Prioritizing the Poor

The philanthropic sector in the U.S. is broken.

According to Jeffery Sach’s estimate in his book, The End of Poverty, just $175 billion dollars annually over the next 20 years, appropriated efficiently, could end global extreme poverty, defined as people living on less than $1.25 a day. The estimate is likely wildly optimistic; nonetheless, it illustrates both the enormous potential of properly applied charitable dollars and the relatively feasible scale of funding needed to fight poverty global poverty.

When Sach’s book was published in 2005, United States philanthropic giving (not including foundations) was a total of $252.2 billion  — a figure which has since increased over 25 percent in less than a decade, to over $316 billion (though lower than the pre-bubble high in 2007 of over $344). Despite the availability of significant funding, the philanthropic sector is structurally inefficient at serving the truly neediest. The scope of the misallocation of charitable dollars owes to three main factors.

The first is rooted in human nature; we are largely driven to give by emotional rather than rational reasons. Donors behave like consumers, rather than investors, meaning that how riveting a charity’s “story” or “brand” is often more important than the evidence of their cost-effectiveness (“return on investment”). This characteristic drastically distorts the efficiency of giving.

Great marketing, not empirical evidence, differentiates non-profits like Charity: Water (3.5 million people served) from Evidence Action’s Safe Water Dispenser project (roughly 1 million people served). Despite its smaller reach, which reflects donor behavior, the later’s intervention is based on better empirical evidence, including multiple Randomized Controlled Trials (RCTs), oversight of academics at the forefront of their fields, and ongoing evaluation of impact on recipients.

Donor issue selection is also often emotionally driven, as in the case of an individual who gives to Lyme disease foundations because they lost somebody dear to the disease, even though most would agree that preventing malaria has the best impact on human welfare on the margin.

The second factor is the lack of information available to both donors and non-profits themselves. It is extremely challenging — and costly — to measure and compare the real-world impact of charities because of unintended consequences and indirect effects. The effects of deworming efforts on long-term education, for example, could very well be better or worse than widely assumed, despite there being multiple RCTs designed to answer this very question.

Sometimes donors lean on false proxies for cost-effectiveness, like  measuring a non-profit’s overhead. Organizations like Charity Navigator are more harmful than helpful in this regard: comparisons of overhead (non-program related operating expenses) distract from the fact that different interventions can be orders of magnitude more impactful than others. In fact, overhead can, depending on context, be the best investment an organization can make, by helping reevaluate programs, and refine best practices in diverse contexts.

The third factor is our misguided charitable tax-deduction policy. Because tax deductions for charitable contributions are tied only to legal classification, rather than impact, tax deductions accrue whether one gives to the Guggenheim or to prevent mother-child transmission of HIV during birth in vulnerable communities. This policy gives legal justification to the ethically flawed rationalization that giving to what we care about matters as much as giving to the neediest. Clearly, the preventing HIV transmission matters more on the margin.

A 2007 study from The Center for Philanthropy at IUPUI found that while households that make less than $100,000 focus around 36 percent of their giving on the needs of the poor, those that make over a million commit less than 22 percent of their giving to such causes, although  the wealthiest (top 10 percent) were the most generous to all causes, donating a greater percentage of their income to charity than any other (there were other divergences too, like the top 10 percent prioritizing the Arts and medical research, while giving a smaller percentage of their donations to religious institutions). These figures reflect substantially different giving priorities between classes, but most relevantly about how most Americans prioritize helping the poor.

In total, only 26 percent of charitable contributions go to the domestic poor, while only 4.6 percent of giving goes to help the poor abroad. In 2005, less than one third of giving in the U.S. directly benefitted poor people, with no reason to believe this trend has reversed from contemporary data (for which I did not have the same detailed analysis).

Here’s the kicker: according to a recent CBO report, of the $39 billion dollars the U.S. spends annually on charitable tax deductions, 80 percent goes in tax relief to the donors who give the least, in proportion to total giving, to the poor: the top 20 percent. Such tax deductions represent only 0.1 percent of after-tax income for families in the middle quintile, but a walloping 1.4 percent of after-tax income amongst the richest 1 percent. Tax deductions accrue disproportionately to people who support the needs of the poor the least, when the policy would be far more efficient if it incentivized intelligent, impactful giving by tying deductions to empirical impact as determined by an unbiased, third party.

The principal problem with giving in the U.S. is not the generosity of Americans, but how that money is distributed. Donors are bad at choosing which charities to give to; charities face distorted incentives to market well, rather than be accountable; and charitable tax deductions incentivize counter-productive behavior with regard to the global poor. These factors contribute to the morally catastrophic distribution of U.S. donations, with those most in need of aid shouldering the burden.