Give Up Blog posts a fascinating map of charitable giving by state, to shoot down a fatuous WSJ op-ed that tries to claim greater charity as a Red State virtue. I certainly don’t see a positive correlation there, do you?
That claim brings to mind another common misconception, that religion is the driving force behind charity. Here’s another map, of the distribution of the ungodly (not just atheists, but agnostics and other people who are not affiliated with any organized religion.)
Again, any correlation between these two maps looks like it would be fairly weak, and isn’t going to support any conclusion that the churchier the population, the more likely it is to support charities*…and please don’t bother to bring up the bogus argument that atheists don’t establish hospitals. I’ll also note that the second largest “religious” group in the US are the nonreligious/secular—while most are uncomfortable with the label “atheist” (something we need to change), freethinkers are still a large but neglected demographic in the US.
*Maybe you could argue that the prominent presence of skinflint callous heathens in a population reminds the godly to donate more, however. I haven’t seen a breakdown of how much the average atheist contributes vs. the average Christian.
In the first map, why is Oregon blue?
It has nothing to do with theism or atheism, I think. For evidence, look at the interior west, which on the one hand has a very high concentration of atheists and on the other a low level of charity. In contrast, the west coast has many atheists and a lot of charity. My guess is that it’s more related to socio-political culture: the northeast and west coast have liberals, who are likelier to donate to charity than the libertarians of the interior west, while at the same time both groups have a higher proportion of atheists than the conservatives who are ubiquitous in the southeast.
In the first map, why is Oregon blue?
giveupblog.com’s maps differentiate between red states and blue states. Oregon is just the only state that both went for Kerry and gives less than $47.80 per capita.
Those red states in the charity map have lower costs of living than the white states: the value of $20 in Alabama is a lot more than the value of $20 in California. The giving values should probably be adjusted accordingly. After that adjustment, it would be closer, though that would not change the conclusion that religious != charitable.
I have a theory about the differential blood-giving noted in the linked posting. Admittedly, it’s a stretch, but maybe more blue-staters would like to give blood, but are more likely to fall under one of the screening rules applied by the Red Cross. It goes beyond sexual history. At one point, you could be ruled out just by having lived in a European country with a mad cow disease problem, but I think rules may have been relaxed since then. It would be nice if there were a medical test to re-certify your blood for donation.
In the first map, why is Oregon blue?
giveupblog.com’s maps differentiate between red states and blue states. Oregon is just the only state that both went for Kerry and gives less than $47.80 per capita.
Oregon is also a very poor state, which would explain the low level of charity. And the state didn’t go for Kerry by much of a margin.
I think if this map proves anything, it’s that charity is less common in poor places, big surprise. And of course, the WSJ has their head up their ass if they think this proves the deep generosity of people who love George Bush.
Hmm, I hate to defend the red states, but total charitable giving per capita doesn’t seem quite as relevant as fraction of income spent on charity. Many of those red states are quite poor; the bottom two categories (under $30) on your map, with the exceptions of Wyoming and Nevada, are among the poorest states in the U.S. See http://tinyurl.com/cu5za .
the WSJ has their head up their ass…
I think you can stop right there. :-)
I agree. They should not measure charity merely in dollars given, but instead in dollars per yearly income. The problem with this is that 10% of your income means a lot more to you if you make $20,000 per year (like you can’t afford to go without that $2,000) than if you make $200,000 (could probably make due on $180,000).
Am I talking out of my ass about this? I guess what I’m getting at is that even standardizing for cost of living or per capita income will still lead to biases because less money means more to a poorer person than more means to a richer one.
I think it is worth considering the relative efficiency of the charities targeted for giving by nontheists and religionists, too. Some people consider giving money to missionaries who do no more than spread biblical flimflam as “charitable giving,” while nontheists would probably only consider giving aimed at addressing real-world problems as charitable.
Religious charities are less inefficient because most of them spend at least some portion of their operating budget on converting people to their favorite fairy tale rather than paying for food and medicine, or sponsoring education and political empowerment, etc.
I think this analysis is almost certainly wrong: it uses a measure of giving that I believe (a) is limited to charitable bequests (i.e., gifts in a will) as opposed to total donations, and (b) does not include many churches, because many of them don’t have to register as 501(c)(3) charities.
The giveaway is Utah, which is filled with tithing Mormons. No way is the per capita donation in that state $20-$30.
(That having been said, I don’t think a correctly performed analysis would necessarily be much different.)
A few random reactions:
Is absolute $ amount the best metric to use? What about % of income?
Be that as it may, a lot of evangelical churches teach that Christians should literally tithe — ie. give 10% of income. Doesn’t sound like many of them are anywhere near that.
Odd about the blood-donor thing — I wonder why? But it’s still a mighty small data point on which to hang sweeping generalizations about politics and social conscience.
It would also be interesting to see what percentage of income tithed to churches actually finds its way to legitimate charitable causes.
I suspect that when you subtract out the costs of building/maintaining the typical suburban McChurch, the staff salaries, car/housing allowances etc., money spent on proselytizing, funding campaigns for “stealth” school-board candidates, etc… there’s not a whole lot left to help out the less fortunate.
Here is a ranking of the states by total itemized charitable contributions as percentage of total adjusted growth income:
(UT) Utah (4.68%)
(DC) Dist. of Col. (2.96%)
(AL) Alabama (2.48%)
(GA) Georgia
(NY) New York
(MD) Maryland
(ID) Idaho
(OK) Oklahoma
(SC) South Carolina
(WY) Wyoming
(NC) North Carolina
(AR) Arkansas
(MS) Mississippi
(VA) Virginia
(MN) Minnesota
(OR) Oregon
(CA) California
(TN) Tennessee
(CO) Colorado (2.00%)
(DE) Delaware
(NE) Nebraska
(KS) Kansas
(FL) Florida
(AZ) Arizona
(MI) Michigan
(KY) Kentucky
(MT) Montana
(MO) Missouri
(NJ) New Jersey
(CT) Connecticut
(IL) Illinois
(TX) Texas
(IA) Iowa
(WA) Washington
(PA) Pennsylvania
(MA) Massachusetts
(IN) Indiana
(HI) Hawaii
(NV) Nevada
(OH) Ohio
(WI) Wisconsin
(NM) New Mexico
(RI) Rhode Island
(LA) Louisiana
(SD) South Dakota (1.53%)
(ME) Maine
(VT) Vermont
(AK) Alaska
(ND) North Dakota
(NH) New Hampshire
(WV) West Virginia (1.14%)
Problems with this analysis:
(1) Assumes itemizers give at same rate as non-itemizers (itemizers are generally richer).
(2) Assumes AGI fairly reflects cost of living differences.
(3) The dataset I used(http://nccsdataweb.urban.org/NCCS/Public/index.php) is several years old.
(4) Assumes it is appropriate in the first place to take into account income/cost-of-living differences. (Massachusetts spends a crapload on public education and supports unions, and consequently its citizens have higher incomes and can give more money. I’m not sure why it makes sense to mark Mass. down for that.)
These problems notwithstanding, I think this is a better ranking than Give Up Blog’s analysis.
caerbannog writes:
It would also be interesting to see what percentage of income tithed to churches actually finds its way to legitimate charitable causes.
I suspect that when you subtract out the costs of building/maintaining the typical suburban McChurch, the staff salaries, car/housing allowances etc., money spent on proselytizing, funding campaigns for “stealth” school-board candidates, etc… there’s not a whole lot left to help out the less fortunate.
This is a fair point, although query how your higher-end colleges, universities and museums would fare on that metric.
I am not sure about the relevance of blood donating, is this purely altruistic or do some states pay donors?
In Canada about 88% of the poplulatin donates to charity (only about 4% of eligible blood donors donagte blood). While religous groups get the most money (52%) they only account for 15% of donations by number. A higher percentage of people in our poorer provinces make donations but there does not seem to be any religous bias to it (our Catholic province of Quebec gives relatively lower amounts.)
I also note that some of the money donated to religous organizations is spent to paay for pastors, missionaires etc and do not contribute to the benefit of the poor (other than givng them spurious comfort).
I am atheist and I have no hesitation in giving to charities becouse I can afford to. I include some religous ones (Salvation Army, at least they don’t require their charges to be overtly religous, they just help em cause they need help).
It is about compassion and a desire to help, not god or politics that determines what most people do I think.
http://www.givingandvolunteering.ca/factsheets/1997_CA_charitable_giving_in_canada.asp
Does it measure those who give of their time instead of money? I mean everything from working in soup kitchens to cleaning abandoned lots to marching on abortion clinics. After Katrina, I gave $50 to a church in east Texas run by a friend of a friend of a friend who took in a displaced group of disabled children.
I like using actual dollar amounts, because it helps make an underlying point. Why are the red states poor? What is cause and what is effect?
home sweet California has one of the largest economies in the world, continues to have huge in-migration [the bit about California being business-unfriendly is largely a canard by republicans trying to get some traction] and is one of the bluest states in the union.
Per Francis, here are the data I used above ranked by total itemized contributions per tax return filed (all returns, including non-itemizers):
(UT) Utah ($1750)
(DC) Dist. of Col. ($1375)
(NY) New York ($1128)
(MD) Maryland ($1097)
(CT) Connecticut ($1025)
(GA) Georgia ($949)
(NJ) New Jersey
(WY) Wyoming
(VA) Virginia
(CA) California
(MN) Minnesota
(NC) North Carolina
(DE) Delaware
(CO) Colorado
(AL) Alabama
(MA) Massachusetts
(ID) Idaho
(SC) South Carolina
(IL) Illinois
(OR) Oregon
(MI) Michigan
(FL) Florida
(OK) Oklahoma
(WA) Washington
(KS) Kansas ($750)
(AZ) Arizona
(NV) Nevada
(TN) Tennessee
(NE) Nebraska
(AR) Arkansas
(PA) Pennsylvania
(TX) Texas
(MO) Missouri
(WI) Wisconsin
(IN) Indiana
(MS) Mississippi
(HI) Hawaii
(RI) Rhode Island
(OH) Ohio
(KY) Kentucky
(IA) Iowa
(NH) New Hampshire
(MT) Montana
(LA) Louisiana
(NM) New Mexico
(VT) Vermont ($511)
(AK) Alaska
(ME) Maine
(SD) South Dakota
(ND) North Dakota
(WV) West Virginia ($352)
Looking at this data, I’m guessing the reason ND did so well in Give Up Blog’s analysis is that it has a lot of people who die relative to the rest of the country (few young people, lots of oldsters) and therefore comes out very well in a bequest analysis.
This is, just like the other side’s spin on it, spurious and disingenuous, as are most statistics without measured analysis.
Some of the dark red states in that chart, like New Mexico and Arkansas, have very high levels of poverty. It’s asinine to ask someone making $12,000 a year while trying to support a family to also donate to … whom? The poor?
Lame.
When you post about politics, Dr Myers, I tend to waiver in respect for your level-headedness. The standard of proof should not change just because it’s not a hard science topic.
Umm, so you think there is a strong correlation somewhere in those maps? My conclusion was that there wasn’t.
Give the fact that the charts address the claims of others, I can’t see why PZ should mention parametres (income levels) outside those claims. I found it nice that alkali took the time to find that information, but they didn’t have anything to do with the original claims – that red states/religious people give more than blue states/non-religious people.
And in other words, you falsified the claims made by others, which is what I was trying to say in my last comment.
According to Time, Texas, Oklahoma, Arkansas and Louisiana have the lowest percentage of donors (61%). Average per capita is 2,204$. There’s no wonder that the West Coast is the most generous because of Hollywood stars. Midwest is the one where people are most likely to give to religious causes. Highest percentage of donors (82%) are found from New England (most libertarian).
All I can do is shake my head in disbelief at this posting.
All of these statistics are pretty useless, the way this should be gone about is to look at the per capita caritable donations as it relates to per capita income. Also I suspect that if we took religious organizations out of the mix, I believe tithing is tax deductable, then the data would be further skewwed.
Why are the red states poor?
One could just as well ask, ‘why are the poor states red’?
I love this site, but this is a rather poor analysis, as some above posters have noted. Social scientists tend to be much more careful with data; flashy maps like this one are popular on the internet, but don’t show much of anything.
Problems here:
1. The quantity of interest is poorly defined. Are we trying to explain, say, “charitable effort”? Then we need to decide what consitutes effort: dollars per capita? dollars/income per capita? do we adjust for publically provided social services crowding out private charity? do we give greater weight to donations from poor people, as in the parable? etc. And then we need to decide what counts as a charity: social services? administration costs for same? churches? fine arts? education?
2. The analysis takes place at an aggregate level, but makes inferences about individuals. It is surely rife with ecological fallacies. Most researchers would dump the state averages, and go straight to individual level data. The effects (if any) of states could be estimated from individual data.
3. The analysis is bivariate, and ignores many confounding effects, including: differences in income, differences in public services, differences in tax rates, etc.
A multivariate analysis of individual data with a well defined measure of charitable effort and appropriate controls for income, policy context, religion of donor, state of donor, etc, could get at the questions here.
The map of aggregate data tells us nothing, really.
I’m scratching my head over this. I didn’t draw any positive conclusions from these maps, and my whole point was that simplistic relationships like the one postulated by the WSJ or theists who claim superior morality simply do not emerge in any obvious way from any of the data I’ve seen.
I think any time “red state,” “blue state,” “money,” and “religion” coincide in one post, you’re asking for knee-jerk responses. :)
However, thanks for posting this – I’m going to use it in my introduction to scientific thinking course, as yet another example of media/special interests feeding us statistics that are basically uninterpretable. (As a resident of Idaho, I can vouch for poverty adversely affecting charity, but I find it totally fascinating that this state is “ungodly.” We have four pages of church listings in our phone book!)
By the way, anybody know which single state is the ungodliest? I wanna move there. ;)
PZ-
I certainly agree with you that the WSJ is almost certainly wrong. I’m cynical enough to believe they didn’t look at any data at all.
That said, while you didn’t drawn any “positive” conclusions, you seem to imply a negative one (ie, that there isn’t a positive correlation between red states and charitable giving).
I honestly don’t know what we’d find if we looked at individual level data, using better concepts and controls for various confounding variables. We might find red state residence increased charitable effort all else equal. We might find the opposite. I agree with you that we can’t tell from simplistic modeling like these maps, but I found the tone of the initial post to suggest the absence of a bivariate state-level correlation among these data was some sort of evidence.
Sigh. It’s just a bit frustrating that the sort of tools used by social scientists—which are designed to cope with problems like confounding variables or ecological fallacies—seldom make it into even the best intellectual/scientific blogs. Biologists are hounded in the press by pseudo-scientists, but we social scientists don’t get anywhere near even media with amateurs playing around with data.
Of course, that could be partly our fault… there isn’t exactly a Pharyngula in the social sciences.
mss,
I don’t know where you got the idea that PZ was asserting “that there isn’t a positive correlation between red states and charitable giving.” It seemed clear to me that he was saying that we can draw no conclusion from these data.
I don’t know where you got the idea that PZ was asserting “that there isn’t a positive correlation between red states and charitable giving.” It seemed clear to me that he was saying that we can draw no conclusion from these data.
I read his post as saying: We can draw no conclusion from these data, therefore the WSJ claim is crap.” That follows if and only if these are the data that the WSJ used to make its claim. If the WSJ based its claim on, for instance, per capita giving as a percentage of total income, then the the data that PZ posted are truly worthless; they’re worthless as an estimate of “generosity,” AND they are worthless as a way to assess the WSJ claim.
The most interesting thing to me is that despite the obvious bias in our culture in most places against atheists, agnostics, godless secular humanists, and our ilk, it is the largest “denomination” outside Christianity. At 13%, it must be larger by far than the number of fundamentalists who try to claim the title “Christian” exclusively for themselves… except when adding mainstream Christians to swell their numbers is required (as when they wish to claim “we are a Christian nation”). It’s also growing faster, I’ll reckon. Despite their claims, I really doubt fundies are growing at a 110% rate over a decade. Despite their efforts to Disneyfy their product – or because of them – there’s only so many adults who will subject themselves to that kind of infantilization.
The non-Judeo-Christian religions are the fastest growing of all, but their numbers are still so slight as to not be as impressive as the Seculars. Although fastest growing of all are Deity-ists which presumably includes some old-fashioned Voltaire-style Deists, the closest relatives to Seculars in the metaphysical evolutionary tree.
I certainly agree that a real investigation of degrees of charity has to account for larger incomes in various places. But the problem with eliminating tithing and other religious contributions is the great diversity in what these support. I suspect that these vary from simply paying clergy’s salaries to support of other organizations. For example, here in Montreal, I am a volunteer at a secular charity that nevertheless receives donations from clergy. Since these clergy in turn have their salaries paid (in part!) by donations from their congregations, this becomes rather messy.
CanuckRob: Quebec is not very Catholic any longer. There are a lot of Catholic institutions and buildings, but it seems to me that its influence on policies and the like is long over. Partication numbers are also falling; more and more old buildings are being sold off for use by universities and other places.
Of course. But if Christianity is fractured into numerous denominations, unbelievers are even worse. If there are 30 million freethinkers in the US, that represents 30 million different views, not one.
I disagree about the proportion of income thing. Yes, blue staters make more money in the first place. But that is not a relevant control for two reasons.
The first is that, like it or not, total charity is what matters to those in need of it, not fractions of anyone’s incomes. Despite the fact that I think he’s something of a slimeball, I can’t deny that Bill Gates has done more for humanity than I or virtually anyone else that I know. The REASON that he COULD do this is because he first made a ton of money. Sure, that makes him more capable of giving than I am, but that’s exactly the point: he did something that made himself more capable of doing good than I have done. He can do more good than I can. Griping about proportions is just my sour grapes.
The second is that making a ton of money, creating wealth, is itself by some estimation a form of charity and positive social good. If you aren’t a devotee of economics or capitalism, you might not agree with me on this one, but I personally think it’s undeniable. While some wealth is just transferred around or earned in interest, the real power of the blue states is that they create wealth, in the economic sense. And wealth cannot be withheld from society: it’s existence is automatically a good for all (whether by lowering prices of all goods, some goods, creating larger capacity for employment, and so on).