In listening to self-described “radical” and “feminist” economists at the American Economic Association conference (example) and also in reading popular articles such as “Pay Gap Is Because of Gender, Not Jobs” (nytimes), the fact that men and women in the U.S. earn different amounts via W-2 wages is accepted as ipso facto evidence that our society is unfair and needs “reform.”
Perhaps we do need reform, and perhaps there is discrimination by employers, but I’m wondering if this is the most important analysis for figuring out if society overall is unfair. Does it matter to Citizen Ruth what she earns, and therefore what she can write down on her 1040 tax return, or what she can spend, and therefore the kind of lifestyle she can enjoy? If spending power is relevant then we may not have any idea how equitable our society is.
Simplest corrections: Do women typically give men more gifts or vice versa? Spending power should be corrected for that. What about paying for vacations and entertainment? If there is a disparity in who pays for these, including among the unmarried, that should go into the accounting. Cash income? Are men or women tending to earn more in the cash or underground economy? Cost of living adjustments? Are there more men versus women who live in high-cost states such as Hawaii, California, New York, or Massachusetts? Are there more men than women who live in high-cost cities versus low-cost rural areas?
For the bigger corrections we need to consider once class at a time.
Let’s consider poor Americans first. A poor woman, like a rich woman, is overwhelmingly likely to be the winner of any kind of custody dispute in the U.S. courts. Thus she is more likely to have possession of a child than would be a poor man. If we assume, as do the child support guidelines, that the child has no value to the parent and is a pure cost, the mother may be better off than the father even if the father is also poor and pays her minimal cash. A poor single parent is entitled to a lot more welfare benefits than a poor childless adult. The single mother may get a free apartment, cash payments, additional food stamps, etc., that are unavailable to the single father. Those should be entered on the female side of the ledger. (“The Work Versus Welfare Trade-Off: 2013” report from Cato Institute runs the numbers for a single mother with two children, roughly equivalent to a pre-tax wage of $50,540 per year in Massachusetts.)
What about married people? A married person, especially if he or she left the workforce for a number of years, may have a lower W-2 wage compared to an equivalently educated spouse. Yet the lower-income spouse may be able to spend some of the higher-income spouse’s wages in addition to his or her own, possibly enjoying a higher ability to consume than does the nominally higher-earning spouse. If it is primarily women who are the lower-earning spouses in marriages then this correction would result in a boost to total female spending power. [Anecdotally when our married friends talk about a financial decision, e.g., whether to do a kitchen renovation desired by the wife or to spend money on a hobby or trip as desired by the husband, it is generally the wife whose preference prevailed.]
How about divorced Americans? Our interviews with divorce litigators nationwide reveal that in many jurisdictions the spouse with a lower earning capacity would be entitled to a larger-than-50-percent share of any marital assets. These money flows can be significant but are not present in any statistics on earnings or wages. (See, for example, the $1 billion that Harold Hamm was ordered to pay the wife who sued him.) In most states the lower-earning spouse can also get alimony, which would appear in household income data (assuming that she voluntarily and accurately reports it to a surveyor) but not in any data on wages (the primary source for complaints that the U.S. is gender-biased). A divorced adult may also collect child support (see below). Again, if the lower-earning spouse tends to be a woman correcting for these effects will boost total female spending power.
Divorce itself can be a significant expense (estimated at roughly $50 billion per year paid to lawyers, psychologists, etc.) and divorce litigators nationwide told us that often men were ordered to pay women’s legal fees. So it would make sense to look at who filed the divorce lawsuit (mostly women) and assume that the person did so for personal benefit, then look at who was ordered to paid the costs (ultimately the children, actually, say the litigators), and treat that spending as a gift from one gender to the other.
What about never-married Americans? Never-married women who live in cities may actually earn higher wages than men (punditfact). The Wall Street Journal says that never-married women nationwide earn 96 percent of what men earned, according to the BLS. Based on our interviews with attorneys and consumers, however, the official stats leave out some spending power. Some single women who became pregnant in states with unlimited child support managed to sell their abortions ($250,000 was a typical number and no W-2s were issued for this “labor”). If a child is produced, a never-married woman could supplement her income by having a child and collecting child support.
How does the child support system affect this analysis? Suppose that an adult American has three children with three dermatologists. If the sexual acts that produced these children occurred in Wisconsin, the parent who wins custody can get 17 percent of each doctor’s pre-tax income, pretty close to 33.3 percent of post-tax income and thus his or her net spending power is the same as any one of the dermatologists. As women have the power to decide whether or not to carry a baby to term, it would most likely be a woman in this situation. If she got bored and decided to take a $20,000/year job at a non-profit organization her W-2 earnings and the doctors’ would go into the comparison cited by political advocates to show that women in the workforce were not being treated fairly.
The adult who gets custody of a child also has a lower tax rate for his or her earnings. Regardless of the amount of child support revenue or how much, if any, is spent on the child, the child is a “dependent” from the IRS’s point of view and the custodial parent gets a tax deduction for that child. If the custodial parent receives gifts or buys things for the child and then gives those things to charity, the adult can take a tax deduction for the value of the items. Once again, this lifts spending power without changing W-2 earnings. (See “The High Price of Being Single in America” (Atlantic) for how small tax advantages can add up.) To the extent that it is primarily women who can choose whether or not to continue a pregnancy and who can obtain custody of a child, this correction would result in an increase in spending power for women.
Readers: What else would be required to create an accurate accounting of the difference in spending power between America’s men and America’s women? And is it more important to focus on spending power or on W-2 wages?
Related:
Full post, including comments 
looks at social mobility, of which income and wealth correlation from generation to generation is a component, across a range of societies including the U.S., England, Sweden, India, China, Japan, and Korea. It turns out that there is no correlation between the ethnic homogeneity of a society and the amount of social mobility, contrary to what one would expect if racism were a factor in keeping sub-sections of a society either in a high-status or low-status condition.