STATA-FORUM.DE

[Anna12345]

I am running a regression where I am trying to identify the influence of the proximity to Tuskegee, Macon County and the percentage of African American population on the Covid-19 vaccination rate on county level.

Code: Alles auswählen

xtreg vaccination_perc_points c.dist_Tuskegee##c.perc_points_afri_am i.urban_rual i.education_low_high median_income perc_points_male social_capital_index perc_points_am_indian_alaka_native perc_points_asian, fe

I have almost 3000 observations and i included state fixed effects. When i am doing so all my reuslts become insignificant. This might be because omitted variables vary across my states are better predictors for the vaccination rate within counties. But i also observed the problem that when i am running the fixed effect regression my VIF of distance to Tuskegee is incredible high, aroung 23. When i am running a regression to estimate the relationship between the distance to Tuskegee and dummy variables for all states, all coefficients are highly significant and the states can explan 96% of the variance. So i guess that multicollinearity between states and the distance to Tuskegee might be a problem which is also intuitively.

Does any one has a idea how i could fix this problem and whether i ahve identified it correctly?

Thanks in advance!

[Staxa]

Thats a difficult question with many parts. I suggest using statalist.org since there are more users with diverse backgrounds to support you.