Full Model
stargazer(plm(lead(log(budget_pc)) ~
log(pop) +
log(gdp_ppp_const) +
log(us_exports_to_country) +
log(us_imports_from_country) +
ideal_point_diff +
polity_score +
log(military_exp + 1) +
log(tradeflow_imf_d_russia + 1) +
log(tradeflow_imf_d_china + 1),
index = c("cow_code", "year"),
model = "within",
data = my_df), type = "text")
##
## ===========================================================
## Dependent variable:
## ---------------------------
## lead(log(budget_pc))
## -----------------------------------------------------------
## log(pop) 0.953
## (0.869)
##
## log(gdp_ppp_const) 0.641
## (0.445)
##
## log(us_exports_to_country) 0.205**
## (0.083)
##
## log(us_imports_from_country) -0.081
## (0.054)
##
## ideal_point_diff -0.589***
## (0.163)
##
## polity_score -0.008
## (0.018)
##
## log(military_exp + 1) 0.034
## (0.141)
##
## log(tradeflow_imf_d_russia + 1) 0.037
## (0.026)
##
## log(tradeflow_imf_d_china + 1) 0.141***
## (0.043)
##
## -----------------------------------------------------------
## Observations 960
## R2 0.061
## Adjusted R2 -0.118
## F Statistic 5.768*** (df = 9; 806)
## ===========================================================
## Note: *p<0.1; **p<0.05; ***p<0.01
Average over all years
my_df %<>%
filter(year != 2013) %>%
group_by(year) %>%
mutate(total_budget = sum(budget, na.rm = TRUE)) %>%
ungroup() %>%
group_by(cow_code, year) %>%
mutate(budget_proportion = budget / total_budget ) %>%
mutate(budget_percent = budget_proportion*100) %>%
ungroup() %>%
group_by(cow_code) %>%
mutate(avg_budget_all_yrs = mean(budget, na.rm = TRUE)) %>%
mutate(avg_budget_pc_all_yrs = mean(budget_pc, na.rm = TRUE)) %>%
mutate(avg_us_expt_all_yrs = mean(us_exports_to_country, na.rm = TRUE)) %>%
mutate(avg_us_impts_all_yrs = mean(us_imports_from_country, na.rm = TRUE)) %>%
mutate(avg_gdp_ppp_all_yrs = mean(gdp_ppp_const, na.rm = TRUE))
stargazer(lm(log(avg_budget_pc_all_yrs) ~
log(avg_us_expt_all_yrs) +
log(avg_us_impts_all_yrs) +
log(avg_gdp_ppp_all_yrs),
data = my_df), type = "text")
##
## =====================================================
## Dependent variable:
## ---------------------------
## log(avg_budget_pc_all_yrs)
## -----------------------------------------------------
## log(avg_us_expt_all_yrs) 0.088***
## (0.026)
##
## log(avg_us_impts_all_yrs) 0.079***
## (0.024)
##
## log(avg_gdp_ppp_all_yrs) -0.487***
## (0.028)
##
## Constant 7.007***
## (0.432)
##
## -----------------------------------------------------
## Observations 983
## R2 0.331
## Adjusted R2 0.329
## Residual Std. Error 0.917 (df = 979)
## F Statistic 161.610*** (df = 3; 979)
## =====================================================
## Note: *p<0.1; **p<0.05; ***p<0.01
my_df %>%
filter(year == 2019) %>%
filter(!is.na(polity_score)) %>%
ggplot(aes(y = log(budget_pc), x = polity_score)) +
geom_point(aes(color = as.factor(freedom_house.x))) +
geom_smooth(method = "loess") +
theme(legend.position = "none")
## `geom_smooth()` using formula 'y ~ x'
Economic Regression - Time fixed effects
stargazer(plm(log(budget + 1) ~
log(gdp_ppp_const) +
log(us_exports_to_country + 1) +
log(us_imports_from_country + 1),
data = my_df,
index = c("cow_code", "year"),
model = "within",
effects = "time"), type = "text")
##
## ============================================================
## Dependent variable:
## ---------------------------
## log(budget + 1)
## ------------------------------------------------------------
## log(gdp_ppp_const) -0.744***
## (0.188)
##
## log(us_exports_to_country + 1) 0.080*
## (0.043)
##
## log(us_imports_from_country + 1) 0.014
## (0.031)
##
## ------------------------------------------------------------
## Observations 983
## R2 0.022
## Adjusted R2 -0.177
## F Statistic 6.039*** (df = 3; 816)
## ============================================================
## Note: *p<0.1; **p<0.05; ***p<0.01
Economic Regression - 2019 only
my_df %>%
filter(year == 2019) -> df19
stargazer(lm(log(budget_pc) ~
log(gdp_ppp_const) +
log(us_exports_to_country + 1) +
log(us_imports_from_country + 1),
data = df19), type = "text")
##
## ============================================================
## Dependent variable:
## ---------------------------
## log(budget_pc)
## ------------------------------------------------------------
## log(gdp_ppp_const) -0.506***
## (0.079)
##
## log(us_exports_to_country + 1) 0.189***
## (0.072)
##
## log(us_imports_from_country + 1) 0.022
## (0.067)
##
## Constant 6.487***
## (1.245)
##
## ------------------------------------------------------------
## Observations 163
## R2 0.278
## Adjusted R2 0.264
## Residual Std. Error 1.074 (df = 159)
## F Statistic 20.383*** (df = 3; 159)
## ============================================================
## Note: *p<0.1; **p<0.05; ***p<0.01
my_df %<>%
mutate(us_exports_pc = us_exports_to_country / pop )
my_df %>%
filter(year == 2019) %>%
filter(!is.na(freedom_house.x)) %>%
ggplot(aes(y = log(budget_pc), x = log(us_exports_pc))) +
geom_point(aes(color = as.factor(freedom_house.x))) +
facet_wrap(~freedom_house.x) +
geom_smooth(method = "glm") +
theme(legend.position = "none")
## `geom_smooth()` using formula 'y ~ x'
Economic Regression - average of all years
my_df %>%
filter(year == 2019) -> df19
stargazer(lm(log(avg_budget_pc_all_yrs) ~
log(avg_gdp_ppp_all_yrs) +
log(avg_us_expt_all_yrs) +
log(avg_us_impts_all_yrs ),
data = my_df), type = "text")
##
## =====================================================
## Dependent variable:
## ---------------------------
## log(avg_budget_pc_all_yrs)
## -----------------------------------------------------
## log(avg_gdp_ppp_all_yrs) -0.487***
## (0.028)
##
## log(avg_us_expt_all_yrs) 0.088***
## (0.026)
##
## log(avg_us_impts_all_yrs) 0.079***
## (0.024)
##
## Constant 7.007***
## (0.432)
##
## -----------------------------------------------------
## Observations 983
## R2 0.331
## Adjusted R2 0.329
## Residual Std. Error 0.917 (df = 979)
## F Statistic 161.610*** (df = 3; 979)
## =====================================================
## Note: *p<0.1; **p<0.05; ***p<0.01
Political / security goals
my_df %<>%
dplyr::mutate(mil_aid_pc = sum_mil_aid / pop) %>%
dplyr::mutate(mil_exp_pc = military_exp / pop)
stargazer(plm(lead(log(budget_pc)) ~
polity_score +
I(polity_score^2) +
ideal_point_diff,
data = my_df,
index = c("cow_code", "year"),
model = "within",
effects = "time"), type = "text")
##
## ============================================
## Dependent variable:
## ---------------------------
## lead(log(budget_pc))
## --------------------------------------------
## polity_score -0.016
## (0.022)
##
## I(polity_score2) -0.001
## (0.003)
##
## ideal_point_diff -0.221
## (0.157)
##
## --------------------------------------------
## Observations 963
## R2 0.004
## Adjusted R2 -0.201
## F Statistic 1.099 (df = 3; 798)
## ============================================
## Note: *p<0.1; **p<0.05; ***p<0.01
Political influence
stargazer(plm(budget_percent ~
log(tradeflow_comtrade_d_china) +
log(distcap_china) +
as.factor(heg_d_china),
data = my_df,
index = c("cow_code", "year"),
model = "within",
effect = "time"), type = "text")
##
## ===========================================================
## Dependent variable:
## ---------------------------
## budget_percent
## -----------------------------------------------------------
## log(tradeflow_comtrade_d_china) 0.040***
## (0.013)
##
## log(distcap_china) -0.499***
## (0.079)
##
## as.factor(heg_d_china)1 -1.497***
## (0.514)
##
## -----------------------------------------------------------
## Observations 849
## R2 0.071
## Adjusted R2 0.063
## F Statistic 21.277*** (df = 3; 841)
## ===========================================================
## Note: *p<0.1; **p<0.05; ***p<0.01
Russian influence
stargazer(plm(log(budget_pc) ~
log(tradeflow_comtrade_d_russia) +
as.factor(soviet_iron_curtain) +
as.factor(soviet_republics),
data = my_df,
index = c("cow_code", "year"),
model = "within",
effect = "time"), type = "text")
##
## ============================================================
## Dependent variable:
## ---------------------------
## log(budget_pc)
## ------------------------------------------------------------
## log(tradeflow_comtrade_d_russia) -0.067***
## (0.008)
##
## as.factor(soviet_iron_curtain)1 1.005***
## (0.139)
##
## as.factor(soviet_republics)1 1.160***
## (0.128)
##
## ------------------------------------------------------------
## Observations 996
## R2 0.130
## Adjusted R2 0.123
## F Statistic 49.307*** (df = 3; 987)
## ============================================================
## Note: *p<0.1; **p<0.05; ***p<0.01