(Original PDF) Real Stats Using Econometrics for

Political Science and Public Policy download

https://ebooksecure.com/product/original-pdf-real-stats-using-
econometrics-for-political-science-and-public-policy/

Download full version ebook from https://ebooksecure.com

We believe these products will be a great fit for you. Click
the link to download now, or visit ebooksecure.com
to discover even more!

(eBook PDF) Public Finance and Public Policy 5th

Edition

http://ebooksecure.com/product/ebook-pdf-public-finance-and-
public-policy-5th-edition/

Public Finance and Public Policy 5th Edition (eBook

PDF)

http://ebooksecure.com/product/public-finance-and-public-
policy-5th-edition-ebook-pdf/

(eBook PDF) Public Policy: Investing for a Better World

http://ebooksecure.com/product/ebook-pdf-public-policy-investing-
for-a-better-world/

(eBook PDF) Public Policy: Investing for a Better World

http://ebooksecure.com/product/ebook-pdf-public-policy-investing-
for-a-better-world-2/
(eBook PDF) Using Econometrics A Practical Guide 7th
Edition

http://ebooksecure.com/product/ebook-pdf-using-econometrics-a-
practical-guide-7th-edition/

Science for Policy Handbook 1st edition - eBook PDF

https://ebooksecure.com/download/science-for-policy-handbook-
ebook-pdf/

(eBook PDF) Public Health Science and Nursing Practice:

Caring for Populations

http://ebooksecure.com/product/ebook-pdf-public-health-science-
and-nursing-practice-caring-for-populations/

(eBook PDF) Using Econometrics A Practical Guide 6th by

Studenmund

http://ebooksecure.com/product/ebook-pdf-using-econometrics-a-
practical-guide-6th-by-studenmund/

(eBook PDF) Public Policy in Action Perspectives on the

Policy Process

http://ebooksecure.com/product/ebook-pdf-public-policy-in-action-
perspectives-on-the-policy-process/
 CONTENTS vii

Conclusion 433 .
Further Reading 434 .
Key Terms 434
. Computing Corner 435 .
Exercises 439

IV ADVANCED MATERIAL 447

13 Time Series: Dealing with Stickiness over Time 449

13.1 Modeling Autocorrelation 450
13.2 Detecting Autocorrelation 453
13.3 Fixing Autocorrelation 457
13.4 Dynamic Models 462
13.5 Stationarity 465
Conclusion 475 .
Further Reading 476 .
Key Terms 477
. Computing Corner 477 .
Exercises 479

14 Advanced OLS 481

14.1 How to Derive the OLS Estimator and Prove Unbiasedness 481
14.2 How to Derive the Equation for the Variance of β̂1 487
14.3 How to Derive the Omitted Variable Bias Conditions 489
14.4 Anticipating the Sign of Omitted Variable Bias 491
14.5 Omitted Variable Bias with Multiple Variables 494
14.6 Omitted Variable Bias due to Measurement Error 495
Conclusion 497 .
Further Reading 497 .
Key Term 498
. Computing Corner 498 .
Exercises 498

15 Advanced Panel Data 501

15.1 Panel Data Models with Serially Correlated Errors 501
15.2 Temporal Dependence with a Lagged Dependent Variable 503
15.3 Random Effects Models 507
Conclusion 509 .
Further Reading 509 .
Key Term 510
. Computing Corner 510 .
Exercises 513
viii CONTENTS

16 Conclusion: How to Be a Statistical Realist 516

Further Reading 520

APPENDICES
Math and Probability Background 521
A. Summation 521
B. Expectation 521
C. Variance 522
D. Covariance 523
E. Correlation 524
F. Probability Density Functions 524
G. Normal Distributions 526
H. Other Useful Distributions 532
I. Sampling 534
Further Reading 537 . Key Terms 537 . Computing Corner 537

Citations and Additional Notes 539

Guide to Review Questions 549

Bibliography/Credits 559

Glossary 568

Index 577
LIST OF FIGURES

1.1 Rule #1 2
1.2 Weight and Donuts in Springfield 4
1.3 Regression Line for Weight and Donuts in Springfield 5
1.4 Examples of Lines Generated by Core Statistical Model (for Review
Question) 7
1.5 Correlation 10
1.6 Possible Relationships Between X, , and Y (for Discussion
Questions) 12
1.7 Two Scenarios for the Relationship between Flu Shots and Health 14

2.1 Two Versions of Debt and Growth Data 25

2.2 Weight and Donuts in Springfield 27
2.3 Scatterplots of Violent Crime against Percent Urban, Single Parent,
and Poverty 31

3.1 Relationship between Income Growth and Vote for the Incumbent
President’s Party, 1948–2012 46
3.2 Elections and Income Growth with Model Parameters Indicated 51
3.3 Fitted Values and Residuals for Observations in Table 3.1 52
3.4 Four Distributions 55
3.5 Distribution of β̂1 58
3.6 Two Distributions with Different Variances of β̂1 62
3.7 Four Scatterplots (for Review Questions on page 63) 64
3.8 Distributions of β̂1 for Different Sample Sizes 66
3.9 Plots with Different Goodness of Fit 72
3.10 Height and Wages 74
3.11 Scatterplot of Violent Crime and Percent Urban 77
3.12 Scatterplots of Crime against Percent Urban, Single Parent, and
Poverty with OLS Fitted Lines 78

ix
x LIST OF FIGURES

4.1 Distribution of β̂1 under the Null Hypothesis for Presidential

Election Example 95
4.2 Distribution of β̂1 under the Null Hypothesis with Larger Standard
Error for Presidential Election Example 99
4.3 Three t Distributions 100
4.4 Critical Values for Large-Sample t Tests 102
4.5 Two Examples of p Values 107
4.6 Statistical Power for Three Values of β1 Given α = 0.01 and a
One-Sided Alternative Hypothesis 110
4.7 Power Curves for Two Values of se(β̂1 ) 113
4.8 Meaning of Confidence Interval for Example of 0.41 ± 0.196 118

5.1 Monthly Retail Sales and Temperature in New Jersey

from 1992 to 2013 129
5.2 Monthly Retail Sales and Temperature in New Jersey with
December Indicated 130
5.3 95 Percent Confidence Intervals for Coefficients in Adult Height,
Adolescent Height, and Wage Models 134
5.4 Economic Growth, Years of School, and Test Scores 143

6.1 Goal Differentials for Home and Away Games for Manchester City
and Manchester United 168
6.2 Bivariate OLS with a Dummy Independent Variable 171
6.3 Scatterplot of Obama Feeling Thermometers and Party
Identification 173
6.4 Three Difference of Means Tests for Review Questions 174
6.5 Scatterplot of Height and Gender 176
6.6 Another Scatterplot of Height and Gender 177
6.7 Fitted Values for Model with Dummy Variable and Control
Variable: Manchester City Example 180
6.8 Relation between Omitted Variable (Obama Vote) and Other
Variables 187
6.9 Confidence Intervals for “Newly Elected” Variable in Table 6.8 190
6.10 Interaction Model of Salaries for Men and Women 192
6.11 Various Fitted Lines from Dummy Interaction Models (for Review
Questions) 194
 LIST OF FIGURES xi

6.12 Heating Used and Heating Degree-Days for Homeowner Who

Installed a Programmable Thermostat 196
6.13 Heating Used and Heating Degree-Days with Fitted Values for
Different Models 199
6.14 Marginal Effect of Text Ban as Total Miles Changes 204

7.1 Average Life Satisfaction by Age in the United States 208

7.2 Life Expectancy and per Capita GDP in 2011 for All Countries
in the World 210
7.3 Linear and Quadratic Fitted Lines for Life Expectancy Data 211
7.4 Examples of Quadratic Fitted Curves 212
7.5 Global Temperature over Time 215
7.6 Hypothetical Investment Data (for Review Questions) 217
7.7 Linear-Log Model for Life Expectancy Data 220

8.1 Robberies and Police for Large Cities in California 250

8.2 Robberies and Police for Specified Cities in California 250
8.3 Robberies and Police for Specified Cities in California with
City-Specific Regression Lines 251
8.4 Robberies and Police for Hypothetical Cities in California 257
8.5 Difference-in-Difference Examples 270
8.6 More Difference-in-Difference Examples (for Review Question) 274

9.1 Conditions for Instrumental Variables 294

9.2 Simultaneous Equation Model 309

10.1 Compliance and Non-compliance in Experiments 334

11.1 Drinking Age and Test Scores 366

11.2 Basic Regression Discontinuity Model, Yi = β0 + β1 Ti + β2 (X1i − C) 369
11.3 Possible Results with Basic RD Model 370
11.4 Possible Results with Differing-Slopes RD Model 374
11.5 Fitted Lines for Examples of Polynomial RD Models 375
11.6 Various Fitted Lines for RD Model of Form
Yi = β0 + β1 Ti + β2 (X1i − C) + β3 (X1i − C)Ti (for Review Question) 377
xii LIST OF FIGURES

11.7 Smaller Windows for Fitted Lines for Polynomial RD Model in

Figure 11.5 379
11.8 Bin Plots for RD Model 380
11.9 Binned Graph of Test Scores and Pre-K Attendance 382
11.10 Histograms of Assignment Variable for RD Analysis 385
11.11 Histogram of Age Observations for Drinking Age Case Study 388

12.1 Scatterplot of Law School Admissions Data and LPM Fitted Line 404
12.2 Misspecification Problem in an LPM 405
12.3 Scatterplot of Law School Admissions Data and LPM- and
Probit-Fitted Lines 407
12.4 Symmetry of Normal Distribution 411
12.5 PDFs and CDFs 412
12.6 Examples of Data and Fitted Lines Estimated by Probit 416
12.7 Varying Effect of X in Probit Model 419
12.8 Fitted Lines from LPM, Probit, and Logit Models 426
12.9 Fitted Lines from LPM and Probit Models for Civil War Data
(Holding Ethnic and Religious Variables at Their Means) 432
12.10 Figure Included for Some Respondents in Global Warming Survey
Experiment 443

13.1 Examples of Autocorrelation 452

13.2 Global Average Temperature since 1880 455
13.3 Global Temperature Data 461
13.4 Data with Unit Roots and Spurious Regression 468
13.5 Data without Unit Roots 469
13.6 Global Temperature and Carbon Dioxide Data 472

A.1 An Example of a Probability Density Function (PDF) 525

A.2 Probabilities That a Standard Normal Random Variable
Is Less than Some Value 526
A.3 Probabilities That a Standard Normal Random Variable
Is Greater than Some Value 527
A.4 Standard Normal Distribution 528
A.5 Two χ Distributions
2
533
A.6 Four F Distributions 535
R.1 Identifying β0 from a Scatterplot 550
LIST OF TABLES

1.1 Donut Consumption and Weight 3

2.1 Descriptive Statistics for Donut and Weight Data 26

2.2 Frequency Table for Male Variable in Donut Data Set 26
2.3 Frequency Table for Male Variable in Second Donut Data Set 27
2.4 Codebook for Height and Wage Data 29
2.5 Descriptive Statistics for State Crime Data 31
2.6 Variables for Winter Olympics Questions 39
2.7 Variables for Height and Weight Data in the United States 40

3.1 Selected Observations from Election and Income Data 51

3.2 Effect of Height on Wages 75
3.3 OLS Models of Crime in U.S. States 77
3.4 Variables for Questions on Presidential Elections and the Economy 86
3.5 Variables for Height and Wage Data in Britain 89
3.6 Variables for Divorce Rate and Hours Worked 89

4.1 Type I and Type II Errors 93

4.2 Effect of Income Changes on Presidential Elections 95
4.3 Decision Rules for Various Alternative Hypotheses 101
4.4 Critical Values for t Distribution 104
4.5 Effect of Height on Wages with t Statistics 104
4.6 Calculating Confidence Intervals for Large Samples 119
4.7 Variables for Height and Wage Data in the United States 123

5.1 Bivariate and Multivariate Results for Retail Sales Data 131
5.2 Bivariate and Multiple Multivariate Results for Height
and Wages Data 133

xiii
xiv LIST OF TABLES

5.3 Using Multiple Measures of Education to Study Economic Growth

and Education 142
5.4 Effects of Judicial Independence on Human Rights 154
5.5 Variables for Height and Wage Data in the United States 161
5.6 Variables for Cell Phones and Traffic Deaths Data 163
5.7 Variables for Speeding Ticket Data 164
5.8 Variables for Height and Wage Data in Britain 165

6.1 Feeling Thermometer Toward Barack Obama 172

6.2 Difference of Means Test for Height and Gender 176
6.3 Another Way to Show Difference of Means Test Results for
Height and Gender 178
6.4 Manchester City Example with Dummy and Continuous
Independent Variables 179
6.5 Using Different Excluded Categories for Wages and Region 183
6.6 Hypothetical Results for Wages and Region When Different
Categories Are Excluded 185
6.7 Difference of Means of Conservatism of Republicans Elected
in 2010 186
6.8 Multivariate OLS Analysis of Republicans Elected in 2010 189
6.9 Interpreting Coefficients in Dummy Interaction Model:
Yi = β0 + β1 Xi + β2 Di + β3 Xi × Di 193
6.10 Data from Programmable Thermostat and Home Heating Bills 197
6.11 Variables for Monetary Policy Data 202
6.12 Variables for Speeding Ticket Data 205

7.1 Global Temperature, 1879–2012 216

7.2 Different Logged Models of Relationship between
Height and Wages 221
7.3 Determinants of Major League Baseball Salaries, 1985–2005 223
7.4 Means and Standard Deviations of Baseball Variables 224
7.5 Means and Standard Deviations of Baseball Variables
for Three Players 225
7.6 Standardized Determinants of Major League Baseball Salaries,
1985–2005 225
7.7 Unrestricted and Restricted Models for F Tests 233
7.8 Variables for Political Instability Data 239
 LIST OF TABLES xv

7.9 Variables for Global Education Data 240

7.10 Variables for Height and Wage Data in Britain 241
7.11 Variables for Speeding Ticket Data 242

8.1 Basic OLS Analysis of Burglary and Police Officers 249

8.2 Example of Robbery and Police Data for Cities in California 255
8.3 Robberies and Police Data for Hypothetical Cities in California 257
8.4 Burglary and Police Officers, Pooled versus Fixed Effects Models 258
8.5 Burglary and Police Officers for Multiple Models Models 265
8.6 Bilateral Trade, Pooled versus Fixed Effects Models 267
8.7 Effect of Stand Your Ground Laws on Homicide Rate per 100,000
Residents 272
8.8 Variables for Presidential Approval Data 280
8.9 Variables for Peace Corps Data 281
8.10 Variables for Instructor Evaluation Data 282
8.11 Variables for the HOPE Scholarship Data 283
8.12 Variables for the Texas School Board Data 284
8.13 Variables in the Cell Phones and Traffic Deaths Data 285

9.1 Levitt (2002) Results on Effect of Police Officers on Violent Crime 289
9.2 Influence of Distance on NICU Utilization (First-Stage Results) 298
9.3 Influence of NICU Utilization on Baby Mortality 299
9.4 Regression Results for Models Relating to Drinking and Grades 300
9.5 First-Stage Reduced Form Regressions for Bush/Iraq War
Simultaneous Equation Model 313
9.6 Second-Stage Results for Bush/Iraq War Simultaneous Equation
Model 314
9.7 Variables for Rainfall and Economic Growth Data 319
9.8 Variables for News Program Data 320
9.9 Variables for Fish Market Data 321
9.10 Variables for Education and Crime Data 322
9.11 Variables for Income and Democracy Data 323
xvi LIST OF TABLES

10.1 Balancing Tests for the Progresa Experiment: Differences of

Means Tests Using OLS 331
10.2 First-Stage Regression in Campaign Experiment: Explaining
Contact 339
10.3 Second-Stage Regression in Campaign Experiment: Explaining
Turnout 340
10.4 Various Measures of Campaign Contact in 2SLS Model for
Selected Observations 341
10.5 First-Stage Regression in Domestic Violence Experiment:
Explaining Arrests 343
10.6 Selected Observations for Minneapolis Domestic Violence
Experiment 344
10.7 Using Different Estimators to Analyze the Results of the
Domestic Violence Experiment 345
10.8 Regression Results for Models Relating Teacher Payment
Experiment (for Review Questions) 352
10.9 Effect of Terror Alerts on Crime 355
10.10 Variables for Get-out-the-Vote Experiment from Gerber and
Green (2005) 359
10.11 Variables for Resume Experiment 361
10.12 Variables for Afghan School Experiment 363

11.1 RD Analysis of Prekindergarten 383

11.2 RD Analysis of Drinking Age and Test Scores 388
11.3 RD Diagnostics for Drinking Age and Test Scores 389
11.4 Variables for Prekindergarten Data 392
11.5 Variables for Congressional Ideology Data 394
11.6 Variables for Head Start Data 396

12.1 LPM Model of the Probability of Admission to Law School 403

12.2 Sample Probit Results for Review Questions 418
12.3 Dog Ownership and Probability of Supporting Obama in 2008
Election 423
12.4 Estimated Effect of Dog Ownership and Ideology on Probability
of Supporting Obama in 2008 Election 425
12.5 Unrestricted and Restricted Probit Results for LR
Test 429
 LIST OF TABLES xvii

12.6 Probit Models of the Determinants of Civil Wars 431

12.7 Variables for Iraq War Data 440
12.8 Variables for Global Warming Data 441
12.9 Variables for Football Coach Data 443
12.10 Variables for Donor Experiment 445
12.11 Balance Tests for Donor Experiment 445

13.1 Using OLS and Lagged Error Model to Detect Autocorrelation 456
13.2 Example of ρ-Transformed Data (for ρ̂ = 0.5) 458
13.3 Global Temperature Model Estimated by Using OLS and via
ρ-Transformed Data 462
13.4 Dickey-Fuller Tests for Stationarity 473
13.5 Change in Temperature as a Function of Change in Carbon Dioxide
and Other Factors 474
13.6 Variables for James Bond Movie Data 480

14.1 Effect of Omitting X2 on Coefficient Estimate for X1 493

14.2 Variables for Winter Olympics Data 499

15.1 Another Set of Variables for Winter Olympics Data 513

A.1 Examples of Standardized Values 530

R.1 Values of β0 , β1 , β2 , and β3 in Figure 8.6 554
 USEFUL COMMANDS FOR STATA

Task Command Example Chapter

Help help help summarize 2

Comment line * * This is a comment line 2
Comment on command line /* */ use "C:\Data.dta" /* This is a comment */
Continue line /* */ reg y X1 X2 X3 /*
*/ X4 X5
Load Stata data file use use "C:\Data.dta" 2
Load text data file insheet insheet using "C:\Data.txt" 2
Display variables in memory list list /* Lists all observations for all variables */ 2
list Y X /* Lists all observations for Y and X */ 2
list X in 1/10 /* Lists first 10 observations for X */ 2
Descriptive statistics summarize summarize X1 X2 Y 2
Frequency table tabulate tabulate X1 2
Scatter plot scatter scatter Y X 2
scatter Y X, mlabel(name) /* Adds labels */ 2
Limit data if summarize X1 if X2 == 1 2
Not equal != summarize X1 if X2!=0 3
Equal (as used in if statement, for example) == summarize X1 if X2 == 1 2
Missing data in Stata . Stata treats missing data as having infinite value, so
list X1 if X2 > 0 will include values of X1 for
which X2 is missing
Regression reg reg Y X1 X2 3
Heteroscedasticity robust regression , robust reg Y X1 X2, robust 3
Generate predicted values predict predict FittedY /* Run this after reg command */ 3
Add regression line to scatter plot twoway, lfit twoway (scatter Y X) (lfit Y X) 3
Critical value for t-distribution, two-sided invttail display invttail(120, .05/2) /* For model with 120 4
degrees of freedom and alpha = 0.05; note that
we divide alpha by 2 */
Critical value for t-distribution, one-sided invttail display invttail(120, .05) /* For model with 120 4
degrees of freedom and alpha = 0.05 */
Critical value for normal distribution, two-sided invnormal display invnormal(.975) /* For alpha = 0.05, note 4
that we divide alpha by 2 */
Critical value for normal distribution, one-sided invnormal display invnormal(.05) 4
Two sided p-values [Reported in reg output] 4
One sided p-values ttail display 2*ttail(120, 1.69) /* For model with 120 4
degrees of freedom and a t-statistic of 1.69 */
Confidence intervals [Reported in reg output] 4
Difference of means test reg reg Y Dum /* Where Dum is a dummy variable */ 6
Create an interaction variable gen gen DumX = Dum * X 6

xviii
 USEFUL COMMANDS FOR STATA xix

Task Command Example Chapter

Create a squared variable gen gen X_sq = X^2 7

Create a logged variable gen gen X_log =log( X) 7
Delete a variable drop drop X7
Produce standardized regression coefficients , beta reg Y X1 X2, beta 7
Produce standardized variable egen egen X_std = std(X) /* Creates variable called 7
X_std */
F-test test test X1 = X2 = 0 /* Run this after regression with 7
X1 and X2 in model */
Critical value for F test invF display invF(2, 120, 0.95) /* Degrees of freedom 7
equal 2 and 120 (order matters!) and alpha =
0.05 */
Generate dummy variables for each unit tabulate and tabulate City, generate(City_dum) 8
generate
LSDV model for panel data reg reg Y X1 X2 City_dum2 - City_dum80 8
De-meaned model for panel data xtreg xtreg Y X1 X2, fe i(City) 8
Two-way fixed-effects xtreg xtreg Y X1 X2 i.year Yr2- Yr10, fe i(City) 8
2SLS model ivregress ivregress 2sls Y X2 X3 (X1 = Z), first 9
Probit probit probit Y X1 X2 X3 12
Logit logit logit Y X1 X2 X3 12
Critical value for chi-square test invchi2 display invchi2(1, 0.95) /* Degrees of freedom = 12
1 and 0.95 confidence level */
Account for autocorrelation in time series data prais tsset Year 13
prais Y X1 X2, corc twostep
Include lagged dependent variable L.Y reg Y L.Y X1 X2 /* Run tsset command first */ 13
Augmented Dickey-Fuller test dfuller dfuller Y, trend lags(1) regress 13
Generate draws from standard normal rnormal gen Noise = rnormal(0,1) /* Length will be same 14
distribution as length of variables in memory */
Indicate to Stata unit and time variables tsset tsset ID time 15
Panel model with autocorrelation xtregar xtregar Y X1 X2, fe rhotype(regress) twostep 15
Include lagged dependent variable L.Y xtreg Y L.Y X1 X2, fe i(ID) 15
Random effects panel model , re xtreg Y X1 X2, re 15
USEFUL COMMANDS FOR R

Task Command Example Chapter

Help ? ?mean 2
Comment line # # This is a comment 2
Load R data file Data = "C:\Data.RData" 2
Load text data file read.table Data = read.table("C:\Data.txt", header = TRUE) 2
Display names of variables in memory objects objects() # Will list names of all variables in memory 2
Display variables in memory [enter variable X1 # Display all values of this variable; enter directly in console 2
name] or highlight in editor and press ctrl-r
X1[1:10] # Display first 10 values of X1 2
Missing data in R NA
Mean mean mean(X1) 2
mean(X1, na.rm=TRUE) # Necessary if there are missing values
Variance var var(X1) 2
var(X1, na.rm=TRUE) # Necessary if there are missing values
sqrt(var(X1)) # This is the standard deviation of X1
Minimum min min(X1, na.rm=TRUE) 2
Maximum max max(X1, na.rm=TRUE) 2
Number of observations sum and is.finite sum(is.finite(X1)) 2
Frequency table table table(X1)
Scatter plot plot plot(X, Y) 2
text(X, Y, name) # Adds labels from variable called "name" 2
Limit data (similar to an if statement) [] plot(Y[X3<10], X1[X3<10]) 2
Equal (as used in if statement, for example) == mean(X1[X2==1]) # Mean of X1 for cases where X2 equals 1
Not equal != mean(X1[X1!=0]) # Mean of X1 for observations where X1 is
not equal to 0
Regression lm lm(Y ~X1 + X2) # lm stands for "linear model" 3
Results = lm(Y~X) # Creates an object called "Results" that 3
stores coefficients, standard errors, fitted values and other
information about this regression
Display results summary summary(Results) # Do this after creating "Results" 3
Install a package install.packages install.packages("AER") # Only do this once for each computer 3
Load a package library library(AER) # Include in every R session in which we use
package specified in command
Heteroscedasticity robust regression coeftest coeftest(Results, vcov = vcovHC(Results, type = "HC1")) 3
# Need to install and load AER package for this command. Do
this after creating OLS regression object called "Results"
Generate predicted values $fitted.values Results$fitted.values # Run after creating OLS regression object 3
called "Results"
Add regression line to scatter plot abline abline(Results) # Run after plot command and after creating 3
"Results" object based on a bivariate regression

xx
 USEFUL COMMANDS FOR R xxi

Task Command Example Chapter

Critical value for t-distribution, two-sided qt qt(0.975, 120) # For alpha = 0.05 and 120 degrees of freedom; 4
divide alpha by 2
Critical value for t-distribution, one-sided qt qt(0.95, 120) # For alpha = 0.05 and 120 degrees of freedom 4
Critical value for normal distribution, two-sided qnorm qnorm(0.975) # For alpha = 0.05; divide alpha by 2 4
Critical value for normal distribution, one-sided qnorm qnorm(0.95) # For alpha = 0.05 4
Two sided p-values [Reported in summary(Results) output]
One sided p-values pt 2*(1-pt(abs(1.69), 120) # For model with 120 degrees of 4
freedom and a t-statistic of 1.69
Confidence intervals confint confint(Results, level = 0.95) # Do after creating OLS object 4
called "Results"
Difference of means test lm lm(Y~Dum) # Where Dum is a dummy variable 6
Create an interaction variable DumX = Dum * X # Or use <- in place of =
Create a squared variable X_sq = X^2 7
Create a logged variable X_log =log( X) 7
Produce standardized regression coefficients scale Res.std = lm(scale(Y) ~scale(X1) + scale(X2) ) 7
Display R squared $r.squared summary(Results)$r.squared 7
Critical value for F test qf qf(.95, df1 = 2, df2 = 120) # Degrees of freedom equal 2 and 7
120 (order matters!) and alpha = 0.05
LSDV model for panel data factor Results = lm(Y ∼ X1 + factor(country)) # Factor adds a 8
dummy variable for every value of variable called country
One-way fixed-effects model (de-meaned) plm library(plm) 8
Results = plm(Y ~X1+ X2+ X3, data = dta,
index=c("country"), model="within")
Two-way fixed-effects model (de-meaned) plm library(plm) 8
Results = plm(Y ~X1+ X2+ X3 data = dta,
index=c("country", "year"), model="within",
effect = "twoways")
2SLS model ivreg library(AER) 9
ivreg(Y ~X1 + X2 + X3 |Z1 + Z2 + X2 + X3)
Generate draws from standard normal distribution rnorm Noise = rnorm(500) # 500 draws from standard normal 14
distribution
Panel model with autocorrelation [See Computing Corner in Chapter 15] 15
Include lagged dependent variable plm with Results = plm(Y ~lag(Y) + X1 + X2, data = dta, index = 15
lag(Y) c("ID", "time"), effect = "twoways")
Random effects panel model plm with Results = plm(Y ~X1 + X2, data = dta, model = "random") 15
"random"
PREFACE FOR STUDENTS:
HOW THIS BOOK CAN HELP YOU
LEARN STATISTICS

“Less dull than traditional texts.”—Student A.H.

“It would have been immensely helpful for me to have a textbook like this in
my classes throughout my college and graduate experience. It feels more like
an interactive learning experience than simply reading equations and facts out
of a book and being expected to absorb them.”—Student S.A.

“I wish I had had this book when I was first exposed to the material—it would
have saved a lot of time and hair-pulling. . . .”—Student J.H.
“Material is easy to understand, hard to forget.”—Student M.H.

This book introduces the statistical tools necessary to answer important questions.
Do antipoverty programs work? Does unemployment affect inflation? Does
campaign spending affect election outcomes? These and many more questions
not only are interesting but also are important to answer correctly if we want to
support policies that are good for people, countries, and the world.
When using statistics to answer such questions, we need always to remember
a single big idea: correlation is not causation. Just because variable Y rises when
variable X rises does not mean that variable X causes variable Y to rise. The
essential goal of statistics is to figure out when we can say that changes in variable
X will lead to changes in variable Y.
This book helps us learn how to identify causal relationships with three
features seldom found in other statistics and econometrics textbooks. First, it
focuses on the tools that researchers use most. These are the real stats that help us
make reasonable claims about whether X causes Y. By using these tools, we can
produce analyses that others can respect. We’ll get the most out of our data while
recognizing the limits in what we can say or how confident we can be.
Our emphasis on real stats means that we skip obscure statistical tools that
could come up under certain conditions: they are not discussed here. Statistics
is too often complicated by books and teachers trying to do too much. This book
shows that we can have a sophisticated understanding of statistical inference with-
out having to catalog every method that our instructor had to learn as a student.
Second, this book works with a single unifying framework. We don’t start
over with each new concept; instead, we build around a core model. That means

xxii
 PREFACE FOR STUDENTS xxiii

there is a single equation and a unifying set of assumptions that we poke, probe,
and expand throughout the book. This approach reduces the learning costs of
moving through the material and allows us to go back and revisit material. As
with any skill, we probably won’t fully understand any given technique the
first time we see it. We have to work at it, we have to work with it. We’ll
get comfortable, we’ll see connections. Then it will click. Whether the skill is
jumping rope, typing, throwing a baseball, or analyzing data, we have to do things
many times to get good at it. By sticking to a unifying framework, we have more
chances to revisit what we have already learned. You’ll also notice that I’m not
afraid to repeat myself on the important stuff. Really, I’m not afraid to repeat
myself.
Third, this book uses many examples from the policy, political, and economic
worlds. So even if you do not care about “two-stage least squares” or “maximum
likelihood” in and of themselves, you will see how understanding these techniques
will affect what you think about education policy, trade policy, election outcomes,
and many other interesting issues. The examples make it clear that the statistical
tools developed in this book are being used by contemporary social scientists who
are actually making a difference with their empirical work.
Real Stats is for people who care about policy, politics, economics, and law.
It can serve as the primary course textbook in applied statistics and research
methods courses or as a supplemental text providing more intuition and context
in more advanced methods course. As more and more public policy and corporate
decisions are based on statistical analysis, this book can also be used outside of
course work. Statistical analysis has infiltrated into every area of our lives—from
entertainment to sports (I no longer spit out my coffee when I come across
an article on regression analysis of National Hockey League players)—and a
working knowledge of basic statistical techniques can help anyone make better
sense of the world around them.

What’s In This Book?

The preparation necessary to use this book successfully is modest. We use basic
algebra a fair bit, being careful to explain every step. You do not need calculus.
We refer to calculus when useful, and the book certainly could be used by a course
that works through some of the concepts using calculus, but you can understand
everything without knowing calculus.
We start with two introductory chapters. Chapter 1 lays out the challenge of
statistical inference. This is the challenge of making probabilistic yet accurate
claims about causal relations between variables. We present experiments as an
ideal way to conduct research, but we also show how experiments in the real world
are tricky and can’t answer every question we care about. This chapter provides
the “big picture” context for statistical analysis that is every bit as important as
the specifics that follow.
Chapter 2 provides a practical foundation related to good statistical practices.
In every statistical analysis, data meets software, and if we’re not careful, we
xxiv PREFACE FOR STUDENTS

lose control. This chapter therefore seeks to teach good habits about documenting
analysis and understanding data.
The five chapters of Part I constitute the heart of the book. They introduce
ordinary least squares (OLS), also known as regression analysis. Chapter 3
introduces the most basic regression model, the bivariate OLS model. Chapter
4 shows how to use OLS to test hypotheses. Chapters 5 through 7 introduce
the multivariate OLS model and applications. By the end of Part I, you will
understand regression and be able to control for anything you can measure. You’ll
also be able to fit curves to data and assess whether the effects of some variables
differ across groups, among other skills that will impress your friends.
Part II introduces techniques that constitute the modern statistical toolkit.
These are the techniques people use when they want to get published—or
paid. These techniques build on multivariate OLS to give us a better chance of
identifying causal relations between two variables. Chapter 8 covers a simple yet
powerful way to control for many factors we can’t measure directly. Chapter 9
covers instrumental variable techniques, which work if we can find a variable
that affects our independent variable but not our dependent variable. Instrumental
variable techniques are a bit funky, but they can be very useful for isolating causal
effects. Chapter 10 covers randomized experiments. Although ideal in theory, in
practice such experiments often raise a number of statistical challenges we need
to address. Chapter 11 covers regression discontinuity tools that can be used when
we’re studying the effect of variables that were allocated based on a fixed rule. For
example, Medicare is available to people in the United States only when they turn
65; admission to certain private schools depends on a test score exceeding some
threshold. Focusing on policies that depend on such thresholds turns out to be a
great context for conducting credible statistical analysis.
Part III covers dichotomous dependent variable models. These are simply
models in which the outcome we care about takes on two possible values.
Examples include high school graduation (someone graduates or doesn’t),
unemployment (someone has a job or doesn’t), and alliances (two countries sign
an alliance treaty or don’t). We show how to apply OLS to such models and
then provide more elaborate models that address the deficiencies of OLS in this
context.
Part IV supplements the book with additional useful material. Chapter 13
covers time series data. The first part is a variation on OLS; the second part
introduces dynamic models that differ from OLS models in important ways.
Chapter 14 derives important OLS results and extends discussion on specific
topics. Chapter 15 goes into greater detail on the vast literature on panel data,
showing how the various strands fit together.
Chapter 16 concludes the book with tips on adopting the mind-set of a
statistical realist. In fact, if you are looking for an overall understanding of the
power and limits of statistics, you might want to read this chapter first—then
read it again once you’ve learned all the statistical concepts covered in the other
chapters.
Another Random Scribd Document
with Unrelated Content
"Of course, you're quite right, baron; go by all means, but conceal
yourself carefully, so that you may not be surprised."
"No, I was wrong," said Canolles, "and you are right. But how can I
warn her?"
"It seems to me that a letter—"
"Who will carry it?"
"I thought that I saw a servant following you. A servant, under such
circumstances, runs the risk of nothing worse than a few blows,
while a gentleman risks his life."
"Verily I am losing my wits," said Canolles; "Castorin will do the
errand to perfection, especially as I suspect that the rascal has allies
in the house."
"You see that the matter can all be arranged here," said the
viscount.
"Yes. Have you writing materials?"
"No; but they have them downstairs."
"Pardon me, I beg you," said Canolles; "upon my word, I can't
imagine what has happened to me this evening, for I say one idiotic
thing after another. No matter. Thanks for your good advice,
viscount, and I shall act upon it immediately."
Without taking his eyes from the young man, whom he had been
examining for some moments with strange persistency, Canolles
backed to the door and descended the stairs, while the viscount
muttered anxiously,—
"How he stares at me! can he have recognized me?"
Canolles meanwhile had gone down to the ground-floor, and having
gazed for a moment with profound sorrow at the quail, partridge,
and sweetmeats, which Master Biscarros was himself packing in the
hamper upon the head of his assistant cook, and which another than
he was to eat perhaps, although they were certainly intended for
him, he asked to be shown to his room, called for writing materials,
and wrote to Nanon the following letter:—

Dear Madame—About a hundred yards from your door, if nature

had endowed your lovely eyes with the power to see in the
dark, you could descry in a clump of trees Monsieur le Duc
d'Épernon, who is awaiting my coming to have me shot, and
compromise you wofully as a consequence. But I am by no
means anxious to lose my life or to cause you to lose your
peace of mind. Have no fear, therefore, in that direction. For my
own part I propose to make use of the leave of absence which
you procured for me the other day that I might take advantage
of my freedom to come and see you. Where I am going, I have
no idea; indeed, I am not sure that I shall go anywhere.
However that may be, recall your fugitive adorer when the
storm has passed. They will tell you at the Golden Calf in which
direction I have gone. You will give me due credit, I trust, for
my self-sacrifice. But your interests are dearer to me than my
own enjoyment. I say my own enjoyment because I should
have enjoyed pommelling Monsieur d'Épernon and his minions
under their disguise. Believe me, dear lady, your most devoted
and most faithful servant.

Canolles signed this effusion, overflowing with Gascon

magniloquence, knowing the effect it would have upon the Gascon
Nanon. Then he summoned his servant.
"Come hither, Master Castorin," said he, "and tell me frankly on what
terms you are with Mademoiselle Francinette."
"But, monsieur," replied Castorin, wondering much at the question,
"I don't know if I ought—"
"Have no fear, master idiot; I have no designs upon her, and you
haven't the honor of being my rival. I ask the question simply for
information."
"Ah! that's a different matter, monsieur, and I may say that
Mademoiselle Francinette has deigned to appreciate my good
qualities."
"Then you are on the best of terms with her, aren't you, monsieur
puppy? Very good. Take this letter and go around by the fields."
"I know the road, monsieur," said Castorin, with a self-satisfied
expression.
"'T is well. Knock at the back door. No doubt you know that door,
too?"
"Perfectly well."
"Better and better. Take that road, knock at that door, and hand this
letter to Mademoiselle Francinette."
"Then, monsieur," said Castorin, joyfully, "I may—"
"You may start instantly; you have ten minutes to go and come. This
letter must be delivered to Mademoiselle Nanon de Lartigues at the
earliest possible moment."
"But suppose they don't open the door, monsieur?" queried Castorin,
suspecting that something had gone wrong.
"Why, you must be a fool in that case, for you should have some
particular way of knocking, which makes it certain that a brave
fellow like you won't be left outside; if that's not the case, I am
much to be pitied for having such a dolt in my service."
"I have a private knock, monsieur," said Castorin, with his most
conquering air; "first I knock twice softly, then a third—"
"I don't ask you how you knock, nor do I care, if the door is opened.
Begone! and if you are taken by surprise, eat the paper, or I'll cut off
your ears when you return, if it's not already done."
Castorin was off like a flash. But when he reached the foot of the
staircase he stopped, and, in defiance of all rules, thrust the letter
into the top of his boot; then he left the inn by the barn-yard door,
and made a long circuit, sneaking through the bushes like a fox,
jumping ditches like a greyhound, until at last he reached the rear
door of the little house, and knocked in the peculiar fashion he had
tried to explain to his master.
It proved to be so effective that the door opened instantly.
Ten minutes later, Castorin returned to the inn without accident, and
informed his master that the letter was in the fair hands of
Mademoiselle Nanon.
Canolles had employed these ten minutes in opening his
portmanteau, laying out his robe de chambre, and ordering his
supper to be served. He listened with visible satisfaction to Monsieur
Castorin's report, and made a trip to the kitchen, giving his orders
for the night in a loud tone, and yawning immoderately, like a man
who is impatient for bed-time to arrive. These manœuvres were
intended to convince the Duc d'Épernon, if he had put spies upon
him, that the baron had never intended to go farther than the inn,
where he had stopped for supper and lodging, like the
unpretentious, inoffensive traveller he was. And the scheme really
produced the result that the baron hoped. A man in the guise of a
peasant, who was drinking in the darkest corner of the public room,
called the waiter, paid his reckoning, rose, and went out
unconcernedly, humming a tune. Canolles followed him to the door,
and saw that he went toward the clump of trees; in a few moments
he heard the receding steps of several horses,—the ambuscade was
raised.
Thereupon the baron, with his mind at rest concerning Nanon,
thought only of passing the evening as agreeably as possible; he
therefore bade Castorin bring cards and dice, and, having done so,
to go and ask the Vicomte de Cambes if he would do him the honor
to receive him.
Castorin obeyed, and found at the vicomte's door an old, white-
haired squire, who held the door half open, and replied surlily to his
complimentary message,—
"Impossible at present; Monsieur le Vicomte is very much engaged."
"Very well," said Canolles, when the answer was reported to him, "I
will wait."
As he heard considerable noise in the direction of the kitchen, to
pass the time away he went to see what was going on in that
important part of the establishment.
The uproar was caused by the return of the poor scullion, more dead
than alive. At the bend in the road he was stopped by four men,
who questioned him as to the objective point of his nocturnal
expedition; and upon learning that he was carrying supper to the
lady at the little house among the trees, stripped him of his cap, his
white waistcoat and his apron. The youngest of the four then
donned the distinctive garb of the victim's profession, balanced the
hamper on his head, and kept on toward the little house in the place
and stead of the scullion. Not long after, he returned, and talked in a
low tone with the man who seemed the leader of the party. Then
they restored his vest and cap and apron, replaced the hamper on
his head, and gave him a kick in the stern to start him in the
direction he was to follow. The poor devil asked for no more definite
instructions; he started off at full speed, and fell half-dead with
terror at the door of the inn, where he had just been picked up.
This episode was quite unintelligible to everybody except Canolles;
and as he had no motive for explaining it, he left host, waiters,
chambermaids, cook, and scullion to cudgel their brains over it;
while they were outdoing one another in wild conjectures, he went
up to the vicomte's door, and, assuming that the first message he
had sent him by the mouth of Monsieur Castorin permitted him to
dispense with a second formality of the same nature, he opened the
door unceremoniously and went in.
A table, lighted and set with two covers, stood in the middle of the
room, awaiting, to be complete, only the dishes with which it was to
be embellished.
Canolles noticed the two covers, and drew a joyful inference
therefrom.
However, the viscount when he saw him standing in the doorway,
jumped to his feet so suddenly that it was easy to see that he was
greatly surprised by the visit, and that the second cover was not
intended for the baron, as he flattered himself for an instant that it
was. His doubts were set at rest by the first words the viscount
uttered.
"May I be permitted to know, Monsieur le Baron," he asked, walking
to meet him ceremoniously, "to what new circumstance I am
indebted for the honor of this visit?"
"Why," rejoined Canolles, somewhat taken aback by this ungracious
reception, "to a very natural circumstance. I am hungry. I thought
that you must be in the same plight. You are alone, I am alone; and
I wished to have the honor of suggesting to you that we sup
together."
The viscount looked at Canolles with evident distrust, and seemed to
feel some embarrassment in answering him.
"Upon my honor!" said Canolles, laughing, "one would say that I
frighten you; are you a knight of Malta, pray? Are you destined for
the Church, or has your respectable family brought you up in holy
horror of the Canolles? Pardieu! I shall not ruin you if we pass an
hour together on opposite sides of a table."
"Impossible for me to go to your room, baron."
"Very well, don't do it. But as I am already here—"
"Even more impossible, monsieur; I am expecting some one."
This time Canolles was disarmed.
"You are expecting some one?" he said.
"Yes."
"'Faith," said Canolles, after a moment of silence, "I should almost
prefer that you had let me go on at any risk, rather than spoil, by
your manifest repugnance for my society, the service you rendered
me, for which I fear that I have not as yet thanked you sufficiently."
The young man blushed and walked to Canolles' side.
"Forgive me, monsieur," he said in a trembling voice; "I realize how
rude I am; and if it were not serious business, family matters, which
I have to discuss with the person I expect, it would be both an
honor and a pleasure to admit you as a third, although—"
"Oh, finish!" said Canolles; "whatever you say, I am determined not
to be angry with you."
"Although," continued the viscount, "our acquaintance is one of the
unforeseen results of mere chance, one of those fortuitous
meetings, one of those momentary relations—"
"Why so?" queried Canolles. "On the contrary, the most sincere and
enduring friendships are formed in this way: we simply have to give
credit to Providence for what you attribute to chance."
"Providence, monsieur," the viscount rejoined with a laugh, "decrees
that I depart two hours hence, and that, in all probability, I take the
opposite direction to that you will take; receive, therefore, my
sincere regrets at my inability to accept, gladly as I would do so if I
could, the friendship you offer me so cordially, and of which I fully
appreciate the worth."
"You are a strange fellow, upon my word," said Canolles, "and the
generous impulse upon which you acted in the first place gave me
quite a different idea of your character. But of course it shall be as
you desire; I certainly have no right to persist, for I am your debtor,
and you have done much more for me than I had any right to expect
from a stranger. I will return, therefore, to my own room, and sup
alone; but I assure you, viscount, it goes against my grain. I am not
addicted to monologue."
Indeed, notwithstanding what he said, and his declared purpose to
withdraw, Canolles did not withdraw; some power that he could not
understand seemed to nail him to his place; he felt irresistibly drawn
to the viscount, who, however, took up a candle and approached him
with a charming smile.
"Monsieur," said he, extending his hand, "however that may be, and
short as our acquaintance has been, I beg you to believe that I am
overjoyed to have been of service to you."
Canolles heard nothing but the compliment; he seized the hand the
viscount offered him, which was warm and soft, and, instead of
answering his friendly, masculine pressure, was withdrawn at once.
Realizing that his dismissal was none the less a dismissal, although
couched in courteous phrase, he left the room, disappointed and
thoughtful.
At the door he encountered the toothless smile of the old valet, who
took the candle from the viscount's hands, ceremoniously escorted
Canolles to his door, and hastened back to his master, who was
waiting at the top of the stairs.
"What is he doing?" the viscount asked in an undertone.
"I think he has made up his mind to take supper alone," replied
Pompée.
"Then he won't come up again?"
"I hope not, at least."
"Order the horses, Pompée; it will be so much time gained. But what
is that noise?"
"I should say it was Monsieur Richon's voice."
"And Monsieur de Canolles?"
"They seem to be quarrelling."
"On the contrary, they are greeting each other. Listen!"
"If only Richon does n't say anything."
"Oh! there's no fear of that; he's very circumspect."
"Hush!"
As they ceased to speak, they heard Canolles' voice.
"Two covers, Master Biscarros," he cried. "Two covers! Monsieur
Richon sups with me."
"By your leave, no," replied Richon; "it's impossible."
"The deuce! so you too propose to sup alone, like the young
gentleman upstairs?"
"What gentleman?"
"The one upstairs, I say."
"What's his name?"
"Vicomte de Cambes."
"Oho! you know the viscount, do you?"
"Pardieu! he saved my life."
"He?"
"Yes, he."
"How was that?"
"Sup with me, and I'll tell you the whole story during supper."
"I cannot; I am to sup with him."
"Ah! yes; he is awaiting some one."
"Myself; and as I am late, you will allow me to leave you, will you
not, baron?"
"Sacrebleu! no, I will not allow it!" cried Canolles. "I have taken it
into my head that I will sup in company, and you will sup with me or
I with you. Master Biscarros, two covers!"
But while Canolles turned his back to see if the order was executed,
Richon darted rapidly up the staircase. When he reached the top
stair a little hand met his and drew him into the viscount's room, the
door of which immediately closed behind him, and was locked and
bolted for greater security.
"In very truth," muttered Canolles, looking about in vain for Richon,
and seating himself at his solitary table, "in very truth, I don't know
what the people of this cursed country have against me; some of
them run after me to kill me, and others avoid me as if I had the
plague. Corbleu! my appetite is vanishing; I feel that I am growing
melancholy, and I am capable of getting as drunk as a lansquenet
to-night. Holé! Castorin, come here and be thrashed. Why, they are
locking themselves in up there as if they were conspiring. Double
calf that I am! of course they are conspiring; that's just it, and it
explains everything. The next question is, in whose interest are they
conspiring?—the coadjutor's? the princes'? the parliament's? the
king's? the queen's? Monsieur de Mazarin's? 'Faith, they may
conspire against any one they choose, it's all the same to me; and
my appetite has returned. Castorin, order up my supper, and give
me some wine; I forgive you."
Thereupon Canolles philosophically attacked the first supper that
was prepared for the Vicomte de Cambes, which Master Biscarros
was compelled to serve up to him, warmed over, for lack of supplies.

IV.

Let us now see what was taking place under Nanon's roof while
Baron de Canolles was vainly seeking some one to sup with him,
until, growing weary of the profitless quest, he decided at last to sup
by himself.
Nanon, whatever her enemies may have said or written—and among
her enemies must be accounted the great majority of the historians
who have devoted any space to her—was, at this period, a charming
creature of some twenty-five or twenty-six years; small of stature,
dark-skinned, but with a supple, graceful figure, bright, fresh
coloring, eyes of deepest black, in whose limpid depths all the
passions and emotions found expression: gay on the surface, in
appearance a laughing siren. But Nanon was very far from giving her
mind to the whims and follies which embroider with fantastic designs
the silky and golden woof of which the life of a petite-maîtresse
ordinarily consists. On the contrary, the most weighty conclusions,
long and laboriously reasoned out in her shapely head, assumed an
aspect no less seductive than clear when enounced by her vibrating
voice, in which the Gascon accent was very marked. No one would
have divined the untiring perseverance, the invincible tenacity, and
the statesmanlike depth of insight which lay beneath that rosy,
smiling mask, behind that look overflowing with voluptuous promise,
and glowing with passion. And yet such were Nanon's qualities, good
or bad according as we look at the face or the reverse of the medal.
Such was the scheming mind, such the ambitious heart, to which her
seductive body served as envelope.
Nanon was of Agen. Monsieur le Duc d'Épernon, son of that
inseparable friend of Henri IV. who was in his carriage when
Ravaillac's knife struck him, and was the object of suspicions which
did not stop short of Marie de Médicis—Monsieur le Duc d'Épernon,
appointed governor of Guyenne, where his arrogance, his insolence,
and his exactions caused him to be generally execrated, was
captivated by the little creature, who was the daughter of a simple
attorney. He paid court to her, and conquered her scruples with
great difficulty, and after a long defence maintained with the skill of
a consummate tactician determined that the victor shall pay the full
price of his victory.
But, as the ransom of her thenceforth ruined reputation, Nanon had
despoiled the duke of his power and his freedom. At the end of the
first six months of her liaison with the governor of Guyenne, she was
the de facto ruler of that fair province, returning with interest the
injuries and insults she had received from all those who had slighted
or humiliated her. A queen by chance, she became a tyrant by
design, shrewdly realizing the advisability of supplementing the
probable brevity of her reign by abusing her power.
As a consequence, she seized upon everything she could reach,—
treasure, influence, honors. She was enormously wealthy, distributed
appointments, received visits from Mazarin and the leading
noblemen at court. With admirable skill she made of the various
elements that she had at her disposal a combination useful to her
credit, and profitable to her fortune. Every service that Nanon
rendered had its stated price. There was a regularly established tariff
for appointments in the army and in the magistracy: Nanon would
procure this position or that for some fortunate individual, but it
must be paid for in hard cash or by a royal gift; so that when she
relaxed her hold upon a fragment of power for the benefit of one
person or another, she recouped the fragment in another form,
giving up the authority, but retaining the money, which is its active
principle.
This explains the duration of her reign; for men, in their hatred,
hesitate to overthrow an enemy who will have any consolation
remaining in his downfall. Vengeance thirsts for total ruin, for
complete prostration. Nations are reluctant to expel a tyrant who
would carry away their money, and depart with smiling face. Nanon
de Lartigues had two millions.
And so she lived in a species of security over the volcano which was
unceasingly shaking everything about her to its foundations. She had
felt the popular hatred rise like the tide, increase in force, and assail
with its waves the power of Monsieur d'Épernon, who, when hunted
from Bordeaux in a day of wrath, had carried Nanon in his wake, as
the ship carries the skiff. Nanon bent before the storm, ready to
stand erect again when it should have passed; she had taken
Monsieur de Mazarin for her model, and, an humble pupil, she
practised at a distance the political tactics of the clever and pliable
Italian. The cardinal's notice was attracted by this woman, who
waxed great and wealthy by the same method which had made him
a prime minister, possessed of fifty millions. He admired the little
Gasconne; he did more than that,—he let her do as she chose.
Perhaps we shall eventually know why.
Notwithstanding all this, and although some who claimed to be
better informed averred that she corresponded directly with
Monsieur de Mazarin, but little was said of the fair Nanon's political
intrigues. Canolles himself, who, however, being young and rich and
handsome, could not understand the need of intriguing, did not
know what to think upon that point. As to love-affairs, whether it
was that Nanon, in her preoccupation by more serious matters, had
postponed them to a more convenient season, or that the gossip
caused by Monsieur d'Épernon's passion drowned whatever noise
any secondary amours might have made, even her enemies were
not lavish of scandalous reports in her regard, and Canolles was
justified in believing, as a matter of personal and national self-
esteem, that Nanon was invincible before his appearance upon the
scene. It may be that Canolles was, in truth, the beneficiary of the
first real passion of that heart, hitherto accessible to ambition only;
it may be that prudence had enjoined upon his predecessors
absolute silence. At all events, Nanon, as mistress, was a fascinating
woman; Nanon, insulted, was like to be a redoubtable foe.
The acquaintance between Nanon and Canolles had come about in
the most natural way. Canolles, a lieutenant in the Navailles
regiment, aspired to the rank of captain; in order to obtain the
promotion, he was obliged to write to Monsieur d'Épernon, colonel-
general of infantry. Nanon read the letter, and replied in the ordinary
way, making a business appointment with Canolles. He selected from
among his family jewels a magnificent ring, worth some five hundred
pistoles (it was less expensive than to purchase a company), and
betook himself to the place appointed for the meeting. But on this
occasion Canolles, preceded by the renown of his previous triumphs,
upset all Mademoiselle de Lartigues' calculations. It was the first
time that he had seen Nanon; it was the first time that Nanon had
seen him; they were both young, handsome, and clever. Their
conversation consisted chiefly of reciprocal compliments; not a word
was said concerning the business which brought them together, and
yet the business was done. The next day Canolles received his
captain's commission, and when the ring passed from his finger to
Nanon's it was not as the price of gratified ambition, but as a pledge
of mutual love.

V.

A few words will suffice to explain Nanon's residence near the village
of Matifou. As we have said, the Duc d'Épernon was intensely hated
in Guyenne. Nanon, who had been honored by being transformed
into his evil genius, was execrated. The popular outcry drove them
from Bordeaux to Agen. But at Agen it began anew. One day the
gilded carriage in which Nanon was driving to join the duke was
overturned upon a bridge. By some unexplained means, Nanon
found herself in the river, and Canolles pulled her out. One night
Nanon's residence in the city took fire, and Canolles it was who
made his way to her bedroom and saved her from the flames. Nanon
concluded that the Agenois might probably succeed at the third trial.
Although Canolles left her side as little as possible, it would be a
miracle if he should always happen to be on hand at a given point to
rescue her. She availed herself of the duke's absence on a tour
through the province, and of an escort of twelve hundred men, of
whom the Navailles regiment furnished its quota, to leave the city at
the same time with Canolles, hurling defiance from her carriage
windows at the populace, who would have liked nothing so much as
to wreck the carriage, but dared not.
Thereafter the duke and Nanon selected, or rather Canolles had
secretly selected for them, the little country-house where it was
decided that Nanon should remain while an establishment was being
prepared for her at Libourne. Canolles procured a leave of absence,
ostensibly in order to attend to some private business at his home,
really so that he might be at liberty to leave his regiment, which had
returned to Agen, and to remain within a reasonable distance of
Matifou, where his protecting presence was more necessary than
ever.
In fact, events were becoming alarmingly serious. The princes of
Condé, Conti, and Longueville, who had been arrested and
imprisoned at Vincennes on the 17th January preceding, afforded an
excellent pretext for civil war to the four or five factions which
divided France at that epoch. The unpopularity of the Duc
d'Épernon, who was known to be entirely devoted to the court,
continued to increase, although it was reasonable to hope that it had
reached its limit. A catastrophe, earnestly desired by all the factions,
who, under the extraordinary conditions prevailing in France at the
time, did not themselves know where they stood, was imminent.
Nanon, like the birds which see the storm approaching, disappeared
from the sky and betook herself to her leafy nest, there to await the
result unknown and in obscurity.
She gave herself out as a widow, desirous of living in seclusion. So
Master Biscarros described her, the reader will remember.
Monsieur d'Épernon paid her a visit, and announced his intention of
being absent for a week. As soon as he took his leave of her, Nanon
sent by the tax-collector, her protégé, a little note to Canolles, who
was making use of his leave of absence to remain in the
neighborhood. But, as we have seen, the original note had
disappeared in the messenger's hands, and had become a copy
under Cauvignac's pen. The reckless young nobleman was making
all haste to obey the summons contained therein, when the Vicomte
de Cambes stopped him four hundred yards from his destination. We
know the rest.
Nanon therefore was awaiting Canolles, as a woman who loves is
wont to await the loved one, consulting her watch ten times a
minute, walking to the window again and again, listening to every
sound, gazing questioningly at the sun as it sank in ruddy splendor
behind the mountain, to give place to the first shades of night. The
first knocking was at the front door, and she despatched Francinette
thither; but it was only the pseudo-waiter from the inn, bringing the
supper for which the guest was lacking. Nanon looked out into the
hall and saw Master Biscarros' false servant, who, for his part, stole
a glance into the bedroom, where a tiny table was set with two
covers. Nanon bade Francinette keep the dishes hot, then sadly
closed the door and returned to the window, which showed her the
road still deserted as far as she could see it in the gathering
darkness.
A second knock, a peculiar knock, was heard, this time at the back
door, and Nanon cried,—
"Here he is!"
But still she feared that it was not he, and stopped in the middle of
the room. The next moment the door opened, and Mademoiselle
Francinette appeared on the threshold in evident consternation,
holding the letter in her hand. Nanon spied the paper, rushed up to
her, tore it from her hand, hastily opened it, and read it in an agony
of fear.
The perusal of the letter was like a thunder-clap to Nanon. She
dearly loved Canolles, but with her, ambition was almost equal to
love, and in losing the Duc d'Épernon she would lose not only all her
hopes of fortune to come, but perhaps her accumulated wealth as
well. However, as she was a quick-witted siren, she began by putting
out the candle, which would have caused her shadow to betray her
movements, and ran to the window. It was time; four men were
approaching the house, and were not more than fifty feet away. The
man in the cloak walked first, and in the man in the cloak Nanon
recognized the duke beyond a peradventure. At that moment
Mademoiselle Francinette entered, candle in hand. Nanon glanced
despairingly at the table and the two covers, at the two arm-chairs,
at the two embroidered pillows, which displayed their insolent
whiteness against the background of crimson damask bed-curtains,
and at her fascinating négligé, which harmonized so well with all the
rest.
"I am lost!" she thought.
But almost immediately her wits returned to her, and a smile stole
over her face; like a flash she seized the plain glass tumbler
intended for Canolles, and threw it out into the garden, took from its
box a golden goblet adorned with the duke's arms, and placed
beside his plate his silver cover; then, shivering with fear, but with a
forced smile upon her face, she rushed down the stairs, and reached
the door just as a grave, solemn blow was struck upon it.
Francinette was about to open the door, but Nanon caught her by
the arm, thrust her aside, and said, with that swift glance which,
with women taken by surprise, serves so well to complete their
thought,—
"I am waiting for Monsieur le Duc, not for Monsieur de Canolles.
Serve the supper!"
With that she drew the bolts herself, and threw herself upon the
neck of the man with the white plume, who was preparing to greet
her with a most savage expression.
"Ah!" she cried, "my dream did not play me false! Come, my dear
duke, everything is ready, and we will go to supper at once."
D'Épernon was dumfounded; however, as a caress from a pretty
woman is always acceptable, he allowed himself to be kissed.
But the next moment he remembered what overwhelming proof he
possessed.
"One moment, mademoiselle," he said; "let us have an
understanding, if you please."
With a wave of his hand to his followers, who drew back respectfully
but did not go away altogether, the duke entered the house alone,
with slow and measured step.
"Pray, what's the matter, my dear duke?" said Nanon, with such well-
feigned gayety that any one might have thought it natural; "did you
forget something the last time you were here, that you look around
so carefully on all sides?"
"Yes," said the duke; "I forgot to tell you that I am not a
consummate ass, a Géronte, such as Monsieur Cyrano de Bergerac
introduces in his comedies, and having forgotten to tell it you, I have
returned in person to prove it to you."
"I do not understand you, monseigneur," said Nanon, with the most
tranquil and sincere expression imaginable. "Explain yourself, I beg."
The duke's eyes rested on the two arm-chairs, and passed thence to
the two covers and the two pillows. There they paused for a longer
time, while an angry flush overspread his face.
Nanon had foreseen all this, and she awaited the result of his
scrutiny with a smile which disclosed her pearly teeth. But the smile
strongly resembled a contraction of the nerves, and her teeth would
have chattered if anguish had not kept them pressed tightly
together.
The duke at last fixed his wrathful gaze upon her.
"I am still awaiting your Lordship's pleasure," said Nanon, with a
graceful courtesy.
"My Lordship's pleasure is that you explain this supper."
"I have already told you that I dreamed that you would return to-
day although you left me only yesterday. My dreams never fail to
come true, so I ordered this supper purposely for you."
The duke made a grimace which he intended to pass off for an
ironical smile.
"And the two pillows?" he said.
"Pray, is it monseigneur's intention to return to Libourne? In that
case, my dream lied to me, for it told me that monseigneur would
remain."
The duke made a second grimace even more significant than the
first.
"And this charming négligé, madame? And these exquisite
perfumes?"
"It is one of those I am accustomed to wear when I expect
monseigneur. The perfume comes from sachets of peau d'Espagne,
which I put in my wardrobes, and which monseigneur has often told
me he preferred to all others, because it is the queen's favorite
perfume."
"And so you were expecting me?" rejoined the duke, with a sneering
laugh.
"Good lack, monseigneur," said Nanon, frowning; "I believe, God
forgive me, that you would like to look in the closets. Are you
jealous by any chance?"
Nanon laughed aloud, whereat the duke assumed his most majestic
air.
"I, jealous? No, no! Thank God, I'm no such idiot as that. Being old
and rich, I know naturally that I was made to be deceived, but I
propose to prove to those who deceive me that I am not their
dupe."
"How will you prove it, pray? I am curious to know."
"Oh! it will be an easy matter. I shall simply have to show them this
paper."
He took a letter from his pocket.
"I don't dream, myself," he said; "at my age one doesn't dream,
even when awake; but I receive letters. Read this one; it's very
interesting."
Nanon shuddered as she took the letter the duke handed her, and
started when she saw the writing; but the movement was
imperceptible, and she read,—

É
"'Monseigneur le Duc d'Épernon is informed that a man who, for six
months past, has been on familiar terms with Mademoiselle Nanon
de Lartigues, will visit her this evening, and will remain to supper
and to sleep.
"'As I do not desire to leave Monseigneur le Duc d'Épernon in
uncertainty, he is informed that his fortunate rival is Monsieur le
Baron de Canolles.'"
Nanon turned pale; the blow struck home.
"Ah! Roland! Roland!" she murmured, "I believed myself to be well
rid of you."
"Am I well informed?" queried the duke, triumphantly.
"Not by any means," retorted Nanon; "and if your political police is
no better organized than your amorous police, I pity you."
"You pity me?"
"Yes; for this Monsieur de Canolles, whom you gratuitously honor by
believing him to be your rival, is not here, and you are at liberty to
wait and see if he comes."
"He has come."
"He?" cried Nanon. "That is not true!"
There was an unmistakable accent of truth in this exclamation of the
accused.
"I mean that he came within four hundred yards, and stopped at the
Golden Calf, luckily for him."
Nanon saw that the duke was not nearly so well informed as she had
supposed at first; she shrugged her shoulders as another idea,
prompted doubtless by the letter, which she was folding and
unfolding in her hands, began to take root in her mind.
"Is it possible," said she, "that a man of intellect, one of the
cleverest politicians in the kingdom, allows himself to be gulled by
anonymous letters?"
"That's all very well; but how do you explain this letter, anonymous
or not?"
"Why, the explanation's very simple; it's simply a continuation of the
generous proceedings of our friends at Agen. Monsieur de Canolles
applied to you for leave of absence on account of urgent private
business, and you granted it; they found out that he had come in
this direction, and this absurd accusation has no other foundation
than his journey."
Nanon noticed that the duke's features did not relax, but that his
scowl became more pronounced.
"The explanation would answer, if the letter you attribute to your
friends had not a certain postscript, which, in your confusion, you
omitted to read."
The young woman shivered with terror; she realized that, if chance
did not come to her assistance, she could not long continue the
struggle.
"A postscript?" she repeated.
"Yes; read it," said the duke; "you have the letter in your hands."
Nanon tried to smile; but she felt that her distorted features would
not lend their aid to any such demonstration; she contented herself,
therefore, with reading aloud, in the firmest tones she could
command,—

"'I have in my possession Mademoiselle de Lartigues' letter to

Monsieur de Canolles, making the appointment I mention for
this evening. I will give up the letter in exchange for a paper
signed in blank by Monsieur d'Épernon, to be handed to me by
a man, alone in a boat on the Dordogne, opposite the village of
Saint-Michel-la-Rivière, at six o'clock in the evening.'"

"And you were so imprudent—" continued Nanon.

Welcome to Our Bookstore - The Ultimate Destination for Book Lovers
Are you passionate about testbank and eager to explore new worlds of
knowledge? At our website, we offer a vast collection of books that
cater to every interest and age group. From classic literature to
specialized publications, self-help books, and children’s stories, we
have it all! Each book is a gateway to new adventures, helping you
expand your knowledge and nourish your soul
Experience Convenient and Enjoyable Book Shopping Our website is more
than just an online bookstore—it’s a bridge connecting readers to the
timeless values of culture and wisdom. With a sleek and user-friendly
interface and a smart search system, you can find your favorite books
quickly and easily. Enjoy special promotions, fast home delivery, and
a seamless shopping experience that saves you time and enhances your
love for reading.
Let us accompany you on the journey of exploring knowledge and
personal growth!

ebooksecure.com