I figured out that margins (from R, mimics Stata) is a good way to get marginal effects (and their significance, CI, ...) for linear models in R.


data("LaborSupply", package = "plm")

# Linear Regression
linear_model <- lm(lnwg ~ kids + age + I(age^2), data = LaborSupply)

summary(linear_model )           

But it does not support any kind of panel model (found several open/unanswered questions on this).

data("Produc", package = "plm")

# Fixed Effects Regression    
FE_model <- plm(log(gsp) ~ log(pcap) + log(pc) + log(emp) + unemp,
          data = Produc, 
          index = c("state","year"),
Error in crossprod(beta, t(X)) : non-conformable arguments

How can I calculate marginal effects for a fixed effects model and test their statistical significance manually? What is the best practice workaround? What is the math behind those calculations? Perhaps I can implement a formula in my FE-case. Worst case would be to transform my data and re-do the analysis in Stata.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.