The individual sources that study uses for number of registered voters is the individual state SOS websites, the same as mine.
Where that study differed was in its data for number of eligible voters, which it says is based on "the Census Bureau’s most recent five-year population estimates, gathered by the American Community Survey (ACS) from 2014 through 2018."
I suspect that the reason they used that was because in order to get down to county-level numbers there's nothing more recent.
But when we compare the numbers gotten from that same source at the state level with more recent estimates of state populations as of July 1, 2020, we see that the differences are enough to matter with respect to the claims Judicial Watch made.
Again, taking Colorado as a test case, for the whole state, the 2014-2018 ACS voting age population is 4,271,322 (available
here). But the July 1, 2020 estimate is 4,557,684 (available
here). An estimate for number of eligible voters can be made from that voting age population, as was done in the first link I gave. This shows that the claim Judicial Watch makes about the state of Colorado as a state is based on comparing recent numbers for number of registered voters with outdated numbers for number of eligible voters, as I suspected.
I can't say for sure if their county-level numbers are also wrong. But this does highlight the weakness of the data they rely on (admittedly in the apparent absence of anything better).
Edit: However, once the 2020 US Census data comes out we will have better county numbers to use. Hopefully Judicial Watch updates their study to take those into account.
Connect With Us