Ed Dolan's Econ Blog

The Charitable Deduction as a Tax Expenditure: What it Buys and What to Do About It (Part 1)

In size, the tax deduction for charitable contributions ranks sixth on the dirty-dozen list of tax loopholes, far behind deductions for employer-paid health care or home mortgage interest. In popularity, though, it probably ranks first. Perhaps that is because so many people think it really is what it purports to be: a reduction in taxes that encourages charitable giving. Since people like both tax cuts and charity, it is not surprising that they like the charitable deduction. Before we accept that reasoning, however, we need to look more closely at what the charitable deduction really is and what it gives us.

Tax Deduction or Tax Expenditure?

When economists look at deductions, credits, exclusions, and other tax preferences, they do not see reductions in the burden of government. Instead, they see disguised forms of spending that, at best, leave the burden of government unchanged, and more likely, increase it. That is why they prefer the term “tax expenditure.” A simple example will explain why the term is apt.

Suppose the government decides it is time to replace the aging 747 that serves as Air Force One with a shiny new Boeing Dreamliner. The list price, without the upgrades the president will need, is about $200 million. How to pay for it? The Treasury could write a check for $200 million to the Boeing Company, in which case it would appear in the budget as an ordinary line-item expenditure. Instead, it could compensate Boeing with a $200 million tax credit. Boeing wouldn’t care which. (I am assuming here that Boeing actually pays taxes, as it claims to do; a premise that some dispute.) People who worry about the burden of government should not care much, either. Whichever way the government pays for the Dreamliner—writing a check or making a tax expenditure—the purchase places a $200 million burden on the public. The burden must take one of three forms, or a combination of them:

  1. Someone will have to pay $200 million in additional taxes
  2. Some other expenditure will have to be cut by $200 million
  3. The government will have to borrow the funds, which means $200 million in future tax increases or expenditure cuts (unless the Treasury intends to default, a possibility we won’t consider here).

Does this reasoning mean that we should classify all tax cuts as tax expenditures? No. It is true that the government would have to resort to one of the three alternatives listed above if it announced a simple tax cut of $200 million, but there is an important difference. A tax cut unconditionally leaves money in the pockets of citizens that they can use to buy something they themselves want. A tax expenditure imposes a condition that citizens spend the money they would otherwise have paid in taxes on something the government wants. The attached conditions that direct spending into certain areas are what makes the charitable deduction a tax expenditure.

Are Tax Expenditures Ever a Good Idea?

Tax expenditures need not always be a worse way for the government to buy what it wants than direct, on-budget spending. There are at least two reasons they could sometimes be better.

One is the possibility that tax expenditures may allow greater consumer choice. For example, suppose the government, worried about climate change, wants to encourage home energy efficiency. A tax credit for replacement windows, insulation upgrades, or replacing appliances would allow homeowners to choose the kind of investment that gave the greatest payoff for their particular home. For a given impact on the budget, it might well be more efficient than, say, hiring government crews to go from home to home to install insulation upgrades. (Note, however, that a carbon tax or some other form of energy signal would arguably be better than either approach.)

A second possibility is that a tax expenditure might induce people to purchase some public good from private suppliers that operate more efficiently than a government agency. For example, if private health insurance companies operated more efficiently than government programs like Medicare, it could make sense to expand health care coverage using tax credits instead of by broadening Medicare eligibility. (Again the example is hypothetical. In practice, it is far from clear that private insurers are more efficient than Medicare administrators.)

Imperfect though they are, these examples show that we should not rule out tax expenditures as a way to encourage the supply of desirable public goods. However, they also remind us that the real test is whether a given tax expenditure encourages the purchase of goods or services that are really worth their cost in terms of public funds. If they are not, then considerations of the relative efficiency of tax expenditures and direct expenditures are beside the point.

What Does the Charitable Tax Expenditure Buy?

The question at hand, then, is, What does the charitable deduction really buy? Much less by way of charity than is often thought, at least if we understand charity in the ordinary way.

Charity, to cite a few random  on-line dictionary  definitions, means “generosity and helpfulness especially toward the needy or suffering,” “provision of help or relief to the poor,” or “generous actions or donations to aid the poor, ill, or helpless.” How do these definitions compare with the actual patterns of giving under the so-called charitable deduction?

The Giving USA Foundation provides an annual tally of tax-qualified giving. Using IRS data, the foundation estimated 2010 giving at $291 billion. This chart shows a breakdown by category of recipient:

The next question is, What percentage of tax-qualified giving  can we reasonably call “charitable,” using the ordinary meaning of the term? Clearly not all of it.

Let’s begin with the largest category, religious giving. Many people think of support for churches as the essence of charity, and spokespeople for the sector do their best to reinforce that view. For example, writing on the website, Chuck Colson tells us that “charities, especially religious charities, provide services that the government won’t or can’t do efficiently.   . . Look just at what Catholic Charities does for AIDS victims, or for the savings religious schools provide municipalities. Or the services provided by homeless shelters, prison ministries (like Prison Fellowship), medical research, and on and on.”

If we look past the rhetoric at the actual numbers, though, we get a different picture. The next chart shows some useful data from a survey of church budget priorities provided by Christianity Today.

What we see is that churches spend the most of their budgets on staff and facilities that have worship as their main purpose. Worship is all well and good, but it is not charity. In saying that, we do not have to adopt the cynical attitude of a Samuel Butler, who, in his utopian novel  Erewhon, lampooned churches as “musical banks,” where parishioners made deposits but never withdrawals. Even if we accept Christian or other theologies at face value, the purpose of worship is to put the participant at rights with God; it is in that sense an essentially self-regarding rather than a charitable activity.

The truly charitable activities in the Christianity Today breakdown fall under the categories of ministries and missions. Those comprise everything from prison ministries, to soup kitchens, to aid for hospitals in Africa. What we conclude is that charitable activities of religious organizations make up not 35 percent of all tax-qualified giving, but rather, something like 17 percent of 35 percent, or about 7 percent of the total.

Sufficiently diligent research could no doubt track down similar data for the charitable and non-charitable components of other giving categories, but I will, instead, resort to  guesswork. The next largest giving category is education. Reasonably construed, the charitable component of giving to education would be the part that provides scholarships for needy students and pupils. Giving money to send your local school’s volleyball team to the state tournament wouldn’t count. Giving money to build a new science building at your alma mater would count only to the extent that students on needs-based scholarship used it, and so on. I am guessing that 20 percent of educational giving might qualify as charitable.

I have equal reservations about the activities of foundations. Foundations support many causes, not all of which are charitable. Not only do they provide many services to people who are not needy; everyone knows that the not-for-profit world is full or organizations that devote 100 percent of their energies to purely political activities. The IRS is notoriously permissive in accepting the slightest fig leaf of educational or scientific purpose to allow partisan organizations to qualify for tax-exempt status.

Much the same reasoning applies to giving for health care. Some of it, certainly, goes to support provision of health care services for the medically indigent, but not all of it. For example, donor-supported medical research results in new treatments that benefit insured as well as uninsured patients. It would be a stretch to suppose that half of this category is truly charitable.

Giving to the arts is very nice, of course, but unless it supports free opera tickets for inner city kids, it is not charity. Much of it caters the personal tastes of wealthy donors, not to the poor and needy.

Suppose we give a pass to donor organizations like United Way, and consider all of their activities to be charitable. Suppose we allow the welfare of animals and endangered plants equal charitable status with the welfare of people. No matter what allowances we make, it is hard to construe much more than a third of tax-qualified giving as truly charitable. Even if you don’t like my guesswork, I challenge you to get that number up to half.

Does the Charitable Deduction Increase Giving?

If the charitable deduction dramatically increased total giving, it might be possible to justify its budgetary cost of $50 to $60 billion per year, even if only a third of the giving were actually charitable in nature. For example, if the charitable tax expenditure generated $3 in new giving for each $1 of cost to the budget, its “charity efficiency” would rise to 100 percent. However, the evidence that the deduction has any effect at all on giving is surprisingly weak.

William C. Randolf of the Treasury department cites several econometric studies in an entry written for the NTA Enclyclopedia of Taxation and Tax Policy. (The entry is available on line here, courtesy of the Tax Policy Center.) He concludes that there is some evidence of a positive effect on giving, but warns that the most recent evidence is not very robust, and offers his opinion that the effect is often overstated.

Neither does historical evidence support the notion that the deduction has much effect on total giving. Writing recently for Nonprofit Quarterly, Jack Shakely examines several episodes in which tax law changes reduced the value of the deduction. One of the first was the 1944 decision to introduce a standard deduction, thereby making itemization unattractive to middle- and lower-income taxpayers. Charities warned that the sky would fall, but the change turned out to have little impact.

The same pattern has repeated itself with every reduction of tax rates for high earners. For example, the Economic Recovery Tax Act of 1981 cut the top rate from 70 percent to 50 percent, almost doubling the out-of-pocket cost of charitable donations to top-bracket taxpayers. Despite forecasts of doom from the nonprofit sector, there was little impact on total donations.

Certainly there is nothing in the empirical evidence to support the idea that a reduction in tax benefits cuts contributions on a dollar-for-dollar basis, as Martin Feldstein claimed in a 2009 editorial for the New York Times. The idea of a “charitable multiplier” greater than one is even farther from reality. Instead, charitable giving has varied surprisingly little over the years, even as top tax rates have risen and fallen dramatically.

The Bottom Line

The bottom line is that the popularity of the charitable deduction rests on a set of false premises. In reality, the deduction is best viewed not as a reduction in taxation but as an increase in expenditures. Surprisingly little of the giving that qualifies for the deduction goes to truly charitable purposes. And warnings that the nonprofit sector would face collapse without the charitable deduction are greatly exaggerated, if not altogether baseless.

My personal conclusion is that Congress should simply abolish the charitable deduction along with the host of other tax loopholes that permeate our rotten tax code. However, that may be too radical for many. For those who cling to the idea that the tax code should in some way encourage charitable giving, I will examine proposals for reform in Part 2 of this post.

Follow this link for Part 2 of this post



13 Responses to “The Charitable Deduction as a Tax Expenditure: What it Buys and What to Do About It (Part 1)”

Ed Dolan EdDolanApril 20th, 2012 at 10:36 am

Thanks, Luis. The research you cite is highly relevant to this post; I recommend that everyone read it.

The research in question, by Sarah Smith of Bristol U., reports that in the British case, each $1 lost to the Treasury from the charitable deduction produces $1.16 in added giving from high earning donors, a number she calls a elasticity of -1.16. On the face of it, an elasticity with absolute value greater than 1 means the government is getting more "bang for its buck" by using the tax expenditure route rather than the direct spending route to support charity.

However, this result needs to be qualified by noting that not all tax-induced donations serve the intended public purpose. As I argue in my post, in the US case, only about 1/3 of tax-qualified donations go to actual charity. If that is true, direct public aid to the poor and needy would produce more charitable benefit per dollar cost in public funds than the tax expenditure. To put it another way, the threshold value for the elasticity is not 1, but 1 divided by the share of tax-qualified donations that hit the intended policy target.

SharonApril 20th, 2012 at 10:33 am

Enlightening. Interestingly, I have never once have I seen my taxes reduced by charitable donations. I do my taxes on Turbotax and I cannot recall a single instance of the amount that I owe being reduced by even a dollar. So, only bigger fish are reaping this reward

Victor MathesonApril 21st, 2012 at 9:09 am

Nice post, Ed. I would comment, however, that I think your definition of what government is trying to encourage through the charitable deduction is too narrow. If we expand the definition to include encouragement of public good provision, then a much higher percentage of contributions are appropriate.

For example, contributions to the World Wildlife Fund don't fit the narrow (and probably technnically correct) definition of "charitable" because the poor don't disproportionately value the environment. But environmental conservation is a classic example of a public good that government has a strong motivation for encouraging since public goods are likely to be undersupplied by a free market. Same reasoning applies to the American Cancer Fund or, arguably, education donations.

While the deduction is called the "charitable donation" deduction it should be more accurately referred to as the "elimination of market failures and undesirable market outcomes" deduction. That just doesn't fit on the Schedule B.

Ed Dolan EdDolanApril 21st, 2012 at 10:30 am

Thanks, Victor. You make a good point. Here is how I would respond:

First, there's a lot in the name when it comes to political support. For example, one could argue that culture is a public good, and many European governments generously support opera houses and museums on that theory. However, in the United States, public support for the arts is very low. In fact, polls on budget cuts seem to put NPR and the National Endowment for the Arts, tiny as they are, on the "cut first list." Same goes for some items in the international component of the charitable deduction, which would never pass the popularity test as a line item for foreign aid. In short, there are some public goods that seem to coat-tail on the "charity" label even though they are not charity, strictly speaking, and would be very unpopular if put up for a Congressional vote on their own.

Second, I do agree with you about environment and wildlife, so I included that as 100%
charity (broadly speaking), which is what I meant with the line about animals and endangered plants. It's only 2 percent of the total, though.

Third, with regard to health and education, I agree that these can be classified as "mixed" public goods that provide some private and some public benefits. However, I would also be narrow about defining a public good. The classic definition of a public good is a market failure that arises from the joint properties of nonexclusion (if you provide it to one person, everyone gets the benefit) and nonexhaustion (more people receiving the benefits does not use it up).

Most health and education spending really doesn't fit that pattern. You can exclude people from hospitals, doctors offices, and college campuses. They also are exhaustible; to serve more patients or students, you need more classrooms and more doctors.

Yes, it can be argued that there are true public good components of both. For health care, suppression of communicable diseases is a public good, and probably some very basic kinds of medical research that can't be recaptured downmarket. However, a heart transplant is not a public good, nor is research that produces products that can later be sold to cover the cost. Similarly, education in basic citizenship and literacy is arguably a public good, but teaching someone to be an accountant or engineer is not, because the student recaptures the cost by charging for services after graduation.

In getting to my 1/3 figure for charity, I allowed 20% of education giving and 50% of health care giving. I stand by my guesswork, although I agree it would be interesting to see some actual research on how the money is spent.

llisa2u2April 21st, 2012 at 10:22 am

I thought your article explained tax expenditure vs. tax deduction very well. Using charitable donation as an example was an interesting comparison tool. How much do different people give to one another? Who gives time and who gives money? To and for what? Here is a link:

For anyone who is truly interested in reading and comparing about "giving" and "need" issues and comparisons, and then real issues about "taxes". The report is primarily a report from the Center for Philanthropy from Indiana University. It was prepared for GOOGLE. It is a very good report primarily because the executive summary is direct and gives a reasonable summary of complex data analysis. Here are the most important 2 sentences from that summary:

"Of the $250 billion in donations, less than $78 billion explicitly targeted those in need.
Only 8 percent of households’ donated dollars were reported as contributions to help meet basic needs–providing food, shelter, or other necessities."

The second most important fact is the graph on page 14, in, in the complete report. In that graph what is important to focus on is the lime green bars, in each of the columns that make up the graph. These colored bars graphically designate the percentage of giving that actually is given with the intent of providing "basic needs" to people that may need that kind of financial support. The percentages are absolutely minimal for all demographic groups and income level groups. It is the most minimal comparative contribution percentage for the income level of households with an annual income of over $1,000,000.

Now, it is also important to recognize that my above statements are very selective and only refer to "basic needs giving". That is only one category of a number of philanthropy categories.

Read the report, make your own conclusions. I have the opinion that the "state" very importantly assumes "basic needs giving" issues when any state establishes governmental financial policies that supports "basic needs" of the majority of it's citizens. When citizens do not acknowledge the facts, it is very easy to confuse issues of "taxation" and exactly what taxes can do to more effectively support the goals of the state as desired by good and reasonable leaders, as the most effective members for all other members of any state.

So, in other words, Don't kid yourself or anybody else about who is willing to do what for
anybody if you really don't need to. I include myself in that little remark. I also think that is why any government needs to spend certain monies for certain things rather than not spending any monies for certain other things.

Earl GrinolsApril 24th, 2012 at 2:02 pm

It is possible to reach any conclusion if one is allowed to start from a false premise. Your premise is that the uses to which charitable giving is put do not qualify as true charity. Even if readers grant your narrow definition, why should we necessarily come to your conclusion that deductible gifts for the broader list of purposes is socially undesirable? A person giving to, say, the opera does not have the money for his or her own use and so it is not income to that giver in the normal meaning of the word income.

Perhaps your real complaint is that you object to the legal existence of untaxed entities such as not-for-profits.

Ed Dolan EdDolanApril 24th, 2012 at 5:02 pm

Earl– I take your point that the more broadly you define public purpose, the stronger the argument for the charitable deduction. It is the same point Victor Matheson makes in his comment above, so please begin by reading my reply to Victor. I would add that it is going a bit far to call my definition of charity a "false premise." We may have a disagreement as to whether subsidies to the opera are a worthy pubic purpose or not, but it is hard for me to see that one of these views is "true" and the other "false."

Next, I would say that even if 100 percent of tax-qualified giving went to legitimate public purposes, that would not be a conclusive argument in favor of the tax deduction. As I will show in Part 2 of this post, the loss of revenue to the government can still be larger than the gain in giving, so it might make more sense for the government to spend the money directly.

Third, you are right that my real objection is to the tax-exempt status of nonprofits. I don't object to the existence of such organizations. I spent years of my life as head of a nonprofit organization. It is a strength of a free-market economy that it provides a framework for schools, hospitals, opera houses, etc., to exist and operate for motives other than profit–motives that may be narrowly charitable or, instead, aesthetic, cultural, religious or whatever. However, if nonprofit organizations do not serve a public purpose, they should get by on nondeductible private contributions, and if they do serve a public purpose, they should seek public aid through a transparent, competitive grant-giving process, not through a tax deduction.

SacrificeApril 25th, 2012 at 12:49 am

Clear & Cogent.

The problem is that 83% of Americans self-identify as 'religious'. 91% of Americans believe in God, and 85% believe in heaven.

Good luck with this track you are on.

raclarkinApril 25th, 2012 at 10:45 am

The problem might be in the paragraph (When citizens do not acknowledge the facts, it is very easy to confuse issues of "taxation" and exactly what taxes can do to more effectively support the goals of the state as desired by good and reasonable leaders, as the most effective members for all other members of any state.) Recent polls do not show and confidence in the "good and reasonable leaders etc."

Jolly12December 11th, 2012 at 10:21 pm

My own experience over the last years as a self-supported volunteer engineer overseas testifies to your conclusions. The US charitable organization tax exemptions have produced a multitude of what can only be described as leech organizations that suck out life and provide little or no benefit; in many cases actually causing more harm than good to the charity they profess to help. There are arguably many valid and valuable charitable organizations that do indeed serve society but I believe those would function equally well or better without a tax-exempted status. I would strongly support a political effort to rescind the charitable tax exemptions in the US tax code. Cutting the exemption would be healthy for our country and be akin to pruning the vine to improve the grape- kill the diseased and corrupted for the benefit of the whole. I appreciate your article Mr. Dolan and my question to you is if you are aware of any serious political efforts underway to attempt such a tax code revision.

Charitable GivingJanuary 29th, 2013 at 9:17 am

Very well explained. However, like you said, getting the government to throw out the tax deduction would be extremely radical and therefore I could really never believe that would ever happen.

TracySchramJune 24th, 2014 at 2:16 pm

Points regarding charitable deductions and related tax deduction are well highlighted here. Tax deductions are no way going to be taken off, since it will definitely going to effect a lot of people. Most of the consist of people having a average income with which they tries to manage their family. Tax deductions are the ways through which they save some dollars from the taxable amount. People are considering smarter was of investing in order to do that.