collapse

[Monicucha Paganini]

I am merging two datasets: one about employment of household members and another one for FDI at the sectoral level. I merged the two by country, year and sector.
my questions is, if I want to regress the employment share in manufacturing on fdi level and the dependent variable is at the country/provincial level, can I collapse the data or is it better to reshape wide at the provincial level?

thanks