Over the last fifty years U.S. average household size has declined dramatically. When making within-year comparisons of household income, it is regarded as good practice to adjust for adult-equivalent household size, often by dividing by the square root of the number of household members or the poverty level for households of that size, or by summing some weighted value of number of adults and number f children.

Assuming (as I do) that this practice is appropriate, should equivalence scales also be used in making intertemporal comparisons? I am very much of two minds about this. Changes in birth control technology, divorce law, and public assistance all suggest that constraints on smaller family sizes have been relaxed. If people took advantage of that relaxation, the usual economic approach would suggest that the change is welfare-improving. On the other hand, if families are shrinking mainly because men are louts and have increased ability to shirk their family support responsibilities, then that conclusion seems insecure at best.

To put the question in more concrete terms, I am doing some work on inequality trends using the CPS, and applying an inverse square root household size adjustment to HH income. I am thinking of normalizing that adjustment with a constant so as to leave median income the same, so that it is applied within but not between years. I can not decide if that is a good idea or not. Is there a literature on this?

I find it remarkable (shocking, actually) that this forum does not have tags for family economics (that's a field, for gosh sake), marriage, fertility, or any related item I could find. Or for intertemporal comparisons of utility or welfare.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.