Back before I started blogging in 2008, I read and then wrote an online review of Arthur Brooks’ book Who Really Cares: The Surprising Truth About Compassionate Conservatism. I’m reposting it here with some updates, because I continue to find people citing Brooks’ study as evidence that conservatives are more compassionate than liberals.
In short, Brooks’ book is a highly flawed diatribe posing as a rigorous statistical analysis. Anyone who has taken basic social science or statistics should easily be able to see through his poor use of numbers and data, but unfortunately it’s become somewhat of a touchstone anyway.
First of all, Brooks states on several occasions that his goal with the book is (if I may paraphrase) to get liberals off their high horse about conservative greed and selfishness, and to motivate them (the liberals) to give more. However, he has so completely intermixed conservative political opinion with his statistics and arguments, that the only ones who are going to be around to hear him are either masochists like me or the already-converted conservatives. If one REALLY wishes to motivate liberals to greater charity, spending a large portion of the book talking about how deluded, misguided, and selfish they are is probably not the most effective strategy. The book as written is far more likely to reinforce conservative Americans in their smug, self-satisfied opinion of their goodness, than it is to spur any liberals to good works.
Secondly, although I suspect that a number of the statistical claims Brooks makes may be true (although any beginning statistics course will emphasize that correlation is not the same as causation), they are compromised by other places where he frankly plays fast and loose with the numbers. One of the most egregious examples is his discussion of the estate tax on page 102 (hardcover edition). Brooks cites as proof that taxation hurts charitable giving, the statistic that someone who inherits $20,000 will give an additional $82 to charity; someone who earns that same $20k in the stock market will give $48, and someone who just gets a $20k salary increase will only give another $18. He lets us digest those figures for a few pages, and then on page 113 announces “Even if the estate tax’s demise did redirect donations toward heirs, these would be especially likely to give much of it away.” (emphasis mine).
Now, I don’t know about your math, but in my calculations $82 is not MUCH of $20,000. Even if the estate tax were only 10% (and it’s much higher than that), the tax revenue (admittedly not the same as charity) would be $2,000. If “generosity” is defined as giving less than 1% of anything, God preserve us from the stingy folks! Of course, Brooks’ argument here also suffers from the reality that no estate in the U.S. is taxed until it exceeds $5.34 million (it was $2 million when Brooks wrote the book), so that $20,000 differential he’s talking about is on top of a bigger chunk of change than most Americans will ever see, but I digress.
Additionally, it has been noted by at least one reviewer I read elsewhere, but it merits repeating, that a good deal of the giving behavior of religious-vs-nonreligious folks comes from a survey of self-reported behavior, not observational studies (which are admittedly harder to do). I doubt that it would fully explain the difference, but I know that religious folks (and I am one) tend to answer survey questions according to what they “know they ought to do” in at least some cases—that is, when self-reported “good” behaviors are compared to observable criteria, there is almost always an over-reporting of the “goodness.” Since non-religious people are (I am assuming, as is Brooks) less likely to feel a moral compulsion to give, they may also be less likely to over-report their giving. I doubt this could explain away the entire difference, but it is a potential source of bias that is worth examining.
But the most wrongheaded elements of the book occur when Brooks leaps from statistics to policy. Even if we give him the benefit of the doubt, that every statistical claim he makes is correct, valid, and supported by the data, his conclusions don’t follow. I am one of his “rare bird” religious liberals. . .actually I’m an even rarer bird, as I’m a morally-conservative but politically-liberal Protestant/Evangelical Christian (yes, we do exist!). Brooks concedes that we “religious liberals” give nearly as much as the religious conservatives (I actually give more than a lot of conservative friends of similar age and income in my own church), and that both give a great deal more than the nonreligious of either stripe (of course, whether even this difference is inside or outside the margin of error, we can’t tell since Brooks has not disclosed his methodology). He makes it abundantly clear that religion trumps every other indicator, and then uses his numbers on the generosity of conservatives to justify and advance a conservative political agenda that is at odds on nearly every point, with the perspective of us “religious liberals.”
He also makes the fallacious implication, when arguing taxation-vs-charity, that taxes are primarily (both in intent and effect) directed at the redistribution of income, and secondly that these taxes are spent in largely the same spheres as charity, such that one would supplant the other. Both implications ignore the fact that the U.S. federal budget includes vast sums—all paid for by taxes—that can’t be classified as charitable causes under the most expansive of definitions. Defense, debt service, agricultural subsidy, commerce, public health (disease control, not medical aid) and even the basic operations of government itself are items that together add up to vastly larger sums than anything that can be construed as being in the same competitive space with charity. These things still require funding, and that funding comes from taxes. Much of the debate over taxes, therefore, is not a question of taxation vs charity, but rather: “Given these public expenses, how are we going to distribute the burden of paying for them?” This is an ethical question, and compassion can and does play into some people’s answer to that question, but it is not necessarily correlated with the compassion of charitable giving. It is not for this reason any less-valid a concern from a compassionate point of view.
Yet another fallacy in Brooks’ reasoning is that all giving that is legally defined as “charity” (a term he never actually defines), is in fact motivated in any way by compassion. The vast majority of many religious people’s giving is, I rather suspect, spent on the institutions and facilities that service their religious practice. This is true in every church I’ve ever attended. To really get at the question of charity, one would have to analyse giving that is actually directed toward charitable purposes … this is an analysis Brooks does not do.
In the final analysis I, too, would like to see charitable giving increase. I do think that Brooks makes a valid point that anybody who’s crowing about compassion should put his/her own money where their mouth is and pony up real donations of time and cash. But particularly if charity is to fill the gap of that minority portion of government spending that IS in the same space, EVERYONE—conservatives, liberals, religious, nonreligious—will have to give a LOT more than they give today to make the slightest dent in the problem. Far from soothing complacency for the “compassionate conservatives,” Brooks should call his own people (a majority, he says) to greater good works and not merely heap scorn on those outside his fold whom he deems hypocritical.