注意：R/RStudio上で行った分析は、文字化けを避けるために英語で実施した。論文執筆の段階で、分析に用いた変数や分析結果を日本語へ訳した。

1 分析準備

1.1 分析に必要なパッケージをインストール

pacman::p_load(tidyverse, summarytools, DT, devtools, estimatr, jtools, broom.mixed, ggstance)

library(tidyverse)
library(summarytools)
library(DT)
library(devtools)
library(estimatr)
library(jtools)
library(broom.mixed)
library(ggstance)
devtools::install_github("JaehyunSong/BalanceR")
library(BalanceR)

1.2 データを読み込む

df <- read_csv("cleaned_mx_NA.csv")

1.3 処置変数をFactor変数に変換

df <- df |>  
  mutate(Treat = factor(Treat,
                        levels = c("Control", "US", "China")))

2 データのクリーニング

df <- df |>  
  select(RESPID, Treat, Q1, Q2, Q4, Q5_1:Q5_5, Q9, Q17:Q19, Q20_4, Q21, Q25, Q32, Q35:Q37, Q39, Q40, Treat, Q3S)
df <- df |>  
  mutate(Female       =  if_else(Q2 == 2, 1, 0),     
         # Binary (female = 1, male =0)            
         # Age divided by 10 (centered by median)
         Age          = (Q1 - median(Q1)) / 10,        
         # Feeling thermometer (Experiments_priming)
         Thermo       =  Q3S,
         # Outcome Variable: Attitude toward international   trade (high scores means greater support: 1 = very bad, 2 = bad, 3 = neither good or bad, 4 = good, 5 = very good)
         Y          =  6 - Q4,
         # Outcome Varialbe: Attitude toward international trade by partners
         Y_EU       =  6 - Q5_1,
         Y_US       =  6 - Q5_2,
         Y_China    =  6 - Q5_3,
         Y_Mercosur =  6 - Q5_4,
         Y_TPP      =  6 - Q5_5,
         # Perception on emigration
         Emig         =  6 - Q9,            
         # Job types (primary sector including agriculture, fishery, forestry, mmanufacturing industry)
         Primary      =  if_else(Q39 == 1, 1, 0),  
         Manufac      =  if_else(Q39 == 3, 1, 0),  
         # Political knowledge
         Pol_know1    =  if_else(Q17 == 2, 1, 0),
         Pol_know2    =  if_else(Q18 == 3, 1, 0),
         Pol_know3    =  if_else(Q19 == 3, 1, 0),
         Pol_know     =  Pol_know1 + Pol_know2 + Pol_know3,
         # Evaluation of government
         Eval_gov     =  5 - Q21,       
         # Partisanship
         PRI          =  if_else(Q25 == 1, 1, 0),
         PAN          =  if_else(Q25 == 2, 1, 0),
         PRD          =  if_else(Q25 == 3, 1, 0),
         MORENA       =  if_else(Q25 == 4, 1, 0),
         # Demographic attributes
         Emigration   =  4 - Q20_4,     # Emigrate to work
         # Ethnicity/race
         Indigenous   =  if_else(Q32 == 1, 1, 0),
         Mestizo      =  if_else(Q32 == 3, 1, 0),
         White        =  if_else(Q32 == 4, 1, 0),
         # Living location
         Urban        =  if_else(Q36 == 1, 1, 0),  
         Rurul        =  if_else(Q36 == 3, 1, 0),
         # Education level centered by median
         Edu          =  Q37 - 3.5,     
         # Income level centered by median
         Income       =  Q40 - 5.5     
  )
# 州ダミーを作成 
df <- df %>% 
  mutate(State_dummy = factor(Q35))

3 記述統計

df2 <- df |> 
  select(Y, Y_US, Y_China, Female, Age, Income, Primary, 
         Manufac, Emigration, Pol_know, PRI, PAN, PRD, MORENA) 
print(dfSummary(df2, 
                style = "grid", plain.ascii = FALSE,    
                graph.magnif = 0.85), 
                method = "render", heading = FALSE)

Data Frame Summary

df2

Dimensions: 1050 x 14
Duplicates: 9
No Variable Stats / Values Freqs (% of Valid) Graph Valid Missing
1 Y [numeric]
Mean (sd) : 4.3 (0.7)
min ≤ med ≤ max:
1 ≤ 4 ≤ 5
IQR (CV) : 1 (0.2)
1:5(0.5%)
2:12(1.2%)
3:93(8.9%)
4:455(43.7%)
5:476(45.7%)
 1041 (99.1%) 9 (0.9%)
2 Y_US [numeric]
Mean (sd) : 3.9 (1)
min ≤ med ≤ max:
1 ≤ 4 ≤ 5
IQR (CV) : 2 (0.3)
1:29(2.8%)
2:76(7.3%)
3:164(15.7%)
4:458(43.8%)
5:319(30.5%)
 1046 (99.6%) 4 (0.4%)
3 Y_China [numeric]
Mean (sd) : 3.9 (1)
min ≤ med ≤ max:
1 ≤ 4 ≤ 5
IQR (CV) : 1 (0.2)
1:20(1.9%)
2:71(6.8%)
3:169(16.2%)
4:470(45.1%)
5:312(29.9%)
 1042 (99.2%) 8 (0.8%)
4 Female [numeric]
Min : 0
Mean : 0.5
Max : 1
0:525(50.0%)
1:525(50.0%)
 1050 (100.0%) 0 (0.0%)
5 Age [numeric]
Mean (sd) : 0.1 (1.3)
min ≤ med ≤ max:
-2.1 ≤ 0 ≤ 2.4
IQR (CV) : 2.3 (13)
 46 distinct values 1050 (100.0%) 0 (0.0%)
6 Income [numeric]
Mean (sd) : -2.1 (2)
min ≤ med ≤ max:
-4.5 ≤ -2.5 ≤ 4.5
IQR (CV) : 2 (-0.9)
 10 distinct values 964 (91.8%) 86 (8.2%)
7 Primary [numeric]
Min : 0
Mean : 0
Max : 1
0:881(98.1%)
1:17(1.9%)
 898 (85.5%) 152 (14.5%)
8 Manufac [numeric]
Min : 0
Mean : 0.1
Max : 1
0:819(91.2%)
1:79(8.8%)
 898 (85.5%) 152 (14.5%)
9 Emigration [numeric]
Mean (sd) : 1.3 (0.6)
min ≤ med ≤ max:
0 ≤ 1 ≤ 3
IQR (CV) : 0 (0.5)
0:6(0.6%)
1:806(76.8%)
2:154(14.7%)
3:84(8.0%)
 1050 (100.0%) 0 (0.0%)
10 Pol_know [numeric]
Mean (sd) : 2.9 (0.4)
min ≤ med ≤ max:
0 ≤ 3 ≤ 3
IQR (CV) : 0 (0.1)
0:2(0.2%)
1:12(1.2%)
2:93(9.2%)
3:901(89.4%)
 1008 (96.0%) 42 (4.0%)
11 PRI [numeric]
Min : 0
Mean : 0.1
Max : 1
0:912(92.3%)
1:76(7.7%)
 988 (94.1%) 62 (5.9%)
12 PAN [numeric]
Min : 0
Mean : 0.2
Max : 1
0:751(76.0%)
1:237(24.0%)
 988 (94.1%) 62 (5.9%)
13 PRD [numeric]
Min : 0
Mean : 0
Max : 1
0:966(97.8%)
1:22(2.2%)
 988 (94.1%) 62 (5.9%)
14 MORENA [numeric]
Min : 0
Mean : 0.4
Max : 1
0:584(59.1%)
1:404(40.9%)
 988 (94.1%) 62 (5.9%)

Generated by summarytools 1.0.1 (R version 4.3.1)
2023-07-12

4 バランス・チェック

Author/Maintainer: Jaehyun Song (https://www.jaysong.net / )

BC <- BalanceR(data = df, group = "Treat",
               cov  = c("Female", "Age", "Income", 
                        "Emigration", "Primary",
                        "Manufac","Pol_know", "PRI", "PAN",
                        "PRD", "MORENA")) |> 
plot()
print(BC, digit = 3)

5 処置効果の比較

5.1 図4.4. 自由貿易に対する態度 (2)メキシコ

推定値をRで計算した後、エクセルで図４を作成した。

5.1.1 国際貿易一般への支持

Support.df <- df  |> 
  group_by(Treat) |>  
  summarise(Y     = mean(Y, na.rm = TRUE),   # remove NAs
            group = "drop")
Support.df
## # A tibble: 3 × 3
##   Treat       Y group
##   <fct>   <dbl> <chr>
## 1 Control  4.33 drop 
## 2 US       4.25 drop 
## 3 China    4.41 drop
Support.df |>  
  ggplot() +
  geom_bar(aes(x = Treat, y = Y), stat = "identity") +
  geom_label(aes(x = Treat, y = Y,
                 label = round(Y, 3)), label.size = 1) +
            labs(x = "Treatment", y = "Support for 
                 International Trade") + 
            coord_cartesian(ylim = c(0, 4.5))

5.1.2 対アメリカ貿易への支持

Support_US.df <- df |> 
  group_by(Treat)   |> 
  summarise(Y_US  = mean(Y_US, na.rm = TRUE),   # remove NAs
            group = "drop")
Support_US.df
## # A tibble: 3 × 3
##   Treat    Y_US group
##   <fct>   <dbl> <chr>
## 1 Control  3.99 drop 
## 2 US       3.92 drop 
## 3 China    3.85 drop
Support_US.df |>  
  ggplot() +
  geom_bar(aes(x = Treat, y = Y_US), stat = "identity") +
  geom_label(aes(x = Treat, y = Y_US,
                 label = round(Y_US, 3)), label.size = 1) +
  labs(x = "Treatment", y = "Support for International Trade 
       with US") +
  coord_cartesian(ylim = c(0, 4.5))

5.1.3 対中国貿易への支持

Support_China.df <- df |>  
  group_by(Treat) |> 
  summarise(Y_China  = mean(Y_China, na.rm = TRUE), 
            # remove NAs
            group = "drop")
Support_China.df
## # A tibble: 3 × 3
##   Treat   Y_China group
##   <fct>     <dbl> <chr>
## 1 Control    3.99 drop 
## 2 US         3.81 drop 
## 3 China      4.03 drop
Support_China.df |> 
  ggplot() +
  geom_bar(aes(x = Treat, y = Y_China), stat = "identity") +
  geom_label(aes(x = Treat, y = Y_China,
                 label = round(Y_China, 3)), label.size = 1) +
  labs(x = "Treatment", y = "Support for International Trade 
       with China") +
  coord_cartesian(ylim = c(0, 4.5))

6 単回帰分析

推定結果は、「補填　表4 A-2　メキシコ：貿易相手別の支持」のモデル1，4，7に相当

6.1 モデル１（国際貿易一般への支持）

Support.fit <- lm_robust(Y ~ Treat, se_type = "stata", 
                         data = df)
summary(Support.fit)
## 
## Call:
## lm_robust(formula = Y ~ Treat, data = df, se_type = "stata")
## 
## Standard error type:  HC1 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|) CI Lower CI Upper   DF
## (Intercept)  4.32948    0.03801 113.911   0.0000  4.25490  4.40406 1038
## TreatUS     -0.07661    0.05477  -1.399   0.1622 -0.18407  0.03086 1038
## TreatChina   0.07974    0.05455   1.462   0.1441 -0.02731  0.18679 1038
## 
## Multiple R-squared:  0.007747 ,  Adjusted R-squared:  0.005835 
## F-statistic: 3.961 on 2 and 1038 DF,  p-value: 0.01933

6.2 モデル４（対アメリカ貿易への支持

Support_US.fit <- lm_robust(Y_US ~ Treat, se_type = "stata", 
                            data = df)
summary(Support_US.fit)
## 
## Call:
## lm_robust(formula = Y_US ~ Treat, data = df, se_type = "stata")
## 
## Standard error type:  HC1 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|) CI Lower CI Upper   DF
## (Intercept)  3.99135    0.05234 76.2509  0.00000   3.8886 4.094068 1043
## TreatUS     -0.06872    0.07325 -0.9381  0.34840  -0.2125 0.075019 1043
## TreatChina  -0.14564    0.07701 -1.8913  0.05887  -0.2967 0.005467 1043
## 
## Multiple R-squared:  0.003554 ,  Adjusted R-squared:  0.001643 
## F-statistic: 1.789 on 2 and 1043 DF,  p-value: 0.1677

6.3 モデル７（対中国貿易への支持）

Support_China.fit <- lm_robust(Y_China ~ Treat, se_type = 
                                 "stata", data = df)
summary(Support_China.fit)
## 
## Call:
## lm_robust(formula = Y_China ~ Treat, data = df, se_type = "stata")
## 
## Standard error type:  HC1 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|) CI Lower CI Upper   DF
## (Intercept)  3.98559    0.04973 80.1439  0.00000  3.88801  4.08317 1039
## TreatUS     -0.17690    0.07597 -2.3286  0.02007 -0.32596 -0.02783 1039
## TreatChina   0.04869    0.06684  0.7286  0.46644 -0.08246  0.17985 1039
## 
## Multiple R-squared:  0.01037 ,   Adjusted R-squared:  0.008469 
## F-statistic: 4.986 on 2 and 1039 DF,  p-value: 0.006998

7 重回帰分析　　　　　　　　　

推定結果は、「補填　表4 A-2　メキシコ：貿易相手別の支持」のモデル2，5，8に相当

7.1 モデル２（国際貿易一般への支持

Support.fit <- lm_robust(Y ~ Treat + Female + Age + Income + 
                     Primary + Manufac + Emigration + 
                     Pol_know +  PRI + PAN + PRD + MORENA + 
                     State_dummy,
                     se_type = "stata", data = df)
summary(Support.fit)
## 
## Call:
## lm_robust(formula = Y ~ Treat + Female + Age + Income + Primary + 
##     Manufac + Emigration + Pol_know + PRI + PAN + PRD + MORENA + 
##     State_dummy, data = df, se_type = "stata")
## 
## Standard error type:  HC1 
## 
## Coefficients:
##                Estimate Std. Error  t value  Pr(>|t|)  CI Lower CI Upper  DF
## (Intercept)    4.713507    0.29039 16.23147 7.291e-51  4.143409  5.28361 735
## TreatUS       -0.092687    0.06175 -1.50103 1.338e-01 -0.213913  0.02854 735
## TreatChina    -0.024345    0.06462 -0.37675 7.065e-01 -0.151205  0.10251 735
## Female        -0.058442    0.05346 -1.09311 2.747e-01 -0.163402  0.04652 735
## Age            0.031711    0.02140  1.48213 1.387e-01 -0.010293  0.07372 735
## Income         0.030106    0.01451  2.07438 3.839e-02  0.001614  0.05860 735
## Primary       -0.070642    0.15874 -0.44501 6.564e-01 -0.382284  0.24100 735
## Manufac       -0.066369    0.10711 -0.61961 5.357e-01 -0.276653  0.14392 735
## Emigration     0.097297    0.04307  2.25914 2.417e-02  0.012746  0.18185 735
## Pol_know      -0.137382    0.06450 -2.12998 3.350e-02 -0.264007 -0.01076 735
## PRI            0.189167    0.10098  1.87335 6.142e-02 -0.009072  0.38741 735
## PAN            0.173156    0.08094  2.13921 3.275e-02  0.014247  0.33206 735
## PRD           -0.202818    0.21813 -0.92982 3.528e-01 -0.631041  0.22541 735
## MORENA         0.090158    0.07674  1.17480 2.405e-01 -0.060504  0.24082 735
## State_dummy2   0.011292    0.24735  0.04565 9.636e-01 -0.474300  0.49689 735
## State_dummy3   0.157078    0.29077  0.54022 5.892e-01 -0.413756  0.72791 735
## State_dummy4  -0.106564    0.32861 -0.32429 7.458e-01 -0.751692  0.53856 735
## State_dummy5  -0.094531    0.21281 -0.44420 6.570e-01 -0.512321  0.32326 735
## State_dummy6   0.233956    0.43935  0.53250 5.945e-01 -0.628577  1.09649 735
## State_dummy7  -0.018388    0.26898 -0.06836 9.455e-01 -0.546453  0.50968 735
## State_dummy8   0.067161    0.27415  0.24498 8.065e-01 -0.471054  0.60538 735
## State_dummy9  -0.029232    0.26345 -0.11096 9.117e-01 -0.546435  0.48797 735
## State_dummy10 -0.164604    0.32807 -0.50174 6.160e-01 -0.808665  0.47946 735
## State_dummy11  0.524118    0.27631  1.89682 5.824e-02 -0.018342  1.06658 735
## State_dummy12  0.056846    0.25154  0.22600 8.213e-01 -0.436972  0.55067 735
## State_dummy13 -0.026358    0.30331 -0.08690 9.308e-01 -0.621808  0.56909 735
## State_dummy14 -0.011629    0.22566 -0.05153 9.589e-01 -0.454644  0.43139 735
## State_dummy15  0.037296    0.21521  0.17330 8.625e-01 -0.385205  0.45980 735
## State_dummy16  0.264115    0.24291  1.08728 2.773e-01 -0.212773  0.74100 735
## State_dummy17 -0.006792    0.28988 -0.02343 9.813e-01 -0.575875  0.56229 735
## State_dummy19 -0.100445    0.22778 -0.44098 6.594e-01 -0.547619  0.34673 735
## State_dummy20  0.003436    0.29347  0.01171 9.907e-01 -0.572694  0.57957 735
## State_dummy21 -0.250277    0.24543 -1.01974 3.082e-01 -0.732111  0.23156 735
## State_dummy22 -0.379623    0.26412 -1.43730 1.511e-01 -0.898145  0.13890 735
## State_dummy23  0.079531    0.24105  0.32994 7.415e-01 -0.393693  0.55275 735
## State_dummy24  0.134118    0.27150  0.49399 6.215e-01 -0.398889  0.66712 735
## State_dummy25 -0.075718    0.30990 -0.24433 8.070e-01 -0.684109  0.53267 735
## State_dummy26  0.040002    0.28222  0.14174 8.873e-01 -0.514060  0.59406 735
## State_dummy27 -0.046953    0.28323 -0.16577 8.684e-01 -0.602991  0.50909 735
## State_dummy28  0.571136    0.22219  2.57046 1.035e-02  0.134929  1.00734 735
## State_dummy29 -0.165187    0.28894 -0.57171 5.677e-01 -0.732425  0.40205 735
## State_dummy30  0.078147    0.23675  0.33008 7.414e-01 -0.386649  0.54294 735
## State_dummy31 -0.072598    0.27976 -0.25950 7.953e-01 -0.621829  0.47663 735
## State_dummy32 -0.359936    0.45723 -0.78722 4.314e-01 -1.257561  0.53769 735
## 
## Multiple R-squared:  0.06882 ,   Adjusted R-squared:  0.01434 
## F-statistic: 16.26 on 43 and 735 DF,  p-value: < 2.2e-16

7.2 モデル５（対アメリカ貿易への支持）

Support_US.fit <- lm_robust(Y_US ~  Treat + Female + Age + 
                        Income + Primary + Manufac +
                        Emigration + Pol_know +  PRI + PAN + 
                        PRD + MORENA + State_dummy,
                        se_type = "stata", data = df)
summary(Support_US.fit)
## 
## Call:
## lm_robust(formula = Y_US ~ Treat + Female + Age + Income + Primary + 
##     Manufac + Emigration + Pol_know + PRI + PAN + PRD + MORENA + 
##     State_dummy, data = df, se_type = "stata")
## 
## Standard error type:  HC1 
## 
## Coefficients:
##                Estimate Std. Error  t value  Pr(>|t|)  CI Lower CI Upper  DF
## (Intercept)    4.428804    0.45408  9.75333 3.209e-21  3.537357  5.32025 737
## TreatUS       -0.129978    0.08683 -1.49701 1.348e-01 -0.300432  0.04048 737
## TreatChina    -0.203467    0.08922 -2.28038 2.287e-02 -0.378632 -0.02830 737
## Female        -0.074197    0.07411 -1.00114 3.171e-01 -0.219693  0.07130 737
## Age            0.061256    0.02971  2.06205 3.955e-02  0.002937  0.11958 737
## Income         0.005149    0.01854  0.27776 7.813e-01 -0.031244  0.04154 737
## Primary       -0.308600    0.24447 -1.26231 2.072e-01 -0.788545  0.17134 737
## Manufac        0.160315    0.12394  1.29353 1.962e-01 -0.082995  0.40362 737
## Emigration     0.108039    0.06466  1.67091 9.516e-02 -0.018898  0.23498 737
## Pol_know      -0.184634    0.10545 -1.75093 8.037e-02 -0.391649  0.02238 737
## PRI            0.425337    0.13997  3.03870 2.460e-03  0.150544  0.70013 737
## PAN            0.358257    0.11041  3.24492 1.228e-03  0.141510  0.57500 737
## PRD            0.083153    0.24277  0.34252 7.321e-01 -0.393449  0.55976 737
## MORENA         0.166320    0.10488  1.58576 1.132e-01 -0.039586  0.37223 737
## State_dummy2  -0.330958    0.37409 -0.88470 3.766e-01 -1.065368  0.40345 737
## State_dummy3   0.082274    0.42868  0.19193 8.479e-01 -0.759295  0.92384 737
## State_dummy4  -0.845219    0.54595 -1.54817 1.220e-01 -1.917017  0.22658 737
## State_dummy5  -0.166525    0.33789 -0.49284 6.223e-01 -0.829869  0.49682 737
## State_dummy6   0.190866    0.63865  0.29886 7.651e-01 -1.062923  1.44466 737
## State_dummy7   0.074285    0.43224  0.17186 8.636e-01 -0.774287  0.92286 737
## State_dummy8  -0.076763    0.43093 -0.17813 8.587e-01 -0.922767  0.76924 737
## State_dummy9   0.125719    0.39003  0.32233 7.473e-01 -0.639990  0.89143 737
## State_dummy10 -0.180788    0.53032 -0.34091 7.333e-01 -1.221898  0.86032 737
## State_dummy11  0.366904    0.40554  0.90474 3.659e-01 -0.429238  1.16305 737
## State_dummy12 -0.213128    0.41332 -0.51564 6.063e-01 -1.024562  0.59831 737
## State_dummy13 -0.045141    0.41644 -0.10840 9.137e-01 -0.862698  0.77242 737
## State_dummy14 -0.068760    0.34738 -0.19794 8.431e-01 -0.750735  0.61321 737
## State_dummy15 -0.324440    0.34672 -0.93573 3.497e-01 -1.005126  0.35625 737
## State_dummy16  0.373701    0.36230  1.03147 3.027e-01 -0.337564  1.08497 737
## State_dummy17 -0.100681    0.45742 -0.22011 8.259e-01 -0.998684  0.79732 737
## State_dummy19 -0.089694    0.35694 -0.25128 8.017e-01 -0.790438  0.61105 737
## State_dummy20 -0.643816    0.38128 -1.68858 9.172e-02 -1.392333  0.10470 737
## State_dummy21  0.012822    0.36368  0.03526 9.719e-01 -0.701159  0.72680 737
## State_dummy22 -0.786130    0.41093 -1.91303 5.613e-02 -1.592873  0.02061 737
## State_dummy23  0.102655    0.41301  0.24855 8.038e-01 -0.708167  0.91348 737
## State_dummy24 -0.041909    0.38808 -0.10799 9.140e-01 -0.803788  0.71997 737
## State_dummy25  0.070251    0.40129  0.17506 8.611e-01 -0.717551  0.85805 737
## State_dummy26  0.283864    0.38349  0.74021 4.594e-01 -0.469007  1.03674 737
## State_dummy27 -0.303569    0.40515 -0.74928 4.539e-01 -1.098952  0.49181 737
## State_dummy28 -3.062538    0.34511 -8.87406 5.291e-18 -3.740056 -2.38502 737
## State_dummy29  0.041666    0.42661  0.09767 9.222e-01 -0.795849  0.87918 737
## State_dummy30 -0.014810    0.36803 -0.04024 9.679e-01 -0.737314  0.70769 737
## State_dummy31  0.167847    0.38786  0.43275 6.653e-01 -0.593601  0.92930 737
## State_dummy32 -0.166592    0.48121 -0.34620 7.293e-01 -1.111291  0.77811 737
## 
## Multiple R-squared:  0.09351 ,   Adjusted R-squared:  0.04062 
## F-statistic: 204.4 on 43 and 737 DF,  p-value: < 2.2e-16

7.3 対モデル８（中国貿易への支持）

Support_China.fit <- lm_robust(Y_China ~  Treat + Female + Age 
                          + Income + Primary + Manufac + 
                          Emigration + Pol_know +  PRI + PAN + 
                          PRD + MORENA + State_dummy,
                          se_type = "stata", data = df)
summary(Support_China.fit)
## 
## Call:
## lm_robust(formula = Y_China ~ Treat + Female + Age + Income + 
##     Primary + Manufac + Emigration + Pol_know + PRI + PAN + PRD + 
##     MORENA + State_dummy, data = df, se_type = "stata")
## 
## Standard error type:  HC1 
## 
## Coefficients:
##                Estimate Std. Error  t value  Pr(>|t|) CI Lower  CI Upper  DF
## (Intercept)    4.165324    0.29422 14.15731 2.038e-40  3.58772  4.742928 736
## TreatUS       -0.291165    0.08568 -3.39847 7.141e-04 -0.45936 -0.122968 736
## TreatChina    -0.054089    0.07728 -0.69991 4.842e-01 -0.20581  0.097628 736
## Female         0.001087    0.06743  0.01611 9.871e-01 -0.13129  0.133464 736
## Age           -0.020703    0.02893 -0.71557 4.745e-01 -0.07750  0.036097 736
## Income        -0.018382    0.01680 -1.09413 2.743e-01 -0.05136  0.014601 736
## Primary       -0.020314    0.19493 -0.10421 9.170e-01 -0.40300  0.362377 736
## Manufac       -0.067534    0.12771 -0.52882 5.971e-01 -0.31825  0.183180 736
## Emigration     0.026622    0.05338  0.49876 6.181e-01 -0.07816  0.131408 736
## Pol_know      -0.065844    0.07866 -0.83703 4.028e-01 -0.22028  0.088590 736
## PRI            0.463441    0.13508  3.43093 6.351e-04  0.19826  0.728624 736
## PAN            0.198771    0.11198  1.77512 7.629e-02 -0.02106  0.418603 736
## PRD            0.290429    0.19502  1.48924 1.369e-01 -0.09243  0.673287 736
## MORENA         0.416678    0.09930  4.19599 3.049e-05  0.22173  0.611631 736
## State_dummy2  -0.235335    0.22723 -1.03567 3.007e-01 -0.68143  0.210761 736
## State_dummy3   0.299151    0.42205  0.70881 4.787e-01 -0.52941  1.127714 736
## State_dummy4  -0.450586    0.20042 -2.24819 2.486e-02 -0.84405 -0.057119 736
## State_dummy5  -0.240958    0.14899 -1.61724 1.063e-01 -0.53346  0.051544 736
## State_dummy6  -0.319852    0.28761 -1.11211 2.665e-01 -0.88448  0.244780 736
## State_dummy7  -0.579897    0.34724 -1.67002 9.534e-02 -1.26159  0.101800 736
## State_dummy8  -0.072566    0.27416 -0.26469 7.913e-01 -0.61079  0.465656 736
## State_dummy9  -0.072294    0.23603 -0.30629 7.595e-01 -0.53567  0.391085 736
## State_dummy10 -0.436022    0.39578 -1.10167 2.710e-01 -1.21302  0.340979 736
## State_dummy11  0.458076    0.35323  1.29682 1.951e-01 -0.23538  1.151535 736
## State_dummy12 -0.377063    0.26704 -1.41199 1.584e-01 -0.90132  0.147196 736
## State_dummy13 -0.581105    0.37353 -1.55572 1.202e-01 -1.31441  0.152203 736
## State_dummy14 -0.057610    0.17205 -0.33485 7.378e-01 -0.39537  0.280149 736
## State_dummy15 -0.486043    0.16529 -2.94054 3.379e-03 -0.81054 -0.161547 736
## State_dummy16  0.105415    0.22342  0.47182 6.372e-01 -0.33320  0.544031 736
## State_dummy17 -0.084177    0.32755 -0.25699 7.973e-01 -0.72723  0.558874 736
## State_dummy19 -0.029126    0.16081 -0.18112 8.563e-01 -0.34482  0.286568 736
## State_dummy20 -0.969073    0.40553 -2.38963 1.712e-02 -1.76521 -0.172935 736
## State_dummy21 -0.404981    0.20125 -2.01228 4.455e-02 -0.80008 -0.009879 736
## State_dummy22 -0.832088    0.29156 -2.85390 4.440e-03 -1.40448 -0.259695 736
## State_dummy23  0.218747    0.18612  1.17528 2.403e-01 -0.14665  0.584142 736
## State_dummy24 -0.202196    0.44192 -0.45754 6.474e-01 -1.06977  0.665377 736
## State_dummy25  0.105334    0.26016  0.40488 6.857e-01 -0.40541  0.616078 736
## State_dummy26  0.326066    0.23364  1.39558 1.633e-01 -0.13262  0.784751 736
## State_dummy27 -0.081308    0.31137 -0.26113 7.941e-01 -0.69258  0.529967 736
## State_dummy28 -0.375791    0.17247 -2.17889 2.966e-02 -0.71438 -0.037200 736
## State_dummy29  0.269151    0.23280  1.15615 2.480e-01 -0.18788  0.726179 736
## State_dummy30 -0.485246    0.23369 -2.07648 3.820e-02 -0.94402 -0.026474 736
## State_dummy31 -0.179804    0.28836 -0.62354 5.331e-01 -0.74591  0.386299 736
## State_dummy32 -0.476943    0.40941 -1.16495 2.444e-01 -1.28070  0.326811 736
## 
## Multiple R-squared:  0.1271 ,    Adjusted R-squared:  0.07609 
## F-statistic: 2.929 on 43 and 736 DF,  p-value: 4.048e-09

7.4 モデル2, 5, 8の推定結果を可視化（図4.5 自由貿易に対する支持の規定要因 (2)メキシコ）

plot_summs(Support.fit, Support_US.fit, Support_China.fit,
           scale = TRUE, robust = TRUE, 
           coefs = c("処置：アメリカ"="TreatUS", 
                     "処置：中国"="TreatChina",  "性別"= 
                       "Female", "年齢"= "Age", 
                     "所得"="Income", "一次産業"="Primary", 
                     "製造業"= "Manufac", 
                     "移民経験"="Emigration", 
                     "政治知識"="Pol_know", "PRI支持"= "PRI", 
                     "PAN支持"="PAN", "PRD支持"="PRD", 
                     "MORENA支持"="MORENA"), 
           model.names = c("国際貿易一般", "対アメリカ貿易", 
                           "対中国貿易"),
           legend.title = "")