The other arbitrary definition at play here is how the country is broken down into regions. There's a lot more homogeneity in the Northeast and the South than there is between, say, Ohio and North Dakota, or California and Wyoming.

It's both beneficial and interesting to break out the secular v. religious organizations here, but I take issue with the fact that the author from the Friendly Atheist lumps all church-related charitable organizations together as "padding someone's pockets". We know that does happen in some organizations, but I also think many churches provide needed services to the community as well through homeless shelters, soup kitchens, and other resources for people with a variety of needs. Definitely some religions and Christian denominations do more social justice-type work than others (think Catholics and Episcopals), but I don't think it's fair to say that ALL church-giving is going to make a few people rich(er). There's a happy medium between the two somewhere and the use of both charts in helpful in seeing that it's there.

This study also only counts charitable deductions reported to the IRS, so I would be recorded as not giving any money, even though I did. H+R Block told me not to bother keeping track of and claiming charitable deductions, as I rent instead of owning a home and itemized deductions don't make sense for renters. I would have to donate much, much more than a home owner before it would make an impact on my taxes, so why bother itemizing? I'm guessing plenty of renters are in the same boat.

Additionally, the money I give to homeless people wouldn't count, as individual homeless people aren't 501(c)(3)s. I'm hardly stingy, but I look like it on paper.

I was more surprised by the graph itself. Before my brain processed the numbers I thought the sections were similar to bar charts. And since the western region doesn't use a darker color to create shading, I thought it was flat, making it the "stingiest" region.

After I read the whole thing it was much clearer, but my first glance really threw things off!

I find it interesting  how the IRS counts church donations.  For instance, yearly synagogue dues count as a charitable donation - despite the fact that it is only sort of a donation.

If we are "measuring generosity", I think we need to further widen the scope from charitable donations (including volunteerism) to collective giving via taxation.  A country might have generous social programs, with a general attitude that they are taking care of people and so individual donations aren't necessary.  I've read such explanations for the relative "stinginess" of Nordic countries compared to the US.  But if we look at which model is better overall at helping people, good safety net with low individual giving vs. poor safety net with high individual giving, the former is superior:
http://thesocietypages.org/socimages/2012/09/14/reducing-poverty-where-theres-a-will/
Individual donations never make up for solid social programs, and I think a country which relies on charity instead of welfare, such as the US, is much less generous.

In other words, simply looking at charitable giving does not accurately measure generosity. 

I just looked at the website http://www.charitywatch.org/toprated.html , which rates charities based on a number of important criteria (e.g. what proportion of donations are used for actual aid), and found that under the headings "Child Sponsorship," "Hunger," and "International Relief and Development," church-sponsored organizations are  not only rated very highly; they in fact comprise the majority of charitable organizations in those areas. While I was still a practicing Catholic, I donated large amounts, through fund drives at my parish church, to numerous causes, including the local food pantry, to Catholic Charities and to "The Bishops' Annual Appeal", both of which use the VAST majority of their funds to do actual good, NOT to pad anyone's pockets (no, not to pay out bribes to pedophilia victims; yes, both or them report where the money goes). The other donations I made directly to my parish went to support numerous programs, including ones that aided parish members. I don't consider that sort of donation to be "spending money on [my]self and calling it charity" (as "fork" accused), since I did not utilize most of the parish programs I was supporting with my weekly donations. If buying Girl Scout cookies can be considered a charitable donation, then why not count when you give money to support your parish's youth group, especially when you don't have a "youth" in it? While the operationalization issue is hugely important (see Minerva's comment, especially), and while, as has been  pointed out by others, there are lots of other issues with how donations get counted, the level of vitriol expressed by the non-religious towards religion, as if all religious charities are in fact mega-churches that enrich the guy at the top, means that the arguments become irrational. Yes, all of those religious charitable organizations report how much, where it goes, etc. Do individual parishes? No, for the most part, which DOES create operationalization problems. At the same time, I suspect that we all might have different ideas about where best to donate. Some give to their local police, fire, and rescue organizations, some to disease/health based charities, others to ones primarily serving children, and yet others to charities that help animals or try to protect the environment (Should we argue that it's not really charitable since we can benefit from saving the environment or from advances made to cancer therapies?). We could all yell at each other about these various priorities, couldn't we? But that wouldn't serve much purpose. Can we instead discuss issues of operationalization without the hate-filled vitriol?

Hell or High Water, I will not sit by while Maryland, Delaware, and DC get called Southern. The Census Bureau needs to get a clue there, but wait they're too busy being confused over race. Pssh.