Sunday, October 17, 2021


the Art Institute of Chicago fired all of its docents, or trained volunteer museum guides and greeters, for being "mostly older white women of above-average financial means."
According to the Wall Street Journal, on Sept. 3, Veronica Stein, an executive director of learning and engagement at the museum, sent an email to the more than 100 docents the museum has, firing all of them. "In gratitude for their long, unpaid service—averaging 15 years each—the Art Institute offered the involuntarily retired guides a two-year free pass to the museum,"

Cheesing off the demographic that provides so much of your donor pool must be the latest management fad.

Wouldn't you expect that volunteers would be older (children grown), above-average financial means (have leisure), women (men usually supporting the family)?

Certainly that reflects the demographic of the secular volunteers I've seen, and it would seem to plausibly explain the dramatic skewing of volunteer groups. It clearly isn't the whole story, though--the same economic/family issue that select for older and better-off women should also select for high-income minority groups as well--Japanese, Chinese, Indian...

So, off we go looking for rabbit holes.

The abstract for this paper is interesting.

... we analyze survey data on volunteerng, which show that whites volunteer more than blacks. We ask how much of this difference is due to the way human capital is distributed in the population. We develop a theory of volunteering that acknowledges that, besided human capital, social and cultural resources play a role in making volunteer work possible. Black Americans tend to be better endowed with these kinds of resources than whites, which partially compensates for their shortage of human capital. However, blacks are less likely than whites to be asked to volunteer, and less likely to accept the invitation if it is made. ... for all kinds of volunteering except the entirely secular, black volunteering is more influenced by church attendance than is white volunteering ... while socioeconomic differences ahve a smaller impact on black volunteering. Among volunteers for secular activities, church attendence has a negative effect on volunteering, but only for whites.

"Volunteering is a collective behavior" "The volunteer role is part of their identity"

The "collective behavior" aspect suggests to me that if social spheres don't overlap enough, people won't get asked.

As for the "church attendence has a negative effect on volunteering" I wonder if there's a "tapped-out" effect at work.

Political attitudes can play a role in some forms of volunteering: "Blacks are more likely than whites to believe that the government should help fund and organize programs for the poor and more likely to believe that charitable organizations are doing work the government should really be doing."

Interestingly, and perhaps not surprisingly, among blacks who attended church less than once a month, those who felt their religion was important to them were less like to volunteer than those who didn't think it was so important. (kidding themselves?) This was not the case among whites--not sure why.

The paper is from 2000, and the data are, of course, quite a bit older. But though my simple resource model is partly ok, it clearly has limits.


Assistant Village Idiot said...

Arthur C Brooks's Who Really Cares? found that conservatives volunteer more than liberals, but that much of this was explained by churchgoing. Churchgoers gave more and volunteered more even to secular causes like Girl Scouts or local cleanups. That seems different from what you are seeing here, and I don't have an answer.

james said...

I'm not sure why either. It might have been a sampling difference, or perhaps Brooks' sample was better suited to the purpose. Here they say they use Roncek's model (never heard of it) to decompose coefficients into a probability of being a volunteer at all "based on the predictor variables among those who did not volunteer" and another showing the effects of the "predictors on the time spent volunteering." I can imagine subtle errors creeping in here.
Googling... Kang complains that McDonald and Moffit's method is easy to use inappropriately, and it sounds like Roncek's.
I'm not familiar with Tobit analysis, so I only get the most general idea of what's going on.

"Black congregations are more likely that white congregations to be involved in civil rights campaigns, meal services, community development, and public education about diseases. In other words, black congregations are more likely to focus on issues of immediate concert to their community. (The data on congregational activism used by Chaves and Higgins are drawn from only 44% of the national sample of congregations and underrepresent smaller congregations.)"

Christopher B said...

Taking a swag just based on the abstract and what you quoted, I wonder if a confounder is capturing time spent in temporary self-organized efforts like painting a neighbor's house. I could see that being something a group from a church might do but not in any official capacity, and it might be hard to capture. Those involved might not even think of it as volunteering.

james said...