In the Nation's Interest

What’s Behind the “47 Percent?”

By Joseph Minarik

It looks like 47 might become the biggest number in U.S. politics since “54º 40′ or fight,” which is coincidental enough given that 47 is the average of 54 and 40.  But beyond coincidence, what’s with 47?

It seems clear that the now famous (or infamous to some) number came from an estimate of the percentage of the population that pays zero or negative federal income taxes.  The precise estimate was 46.4 percent of all “tax units” (individuals or families that file income tax returns) as of 2011.  The estimate came from Roberton Williams of the non-partisan Tax Policy Center (TPC, an organization created jointly by the Brookings Institution and the Urban Institute).

When that measurement concept is raised, one usually hears that 47 (or 46.4) is too high a percentage of the population to pay no income tax.  One concern is that bearing at least some of the burden of government is a part of citizenship.  Another is that persons who pay no taxes can vote for candidates who will give them benefits without bearing any of the cost. However, this coin has two sides.

The same data source reports that a smaller proportion, 27.6 percent, paid no income or payroll taxes.  So the concept of the sharing of the burden is sensitive to the tax universe considered.  On the political risk, there is no evidence of avid voter turnout on the part of low-income persons (or any other group, for that matter — and certainly not for the population as a whole, sadly).

Another consideration is that the “47″ number (returning to the federal income tax alone) is highly volatile.  Turn the clock back to 2007, and the magic number was 39.9 percent (and only 21.9 percent paid no income and payroll taxes).  With the financial collapse in 2008, the number leapt to 50.8 percent (27.5 percent for both taxes).  So if there is an element of sticker shock in the 47 number, it is a function of the economic times, and for obvious reasons.  When people are out of work, their incomes are down, and their tax liabilities decline or evaporate.  Similarly, tax liabilities move with the trend of income as well as its cycle, and wage growth has been next to nil since years before the financial crisis — and obviously worse since.  But with the anticipated economic recovery and moderate wage growth, the TPC projects that the population paying no income taxes will decline to 36.1 percent by 2022, even if all of the 2001 and 2003 tax cuts are extended.

Still, with all of that variability aside, and apart from the current financial troubles, why are there so many “tax units” — for simplicity, we might say “households” — that pay no income taxes?  There have been several policy developments that have pushed in that direction.

“Don’t tax people into poverty.”  As long as 40 years ago, in the troubling economy of the early 1970s, there was significant national concern about an upswing in poverty.  And at that time, households with incomes below the official poverty definition did pay positive amounts of income tax.  Many Americans reportedly believed that this was counterproductive, or even inhumane.  The result was straightforward increases in the personal exemption and standard deduction such that those in officially defined poverty paid no income tax.

“Reduce the filing and administrative burden.”  At the same time, there was concern that low-income people were filing, and the Internal Revenue Service (IRS) was processing, large numbers of returns that yielded next to no tax liability.  Although low-income returns are usually the simplest in a purely technical sense, low-income working people likely would have the most difficulty bearing that burden.  And the IRS found that such returns were among the most difficult to process, because they had the most frequent illegible entries, arithmetic errors, and the like.  Although this issue did not reach as high a profile as the “don’t tax people into poverty” concern, it was another motivation for the increases in the personal exemption and the standard deduction, which also raised the income threshold below which households were not required even to file returns.

Note that improved optical scanning technology and personal-computer software probably reduce, but do not eliminate, this concern.  And note also that the following consideration has substantially reversed the simplification issue.

“Make work pay.”  Again in the 1970s, concern arose that people with limited skills had little incentive to work, partly because of their low potential wages, and partly because they would lose government benefits as their wages rose.  One option was an increase in the minimum wage.  However, in the weak economy and the high inflation of the 1970s, a higher minimum wage could have induced employers both to cut down on their use of low-wage labor and to pass any remaining higher labor costs along as higher prices.  So instead of increasing the minimum wage, policymakers chose to create a wage supplement through the tax system — which was the genesis of the earned income tax credit (EITC).  The EITC was originally restricted to families with children, in keeping with the commonly held public priority of targeting support to children.  There is now a credit for persons without children, but it is much smaller.

Though in concept simple, the EITC is more complicated than it may seem.  Among the issues:

In the not-infrequent instance of divorce, a father and a mother each typically files his or her own tax return.  Because the EITC is much more generous for families with children, it can become contentious and complex to determine which parent can claim the children for purposes of the EITC.

Also, the EITC was designed to be a supplement to wage income.  The law has become fairly complex to deal with the receipt of portfolio income, intending not to discourage saving, but also not to provide a benefit to those who might choose to live off of a portfolio rather than to work.

Furthermore, the EITC is based on annual income, which is not knowable until the end of a calendar year, and may be unpredictable to the taxpayer.  Beyond complexity and uncertainty, the major consequence for the typical EITC beneficiary is that the credit is determined only upon the filing of the tax return reporting the income for one year in the early months of the following year.  The result is that a tax credit intended for households with modest incomes, in many instances living paycheck to paycheck, becomes a kind of Christmas club that bears no interest — arguably less than fully effective in fulfilling a reasonable conception of its mission.

And finally, although it is in broad terms a work incentive, the EITC does phase out at rates of either 16 percent or 21 percent (for families with children; a lesser 7.65 percent rate applies to the much smaller credit for taxpayers without children).  For those workers in the phase-out range, that could be a disincentive to work additional hours, or to accept a modest pay raise that carries additional responsibility.

So while many would consider the EITC to be a public-policy home run, it is not a grand slam.  It is arguably one of the most complex features of the income tax, and is an important reason why many low-wage workers now rely on paid tax-return preparers.

“Offset the payroll tax.”  Moving beyond its original conception as a wage supplement, the EITC came to be conceived as an offset for low-wage workers’ payroll tax liabilities.  At a first cut, the payroll tax is regressive, and some have argued that this burden should be reduced for those with modest wage income.  This picture is a bit cloudy, though, given that payroll taxes do earn future Social Security benefits based on a formula that is more generous for low-wage workers.  However, although there is no definitive parsing of what part of which statutory increase in the EITC was intended to serve what purpose, offsetting some of the burden of the payroll tax has been used as an argument in the debate over proposed increases.

“Support modest-income families with children.”  Independent of the EITC, the Congress added a child credit to the tax law.  This was in addition to the personal exemption that had been available for each member of the family, including children, since time immemorial.  The choice of a credit, rather than a simple increase in the personal exemption either for all persons or just for children, was to avoid the effect in a progressive system of the exemption giving a larger dollar benefit to upper-income taxpayers in the higher-rate brackets.  Then, in 1997, the tax law changed to make the child credit partially refundable under a highly complex set of circumstances, including any excess of Social Security payroll taxes over the EITC.  The effect was to increase the number of families who pay negative income taxes (that is, who receive all of their withholding (if any) back, and more dollars besides).

Net effect.  So how does all of this add up?  Who, in fact, pays no federal income taxes?

There probably are three main groups who pay no income tax.  The first and most obvious is those who engage in no market activity, and have no income from either capital or labor.

The second is the elderly.  Consider a retired couple who live off of interest income only.  In 2011, each elderly person received an exemption of $3,700 ($7,400 for the two of them), a $13,900 standard deduction, and a tax credit created for the elderly and disabled.  Such a couple with income of as much as $22,888 would pay no income tax.  Just for perspective, if that couple were investing at a 1 percent yield — which probably is not the lowest available for bank savings accounts that allow immediate access to interest earnings — they would have a portfolio worth $2,288,800, which is certainly far above the average wealth for retirees.  Any Social Security benefit on top of that amount of interest income would be partially taxed, however.  But any pension income that the couple would receive, if matched by a reduction of interest income, would leave the federal income tax liability at zero.  Thus, for example, if the couple had a $10,000 pension income, they could receive $12,888 (from a portfolio of perhaps $1,288,800) and still pay no income tax.  Putting all of these facts together suggests that the percentage of elderly persons who currently pay no income tax probably is quite large.  (Calculations by the TPC suggest that the elderly share of the nontaxable population could be about half.)  For perspective, the 2011 official poverty standard for a couple is $14,710, so such an elderly couple would have an income of just over 1.5 times the poverty line.

Source Data: Urban-Brookings Tax Policy Center Microsimulation Model (version 0509-4)

The third group is families with limited labor income.  Because each child in a family receives a personal exemption and a child credit, and the EITC is most generous for families with children, the amount of income a family can earn without paying income tax increases significantly with family size.

To provide a yardstick, consider that the current federal minimum wage is $7.25 per hour.  Thus, an individual who works full time for an entire year (for simplicity, let’s say 50 40-hour weeks, or 2000 hours) would earn $14,500 per year.  And for simplicity, let’s look at husband-and-wife families where one spouse works, and assume that the couple does not itemize deductions.

A couple without children, and therefore not eligible for the EITC or for the partially refundable child credit, would have relatively simple tax circumstances.  That couple in 2011 paid no tax with an income of no more than $19,000 — about 31 percent above the minimum wage, again assuming that only one spouse worked full time and full year.

With children added to the family, the EITC and the child tax credit come into play, circumstances become much more complex — and the amount of income at which federal tax liability is zero can increase substantially.  With two children, the family breadwinner could earn $45,400 — a little more than triple the minimum wage — and pay no income tax.  With four children, the tax threshold grows to $66,135 — more than four-and-one-half times the minimum wage.  And with six children, the threshold grows by another leap and a bound — to $86,865, or almost six times the minimum wage for one breadwinner working full time and full year.

Clearly, all of these circumstances are rifle shots.  There could be any number of variations.  In today’s America, perhaps the most important would be if the two spouses both worked.  In that instance, the $86,865-per-year couple with six children could each work at only three times the minimum wage — much more tedious circumstances.  They would be eligible for a tax credit for child-care expenses, but the amount of that credit would be limited to $6,000 per year (a fixed number in the law for two or more children; $3,000 for one child).  That would be $1,000 per child per year, which would not buy very much child care in most of the country.  The numbers for smaller families would be affected similarly.

Conclusion.  So what do we take from all of this, beyond a lot of numbers?

One salient point is that elderly people living off of accumulated assets invested conservatively, or with modest pensions, are very likely to be a significant part of the population that pays no income tax.  They are not eligible for refundable tax credits such as the EITC, however.  If those past savers must invade their principal to meet what they consider to be their minimum living standards, their future incomes will deteriorate, and they therefore may continue to live in the zero tax zone.  If interest rates rebound quickly, or if they manage to adjust their expenses downward to preserve the corpus of their savings, however, they will have taxable incomes in the future.

The situation of younger working families is much more complicated.  Several individually complex tax provisions interact to produce results far different from those of the elderly.  The notion that a family can earn almost $87,000 per year and pay zero income tax might seem offensive.  Add that this family raises six children and there might be a different reaction.  But note in addition that this result comes from no purposeful manipulation, but merely a plain-vanilla tax return that uses provisions intended to deliver relief to all families with children.  The income-tax-threshold numbers for smaller families are significantly lower.

Some might argue that this tax relief for families with numerous children is excessive.  With multiple tax provisions expiring at the end of 2012, and with many elected policymakers clamoring for fundamental tax reform, the opportunity for change may well present itself.  However, recognize just one additional reality:  A tax code with less relief for families with children, including those with many children, would not be dropped into a totally new society, like the first running of a Sims computer game.  It would, rather, be a change in the rules that would apply to real-world families who now are subject to the tax code described above.  That family with six children might have a mortgage (perhaps an underwater mortgage that it continues to pay out of honor) and a car loan to replace the old clunker with a not-quite-so-old clunker.  Those loans must be paid, and a tax code that reduces the relief for families with children would impinge on the income needed to do so.

So that is where the “47 percent” number comes from.  The question of whether those 47 percent are appropriately motivated and energetic can be answered by someone else.

STAY UP TO DATE