-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathdifference-in-differences.html
426 lines (384 loc) · 20.8 KB
/
difference-in-differences.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
<!DOCTYPE html>
<html lang="" xml:lang="">
<head>
<meta charset="utf-8" />
<meta http-equiv="X-UA-Compatible" content="IE=edge" />
<title>Chapter 6 Difference-in-Differences | Impact Evaluation in Practice</title>
<meta name="description" content="Solutions for Impact Evaluation in Practice (2nd Edition)." />
<meta name="generator" content="bookdown 0.20 and GitBook 2.6.7" />
<meta property="og:title" content="Chapter 6 Difference-in-Differences | Impact Evaluation in Practice" />
<meta property="og:type" content="book" />
<meta property="og:description" content="Solutions for Impact Evaluation in Practice (2nd Edition)." />
<meta name="twitter:card" content="summary" />
<meta name="twitter:title" content="Chapter 6 Difference-in-Differences | Impact Evaluation in Practice" />
<meta name="twitter:description" content="Solutions for Impact Evaluation in Practice (2nd Edition)." />
<meta name="author" content="Liam F. Beiser-McGrath" />
<meta name="author" content="Lecturer (Assistant Professor) in Politics, Royal Holloway, University of London" />
<meta name="author" content="www.liambeisermcgrath.com" />
<meta name="author" content="liam@liambeisermcgrath.com" />
<meta name="date" content="2020-07-29" />
<meta name="viewport" content="width=device-width, initial-scale=1" />
<meta name="apple-mobile-web-app-capable" content="yes" />
<meta name="apple-mobile-web-app-status-bar-style" content="black" />
<link rel="prev" href="regression-discontinuity-designs.html"/>
<link rel="next" href="matching.html"/>
<script src="libs/jquery-2.2.3/jquery.min.js"></script>
<link href="libs/gitbook-2.6.7/css/style.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-table.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-bookdown.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-highlight.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-search.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-fontsettings.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-clipboard.css" rel="stylesheet" />
<script src="libs/accessible-code-block-0.0.1/empty-anchor.js"></script>
<script src="libs/kePrint-0.0.1/kePrint.js"></script>
<style type="text/css">
code.sourceCode > span { display: inline-block; line-height: 1.25; }
code.sourceCode > span { color: inherit; text-decoration: inherit; }
code.sourceCode > span:empty { height: 1.2em; }
.sourceCode { overflow: visible; }
code.sourceCode { white-space: pre; position: relative; }
pre.sourceCode { margin: 0; }
@media screen {
div.sourceCode { overflow: auto; }
}
@media print {
code.sourceCode { white-space: pre-wrap; }
code.sourceCode > span { text-indent: -5em; padding-left: 5em; }
}
pre.numberSource code
{ counter-reset: source-line 0; }
pre.numberSource code > span
{ position: relative; left: -4em; counter-increment: source-line; }
pre.numberSource code > span > a:first-child::before
{ content: counter(source-line);
position: relative; left: -1em; text-align: right; vertical-align: baseline;
border: none; display: inline-block;
-webkit-touch-callout: none; -webkit-user-select: none;
-khtml-user-select: none; -moz-user-select: none;
-ms-user-select: none; user-select: none;
padding: 0 4px; width: 4em;
color: #aaaaaa;
}
pre.numberSource { margin-left: 3em; border-left: 1px solid #aaaaaa; padding-left: 4px; }
div.sourceCode
{ }
@media screen {
code.sourceCode > span > a:first-child::before { text-decoration: underline; }
}
code span.al { color: #ff0000; font-weight: bold; } /* Alert */
code span.an { color: #60a0b0; font-weight: bold; font-style: italic; } /* Annotation */
code span.at { color: #7d9029; } /* Attribute */
code span.bn { color: #40a070; } /* BaseN */
code span.bu { } /* BuiltIn */
code span.cf { color: #007020; font-weight: bold; } /* ControlFlow */
code span.ch { color: #4070a0; } /* Char */
code span.cn { color: #880000; } /* Constant */
code span.co { color: #60a0b0; font-style: italic; } /* Comment */
code span.cv { color: #60a0b0; font-weight: bold; font-style: italic; } /* CommentVar */
code span.do { color: #ba2121; font-style: italic; } /* Documentation */
code span.dt { color: #902000; } /* DataType */
code span.dv { color: #40a070; } /* DecVal */
code span.er { color: #ff0000; font-weight: bold; } /* Error */
code span.ex { } /* Extension */
code span.fl { color: #40a070; } /* Float */
code span.fu { color: #06287e; } /* Function */
code span.im { } /* Import */
code span.in { color: #60a0b0; font-weight: bold; font-style: italic; } /* Information */
code span.kw { color: #007020; font-weight: bold; } /* Keyword */
code span.op { color: #666666; } /* Operator */
code span.ot { color: #007020; } /* Other */
code span.pp { color: #bc7a00; } /* Preprocessor */
code span.sc { color: #4070a0; } /* SpecialChar */
code span.ss { color: #bb6688; } /* SpecialString */
code span.st { color: #4070a0; } /* String */
code span.va { color: #19177c; } /* Variable */
code span.vs { color: #4070a0; } /* VerbatimString */
code span.wa { color: #60a0b0; font-weight: bold; font-style: italic; } /* Warning */
</style>
<link rel="stylesheet" href="style.css" type="text/css" />
</head>
<body>
<div class="book without-animation with-summary font-size-2 font-family-1" data-basepath=".">
<div class="book-summary">
<nav role="navigation">
<ul class="summary">
<li><a href="./">Impact Evaluation in Practice Solutions</a></li>
<li class="divider"></li>
<li class="chapter" data-level="" data-path="index.html"><a href="index.html"><i class="fa fa-check"></i>Preamble</a></li>
<li class="chapter" data-level="" data-path="about-the-author.html"><a href="about-the-author.html"><i class="fa fa-check"></i>About the Author</a></li>
<li class="chapter" data-level="1" data-path="intro.html"><a href="intro.html"><i class="fa fa-check"></i><b>1</b> Introduction</a><ul>
<li class="chapter" data-level="1.1" data-path="intro.html"><a href="intro.html#variable-definitions"><i class="fa fa-check"></i><b>1.1</b> Variable Definitions</a><ul>
<li class="chapter" data-level="1.1.1" data-path="intro.html"><a href="intro.html#outcome-variable"><i class="fa fa-check"></i><b>1.1.1</b> Outcome Variable</a></li>
<li class="chapter" data-level="1.1.2" data-path="intro.html"><a href="intro.html#control-variables"><i class="fa fa-check"></i><b>1.1.2</b> Control Variables</a></li>
<li class="chapter" data-level="1.1.3" data-path="intro.html"><a href="intro.html#other-variables"><i class="fa fa-check"></i><b>1.1.3</b> Other Variables</a></li>
</ul></li>
</ul></li>
<li class="chapter" data-level="2" data-path="causal-inference-and-counterfactuals.html"><a href="causal-inference-and-counterfactuals.html"><i class="fa fa-check"></i><b>2</b> Causal Inference and Counterfactuals</a><ul>
<li class="chapter" data-level="2.1" data-path="causal-inference-and-counterfactuals.html"><a href="causal-inference-and-counterfactuals.html#before-after-designs"><i class="fa fa-check"></i><b>2.1</b> Before-After Designs</a></li>
<li class="chapter" data-level="2.2" data-path="causal-inference-and-counterfactuals.html"><a href="causal-inference-and-counterfactuals.html#enrolled-vs.-non-enrolled"><i class="fa fa-check"></i><b>2.2</b> Enrolled vs. Non-Enrolled</a></li>
</ul></li>
<li class="chapter" data-level="3" data-path="randomised-assignment.html"><a href="randomised-assignment.html"><i class="fa fa-check"></i><b>3</b> Randomised Assignment</a></li>
<li class="chapter" data-level="4" data-path="instrumental-variables.html"><a href="instrumental-variables.html"><i class="fa fa-check"></i><b>4</b> Instrumental Variables</a></li>
<li class="chapter" data-level="5" data-path="regression-discontinuity-designs.html"><a href="regression-discontinuity-designs.html"><i class="fa fa-check"></i><b>5</b> Regression Discontinuity Designs</a></li>
<li class="chapter" data-level="6" data-path="difference-in-differences.html"><a href="difference-in-differences.html"><i class="fa fa-check"></i><b>6</b> Difference-in-Differences</a></li>
<li class="chapter" data-level="7" data-path="matching.html"><a href="matching.html"><i class="fa fa-check"></i><b>7</b> Matching</a></li>
<li class="chapter" data-level="8" data-path="design-and-sampling.html"><a href="design-and-sampling.html"><i class="fa fa-check"></i><b>8</b> Design and Sampling</a><ul>
<li class="chapter" data-level="8.1" data-path="design-and-sampling.html"><a href="design-and-sampling.html#power-calculations-with-clusters"><i class="fa fa-check"></i><b>8.1</b> Power Calculations with Clusters</a></li>
</ul></li>
<li class="chapter" data-level="9" data-path="session-info.html"><a href="session-info.html"><i class="fa fa-check"></i><b>9</b> Session Info</a></li>
<li class="divider"></li>
<li><a href="www.liambeisermcgrath.com" target="blank">www.liambeisermcgrath.com</a></li>
</ul>
</nav>
</div>
<div class="book-body">
<div class="body-inner">
<div class="book-header" role="navigation">
<h1>
<i class="fa fa-circle-o-notch fa-spin"></i><a href="./">Impact Evaluation in Practice</a>
</h1>
</div>
<div class="page-wrapper" tabindex="-1" role="main">
<div class="page-inner">
<section class="normal" id="section-">
<div id="difference-in-differences" class="section level1">
<h1><span class="header-section-number">Chapter 6</span> Difference-in-Differences</h1>
<!-- ```{r, warning=FALSE, message=FALSE} -->
<!-- library(broom) -->
<!-- library(estimatr) -->
<!-- library(haven) -->
<!-- library(kableExtra) -->
<!-- library(MatchIt) -->
<!-- library(texreg) -->
<!-- library(tidyverse) -->
<!-- df <- read_dta("./data/evaluation.dta") -->
<!-- ``` -->
<p>Difference-in-differences can be used to evaluate our Health Insurance Subsidy Program (HISP). In this scenario, you have two rounds of data on two groups of households: one group that enrolled in the program, and another that did not. Remembering the case of the enrolled and non- enrolled groups, you realize that you cannot simply compare the average health expenditures of the two groups because of selection bias. Because you have data for two periods for each household in the sample, you can use those data to solve some of these challenges by comparing the change in health expenditures for the two groups, assuming that the change in the health expenditures of the non-enrolled group reflects what would have happened to the expenditures of the enrolled group in the absence of the program. Note that it does not matter which way you calculate the double difference.</p>
<p>Next, you estimate the effect using regression analysis. Using a simple linear regression to compute the simple difference-in- differences estimate, you find that the program reduced household health expenditures by US$8.16. You then refine your analysis by adding additional control variables. In other words, you use a multivariate linear regression that takes into account a host of other factors, and you find the same reduction in household health expenditures.</p>
<div class="sourceCode" id="cb23"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb23-1"><a href="difference-in-differences.html#cb23-1"></a>out_did <-<span class="st"> </span><span class="kw">lm_robust</span>(health_expenditures <span class="op">~</span><span class="st"> </span>round <span class="op">*</span><span class="st"> </span>enrolled, </span>
<span id="cb23-2"><a href="difference-in-differences.html#cb23-2"></a> <span class="dt">data =</span> df <span class="op">%>%</span><span class="st"> </span><span class="kw">filter</span>(treatment_locality <span class="op">==</span><span class="st"> </span><span class="dv">1</span>),</span>
<span id="cb23-3"><a href="difference-in-differences.html#cb23-3"></a> <span class="dt">clusters =</span> locality_identifier)</span>
<span id="cb23-4"><a href="difference-in-differences.html#cb23-4"></a></span>
<span id="cb23-5"><a href="difference-in-differences.html#cb23-5"></a>out_did_wcov <-<span class="st"> </span><span class="kw">lm_robust</span>(health_expenditures <span class="op">~</span><span class="st"> </span>round <span class="op">*</span><span class="st"> </span>enrolled <span class="op">+</span></span>
<span id="cb23-6"><a href="difference-in-differences.html#cb23-6"></a><span class="st"> </span>age_hh <span class="op">+</span><span class="st"> </span>age_sp <span class="op">+</span><span class="st"> </span>educ_hh <span class="op">+</span><span class="st"> </span>educ_sp <span class="op">+</span><span class="st"> </span></span>
<span id="cb23-7"><a href="difference-in-differences.html#cb23-7"></a><span class="st"> </span>female_hh <span class="op">+</span><span class="st"> </span>indigenous <span class="op">+</span><span class="st"> </span>hhsize <span class="op">+</span><span class="st"> </span>dirtfloor <span class="op">+</span><span class="st"> </span></span>
<span id="cb23-8"><a href="difference-in-differences.html#cb23-8"></a><span class="st"> </span>bathroom <span class="op">+</span><span class="st"> </span>land <span class="op">+</span><span class="st"> </span>hospital_distance, </span>
<span id="cb23-9"><a href="difference-in-differences.html#cb23-9"></a> <span class="dt">data =</span> df <span class="op">%>%</span><span class="st"> </span><span class="kw">filter</span>(treatment_locality <span class="op">==</span><span class="st"> </span><span class="dv">1</span>),</span>
<span id="cb23-10"><a href="difference-in-differences.html#cb23-10"></a> <span class="dt">clusters =</span> locality_identifier)</span></code></pre></div>
<div class="sourceCode" id="cb24"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb24-1"><a href="difference-in-differences.html#cb24-1"></a><span class="kw">htmlreg</span>(<span class="kw">list</span>(out_did, out_did_wcov), <span class="dt">doctype =</span> <span class="ot">FALSE</span>,</span>
<span id="cb24-2"><a href="difference-in-differences.html#cb24-2"></a> <span class="dt">custom.coef.map =</span> <span class="kw">list</span>(<span class="st">'enrolled'</span> =<span class="st"> "Enrollment"</span>,</span>
<span id="cb24-3"><a href="difference-in-differences.html#cb24-3"></a> <span class="st">'round'</span> =<span class="st"> "Round"</span>,</span>
<span id="cb24-4"><a href="difference-in-differences.html#cb24-4"></a> <span class="st">'round:enrolled'</span> =<span class="st"> "Enrollment X Round"</span>),</span>
<span id="cb24-5"><a href="difference-in-differences.html#cb24-5"></a> <span class="dt">caption =</span> <span class="st">"Evaluating HISP: Difference-in-Differences with Regression"</span>,</span>
<span id="cb24-6"><a href="difference-in-differences.html#cb24-6"></a> <span class="dt">caption.above =</span> <span class="ot">TRUE</span>,</span>
<span id="cb24-7"><a href="difference-in-differences.html#cb24-7"></a> <span class="dt">custom.model.names =</span> <span class="kw">c</span>(<span class="st">"No Covariate Adjustment"</span>, <span class="st">"With Covariate Adjustment"</span>))</span></code></pre></div>
<table class="texreg" style="margin: 10px auto;border-collapse: collapse;border-spacing: 0px;color: #000000;border-top: 2px solid #000000;">
<caption>
Evaluating HISP: Difference-in-Differences with Regression
</caption>
<thead>
<tr>
<th style="padding-left: 5px;padding-right: 5px;">
</th>
<th style="padding-left: 5px;padding-right: 5px;">
No Covariate Adjustment
</th>
<th style="padding-left: 5px;padding-right: 5px;">
With Covariate Adjustment
</th>
</tr>
</thead>
<tbody>
<tr style="border-top: 1px solid #000000;">
<td style="padding-left: 5px;padding-right: 5px;">
Enrollment
</td>
<td style="padding-left: 5px;padding-right: 5px;">
-6.30<sup>*</sup>
</td>
<td style="padding-left: 5px;padding-right: 5px;">
-1.51<sup>*</sup>
</td>
</tr>
<tr>
<td style="padding-left: 5px;padding-right: 5px;">
</td>
<td style="padding-left: 5px;padding-right: 5px;">
[-6.69; -5.91]
</td>
<td style="padding-left: 5px;padding-right: 5px;">
[-1.77; -1.25]
</td>
</tr>
<tr>
<td style="padding-left: 5px;padding-right: 5px;">
Round
</td>
<td style="padding-left: 5px;padding-right: 5px;">
1.51<sup>*</sup>
</td>
<td style="padding-left: 5px;padding-right: 5px;">
1.45<sup>*</sup>
</td>
</tr>
<tr>
<td style="padding-left: 5px;padding-right: 5px;">
</td>
<td style="padding-left: 5px;padding-right: 5px;">
[ 0.79; 2.24]
</td>
<td style="padding-left: 5px;padding-right: 5px;">
[ 0.73; 2.17]
</td>
</tr>
<tr>
<td style="padding-left: 5px;padding-right: 5px;">
Enrollment X Round
</td>
<td style="padding-left: 5px;padding-right: 5px;">
-8.16<sup>*</sup>
</td>
<td style="padding-left: 5px;padding-right: 5px;">
-8.16<sup>*</sup>
</td>
</tr>
<tr>
<td style="padding-left: 5px;padding-right: 5px;">
</td>
<td style="padding-left: 5px;padding-right: 5px;">
[-8.81; -7.52]
</td>
<td style="padding-left: 5px;padding-right: 5px;">
[-8.81; -7.52]
</td>
</tr>
<tr style="border-top: 1px solid #000000;">
<td style="padding-left: 5px;padding-right: 5px;">
R<sup>2</sup>
</td>
<td style="padding-left: 5px;padding-right: 5px;">
0.34
</td>
<td style="padding-left: 5px;padding-right: 5px;">
0.55
</td>
</tr>
<tr>
<td style="padding-left: 5px;padding-right: 5px;">
Adj. R<sup>2</sup>
</td>
<td style="padding-left: 5px;padding-right: 5px;">
0.34
</td>
<td style="padding-left: 5px;padding-right: 5px;">
0.55
</td>
</tr>
<tr>
<td style="padding-left: 5px;padding-right: 5px;">
Num. obs.
</td>
<td style="padding-left: 5px;padding-right: 5px;">
9919
</td>
<td style="padding-left: 5px;padding-right: 5px;">
9919
</td>
</tr>
<tr>
<td style="padding-left: 5px;padding-right: 5px;">
RMSE
</td>
<td style="padding-left: 5px;padding-right: 5px;">
7.91
</td>
<td style="padding-left: 5px;padding-right: 5px;">
6.54
</td>
</tr>
<tr style="border-bottom: 2px solid #000000;">
<td style="padding-left: 5px;padding-right: 5px;">
N Clusters
</td>
<td style="padding-left: 5px;padding-right: 5px;">
100
</td>
<td style="padding-left: 5px;padding-right: 5px;">
100
</td>
</tr>
</tbody>
<tfoot>
<tr>
<td style="font-size: 0.8em;" colspan="3">
<sup>*</sup> Null hypothesis value outside the confidence interval.
</td>
</tr>
</tfoot>
</table>
<p><strong>What are the basic assumptions required to accept this result from difference-in-differences?</strong></p>
<p>To accept this result, we assume that there are no differential time varying factors between the two groups other than the program. We assume that the treatment and comparison groups would have equal trends or changes in outcomes in the absence of treatment. While this assumption can’t be tested in the postintervention period, we can compare trends before the intervention starts.</p>
<p><strong>Based on the result from difference-in-differences, should HISP be scaled up nationally?</strong></p>
<p>No, based on this result, the HISP should not be scaled up nationally because it has decreased health expenditures by less than the $10 threshold level. Taking the estimated impact under random assignment as the “true” impact of the program suggests that the difference in difference estimate may be biased. In fact, in this case, using the nonenrolled households as a comparison group does not accurately represent the counterfactual trend in health expenditures.</p>
</div>
</section>
</div>
</div>
</div>
<a href="regression-discontinuity-designs.html" class="navigation navigation-prev " aria-label="Previous page"><i class="fa fa-angle-left"></i></a>
<a href="matching.html" class="navigation navigation-next " aria-label="Next page"><i class="fa fa-angle-right"></i></a>
</div>
</div>
<script src="libs/gitbook-2.6.7/js/app.min.js"></script>
<script src="libs/gitbook-2.6.7/js/lunr.js"></script>
<script src="libs/gitbook-2.6.7/js/clipboard.min.js"></script>
<script src="libs/gitbook-2.6.7/js/plugin-search.js"></script>
<script src="libs/gitbook-2.6.7/js/plugin-sharing.js"></script>
<script src="libs/gitbook-2.6.7/js/plugin-fontsettings.js"></script>
<script src="libs/gitbook-2.6.7/js/plugin-bookdown.js"></script>
<script src="libs/gitbook-2.6.7/js/jquery.highlight.js"></script>
<script src="libs/gitbook-2.6.7/js/plugin-clipboard.js"></script>
<script>
gitbook.require(["gitbook"], function(gitbook) {
gitbook.start({
"sharing": {
"github": false,
"facebook": true,
"twitter": true,
"linkedin": false,
"weibo": false,
"instapaper": false,
"vk": false,
"all": ["facebook", "twitter", "linkedin", "weibo", "instapaper"]
},
"fontsettings": {
"theme": "white",
"family": "sans",
"size": 2
},
"edit": {
"link": null,
"text": null
},
"history": {
"link": null,
"text": null
},
"view": {
"link": null,
"text": null
},
"download": ["iep_solutions.pdf", "iep_solutions.epub"],
"toc": {
"collapse": "subsection"
}
});
});
</script>
</body>
</html>