It's a census, so they're taking a sample. I would assume that 75% of counties had statistical outcomes that weren't statistically significantly different from

It's a census, so they're taking a sample. I would assume that 75% of counties had statistical outcomes that weren't statistically significantly different from the previous measurement. Consider if you sample the same county on the same day 10 times, you'll get 10 different outcomes, that doesn't mean income changed that day. In addition, they may have used a threshold value. But it's not noted anywhere. I'd also assume indeed there was inflation-adjustment.