I'm trying to match towns from a dataset to Census tracts in the year 2000, with FIPS code as a place identifier. I'm aware of http://mcdc2.missouri.edu/websas/geocorr2k.html#GFOPTS, the GeoCorr service provided by U Missouri, but this crosswalk only has places at the CDP (Census designated place) and borough level for many municipalities in New England. The data I am trying to match has places at the town and township level. For example, in GeoCorr, only Adams CDP MA, with FIPS code 590, is listed, whereas my dataset lists Adams Town MA, with FIPS code 555. Google maps indicates they occupy approximately the same location, but Census Bureau Quickfacts suggests these two places are in fact different.
For future reference, I just figured out the solution to this question, with help from a statistical research/data coordinator. The town and township level (which in the example given is Adams Town MA with FIPS code 555) is of the summary level Minor Civil Division (MCD) or county subdivision. It turns out the US Board on Geographic Names gives different FIPS class code definitions for different summary levels, for the same place. Since Adams CDP (census designated place) is a different level, it turns out on U Missouri's GeoCorr service, both County Subdivision (MCD) and Place must be selected as target output.