The data were reported at different points in the month. Nevada, for instance, provides numbers through Oct. 10, while Colorado’s reporting is fairly up-to-date (it’s current as of the 28th). Regardless, I prorated the numbers through the 31st. This probably understates the growth of enrollment, given that some states have worked out the bugs in their websites and people have become more aware of the programs.

Next, I looked at each state’s share of the national population. I then divided total enrollment in the state by the state’s share of the national population. This allows us to extrapolate how many people might have enrolled if a given state’s exchange had been a national one (or, if you prefer, if a given state had the population of the country as a whole).

The numbers vary by a fair amount. If exchanges had functioned and attracted customers nationally at the same rate as Minnesota, they would have attracted some 53,000 customers. On the other hand, if they had functioned like Kentucky’s, they would have attracted around 450,000, which would be close to the administration’s low-end estimate. On average, it works out to about 280,000 customers nationally.