Running regressions with an index I created

[hugo dupont]

Hi all,

I am running a regression on the impact of freedom in a country on socioeconomic development during COVID-19 (171 countries).

For this, I have created a socioeconomic development index similar to that of the Human Development Index and a COVID-index.

My variables:
Socioeconoimc Index 2020: Ind_2020_Full
Socioeconomic Index 2019: Ind_2019_Full
Covid Index: ind_cov_stri_nd_2020
Freedom Score: fhi_total

I want to find the change in the socioeconomic index between 2019 and 2020 and I was wondering if I could run two regressions and do a simultaneous equation with the two models?

Note: COVID_Index_2020 is not included in the first regression, is this a problem?

Regression #1: reg Index_Socioeconomic_2019 Freedom_Score
Regression #2: reg Index_Socioeconomic_2020 COVID_Index_2020 Freedom_Score

Attached are the regressions I have ran.

[Maria Boutchkova]

What if you go:

rename (Ind_2019_Full  Ind_2020_Full ind_cov_stri_nd_2020) (ind2019 ind2020 ind_cov2020)
reshape long ind ind_cov, i(country) j(year)
replace ind_cov = 0 if year == 2019
reg ind ind_cov fhi_total i.year

Your coefficient of interest is 1.year. Note that use I country as your variable that identifies the countries, so change the name accordingly.