Lies, damned lies, and statistics

This newly published study claims that red state people are more charitable than blue state people with the unequivocal headline, “Red States more charitable“. See there, bunch of stingy commie Others who voted for the Nigerian vs genuine Americans who loved Sarah. But as with all statistics if we tease this a little bit might there be some other possible conclusions.

While the article only talks about a few states the website of the survey has the complete list. [Hint: In articles, esp. from political blogs ALWAYS go look at the original source] Just among red states, the red states that had lower ratios (less “giving”) are mostly plains states and the red states that had higher ratio (more “giving”) are mostly southern states, certainly a statistically significantly conclusion to the casual eye. Another study, that I can’t find at the moment but that I quoted recently, showed a similar pattern, religiosity is lower in the plains states than southern states. There is not as strong a pattern in the blue states but generally the highest giving states are more southern or border, but also some of the least giving are also the least religious. Now both the article and the original source say this as well, but the implication they leave is that religious give more – is that really true? Let’s dig deeper.

Now another obvious pattern, strong but not completely correlated is that the states with the most giving are generally the poorer states, good for them, charity starts at home. But having Utah on the very top of the list, Idaho near the top, but very similar states of Montana and Wyoming near the bottom, hum, are we seeing a pattern yet.

A few more tidbits before it’s time for the final exam.

  1. the definition of most charitable is portion of income not absolute value, IOW, a dollar given in a poor state gets a higher rating in this survey than a dollar given in a richer state
  2. Two high giving states are also home to large Mormon populations where large tithing is effectively mandatory (Romney may not release his tax returns to the voters but you can be sure the church knows.
  3. We’ve already seen a correlation between higher income and reduced religiosity.

Still don’t get it. Well, folks, this is from IRS data (which must be voluntary self-selecting survey since this is not the IRS saying this but a private group, which might have some influence; nor is there any claim that the samples by state are proportional to the population of those states; and the organization doing the survey is one that promotes charitable giving which hardly makes them an impartial group).

Still don’t get it, well, maybe that’s because you don’t do that section of Schedule A in your tax return. Donations to religious organizations, not quite the same as United Way or Goodwill, are tax deductible as “contributions”. So in states with high levels of church attendance and especially substantial pressure for tithing, hum, I wonder if that has any effect on these statistics.

I was looking at the original source, always a good thing to do instead of just looking at the web article. I noticed in the sidebar there are multiple articles (different cuts of the data). Naturally Politco only focused in on the article titled, “The Politics of Giving”, but the article, “American’s Generosity Divide” has some good bits:

Religion has a big influence on giving patterns. Regions of the country that are deeply religious are more generous than those that are not.

Already figured that out, but here’s the more telling part which I was hinting at; just in passing they say:

When religious giving isn’t counted, the geography of giving is very different. Some states in the Northeast jump into the top 10 when secular gifts alone are counted. New York would vault from No. 18 to No. 2, and Pennsylvania would climb from No. 40 to No. 4.

My oh my. So we’re finally at the bottom of this. It’s not that religious give more to charity, they give more to their churches. And churches are under no obligation to spend a penny of that giving on charity. They can build ever more buildings with fancier windows and gold inlay bibles or better organs, they can upgrade the parsonage and the pay of various workers, they can spend the money on various kinds of political activism (now allowed, always secret before), they can buy billboards and ads to advertise religion or denounce abortion, they can spend it on retreats and xmas displays, they can spend it to evangelize and recruit. In fact, the mormon church owns a 4B$ shopping complex in Florida which I doubt gives away food to the poor. How much of the contributions go to charity, well, no one knows, because that’s are secret compared to most charitable organizations who are often under the microscope to see what they do with their contributions.

Now compare to United Way or GoodWill or American Cancer Society or even Planned Parenthood, sure they keep some for administration and fund raising but they’re not spending much on obtaining new converts since they don’t actually have any converts. They provide services or research to help all people, not just members of their congregation.

From yet a different breakdown, we see a glimpse at another pattern.

For instance in Alabama the third “highest” givers

The state has only 0.7 registered nonprofits per 1,000 residents—the seventh-lowest rate in the country.

Hum, no charities to give to, but still high giving – I wonder where it’s going. But even with a deep dive we still don’t have enough of the raw data to really see what’s going on and we’re at the mercy of the analysts or reporters who tell us what they want to tell us.

Various other details suggest a far different pattern than the Politco article implied. But we’re not done, there is more:

The rich aren’t the most generous. Middle-class Amer­i­cans give a far bigger share of their discretionary income to charities than the rich.

The 1 percent really are different. Rich people who live in neighborhoods with many other wealthy people give a smaller share of their incomesto charity than rich people who live in more economically diverse communities.

Sounds kinda like the way the rich pay taxes too.

Anyway, I think even looking at what the original source provided a different picture emerges that what Politco reported. But how about this, what if the original source published the complete details of their data (with proper care about privacy) so that any of us could download load into a spreadsheet and do our own analysis. I bet all sorts of things could be teased out of the data that weren’t provided in these brief reports.

The Internet is supposed to empower this kind of thing, but how many people went beyond the Politico article to the original source or figured out that much of the giving data was distorted by income of the states or amount of religious giving. No, the ten second soundbite was that red states are more generous and I’ll bet 99% of readers left it at that and walked away with that conclusion. Whereas I’ll also bet that an actual critical study of the raw data, esp. subtracting out the giving to church improvement funds, would lead a completely different result.

So read these things critically and don’t just take them at face value, including my blogs too.

Advertisements

About dmill96

old fat (but now getting trim and fit) guy, who used to create software in Silicon Valley (almost before it was called that), who used to go backpacking and bicycling and cross-country skiing and now geodashes, drives AWD in Wyoming, takes pictures, and writes long blog posts and does xizquvjyk.
This entry was posted in whine and tagged , . Bookmark the permalink.

6 Responses to Lies, damned lies, and statistics

  1. If you scroll a little further down to the comments, one of the blogger’s commentd shared THIS website at about 7:30 p.m. on 8-20-12:

    http://philanthropy.com/article/FaithGiving/133611/

    Again we can’t look at the data directly, but at the very least, they pulled out the religious giving part, which showed just how much the percentages changed making the Northeast the most charitable to non-secular organizations. Hum…I wonder if this includes universities and colleges—see more below…OR whether it includes what the so-called lobbyists do under the current campaign financing scheme—I wonder whether that can technically be “charitable giving.” I know I cannot write-off contributions for political concerns on MY taxes, but I’m sure that doesn’t stop the 1% American elitists.

    A further website shows a partial criterion that was used to formulate their numbers, but it’s still not enough. It shows that “…the newspaper’s approach was reviewed by” exactly (2) educationally titled individuals from MA & MD, and (2) others who “provided advice” from Duke and Boston colleges located in NC & MA (another hummm….):

    http://philanthropy.com/article/How-The-Chronicle-Compiled-Its/133667/

    Here’s a part I found interesting from this site:

    “The IRS releases total amounts donated, but to protect privacy, the agency does not provide data about the specific charities people supported. Because of discrepancies in the data for people with income below $50,000, The Chronicle’s study includes only taxpayers who reported incomes of $50,000 or more.”

    With this statement alone, their study excluded at least 50% of the American population. OMG!!! How do these people sleep at night? Sloppy study = professional achievement? Arghhh…

    Thanks SO much for providing so much insight AND the logic that goes into providing ACCURATE statistics. Unfortunately, I’ve forgotten more than I retained from my college statistics courses. I’m most appreciative for it gave me something else to think about.

  2. dmill96 says:

    Thanks for digging deeper at the source and pointing out these additional issues.

    The part that is quite significant from this page, http://philanthropy.com/article/FaithGiving/133611/, are the actual values. From the map a crude average total giving in 4.5%, but after religious contributions are subtracted it drops to 1.075%. Wow! That means over 75% of the “charitable” contribution are just church tithes. Furthermore on the map of the after-removing-religious I would say there is no statistically significant difference between the regions, but the pattern that does show indicates red states, aka, higher religiosity states, do actually give a little less (IOW, I would interpret that since they’ve given their money to churches they don’t have any left over for real charities and/or they’ve assuaged their guilt by their church contributions and don’t believe they owe any more).

    But the interpretation of this data is difficult and in fact without breaking out the actual deductions and having all the detail conclusions are difficult at best. Certainly the glaring headline that red states are more charitable is invalid but it’s also not clear what conclusions are valid.

    Bringing up the college contributions is a good point. If one is going to pick on the other side one has to admit their own no-clothes. Since religion is secret and doesn’t report where their contributions go it is easy to assume no a great deal goes outside the church itself. In fact, recently, some articles tried to understand this specifically related to mormons who seem to be particularly oriented toward spending contributions on their church business. But at the same time I know here that Catholic charity is absolutely directed to some of the most disadvantaged and totally qualifies as charity. So without the data it’s difficult to say how much of the contributions to churches passes on to actual charity. The non-profits, OTOH, typically do provide reports and get a lot of criticism when they don’t pass on enough. But, colleges hardly qualify as charity at all. If all the contributions went to scholarships for disadvantaged, maybe that would count, but paying for football stadiums (here in the midwest) or research buildings (east and pacific) hardly counts.

    I’d noted the 50K$ limit and that does seem to be problematic, although I’ll bet that the usual distribution of anything economic in the U.S., those under 50K$, while possibly 50% in number, are probably a relatively small fraction of nation’s total income. Plus it’s probably very difficult to use IRS data for under 50K$ since relatively fewer of those people are going to itemize. And the cost-of-living adjustment methodology seems dicey too.

    So I wonder if anything at all can be deduced from this study!

  3. Excellent analysis. Some people take every statistic they see at face value; this should NEVER be done.

    • dmill96 says:

      I may be critical of the study but I also know it’s very difficult to do studies well. I think the original source tried to do it right but fell short. But once they publish the “popular” press (now mostly meaning blogs) will condense their findings even more into whatever soundbite will attract attention.

      When one sees something that doesn’t seem quite right then those of us who have some time can dig a bit more deeply. That may do little more than satisfy us, but also if we can generate some waves in a vast turbulent ocean at least a few people may hear and so see things a little differently.

      Most people don’t have the time, and sometimes the skill, to check out everything they hear, which is unfortunate because a lot of what they hear, in this polarized world, will be wrong. It’s a irony that we live in a time when never has so much information been available, but we’re so overwhelmed by it we rarely check anything out.

      Thanks for your comment.

  4. You said it, sir—“Certainly the glaring headline that red states are more charitable is invalid but it’s also not clear what conclusions are valid.” In other words, the article is yet another piece of junk sent into cyberspace for people to read and form their opinions based upon what…”invalid” conclusions? And as you said 99% will read it, take it at face-value, and never click on any of the links to determine whether the articles “conclusions” were even close to being correct. Damn…ignorance reigns, intellect declines!

    • dmill96 says:

      The more I looked at the original source I think they’re genuine and trying to figure it out. Of course they have an obvious vested interest, increasing charitable giving, so understanding who already does it and why is useful. The trouble is that analyzing data is tough. It seems like they tried, but they also made a few mistakes. I’m not sure what they think they found. But that’s actually why, sometimes, I’m a fan of crowdsourcing – put a bunch of data online and people will dig into it, sometimes the people who do that are experts who enjoy a challenge and thus you, perhaps, get a much better analysis.

      But the reason I pick up on articles like this is that someone does a study, it’s long and complicated, but the “popular” press (now days meanings blogs as much as actual media) pick up on the study, spend their 5 minutes of attention, write it up to attract viewers, and that’s that. Conveying information is secondary to churning out volume and also working for popularity (when sites are paid by the click, anything to increase clicks).

      I happen to have some time now and I like to read stories, but I actually like to find stories that don’t sound quite right and then try to investigate them more thoroughly. Sometimes that leads somewhere, sometimes it doesn’t. If real journalism is disappearing then this pass-along-what-I-heard style journalism of blogs and popular “press” on the Net have an opportunity to actually dig a little deeper and do real stories. But I don’t count on it.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s