Trust, ability‐to‐pay, and charitable giving

• Author: Paul Missios,Ida Ferrara
• DOI: http://doi.org/10.1111/jpet.12393
• Date: 01 June 2020
• Published date: 01 June 2020

J Public Econ Theory. 2020;22:583–629. wileyonlinelibrary.com/journal/jpet © 2019 Wiley Periodicals, Inc.

|

583

Received: 23 August 2018

|

Accepted: 2 August 2019

DOI: 10.1111/jpet.12393

ORIGINAL ARTICLE

Trust, ability‐to‐pay, and charitable giving

Ida Ferrara

1

|

Paul Missios

2

1

Department of Economics, York

University, Toronto, Ontario, Canada

2

Department of Economics, Ryerson

University, Toronto, Ontario, Canada

Correspondence

Paul Missios, Ryerson University, 350

Victoria Street, Toronto, ON M5B 2K3,

Canada.

Email: pmissios@ryerson.ca

Abstract

In the literature on privately provided public goods,

altruism has been motivated by what contributions can

accomplish (public goods philanthropy), by the pleasure

of giving (warm‐glow philanthropy), or by the desire to

personally make a difference (impact philanthropy).

Underlying these motives is the idea that individuals trust

that their donations reach their goal. We revisit these

models but allow for distrust in the institutional structures

involved. An important result we derive is that trust

considerations determine whether crowding out is less or

more than complete, and we thus open up possibilities in

terms of the extent of crowding out not currently available.

We also model socially motivated philanthropy when

income‐heterogeneous donors take trust and ability‐to‐pay

into account. With ability‐to‐pay in social motivation, an

important result we obtain is that low‐income donors

maycontributemorethanhigh‐income donors (in both

absolute and percentage‐of‐income terms), giving a

potential theoretical foundation to the frequently observed

“U‐shaped”pattern of giving.

1

|

INTRODUCTION

Charitable giving comes from different sources: individuals, foundations, corporations, and

bequests. According to the World Giving Index (WGI) of the Charities Aid Foundation (CAF),

28% of the global population (or 1.4 billion people) is estimated to have donated money in 2013,

21% (or 1.0 billion people) to have volunteered time, and 48% (or 2.3 billion people) to have

helped a stranger (CAF, 2014). At the country level, more than 50% of the populations of the

top 20 income countries are estimated to have donated money in all but three cases. At the

individual giving level, donations by individuals represent 72% of the $335.17 billion collected in

charitable giving in 2013, consistent with the 72% of the $316.23 collected in 2012 and the 73%

of the $298.42 collected in 2011 (Giving USA, 2014). When we examine individual giving as a

percentage of individual income in the United States (Figure 1a) and as a percentage of

household income in Canada (Figure 1b),

1

we find a “U‐shaped”pattern of giving: the share of

0.00%

1.00%

2.00%

3.00%

4.00%

5.00%

6.00%

7.00%

8.00%

9.00%

Giving as a Percentage of Total Income

(U.S.A., 2010/2011/2012)

2012 2011 2010

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%

Under $20,000 $20,000 under

$40,000

$40,000 under

$60,000

$60,000 under

$80,000

$80,000 under

$100,000

$100,000 under

$120,000

$120,000 or more

Giving as a Percentage of Household Income

(Canada, 2007/2010)

2010 2007

(a)

(b)

FIGURE 1 (a) Giving in the United States as a percentage of total income by income level from 2010 to

2012. Data from the 2010, 2011, and 2012 IRS Statistics of Income files, “Individual Complete Report

(Publication 1304),”Table 2.1, available at http://www.irs.gov/uac/SOI‐Tax‐Stats—Individual‐Statistical‐Tables‐

by‐Size‐of‐Adjusted‐Gross‐Income. (b) Giving in Canada as a percentage of household income by income level

for 2007 and 2010. Data from Martin Turcotte (2012), available at http://www.statcan.gc.ca/pub/11‐008‐x/

2012001/article/11637‐eng.htm

1

For the United States, we compu te the shares of individual income de voted to contributions as the ratio be tween average contribution (tot al contributionsof

individuals within each adjus ted gross income bracket divided by th e number of returns) and average total in come (aggregated income of individ uals within

each adjusted gross income bra cket divided by the number of returns). For Canad a, we compute the shares of household income cont ributed to charitable

giving as the ratio between the avera ge annual donation and the mid‐point of the househol d income range. For the lowest income range, we use the up per

bound of the range; for the highe st income range, we use the lower bound of the range.

584

|

FERRARA AND MISSIOS

income donated initially decreases in income but eventually increases, consistent with findings

in Auten, Clotfelter, and Schmalbeck (2000). To date, there are no theoretical models explaining

this U‐shaped pattern of giving but some anecdotal evidence suggests that the distribution of

contributions across different causes may offer a possible explanation.

2

In addition to providing

an explanation for this phenomenon of low‐income individuals giving more than high‐income

individuals as a share of income (ability‐to‐pay based social motivation), we also analyze the

crowding out implications of differences in the trust potential givers hold toward the

institutions providing the charitable good.

Although individuals continue to donate their time and money, a number of high‐profile

scandals involving embezzlement, misuse of donations, slow disbursements of disaster

relief, and generally inefficient operations, such as the United Way in the 1990s, the

American Red Cross Liberty Fund after 9/11, and,morerecently,theBreastCancerSociety

(FTC, 2015), have generated distrust among potential philanthropists. It is not unusual to

come across examples of charities spending only a small fraction of money raised on the

activities they promise. In Light (2008), survey data from 2008 indicate that only 25% of

Americans believe charitable organizations to be “very good”at helping people and 70%

of Americans report charitable organizations to waste “a great deal”or “afairamount”of

money. Accordingly, as public confidence in charities falls, we can expect donations to

decline,

3

but the extent of this decline should ultimately depend on the reasons why

individuals donate in the first place. Experimental evidence that charity quality can

influence contributions is available in Landry, Lange, List, Price, and Rupp (2010).

4

Along

with considering the trustworthiness of potential charities, individuals must also take into

account how governments use tax revenues. Given that tax revenues are to an extent

intended to provide certain charitable goods, and it is further reasonable to expect them to

be targeted toward alleviating poverty and other social ills, we can view taxation as a

substitute for private donations. The corruption or general disinterest by governments to

provide for such goods may lead to changes in private giving: individuals who trust the

government to spend their taxes on social programs may be less willing to give to a charity

than individuals who believe the government is likely to “waste”their taxes. Estimates of

the deadweight losses associated with taxation (e.g., Becker and Mulligan, 2003) suggest

that tax efficiency varies substantially across countries and that the deadweight losses are

significant. At the same time, administrative costs for charities also vary considerably

(for a summary of studies on charity costs across countries, see Burt, 2014).

5

As such,

trust underlies both the act of giving and the potential substitutability of different

modes of giving.

But what drives individuals to donate money in the first place? In the pioneering work of

Bergstrom, Blume, and Varian (1986), individuals give because they value what their

contributions can accomplish (public goods philanthropy). In a model of pure public goods,

2

For example, religious or ganizations tend to receive the great est support through individual contri butions and low‐income people tend to give more to

religious causes (Turc otte, 2012). However, Hoge (1994) and Ros borough (2015) suggest that the core giv ers for congregational giving are rela tivelyhigh

income. Supporting this , in the United States, the U‐shape exists acro ss all states, some of which have signif icantly lower congregational atte ndance than

others. Religious orga nizations may also receive higher dona tions because individuals place highe r trust in these institutions than in priv ate charities.

3

For example, charitable giving to the Catholic Church declined after the U.S. Catholic clergy scandals (Bottan & Perez‐Truglia, 2015).

4

In this paper, we relate giving to trust in institutions (i.e., charity and government). For trust among individuals in a society, see Glaeser, Laibson, Scheinkman,

and Soutter (2000), Alesina and La Ferrara (2002), and Butler, Guiso, and Giuliano (2016).

5

Administrative costs are just one piece of an individual’s trust in a charity, but, other things equal, an individual is more likely to trust a charity with lower

costs than a similar‐goaled charity with high costs.

FERRARA AND MISSIOS

|

585