Reviewing all the welfare spending.

I’ve been reading an interesting report on means-tested welfare spending put out by Robert Rector and some friends at the Heritage Foundation back in September of 2009. Although the report has some very misleading language and some unfair presentation of a numbers game, it does indeed provide some very good observations and suggestions. The most fundamental problem with the whole report is that it seems to take a position that America has two classes: the low-income persons who tend to be unmarried and irresponsible and lack a strong work-ethic, who receive hundreds of billions of dollars in welfare; and the taxpayers who work and earn an income and pay taxes to support the low-income class. The flaw with this whole framing of the issue is that about 70% of the low-income class that receives means-tested benefits is in fact destined to become part of the taxpaying, income-earning class. That is, most people who are poor and receive means-tested benefits are only doing so for a few years while they are children or young parents, and after a few years in poverty they transition out of poverty and generally stay out of poverty unless they go through a life-changing thing like a divorce or an injury or illness, and even then they tend to bounce back out of poverty again. Of the remaining 30% or so who spend more years in poverty you have about half who are on the upper end of the low-income spectrum, and they tend to be low-skilled persons whose incomes bounce around between 50% of poverty and 200% of poverty, depending on life events or labor market conditions. The small percent of the population that spends almost their whole lives in poverty include those who have chronic illnesses, disabilities, developmental disabilities, chronic mental illness, personality disorders, or substance dependence problems. There just aren’t many able-bodied persons in good mental health who are free of addictions and remain poor, receiving means-tested benefits for decades.

One of the best points made by the Heritage Foundation report is that if you combine all the means-tested welfare spending and divide it by the number of persons who live in households earning under 200% of the poverty line, the benefits would equal about $7,000 per person. For a family of four that would equal $28,000, and would bring the family over the poverty threshold (which for a family of four is under $23,000). So, with all this spending on means-tested benefits, we taxpayers ought to expect that after tax and after benefit incomes in the USA would leave no one in poverty. At least, this is what an efficient welfare system would do.


By my calculations, all the benefits and means-tested programs bring the poverty rate from 13.2% (the pre-tax, pre-transfer poverty rate in 2008) down to about 5%-6%. This means that after taxes and transfers and benefits we still have about 5% to 6% of Americans consuming housing, food, energy, medicine, clothing, and such necessities at a level below the poverty threshold. We still have, for example, about 3% or 4% of the population experiencing significant food insecurity (skipping meals, or making choices between eating, paying for housing, purchasing medicine, or heating/cooling their homes). Each year over 100,000 Americans die from medical problems because they can’t afford medicines or medical treatment (see the old 2003 study in the New England Journal of Medicine by David Himmelstein and Steffie Woolhandler). That’s obscene, but it’s about what you would expect if about 4% of Americans were still poor even after taxes and income transfers and benefits, and if the Americans who were poorest were also the Americans with the most health problems who needed more medical care. So, I think the Heritage Foundation does make a good case for re-thinking how we do our means-tested welfare system so that our fairly generous benefits (and I’ve lived in poor neighborhoods within poor nations, so I’m going to admit that American benefits to the poor are often generous) were more efficiently distributed to really eliminate poverty.


I think Robert Rector and the Heritage Foundation may suppose that many taxpayers will feel outraged by average transfers of income from us to our fellow Americans who earn only 200% of poverty or less. For example, my household is solid middle class with a gross pre-tax income of about 236% of poverty, and in 2008 about 14.4% of our income went to all forms of taxes (about $7,500). If we earned $10,500 less than we do, we would be earning 190% of the poverty threshold, and according to the Heritage report, the average benefits to persons in households earning under 200% of poverty is $7,000. That makes it sound as if taxpayers like my family are being brought down from $236% of poverty to 204% of poverty by taxes in order to raise up families earning, for example, 120% of poverty (about $26,500 for a family of four) to 247% of poverty. In other words, means-tested benefits would be adding $7,000 in means-tested benefits per person in the four person household to raise their post-tax post-transfer income (actually a consumption level rather than an income) to $54,500, a couple thousand more than we earned before taxes, and more than $10,000 more than we consumed after taxes and transfers.

Indeed, if things really worked that way, we would have middle-class persons paying $7,500 in taxes so that families making around half of what we do before taxes and benefits could end up living lives at higher consumption levels than we do. You see this perception informally when people at lower-middle-class incomes and working-class incomes, say $30,000 to $40,000, complain about persons with SNAP benefits purchasing better food at the grocery store or driving better cars or living in better houses than they can afford.

Since the American Gross Domestic Product per capita is about $48,000, it seems reasonable to me that we would be transferring $6,000 to $7,000 per capita to the Americans who have incomes under 200% of the poverty level. After all, American spending on medical care, when divided by the population, equals about $7,000 to $8,000 per capita, so I’d think that the poor and near-poor would get income transfers at a level similar at least to this level.

The odd thing is that my household, which is right near the middle of household income distribution in my state (slightly under it), earns only about 27% of what we would if the American GDP was spread equally over everyone (so that each of the four of us had $48,000 per year). Shouldn’t middle-class Americans near the middle of the income distribution be earning more like 40% of what the GDP per capita would be for their household? Why doesn’t the Heritage Foundation publish a paper that looks at that issue of unequal distribution and how the households making over $100,000 are screwing all of us in the middle class who work our 1,900 to 2,000 hours per year for a mere $30,000 to $75,000?

Another thing I love about this Heritage Foundation report is that it puts all the welfare policies together in one place, so you can get an idea of the scope of each type of welfare. I’ll run through them here, sharing my notes.

Medical care. Medicaid. Maternal and Child Health Block Grant. S-CHIP.
Combined state and federal spending was $372.1 billion in FY-2008.
This was 52% of the public effort toward the general welfare through means-tested programs.

Cash Aid. Temporary Assistance to Needy Families (TANF), Supplemental Security Income (SSI), the Earned Income Tax Credit (EITC), and the Additional Child Tax Credit (ACTC). Note that the EITC and ACTC are tax expenditures, so they don’t show up as government spending in the budget. They just show up as less revenue taken in by the IRS (and the IRS actually has the treasury send money back out to people with the EITC and ACTC - a refundable tax credit). According to the Heritage Foundation, total State and Federal spending on the means-tested cash aid was $153.8 billion in FY 2008. Hmm, the TANF budget for FY 2010 is supposed to be $18.6 billion. The EITC was recently still paying out, I think, less than $50 billion. I believe even now SSI is paying out slightly over $50 billion. Many states have EITC, which must explain why this number is so much higher than the federal government's estimate of its spending. I guess maybe if the ACTC is running at about $18 billion and state-run EITC programs pay out $17 billion, the Heritage Foundation numbers are probably about right.

Means-tested food aid comes through Supplemental Nutrition Assistance Program (SNAP, which is still called “Food Stamps”), which in Illinois is delivered through a LINK card. The Women Infants and Children (WIC) program serves almost half the infants born in the USA (about 8.7 million each month in 2008), at a cost of $6.2 billion in the fiscal year 2008. In 2009 SNAP helped 33.7 million persons (summing up all persons living in households that received SNAP benefits), and the average amount received was about $125. In Illinois the maximum benefit for a family of four persons was $668. The program cost about $54 billion that year, at least from the federal level. State programs that administered SNAP added significantly to that level of aid. About 1.5 million persons in Illinois lived in households receiving SNAP benefits. There are also federal school lunch programs (low-cost or free lunches for 30 million children per day at a cost of $9.3 billion), school breakfast programs (breakfasts to 10.6 million children each school day at a cost of $2.4 billion), summer food service programs (meals to 2.1 million children in the summer at a cost of $0.33 billion).

Families following the USDA’s Thrifty Food Plan should be spending 40% to 50% of their food budget on fruits and vegetables, but spending behavior studies in both poor and wealthy families show that everyone tends to spend about 16% to 18% of their food budget on fruits and vegetables. The American Time Use Survey suggests “that low-income women who work full-time spend about 46 minutes per day on meal preparation,” but if people are buying cheaper whole foods rather than prepared and highly processed foods or restaurant foods, their time on food preparation probably ought to average closer to one hour per day on food preparation. Quite possibly some of the resentment against families using SNAP or WIC is because those families buy highly-processed foods and get lots of meat, when they (and all of us) should instead by getting lots of fresh produce instead.

The Heritage Foundation claims that if you combine federal and state spending on means-tested food aid the total cost was $62.8 billion in FY 2008. This is one Heritage figure that seems way too low. I estimate that in 2010 we'll be spending over $70 billion, and perhaps as much as $75 billion on means-tested food aid.

Housing and utility assistance is difficult to calculate at the federal, state, and local level, because it seems almost every power company district has a variety of local programs that get mixed in with the federal Low Income Home Energy Assistance Program (LIHEAP). With housing the main programs are public housing and rental assistance (which is a diverse mix of programs, including the Section 8 Housing Voucher). The Heritage Foundation sums up LIHEAP and the HUD programs for means-tested housing assistance and comes up with a figure of $45.1 billion for Fiscal Year 2008 in housing and utility support. That seems pretty high, as the entire HUD budget in FY 2009 was only about $39 billion and LIHEAP gets between $3.5 and $5.5 billion in funding. I guess the Heritage Foundation must have calculated how much states spend on their state housing authorities and various state and local programs to end homelessness or provide affordable housing to persons with low incomes.

Social Services provided by federal and state governments to persons who qualify through an income test account for $11.6 billion (in FY 2008), according to the Heritage Foundation, and this includes the Social Service Block Grant (SSBG), some TANF funding (I’m assuming Heritage split TANF funding between their category of cash aid and this social service aid category), and Community Services Block Grants (CCDBG).


Child Development and Child Care programs that are means tested include Head Start and Child Care Development Block Grants (CCDBG). The Heritage Foundation says that federal and state funding for these programs was at least (and probably exceeded) $17.7 billion in FY 2008. Some of the means-tested cash aid and social services help was probably spent on child development and child care, so some estimates for those categories may be a bit high while the estimate for this category may be a bit low.

Jobs and job training are frequently available to people without any income test or qualification, but there are some means-tested jobs and job training programs, and the Heritage Foundation says these cost $6.3 billion in FY 2008. They are including stuff like the Workforce Investment Act (WIA) program for adults, Job Corps, Workforce Investment Act Opportunity Grants for Youth, and once again, TANF (which was counted already in cash aid and social services, and contributes to this category as well).

The Heritage Foundation also has a means-tested category for community development, in which they include the Community Development Block Grants (CDBG). These funds are supposed to be spent in areas with high concentrations of persons with low incomes, and much of the money actually goes to help middle-class people who improve infrastructure or offer services or open businesses in areas of concentrated poverty, but at merely $8.2 billion, it’s not worth debating whether CDBG are a means-tested program for the poor; as it certainly sends money into poor communities. I’m puzzled how HUD’s entire budget is under $40 billion, but the Heritage Foundation sums up housing assistance and low-income community grants to get a figure exceeding $53 billion (even subtracting $5 billion for LIHEAP, there still seems to be an extra $8 billion coming from state spending or matching grants, I guess).

The Heritage Foundation also created a category for means-tested educational spending. This is mainly made up with Pell grants and Title 1 Education grants going to low-income communities. By their estimate, this means-tested educational spending was $35.5 billion.
Welfare Programs that Are Not Means Tested

Those are the means-tested welfare benefits. The larger portion of public spending on the general welfare comes from public spending on medical help for persons injured on jobs, unemployment payments to temporarily unemployed persons while they move between jobs, benefits for disabled workers, retirement benefits and health benefits for retired persons, and public education. But I'm going to pretend education is it's own thing, apart from general welfare provision.

The mortgage interest deduction (interest you pay on a home loan for your primary residence is deducted from your taxable income). For example, our household makes a bit less than the median Illinois household income, and our home is valued just slightly under the median home value in Illinois, and we pay several thousands of dollars in mortgage interest each year, and this reduces our taxable income to a point where we end up paying about $450 less in federal income taxes. In other words, the federal government is helping our middle-class family pay for our median-value home to the tune of about $37 per month. If we lived in a real estate market where home costs were much higher, or if we were wealthier and could afford a much more valuable home, then this mortgage interest deduction might save us thousands of dollars in taxes rather than a few hundred dollars. This costs the federal government about $80 billion per year.

The government gave away $10 billion in tax revenue with the first time home buyer credit. But this was a part of the economic stimulus package to save us from the Great Recession becoming the Great Collapse.

The Heritage Foundation did not even mention the $16 billion in federal farm program payments that go out to about 2 million farmers or farms. Persons or farms with incomes over $2.5 million who derive less than 75% of their income from farming, ranching, or forestry aren’t supposed to qualify, but the GAO thinks nearly $50 million is paid out to such persons or farms each year. Anyway, I have family who farm, and since they mainly grow fruits and vegetables, they get hardly any subsidies at all. These subsidies go to cotton farmers and rice farmers and sugar farmers and corn or soybean farmers. They're a real mess, too.

Social Security is nearly $600 billion ($550 billion in FY 2009). for old age and survivors benefits.
Social Security Disability Insurance costs about $120 billion ($117 billion in FY 2009).
In Calendar Year 2009 Unemployment Insurance and various emergency unemployment benefits paid out about $130 billion in benefits. In normal years benefits from unemployment insurance and related programs usually $25 to $30 billion, and until the Great Recession the benefits would typically double over normal rates during a recession.

Medicare is about $450 billion.


So, with subsidies to home-buyers (generally not low-income), social security, Medicare, and unemployment, we are spending about $1.4 trillion in non-means-tested benefits, and about $0.7 trillion in means-tested benefits. So, it seems to me income transfers from the top 80% of the population back to the top 80% of the population must represent transfers at least double the level of transfers from the top 50% of the population to the bottom 20% of the population. Robert Rector seems really alarmed by this, and he wants people to pay more attention to the means-tested welfare programs. But to me, the spending seems reasonable.