First, and most importantly, we caution against comparing any of our products to prior vintages. Products are updated every year with the best available data and methods so it is entirely possible that high variability exists when comparing vintages. In this case, the variables that exhibit high rates of change from last year tend to be variables with very low sample from the raw data (these variables contain a “(!)” in the description) or variables that have had a large increase in their sample size from last year. Both scenarios would lead to high variability in reported variable counts.