He said 1:1 correlated, not 1:1 equal. I.e. if for the sake of argument one in five unemployed are shut-ins but nearly no employed people are, then you have a g

He said 1:1 correlated, not 1:1 equal. I.e. if for the sake of argument one in five unemployed are shut-ins but nearly no employed people are, then you have a good starting point to estimate populations.