predict()
dy/dx w.r.t. : age 1.female

Delta-method
dy/dx Std. Err. t P>|t| [95% Conf. Interval]

age 81.38701 5.338745 15.24 0.000 70.92247 91.85156

female
Female 523.2414 171.3697 3.05 0.002 187.3375 859.1454

Note: dy/dx for factor levels is the discrete change from the base level.

The average marginal and incremental effects calculated with margins, dydx() do
not fully illustrate the complex relationship between age, gender, and total expenditures
because of the interaction between them. One way to improve the interpretation is
to calculate the marginal effect of age separately for men and for women (using the
at(female=(0 1)) option). The marginal effect of age for men is about $23 higher
than for women ($95 compared with $72). This means that as men age, their spending
increases faster than that of women.

. *** Marginal effect of age by gender

. margins, dydx(age) at(female=(0 1))
Average marginal effects Number of obs = 15,946
Model VCE : Robust
Expression : Linear prediction, predict()
dy/dx w.r.t. : age
1._at : female = 0
2._at : female = 1

Delta-method
dy/dx Std. Err. t P>|t| [95% Conf. Interval]

age
_at
1 94.70939 8.664488 10.93 0.000 77.72601 111.6928
2 72.14794 6.52578 11.06 0.000 59.35668 84.93921

Similarly, we calculate the incremental effect of gender at different ages (using the
at(age=(20 45 70)) option). The incremental effect of gender is different at each age,
being more than $1,140 at age 20, and close to 0 around age 70.
 5.3 GLM examples 75

5.2.2 Parameter estimation

As alluded to above, GLM estimation requires two sets of choices. The first set of choices
determines the link function and the distribution family. In section 5.8, we discuss how
we chose based on rigorous statistical tests.
For the parameter estimates in the model to be consistent, it is only necessary to
correctly specify the link function, g, and how the covariates enter the index function.
The choice of the distribution family affects the efficiency of the estimates, but an incor-
rect choice does not lead to inconsistency of the parameter estimates. An inappropriate
assumption about the distribution family can lead to an inconsistent estimate of the
inference statistics, but this inconsistency can be remedied using robust standard errors.
The other choice is whether to estimate GLMs by quasi–maximum likelihood or
iteratively reweighted least squares. In Stata, the default is quasi–maximum likelihood,
which does not require specification of the full log likelihood. The choice between these
two methods does not seem to matter much in practice for typical models and datasets.
After fitting a GLM, one can easily derive marginal and incremental effects of specific
covariates on the expected value of y (or other treatment effects).

5.3 GLM examples

We now show how to estimate GLMs for healthcare expenditures with a few choices of
link functions and distribution families, using the MEPS data introduced in chapter 3.
Specifically, we estimate the effect of age (age) and gender (female is a binary indicator
for being female) on total healthcare expenditures for persons with any expenditures
(exp tot > 0).
Our first example is a model with a log link (option link(log)) and a Gaussian
family (option family(gaussian)). This is equivalent to a nonlinear regression model
with an exponential mean. The results show that healthcare expenditures increase with
age and are higher for women, but the coefficient on female is not statistically significant
at the 5% level. Because the conditional mean has an exponential form, coefficients can
be interpreted directly as percent changes. Expenditures increase by about 2.6% with
each additional year of age after adjusting for gender. Women spend about 8% more
than men [0.080 = exp(0.0770) − 1] after controlling for age.
76 Chapter 5 Generalized linear models

. *** GLM of total expenditures, log link and Gaussian family

. glm exp_tot age female, link(log) family(gaussian) vce(robust)
Iteration 0: log pseudolikelihood = -191206.56
Iteration 1: log pseudolikelihood = -183815.74
Iteration 2: log pseudolikelihood = -172074.86
Iteration 3: log pseudolikelihood = -170297.76
Iteration 4: log pseudolikelihood = -170068.69
Iteration 5: log pseudolikelihood = -170067.92
Iteration 6: log pseudolikelihood = -170067.92
Generalized linear models No. of obs = 15,946
Optimization : ML Residual df = 15,943
Scale parameter = 1.07e+08
Deviance = 1.71351e+12 (1/df) Deviance = 1.07e+08
Pearson = 1.71351e+12 (1/df) Pearson = 1.07e+08
Variance function: V(u) = 1 [Gaussian]
Link function : g(u) = ln(u) [Log]
AIC = 21.33086
Log pseudolikelihood = -170067.9162 BIC = 1.71e+12

Robust
exp_tot Coef. Std. Err. z P>|z| [95% Conf. Interval]

age .0257328 .00101 25.48 0.000 .0237533 .0277123

female .0769791 .0420732 1.83 0.067 -.0054828 .1594411
_cons 7.041415 .0652623 107.89 0.000 6.913503 7.169327

The second example also specifies a log link but assumes that the distribution fam-
ily is gamma (option family(gamma)), implying that the variance of expenditures is
proportional to the square of the mean. This is a leading choice in published models
of healthcare expenditures, but we will return to the choices of link and family more
comprehensively in section 5.8.
The results show that healthcare expenditures increase with age and are higher for
women. Both coefficients are statistically significant, with p < 0.001. Expenditures
increase by about 2.8% with each additional year of age, which is quite close to the
effect fit by the model with the Gaussian family. However, now we find that women
spend about 23% more than men (0.23 = exp(0.2086)−1), after controlling for age. This
is almost three times as large as the effect estimated in the model with the Gaussian
family. A small change in the model leads to a large change in interpretation.
 5.4 GLM predictions 77

. *** GLM of total expenditures, log link and gamma family

. glm exp_tot age female, link(log) family(gamma) vce(robust)
Iteration 0: log pseudolikelihood = -150164.47
Iteration 1: log pseudolikelihood = -148063.58
Iteration 2: log pseudolikelihood = -148047.8
Iteration 3: log pseudolikelihood = -148047.79
Generalized linear models No. of obs = 15,946
Optimization : ML Residual df = 15,943
Scale parameter = 7.097257
Deviance = 33478.20882 (1/df) Deviance = 2.099869
Pearson = 113151.5612 (1/df) Pearson = 7.097257
Variance function: V(u) = u^2 [Gamma]
Link function : g(u) = ln(u) [Log]
AIC = 18.56902
Log pseudolikelihood = -148047.791 BIC = -120801.6

Robust
exp_tot Coef. Std. Err. z P>|z| [95% Conf. Interval]

age .0279516 .0012071 23.16 0.000 .0255856 .0303176

female .2086064 .0478756 4.36 0.000 .114772 .3024408
_cons 6.835683 .0856778 79.78 0.000 6.667757 7.003608

Our primary intent was to use these examples to demonstrate the use of the glm
command and explain how to interpret coefficients. However, these examples also show
that the estimated effects in a sample can be quite different across distribution family
choices when the link function is the same, even though the choice of family has no
theoretical implications for consistency of parameter estimates.
We could run many other GLM models, changing the link function or the distri-
butional family. For example, we could fit a GLM with a square root link (option
link(power 0.5)) and a Poisson family (option family(poisson)). Or we could fit a
GLM with a cube root link (option link(power 0.333)) and an inverse Gaussian family
(option family(igaussian)).

5.4 GLM predictions

For all GLM models with a log link, the expected value of the dependent variable, y, is
the exponentiated linear index function:

E (yi |xi ) = µi = g −1 (x′i β) = exp (x′i β) (5.1)

The sample average of the expected value of total expenditures is the average of µi
over the sample. We calculate its estimate using the margins command. The predicted
mean of total expenditures is $4,509, less than 1% from the sample mean of $4,480.
 78 Chapter 5 Generalized linear models

. *** Predicted mean of total expenditures from GLM with log link and gamma
> family
. quietly glm exp_tot age female, link(log) family(gamma) vce(robust)
. margins
Predictive margins Number of obs = 15,946
Model VCE : Robust
Expression : Predicted mean exp_tot, predict()

Delta-method
Margin Std. Err. z P>|z| [95% Conf. Interval]

_cons 4508.963 81.41072 55.39 0.000 4349.401 4668.525

When we compare predictions from log transformation models in chapter 6 with

the sample mean, we will find that those predictions are much further off. They will
be anywhere from 10% to 20% too high. GLM is generally better than log models at
reproducing the sample mean of the outcome.

5.5 GLM example with interaction term

Before computing marginal effects, we extend our simple specification to include the
interaction of age and gender as a covariate. That is, we allow for the effect of gender
to vary by age (or equivalently, the effect of age to vary by gender). The results with
an interaction term are harder to interpret, but more realistic, and will help show the
power of several Stata postestimation commands.
When including interaction terms, one must use special Stata notation, so that
margins knows the relationship between variables when it takes derivatives. Therefore,
we use c. as a prefix to indicate that age is a continuous variable, i. to indicate that
female is an indicator variable, and ## between them to include not only the main
effects but also their interaction.