I have some cross-sectional data, where I'm interested in estimating a relationship between the "number of items in use" in sector i of country j and the size of that sector (represented by spending power). There are other control variable. However, I have six (6) sectors and 52 countries. This implies 52 observations for each sector, and therefore 312 observations in total. Running a straight regression feels weird. Also, there are many cases where the same "item in use" appears in all or many of the 52 countries, making a choice of appropriate model a little complex. Any ideas on how to handle this? Thanks in advance!