I have a subnational-level dataset that includes census and election data. The census years don't match up with election years, so I want to impute data for some important variables (e.g. education, poverty, etc.). I have three decades of census data, so three points. I know I could estimate a parabolic curve with those three points (in some cases it's almost a linear relationship) and from that impute vector (x,y coordinate) data for the missing non-census years. I can do the match. But I want to be able to do it automatically (preferably either in Excel or Stata) in such a way that I can easily update the data if any corrections are necessary without having to redo all the math. (I also don't want to brute force calculate hundreds of different curve estimates. However, I'm at a loss as to how to do this. Any ideas?