Measuring friendliness and like-mindedness

In his last post, Cullen S. Hendrix from the PIIE proposed to measure friendliness and like-mindedness using mean absolute distance with the US for two variables, the polyarchy score and the ideal points for votes in the UNGA.

https://www.piie.com/blogs/realtime-economics/2024/united-states-wants-work-friendly-and-minded-countries-so-who-are

I made a small Stata code to reproduce the results using the data provided in Cullen’s blog:

****************************************************************
/*
https://www.piie.com/blogs/realtime-economics/2024/
 united-states-wants-work-friendly-and-minded-countries
 -so-who-are

https://www.nato.int/cps/en/natohq/
 topics_52044.htm?selectedLocale=en
*/

use Hendrix_FriendlyLikeMinded.dta, clear

label variable idealpointdistance                            ///
 "UNGA ideal point, mean absolute distance to the US, 2020–22"
 
label variable polyarchy_absdistance                         ///
 "Polarchy score mean absolute distance to the US, 2020–22"

*NATO countries
// Loop to select countries
gen nato = 0

foreach v in TUR CZE SWE SVN SVK GBR ROU PRT POL NLD         ///
 NOR MNE MKD NLD MNE LUX LTU LVA ITA ISL HUN GRC DEU         ///
 FRA FIN EST DNK HRV CAN BGR BEL {
  replace nato = 1 if iso3c == "`v'"  	
}

 
*Reproduce Figure 1*
label define natolab 0 "Non-NATO" 1 "NATO"
label values nato natolab

kountry iso3c, from(iso3c) to(iso2c)
rename _ISO2C_ iso2c
 
twoway (                                                     ///
  scatter polyarchy_absdistance idealpointdistance,          ///
  mlabel(iso2c) colorvar(nato) colordiscrete legend(off)     ///
  zlabel(, valuelabel) coloruseplegend                       ///
  ytitle("") xtitle("")                                      ///
  title("Mean absolute distance with the US")                ///
  name(piie, replace)                                        ///
  text(.7 .5 "{bf:Polyarchy score}", size(small))            ///
  text(.1 4.5 "{bf:UNGA ideal point}", size(small))          ///
  colorlist(blue*0.4%80 red*0.5%80)                          ///
  xline(2.5) yline(.4)                                       ///
  )

*Generate Factor Variable*
factor polyarchy_absdistance idealpointdistance
predict f_friendlylikeminded

sort f_friendlylikeminded

graph export piie.png, width(4000) as(png)                   ///
 name("piie") replace

****************************************************************

It produces the picture below, and I can observe that the upper-left quadrant is empty. We have almost no examples of alignment with the US when a country has a better score in terms of electoral democracy. It means that more democratic societies seem to be aligned with the US.

Following his simple component analysis produces the following ranking for the degree of friendliness to the US:

On my GitHub, codes and data are available here. In the following two blogs, you will find more on UNGA vote data:

Bonus: a nice website to visualize polyarchy scores:

https://ourworldindata.org/grapher/polyarchy-lexical