BHI Policy Study Executive Summary

Giving Credit Where
Credit is Due:
A New Approach to Welfare Funding

December 1995

Executive Summary

The emergence of welfare reform as one of the leading issues of the day symptomizes the failure of the welfare state that emerged out of the Great Society programs of the sixties. In light of the growing unpopularity of those programs, welfare reform is inevitable.

However, welfare reform of the kind being considered or implemented in Washington, DC and in various states will not solve the problem. As long as the existing welfare-delivery mechanism remains in place, reforms such as limiting benefits or mandating school or work by welfare recipients will do little to reverse the trend of the last 30 years toward the creation of a large, costly dependent class.

The reason lies in the method by which the government funds welfare programs. Under the existing system, the taxpayer (who is a potential donor to charitable, nonprofit organizations) sends his money to Washington, DC, where part flows into the welfare bureaucracy, and where a fraction of that part eventually flows into the pockets of welfare recipients.

In this process, the taxpayer/donor has no incentive to take an interest in how his money is spent. First, he has no clear idea (indeed, it turns out that no one has a clear idea) of how much of his tax dollar finally makes it to the welfare recipient. Second, he has no control over the disposition of that tax dollar or, therefore, over the behavior of the welfare recipient.

When the citizen casts a vote for president or congressman, he does not have the slightest hope that, in doing so, he can directly allocate some fraction of his tax dollar to those welfare recipients who are deserving and worthy, and away from those who are not. Legislators and faceless bureaucrats make those decisions for him.

As long as welfare is thus taxpayer funded but bureaucrat controlled, the seeds of poverty are left freely to sprout new generations of poor. The only solution is to make welfare taxpayer funded and taxpayer controlled.

The federal income tax provides a mechanism for effecting this solution. In order for taxpayers to fund a system of social welfare, it is not necessary for their tax payments to flow into government offices. The same funds can flow, instead, into private charitable organizations, which can help recipients lead productive, independent lives while offering a cost saving to the taxpayer.

A Tax Credit for Charitable Contributions

This study proposes a federal tax credit for charitable contributions. Under the proposal formulated here, taxpayers will be able to deduct 100% of their contributions to eligible charities to up to 25% of their federal tax liability. The creation of a tax credit effectively gives the taxpayer/donor a choice between sending his welfare dollar to Washington, DC or to the charity of his choice. It also reduces federal spending and brings spending decisions closer to the taxpayer/donor.

By giving the taxpayer control over where his dollar is spent, the tax credit creates an incentive for the taxpayer to take an interest in how that dollar is spent, to monitor the work of the receiving organization and, perhaps, to spend time volunteering. It correspondingly creates an incentive for eligible charitable organizations to vie for contributions from donors and to do so by offering superior services to their clients.

The proposed tax credit will serve as a catalyst for the creation of thousands of new humanitarian nonprofit organizations, some of which will be created by people leaving government who hope to use their entrepreneurial skills to help make a difference. This group of new organizations will respond to the increased opportunities presented by the tax credit to apply private initiatives to the solution of social problems. Since competition leads to economic efficiency, we expect these organizations to be more innovative and successful than government in providing social services.

Research Questions

The tax-credit proposal offered here raises four types of research questions:

1. How, in light of history and evidence, are taxpayers/donors likely to respond to the credit? Will they give and volunteer more, or not?

2. How, given the response by taxpayers/donors, should the tax credit be structured? What limits should be placed on the credit to assure (1) the appropriate flow of welfare services and (2) "revenue-spending neutrality" (equality between the money lost by government minus governmental overhead and the money directed to eligible charities)?

3. How will the tax credit work in practice? Will taxpayers/donors change their giving as predicted? Will new donors come forward once their tax incentive is increased? Will donors volunteer and take an interest in how their money is spent? How will new charitable organizations arise? What services will they provide? How will charitable organizations and their clients adapt? What federal agency will perform oversight of eligible charitable organizations?

4. How will charitable organizations measure success in shifting from public to private provision of welfare? Will it be in terms of their ability to provide services efficiently with low overhead or in terms of measurable success in helping the poor: a reduced incidence of teenage pregnancy, increased success in job placement, and so forth?

For Andy Shemeline, it was a chunk of iron that inspired him to move to Idaho and raise cattle to feed hungry people. While laboring as a construction worker in Chicago, he sustained a head injury after a crane operator dropped some iron on him. The ensuing epiphany moved Shemeline, in his own words, to "do something for somebody else."

Subsequently, Shemeline started HAPPY (Help American People Prosper Yearly), a small cattle farm to which he devotes most of his money and energy. The 64-year-old rancher struggles to keep his ranch solvent while donating its output of hamburger, roasts, and steaks to area food banks. He provides the brunt of the money needed to maintain his operation.

Although Shemeline works 14-hour days, escalating costs, rent, and scarce water threaten his endeavor. While he wants to continue, he acknowledges the factors that jeopardize the ranch and the possibility of having to shut it down. An infusion of donations, however, would enable him to continue helping feeding the hungry in the Pacific Northwest.

This study addresses questions 1 and 2. The only satisfactory way of addressing questions 3 and 4 is through a model program that provides an experimental field test of the tax credit idea. We suggest a matching-gift experiment that would provide a test of this kind.

Welfare in Theory and Practice

Economic theory recognizes an argument for classifying welfare as a "public good." This argument implies that, because the provision of welfare confers benefits on taxpayers as well as recipients, the welfare system is an appropriate object of taxpayer support. But this argument implies only that taxpayers should support welfare, not that they should cede the provision of welfare to government bureaucrats. Welfare can be taxpayer controlled as well as taxpayer funded.

The history of U.S. welfare reflects a tension between two views regarding the responsibility for providing welfare, one that puts this responsibility on the individual and another that puts it on government. In early times, the first view dominated. In current times, the second view dominates and will continue to dominate under the welfare reform now under consideration.

Although the poverty rate declined from 1959 to 1994, there are still over 38 million Americans living in poverty. Poverty affects women and children more than other groups. The poverty rate was 14.5% for the total population and 54% for single, female heads-of-household in 1994. The poverty rate for children rose from 14.9% in 1970 to 21.8% in 1994. The percentage of households headed by single females rose from 9.8% in 1968 to 21.7% in 1993. The percentage of families headed by never-married individuals rose from .7% in 1968 to 8.1% in 1993.

These statistics stand out against a backdrop of a vastly expanded U.S. welfare system. In 1990, 11.5 million people received benefits under the AFDC program. Government spending on direct public aid for AFDC, food stamps, housing, and so forth increased tenfold from $15 billion to $142 billion from 1970 to 1990. Yet, only 41% of all poverty-level families receive any government benefit at all. Only 23% of all poverty families live in public housing or receive housing benefits, and almost half of those receiving benefits are not poor. In other words, government spending is tremendously inefficient. More than 2/3 of all federal welfare spending ultimately ends up in the pockets of people who are not poor.

Private Charity in Perspective

Individuals gave $102 billion to private charity and $17 billion to nonprofit human-services and public-benefit organizations in 1992. Between 38 million and 98 million people volunteered (the number depends upon how one defines volunteering). Individuals at opposite ends of the income distribution gave higher proportions of their income to charity. Religious organizations received close to two-thirds of all contributions. Social services received 10.5%, while public benefit organizations received 6%.

There are 85,800 social and legal services nonprofit organizations that provide individual and family services, job training and vocational rehabilitation, residential care, day care, and legal aid to the poor. Social and legal services nonprofit organizations receive 42% of their revenue from government, 35% from private giving, and 23% from fees for services. The evidence suggests that the 42% provided by government can be more efficiently donated directly by private givers to the nonprofit sector. This removes a layer of government bureaucracy and increases the value of contributions.

What Does the Literature Say?

We reviewed the literature to see what light it sheds on the questions of (1) why people donate time and money for the betterment of others, (2) how the amount of giving relates to the price of giving, as determined by the tax deductibility of giving, (3) how tax policy affects voluntarism and (4) whether government spending crowds out private giving. The literature points toward a strong positive relationship between tax incentives and the willingness of people to give and to volunteer -- an important point if we are to shift the welfare system from the public to the nonprofit sector.

The literature also suggests that the standard economic model, in which individuals act only out of self interest, can be generalized to incorporate altruism. The generalized model accommodates both pure altruism and a more paternalistic form of altruism.

Of 27 studies that we reviewed, all 27 found that giving increased when the net price of giving fell. Seventeen of these studies found that the estimated coefficient on giving was greater than one, implying that increased tax incentives yield more in additional giving than they cost in lost revenue. Government can both increase the amount of money going to the poor and reduce tax burdens by making the tax incentives for giving more generous.

A related issue is whether giving and volunteering are complementary activities, that is, whether the individual who gives more to charity is more likely to take a personal interest in the charity and to help by volunteering. Some (but not all) of the evidence supports this expectation. A recent Gallup poll found that 29% (some 70 million) of all Americans volunteered their services to their communities. We expect the creation of a tax credit to add many more people to this pool of volunteers.

There is the expectation, finally, that government spending "crowds out" private giving. Thirteen studies found estimated crowd-out to range from - .5% to - 35%, that is, a one-dollar increase in government spending reduces private giving by between .5 and 35 cents. Of nearly 50 published papers, 40 found that private enterprise is unequivocally more efficient than government at delivering services.

In sum, the literature is instructive on the subject of the feasibility of a tax credit for charitable contributions:

1. Both theory and evidence strongly support the idea that tax incentives yield a net gain to society: The increased giving to charity exceeds the revenue lost to government.

2. There is evidence, in addition, that, by encouraging giving, tax incentives encourage volunteering.

3. The private sector can provide services more efficiently than the public sector.

4. Government spending crowds out private giving.

These findings suggest that tax incentives aimed at transferring the provision of welfare services to private charitable organizations represent a cost-effective vehicle for welfare reform. Because, in offering them, government produces more in giving than it loses in revenue, tax incentives can yield a bonus in the form of increased welfare services to the poor.

Insofar as the same tax incentives indirectly encourage taxpayers to volunteer, they also offer a vehicle for lowering the costs of delivering welfare services. Finally, the tax revenue lost by offering a tax incentive is recovered in part by the reduced crowd-out that results.

Dr. Peggy Brown runs The Mandela Town Hall Health Spot and Youth Program, in Lower Roxbury, Massachusetts, an economically disadvantaged section of Boston. Young people congregate in her facility for socializing and a snack at the "feasting table." In winter, she distributes warm clothing.

But Dr. Brown does not give handouts. "It's simple," she says. "To be successful today, we must work hard, have self-discipline and learn marketable skills. My young people will go to college, if that's what they want, because they will make it happen for themselves."

Dr. Brown coaches the Mandela Crew Team. This group of young people races teams from throughout the area. "If these young people can be motivated enough to be on the Charles River in Boston at 6:00 a.m. in November, then they can be motivated enough to accomplish anything. That's a metaphor for life."

Structure of the Credit

Yet, in order to avoid "underpricing" our proposal, we did not count any of the expected cost saving in structuring the tax credit. As mentioned, we aimed, in designing the tax credit, to achieve revenue-spending neutrality. In order to make the credit neutral in this sense, it is necessary (1) to identify the kinds of government services that will be shifted to private charitable organizations, (2) to identify the amount that is spent on those programs under current law and how much of that amount is caught - wasted - by the bureaucratic "filter" and (3) to calculate the upper limit on the tax liability that must, in the light of (1) and (2), be imposed in order to bring about revenue-spending neutrality.

Following these guidelines, we determined that only qualified 501(c)(3) organizations that provide services to the poor will be eligible to receive contributions. Assistance to the poor means assistance for basic human needs. This includes food, shelter, clothing, housing and job training. Organizations providing medical care and intermediaries like the United Way will not be eligible for the credit. No more than 30% of the organization's annual expenditures can be for expenses other than exempt-purpose expenses, such as administration and fundraising.

The proposed credit will be limited to contributions of cash by individuals only. Pass-through entities such as partnerships and S corporations will be permitted to take tax credits because individuals are ultimately responsible for tax and income originating from these sources. Nonitemizers will be eligible for the credit. This removes an inequity in the current law, spurring greater interest in volunteering and contributing on the part of these taxpayers. Contributions eligible for the credit will be allowed for contributions made during the year, and up until the date the return is filed.

The credit will have different effects on different states and regions owing to the expected inclination of taxpayers to give locally. The resulting disparities show that welfare recipients stand to gain in some states and to lose in others. Migration and state-government initiatives will mitigate the losses.

A substantial amount of taxpayer money flowing into the welfare system gets caught in an administrative filter. The amount caught in this filter is estimated at $17 billion, or 15% of the total eligible welfare expenditures. After adjusting for the savings from elimination of the administrative filter, the disallowed tax deduction for credit contributions, and nonprofit organization and governmental administrative overhead, the total tax credit allowed will be $115 billion to be revenue neutral. Based on a static model of giving, the maximum percentage of a taxpayer's total tax that will be eligible for the credit is 25%.

The Buttenweiser family of Belmont, Massachusetts founded the Family-to-Family Project, the purpose of which is to help homeless families navigate the various obstacles that often frustrate their acquisition of permanent housing. The philosophy of the Project requires participating families to assume an active role in this process. While Project members will usually limit direct financial assistance to $4,000, Ann Marie Healey, Executive Director of the Project, avers that "most families don't even need that much."

According to Healey, minimal assistance in the form of a refrigerator or seed money for rent may suffice to provide homeless families with housing. With its emphasis on family involvement and staff assistance, the Project incurs very little in administrative costs.


We developed an econometric model of charitable giving and estimated it using a 1991 data set comprised of information from individual taxpayer returns. We investigated several possible scenarios, using tax credits of different sizes.

We found that charitable giving is related to the price of giving, as measured by the after-tax cost of a dollar's worth of donations. A 10% decrease in the price of giving increases charitable giving by 11.1%. Charitable giving is related to income. We estimate that an increase in income of 10% increases giving by 3.5%, implying that giving is a "normal good."

Taxpayers who itemized in 1991 made $63 billion in charitable contributions. Nonitemizers made approximately $39 billion for a total of $102 billion. The average amount of giving for taxpayers with adjusted gross incomes over $200,000 was $14,040. This comprised 2.75% of income. The average amount of giving for taxpayers with adjusted gross income below $15,000 was $399. This comprised 5.99% of income.

Giving for programs eligible for the tax credit totaled $17 billion in 1991. Of this amount, an estimated $11.2 billion was given by itemizers and $5.8 billion by nonitemizers. We estimated the levels of giving under several different tax credit scenarios and, on the basis of our findings, recommend that taxpayers be allowed a credit of 100 percent for eligible charitable contributions up to 25 percent of total personal tax liability. Under this scenario, taxpayers will contribute more than enough money to transfer welfare responsibility from the public sector to private nonprofit organizations.

Combined with a deduction for contributions above the limit, the law will achieve tax-and-spending neutrality where the amount of the lost revenues is compensated for by reduced government spending on welfare programs.


Because volunteers represent an important method by which charitable organizations are able to deliver services at low cost, we investigated the effects of the tax credit on donors' and other individuals' willingness to volunteer.

We found that volunteer labor is most important to religious, health, and education organizations. Volunteer labor was twice as large as paid labor for religious and cultural organizations. Most volunteers performed clerical, manual, leadership, or fundraising roles. People who volunteer their labor gave more in charitable contributions.

We estimate that approximately 3.76 million volunteers donate time today to organizations that would be eligible for the tax credit. We estimate that enactment of a 100% tax credit will produce an additional 2.06 million volunteers. We base our estimates on research that shows volunteering and giving to be "complementary goods."

Here too our estimate does not take into account the expected growth in the nonprofit sector. It is quite possible that this growth will induce another 6 million people to become volunteers.

Thanksgiving in America means sharing. Families congregate, friends meet, coworkers exchange good tidings, and community organizations plan special events and meals to help the less fortunate. Government, too, helps by distributing food to these community organizations. In 1995, however, for seven days, including some of the week before Thanksgiving, all nonessential federal government services stopped because President Clinton and Congress could not agree on a budget plan. One program deemed nonessential was the distribution of turkeys by the Departments of Agriculture and Education.

With no budget agreement in sight, many less fortunate individuals faced a turkey-less Thanksgiving. The County Food Bank in Worcester, Massachusetts immediately sent out a plea for help since they could no longer rely on the federal government. The response was overwhelming. Small and large businesses agreed to donate large numbers of turkeys. Unions donated labor and worked phone banks while individuals chipped in with their time and brought in single turkeys. A nine-year-old girl donated $30 of her allowance. According to Janet Ward, director of the food bank, "This is the greatest thing that ever happened to the food bank. We've had countless people saying I've got a turkey. Where and when can I deliver it?"

Eventually, the President and Congress agreed to a stop-gap measure that allowed the federal government to resume its operations, including turkey distribution. Yet for a short time the people of Worcester County had a chance to show their community spirit, to show what they could accomplish if the need arose, to show what could happen if the federal government returned some of its functions back to the communities. The spirit of voluntarism soared. Community action, powered by the feeling that comes from helping, ensured that the needy would not be forgotten.

The Nonprofit Sector

We assessed the relative efficiency of nonprofit organizations, particularly schools. There is strong evidence that private schools are able to offer a superior education for less money.

While public elementary schools in Texas spend $3,686 for each pupil, Catholic private elementary schools spend a mere $2,137. At the high-school level, Texas spends $4,282 for each public-school pupil while Catholic high schools spend only $3,666. Boston and New York spend over $7,000 for each public-school student, most of whom achieve poor results on standardized tests. By comparison, the Marcus Garvey School, a private school in South Central Los Angeles, teaches sixth-graders calculus while spending $3,600 for each student.

We also found that nonprofit organizations rely more on volunteers than either government agencies or private sector businesses. Salaries and wages dominate the expense reports of all sizes of nonprofit organizations and comprise the largest class of expenditure. Yet spending is lower for this category than comparable private and public sector industries because of the use of volunteers. Compensation of employees dominates the uses of resources as a percentage of total resources used, no matter what the organizational form.

Larger nonprofit organizations, measured by aggregate contributions received, pay a higher proportion of their expenses as salaries and wages, implying a more corporate, more formal atmosphere with less reliance on volunteers. Larger nonprofit organizations give more money out as grants and allocations. Smaller nonprofit organizations give out more assistance to specific individuals. This suggests smaller, more entrepreneurial organizations will spring up to meet the needs of the displaced welfare state.

Examining input-output tables, we found that compensation of employees and real estate are the two biggest resource uses for most areas of production or service. Social-services industries use more resources for advertising, as evidenced by higher percentages of spending (between 1% and 5% of total spending) on this item than comparable groups. This implies that these industries are willing to perform outreach to raise funds. Oversight by donors and the need to be effective in providing welfare services will deter eligible organizations from overspending on fundraising.


We include two appendices that provide ancillary information. These include: (1) an overview of the major means-tested welfare programs in existence today and (2) a sample of existing nonprofit organizations, some eligible and some ineligible.