4 The two estimates, for example, use similar definitions of children (as those under 19), and the Rockefeller Institute researchers included the state earned income tax credit, in part to be consistent with the federal analysis. However, much of the state and local expenditure data cover a July–June rather than an October–September fiscal year. Moreover, be- cause of the challenge of collecting data across 50 states, the Rockefeller report focuses on only a dozen major programs, including elementary and secondary education, state programs associated with major federal programs (Medicaid, SCHIP, Maternal and Child Health Bureau, TANF, child support enforcement, child care, child welfare, etc.), and state earned income tax credits.
5 The data appendix is available at http://www.urban.org/publica- tions/412110.html.
6 In conjunction with analyzing health reform and children’s health expen- ditures in next year’s report, we will consider expanding our estimate of children’s tax expenditures to include an estimate of the children’s share of the tax exclusion for employer contributions for health insurance. Since premiums often are larger for workers with children than workers without children, it would be useful to add such an exclusion in future reports, although we have had difficulty identifying the data to do so.
7 To calculate the children’s share of the tax expenditure budget, we first have to determine a total tax expenditure budget. To do this we sum up OMB’s estimates of tax provisions for individuals, even though such provisions are not strictly additive because of interaction effects. For this analysis, we do not include corporate tax expenditures, but we do add in our estimate of the dependent exemption so it would be included in our tax expenditure budget, to be consistent with our children’s estimate. (The dependent exemption is not in OMB’s estimate of total tax expenditures because OMB does not classify the dependent exemption as a special tax provision resulting in a tax expenditure but instead views it as part of the overall tax structure.) Our resulting total is $909 billion, including $875 billion in individual tax expenditures and $34 billion for the dependent exemption.
8 Our Urban Institute colleague Ajay Chaudry and his coauthor J. Law- rence Aber also found that 18 percent of ARRA went to children, although they used the earlier CBO estimate and somewhat different methodology (for example, they included Pell grants but did not include immunization funding). See“Low-Income Children, Their Families, and the Great Reces- sion: What’s Next in Policy?,” http://www.urban.org/url.cfm?ID=412069.
9 State funds on elementary and secondary education declined by 0.5 percent (about $1 billion) between 2008 and 2009, even before adjusting for inflation, according to the Fiscal ear 2008 State Expenditure Report of the National Association of State Budget Officers. We estimate that state
Medicaid spending also declined by $4 billion nominally. Federal increases in education and Medicaid were larger than the state declines in these two areas, but we cannot say if total investments increased between 2008 and 2009 because we do not yet have information on local spending on educa- tion in 2009.
10 The state estimate does include the value of state earned income tax credit in states that have such credits. Recall that the bulk of the federal earned income tax credit (the refundable portion) is included in the fed- eral estimate.
11 See Macomber et al. (2010). The proportion of funding provided by the federal versus state and local governments also varies significantly by age, ranging from 77 percent federal for infants and toddlers in 2004 to 27 percent federal for those age 6 to 11.
12 The estimate of elderly spending is not as comprehensive as the esti- mate of spending on children and thus may be conservative. The federal spending includes five major federal programs (Social Security, Medicare, Medicaid, Supplemental Security Income, and Supplemental Nutrition Assistance) and a rough estimate of spending on a dozen other programs based on CBO estimates of federal spending on the elderly in 2000 inflated to 2007. The state estimate is based on elderly spending on Medicaid and state supplementation of SSI payments, ignoring any other small sources of state spending on the elderly.
13 As explained in the methods section, unemployment benefits are not counted as spending on children because benefits to unemployed parents are generally the same size as benefits to individuals without children, and thus there is no marginal benefit attributable to children. However, see footnote 3.
14 Tax reductions benefitting children are included in this analysis, but not tax reductions benefitting the elderly and disabled. Tax reductions are consistently excluded in the comparison of spending on children and the elderly provided in figure 6. Note that between 1960 and 2009, the elderly population more than doubled (from 16.7 million to 39.6 million) while the child population grew by 17 percent (67.1 to 78.4 million).
15 For example, a cut in tax rates from 28 to 25 percent would reduce the value of a $3,500 exemption from $980 to $875, thereby reducing the tax advantage of being a taxpayer with a child (relative to taxes for childless taxpayers) and, thus, child-related tax expenditures. This does not mean, however, that families with children were paying higher taxes than before the tax cut, just higher taxes relative to childless taxpayers.
16 Information on how we classified each program by benefit type (cash vs. in-kind), eligibility limitation (means-tested or not) and spending type (mandatory vs. discretionary) is provided in the data appendix.