In another project, I found an interesting correlation between the IMF’s “Aggregate Measures of Trade Restriction” (MATR) (Estafania-Flores, Furceri, Hannan, Ostry and Rose (2022)) and Chinn-Ito (go year 2006) measures financial openness (KAOPEN).
figure 1: Chinn-Ito KAOPEN and MATR (inverted) for Emerging Markets/Developing Economies (EMDE). Higher values for each variable indicate more openness. source: ito townand MATR.
Every unit change in KAOPEN is associated (inverted) with a 2.5 unit change in MATR. This estimate was relatively unchanged (2.4, t-stat = 98), including time fixed effects (adj-R2 = 0.69). With this level of correlation, it is sometimes difficult to distinguish which variable is important in a multiple regression (i.e. multicollinearity) when both are included on the right.
The time series patterns of the mean (cross-country) values of the two variables are also interesting.
figure 2: Y – mean of Chinn-Ito KAOPEN, X10 – mean MATR (reciprocal, divided by 10) of emerging market/developing economies (EMDEs). Higher values for each variable indicate more openness. source: ito townand MATR.
