To sum up, any differences in unobserved levels of average motivation between single and married mothers complicate using much of the existing empirical litera- ture to estimate the taxpayer cost of family fragmentation. The assumptions that underlie this analysis, however, are extremely cautious in an attempt not to over- state the taxpayer costs. Specifically, by assuming no beneficial behavioral effects of marriage on adults or children, we are likely underestimating taxpayer costs.
Is $112 Billion Too Low?
In this section, we consider four arguments that suggest that the $112 billion esti- mate is too low:
Ignoring the EITC, public education, and other government programs underestimates the true taxpayer costs of family fragmentation.
Ignoring the direct impact of family fragmentation on crime (independent of poverty) underestimates the taxpayer costs.
Ignoring any impact of marriage on single fathers understates the tax- payer cost of family fragmentation.
Ignoring the fact that, given income-eligibility, single-mother households are much more likely than married households to take up subsidies from transfer programs underestimates the likely taxpayer costs of family frag- mentation.
Ignoring the EITC, public education, and other government programs underesti- mates the true taxpayer costs of family fragmentation.
We ignore EITC expenditures largely because of the lack of empirical information needed to make reasonable assumptions about how marriage will affect usage of the EITC and related programs in our complex tax code. But ignoring potential tax- payer savings produced by marriage on EITC expenditures means ignoring a very expensive government program that is almost certainly affected by marriage rates. Taxpayers spend approximately $40 billion on cash assistance to the working poor under the EITC. As shown in table A.2, using the assumptions in this study, family fragmentation would lead to about $12.68 billion in higher taxpayer costs on the EITC. Adjusting this estimate based on the results of Acs and Maag, as discussed under the first argument in the previous subsection, would reduce that amount by about $0.5 billion, leaving a net taxpayer cost of about $12.18 billion.
We have chosen to ignore the EITC expenditures (including potential savings of an additional $12.18 billion each year) because the consequences of marriage for the EITC are complex and would involve multiple assumptions of how marriage would affect men’s and women’s earnings.
In addition to the EITC, this analysis does not assume any costs of family fragmen- tation to the public school system, which is almost certainly not true. Considerable research suggests that children raised outside of intact marriages are more likely to