ECM with two lagged variables - how to set up the equation?

Question

I estimated an ecm with two variables: GDP-growth and public expenditure growth with the aim to investigate dynamic effects. I already checked stationarity with the KPSS-tes