Diversity and Charity: An Inverse Relationship?

Our latest Freakonomics Radio on Marketplace podcast, “What Makes Donors Donate?” looks at what works (and what doesn’t) to incentivize people to give. A new NBER working paper studies the relationship between religious and ethnic diversity and charitable donations by looking at Canadian census data and tax records. Authors James Andereoni, Abigail Payne, Justin D. Smith and David Karp argue that the two are inversely related, that is to say that the more diverse a neighborhood, the lower its charitable donations. From the abstract:

A 10 percentage point increase in ethnic diversity reduces donations by 14%, and a 10 percentage point increase in religious diversity reduces donations by 10%. The ethnic diversity effect is driven by a within-group disposition among non-minorities, and is most evident in high income, but low education areas.  The religious diversity effect is driven by a within-group disposition among Catholics, and is concentrated in high income and high education areas.

In terms of ethnic diversity, the same number of households make a donation, but they make a smaller donation.  The authors write:

This effect is mainly driven by non-minorities, who contribute $9 less for each percentage point their group share drops, and Blacks, who contribute $39 less.  The average effect of ethnic diversity on donations occurs most strongly in high income, and in low education neighborhoods.

Our first instinct is to cry causation/correlation foul. But does anyone have thoughts as to why (if this is in fact a causal relationship) diversity might have an adverse impact on charity donations?  


Seems pretty simple to me. People are more likely to give when the feel empathy towards the group. They are more likely to feel empathy with people who look like them and share their culture and beliefs.

Joe in Jersey

The US has long been the most charitable (in terms of donations) and also the most ethnically diverse country on the planet. Is there anyway to track charitable giving in the US over time? Has it gone up or down?

It appears to me that charitable giving goes up and down with the economy more than anything else (purely an observation no data to back it up). Big disasters (tsunami, earthquakes, etc.) also spur giving. It would be interesting to see what the impact has been over time in the US.


Nowhere near the most charitable.


My first thought is one of exposure. If I am a non-minority some of what I give to charities may focus on helping disadvantages minorities. If I now live among many, I may not see the urgency. If I am a minority I may give to local charities to help my neighbors but not if I'm not confronted with it in my higher income and more racially diverse neighborhood. It isn't a great idea but it is my knee-jerk reaction.


I tend to agree with this line of thinking. If I live in a wealthy neighborhood and never see someone of another color, I might reach the conclusion "those who are not like me need lots of help." On the other hand, if I look around and see Asians, Blacks, Hispanics, and Whites all doing well, I might internalize the thought "hey, everyone is really doing pretty good these days!" or "People who want to work can make it - as evidenced by all those around me - so let them work for their own money."

Justin Dearing

The majority of donations in America are religious, is the same true in Canada? (Source xkcd http://xkcd.com/980/)

So the first thing I'd want to examine is what the numbers look like if you remove donations to religious institutions. I think you'd have to break it out a bit further of donations to my house of worship and donations to anything faith based. I say that because I donate to my Catholic high school, and I don't consider that a purely religious donation. The same would be true of someone that donates to a soup kitchen with religious affiliation.

On another note, is a church in a more ethnically and religiously united neighborhood provide a greater part of the social infrastructure? If church is something I just go to for spiritual/religious reasons, then I might give less to the church, then if I also spent 2-3 evenings a week in the church hall basement for social reasons.



> The majority of donations in America are religious, is the same true in Canada? (Source xkcd http://xkcd.com/980/)

Actually, you have this completely wrong according to the xkcd source. Donations to Religious organizations have the *plurality* of listed organizations as it's the largest group at 34.6%, but non-religious donations (everything else by definition) have the majority (100% - 34.6% = 65.4%).


I think on an individual level, maybe you feel that since you live in a diverse area or neighborhood, you have done your "due diligence" as a member of society. White people may feel guilty for living in an all-white neighborhood, as a result, the members of that area may have increased charitable donations. However, members of a more diverse neighborhood either don't feel guilty or directly aid members of a different group of people. Charitable donations by individuals are rarely intended for people of the same ethnic group. Again, this assumes that charitable donations are tied to feelings of guilt, but I figure this seems like an interesting way to look a this.


Why would you assume that charity is derived from guilt and what is your evidence for those claims?


I would guess that humans innate ethnic chauvinism causes a decrease in social empathy in diverse societies. When donors see beneficiaries as less similar to them, fewer prosocial feelings are triggered, which lowers giving. From an evolutionary perspective, it would make sense. More diversity means less shared genetic material which would lower altruism.

Joe in Jersey

Was income constant or did income drop as the neighborhood changed?
Are all ethnic groups as likely to report charitable giving on their taxes?
Were any measures attempted to track disposable income?
Were changes traced as people moved from one neighborhood to another, did they contribute more before moving?
Were ethnic attitudes towards charitable donations factored (such as giving to friends & family vs giving to organizations)?
Were economic changes in status factored?
Is someone who was raised in poverty or near poverty less or more likely to give to charity if they now earn a high income?
What was the rate of secularism in the different neighborhoods, and does that have a correlation?

You don't give us a lot to go on, what variables were they statistically able to discount?
What was the time frame covered?

S. Hussain

Is it possible that some of these observations may be resulting from the fact that many immigrants and/or visible minorities send charity monies to their home countries? I know that's what we do in my family and I have noticed the same in many of my relatives. We still give locally but a big chunk also goes to charities/independent efforts at helping poor in our countries of origin. Just a thought.


Seems pretty simple to me: the anomaly, from a global perspective, is the high rate of charitable giving among whites, not the low rate among non-whites. Whites in North America have developed, over time, a cultural sense of obligation (noblesse oblige, if you will) towards lower, oppressed classes, which carries forward to the present day in the former of a relatively high level of charitable giving. As society (and economic opportunity) flatten out, you will see these rates decrease.


I would like to believe that this "noblesse oblige" is a prominent feature of North American upper classes, but I don't. To me, conservative people are indeed extremely generous, but only to those whom they identify as part of their "in group". So to me it signals more of an us-vs.-them mentality than a "lift up the poor" mentality.


Probably for the same reason that we self-segregate.
Regarding religion, it's seem pretty intuitive that religious homogeneity in a community would indicate a strong commitment to that religion and lots of giving, whereas lots of inhomogeneity would indicate disparate levels of commitment.


I think Putnam's work E Pluribus Unum provides an interesting correlarry with looking at how trust goes down as homogeneity decreases in communities. More diversity -> less trust.


This feels related to Putnam's follow up findings (post "Bowling Alone") that diminishing social capital is correlated with increasing diversity.

Perhaps it's that the less we feel connnected or bound to those around us, the less we believe that our charity is benefitting our in-group.


There are many Catholic, Presbyterian and Jewish hospitals in the US, catering to and not imposing their faiths on their patients.

Are there ANY Muslim hospitals founded in non-Muslim countries that don't require patients to keep sharia? What is the ratio of zakat (Islamic charity) bestowed in dar al harb on us infidel dhimmies? Has religious non-diversity generated a disproportionately higher level of Islamic charity?


The authors have made a classic statistical error of selection bias towards ethnic groups. Most immigrants come from countries that are impoverished or at war and left their families behind in hopes of a better life. These people send a chunk of their non tax deductible income outside of the country to families on the brink of starvation. The money is use towards food, shelter, education and health. After all that, do you really think they are able to donate money to registered taxable charities? Sorry they don't sponsor individual children, they sponsor whole families! Keep that in mind before make such judgements.