Chapter Six
Our Trillion Dollar Welfare System
In this and the following chapters, our focus shifts to an examination of the government policies that transfer resources from some people to others. Of course, we have touched on some implications of these policies in earlier chapters, especially regarding how they affect measures of poverty and inequality. But now we want to consider more carefully how the policies operate, what consequences they have, and whether alternative policies might produce better results.
This chapter deals with the U.S. welfare system. To many people, “welfare” means a single program, Temporary Assistance for Needy Families (TANF, formerly known as Aid to Families with Dependent Children, or AFDC). Actually, that program is not the only, and is far from being the largest, welfare program in the U.S. The authoritative 2004 Green Book identifies eighty-five separate programs that “…provide cash and noncash aid that is directed primarily to persons with limited income”. These programs, together with Social Security, Medicare, and public schools, constitute the welfare system in the United States.
We begin by examining how much is transferred to low income persons through these programs.
How Generous (or Stingy) Are We?
Egalitarians continually complain that Americans don’t do enough to help those with low incomes. Are they correct? Let’s look at the facts.
In 2005, total (federal, state, and local government combined) expenditures on the 85 welfare programs listed in the 2004 Green Book were $620 billion.[ii] (Of this total, about 70 percent was financed by the federal government.) Is this a miserly sum? On the one hand, it is “only” 5.0 percent of our nation’s total income (GDP). On the other hand, it is larger than total spending on national defense ($495 billion) or on public schools ($472 billion).
Perhaps a more appropriate sort of comparison is to relate the $620 billion expenditure to the size of the problem it is intended to deal with: poverty. In 2005, the official poverty count was 37 million persons. Expenditure per poor person thus turns out to be $16,750. For a poor family of three persons (the average poor family has 3.4 persons), welfare expenditures are over $50,000, more than triple the poverty line for a family of three in 2005 ($15,577). Apparently, we are spending more than enough to completely eliminate poverty, even if the poor have zero earnings or other sources of income on their own.
This raises an obvious question: How can we be spending so much and still have 37 million poor persons? One answer, beloved of conservatives, is that the welfare bureaucracy siphons off most of these funds and they never reach the intended beneficiaries. Administrative costs are included in the $620 billion figure, and it is true that the resources available to help the poor are therefore less than suggested by the total outlays. However, administrative costs are not as large as many believe. The largest program of all, Medicaid, has administrative costs of only 4 percent of outlays. Overall, I estimate that administrative costs are only about 10 percent of total outlays on these programs.
There are two other reasons which, together with administrative costs, explain the coexistence of these huge (relative to the number of poor) outlays and continued poverty. One is that many of these programs provide benefits to people whose incomes, while low, are above the poverty thresholds. The other reason is perhaps more important, and we have already discussed it in the previous chapter. It is that most of these programs provide benefits to recipients that are not counted as income when the government counts the poor. In fact, only about 15 percent of this $620 billion is in the form of cash assistance that gets counted as income when the official poverty rate is estimated.
So it turns out not to be paradoxical that we have large welfare spending (relative to the number of poor persons) and still have many counted as poor. Of course, a proper count, as we explained in the previous chapter, would find very few truly poor persons.
This $620 billion figure actually understates the resources transferred to the low income population because it considers only those expenditure programs that meet the definition of welfare. A welfare program is one which provides benefits only to those with low incomes. They are sometimes referred to as means-tested programs, implying that you must have limited means to qualify for benefits. But the poor also receive benefits from other policies that are not means-tested. Prominent among these programs are public schools, Social Security, and Medicare. Low income households receive benefits (at taxpayer expense) from these programs, but they are not considered welfare programs because benefits do not go only to those with low incomes.
What share of the benefits from these three programs go to those with low incomes is not so straightforward to determine, but I give my rough estimates below. For Social Security, Medicare, and public schools, these figures are the benefits received by the lowest income quintile, not total spending on these programs.[iii] I also include two other sources of benefits accruing to low income persons, private charity and uncompensated medical care for the uninsured. Here are the figures for 2005:
Social Security $100 billion
Medicare $115 billion
Public Schools $105 billon
Private Charity $78 billion
Uncompensated Medical Care $40 billion
These programs therefore contribute in total some $438 billion a year in benefits to those with low incomes. Combined with the welfare programs discussed earlier, the grand total is $1.06 trillion. To be conservative, let’s say that each year Americans transfer roughly a trillion dollars in resources to those in need. This, then, is our trillion dollar welfare system.
If this sum were divided equally among the 20 percent of the population (60 million persons, larger than the lowest quintile) with the lowest incomes, each person would receive about $17,500. Coincidentally, a trillion dollars equals the total before-tax money income of middle income households (the aggregate income of the middle quintile of households, according to the Census Bureau)! It is difficult to see how transferring this amount to the low income population could be characterized as inadequate, but that’s how egalitarians view it.
[i] U.S. Congress, Committee on Ways and Means, 2004 Green Book (Washington, D.C.: U.S. Government Printing Office, March 2004), pp. K1-K13.
[ii] This number is extrapolated from the $583 billion figure for 2004 by assuming welfare expenditures are the same percentage of GDP in 2005 as in 2004. The source for the 2004 figure is: Karen Spar, “Cash and Noncash Benefits for Persons with Limited Income: Eligibility Rules, Recipient and Expenditure Data, FY2002-FY2004,” Congressional Research Service, March 27, 2006.
[iii] The estimates are made as follows. The lowest quintile of families contains 22.3 percent of all children under eighteen, so it is assumed that 22.3 percent of public school spending benefits this quintile. The lowest quintile contains 38.5 percent of those over 65, so it is assumed that they receive 38.5 percent of Medicare spending. For Social Security, which is related to previous earnings which are lower for those in the bottom quintile, I assume that the lowest quintile receives half the average benefit. The figure for charity is from Sheryl Sandberg, “The Charity Gap”, The Wall Street Journal, April 4, 2007, and pertains to only that fraction of total charitable donations that targets those in need. The figure for uncompensated care is equal to 2 percent of total health care spending, which is the share found in a study for New York, Randall R. Bovbjerg et al, Caring for the Uninsured in New York (October 20, 2006): http://www.urban.org/url.cfm?ID=311372