Willard. She was a graduate student when I was a graduate student at the University of

British Columbia. If I understand correctly, she’s coming here as a postdoc next year.

Stephanie Kramer’s a graduate student of mine at the University of Oregon, and of course

thank you, Cristine, for organizing this. It’s great so far. Cristine assigned me

this topic [Psychological and social consequences of religious (dis)belief] to talk about. I

wasn’t sure exactly what she wanted, so I changed it to this [Moral consequences of

religious (dis)belief], which is something I’ve been thinking about a lot recently

and running a number of analyses on. Over dinner, I realized that this was a little

irrelevant to what we’re talking about today, and so I woke up at 5 AM and wrote a new talk.

I’m actually going to go with the first talk, but if anybody’s interested in what

the new talk was about, which was about how stereotypes about Christians being unintelligent

and unscientific end up undermining their interest in science and pushing them out of

the field. That’s another topic I can talk about, either in the question-answer period

or you can come find me afterwards. Today I’m going to talk about this question [“Does

God make you good?], because this question seems like one people are interested in. People

have been interested in it for a very long time. Back in the time of Socrates, we know

at least back then they were debating this question. Twenty-five centuries ago, if we

fast-forward to the tail-end of the Enlightenment, you have common arguments being made. This

is Voltaire. He said, “This sublime system is necessary to man. It is the sacred tie

that binds society, the first foundation of holy equity, the bridle to the wicked, the

hope of the just…if God did not exist, it would be necessary to invent him.” We fast-forward

a little further. This is the late great Christopher Hitchens. I like Christopher Hitchens a lot,

but he’s a little off the mark in this statement. So he says, “We keep on being told that

religion, whatever its imperfections, at least instills morality. On every side, there is

conclusive evidence that the contrary is the case and tat faith causes people to be more

mean, more selfish, and perhaps above all, more stupid.” Now he’s right about this

first bit. We keep hearing about this. But he’s wrong about everything after that.

There is not conclusive evidence about religion making people mean or selfish. There are many,

many anecdotes that you can pull to make that claim. But in terms of there being evidence

for this, there is evidence, it doesn’t lean in that direction, it’s not nearly

conclusive. That’s the topic that I’ve undertaken to try to answer, to try to find

an answer to that question—does God make you good? It’s a nice question, because

you have this nice alliteration between “God” and “good”, and they kind of look like

homographs of each other. It’s also a question that’s fraught with ambiguities. It really

matters what you mean by good, and it matters what you mean by God, and it sort of matters

what you mean by make. Let’s focus on this one right now. I think this is possibly the

most important. You do find that if you’re looking at how religiosity affects how people

act morally, there are systematic differences in between what religious and nonreligious

people think is moral. In particular, one of the recent and popular theories on moral

pluralism has been moral foundations theory, which fractionates our moral concerns into

5 categories. Harm and care, so issues of compassion, issues about suffering. Justice

and fairness, that’s another concern. Concerns about authority, respecting people who are

in authority. Loyalty to the group and purity and disgust—another concern for what’s

moral and what isn’t. If you look at how this varies across the religious spectrum,

you see that the first two there’s agreement on. There’s not much variance in how much

religious people versus nonreligious people believe that this is an important concern.

But on the other 3 is where you see the difference. It’s the difference you’d expect if you’re

familiar with research on political conservatism and moral foundations. These results exist

after you control for political conservatism, and in many cases they’re actually stronger

than the political conservative differences between conservatives and liberals. There

is disagreement on what it is to be moral between these groups. Nonreligious people

also—this is work by Jared Piazza—tend to base their morals on utilitarian consequentialism.

So what are the benefits and costs that happen based on the action that you should or should

not enact. They’ll be more likely than religious people to agree to a statement like this:

Should you break a promise if it may lead to a greater good or prevent further suffering?

They’ll say yes, if it’s going to create greater good, sure let’s do that. Religious

people are going to be much more deontic with this. They’re going to feel that they should

not break the rule, even if the consequence of not breaking the rule is greater suffering.

There are these important foundational meta-ethical differences. But there are areas of agreement.

Those are areas in which we can make direct comparisons. Things like generosity or not

cheating, charity, helping, or volunteerism. These are all situations that are not pitting

2 different moral foundations, say freedom vs. equality or justice vs. mercy, against

each other. Instead you’re pitting selfish behavior against non-selfish behavior, or

what you could call prosociality vs. antisocial selfishness. It’s on this measure, prosociality,

this constellation of moral behaviors that we can actually look at differences across

the religious spectrum. So we conducted a meta-analysis of as many of the studies as

we could find that’s been done on this. We pulled this all together. This includes

31 studies, over about 4 decades, with almost 31,000 participants. We included studies if

they had people reporting their religiosity and if they had some measure, either a self-report

measure or a behavioral measure, of that constellation of moral behaviors that we talked about before—prosociality.

The reason that we split these apart is because there are limits to some of these measures.

So self-report measures, I think most of you guys are going to be familiar with the limitations

of asking people how moral they are. Here’s an example. So this is a common self-report

measure. This is the satisfaction of life scale, and if you say that the conditions

of my life are excellent, that you’re satisfied with your life, I will generally take your

word for that. Imagine if you switched this to the satisfaction with yourself scale, and

then you made a claim that I’m a good mother, I’m not racist, I’m charitable and a moral-abiding

person. Those are situations where maybe I don’t want to take your word as much because

you could be self-inflating. You could be overemphasizing how good you are compared

to how good you actually are. So, we also included behavioral measures, and we made

a distinction between these. Behavioral measures are measures that social psychologists and

experimental economists have concocted to measure actual behavior in which there’s

a direct cost for the participant to actually demonstrate that they are pro-social in this

situation. One of these examples is they create a spontaneous helping situations. Someone

will drop a bunch of papers accidentally, and they’ll see how many participants in

the study actually pick them up. Other examples are economic games. A common one is the Dictator

Game. A lot of you will have heard of this Dictator Game. Basically it’s very simple.

You take 2 participants into your lab room. You put them into separate rooms. They don’t

meet each other, they don’t have reputational concern with each other, and then you give

one of them $10. You give the other one, the receiver, nothing. Then you say to the dictator,

the person who got the $10, you can dictate how much each of you gets. The other person

got nothing; do you want to split the $10 that you have? So you could split it 50-50,

which would be the equitable thing to do since you were both randomly put into these roles.

You could be more selfish. You could only give them $1, or you could do what most people

in the game do, which is to give them nothing and keep the $10 themselves. So these are

examples of behavioral measures. When we put all these together…let me just orient you

to this graph. This is a forest plot of the 31 studies that I was talking about. When

the squares and the lines go towards that side, that indicates that there’s a negative

association between religiosity and pro-social behaviors. So religious people in that study

ended up being less pro-social than the nonreligious people. On the other hand, if it’s passing

that line, that means it’s significantly more pro-social for religious people. There’s

a positive relationship between religiosity and pro-sociality. So what you can see here

is there is small but consistent positive relationship, in terms of pro-sociality with

religious people. If you were to break it down between these self-report measures, these

behavioral measures, there you see a difference. There are the self-report measures. It’s

a correlation of about .2, in terms of the self-report measures—questions about how

much do you volunteer, how much do you give to charity. And if you look at the behavioral

measures, they’re actually 0. There’s no hint of an effect. There’s actually a

substantial difference between what is reported when you’re asked how pro-social you are

and, when you test in the lab, how pro-social religious people actually demonstrate they

are. So there are a couple reasons that could explain this discrepancy, as far as I can

think of. One, is the one that might immediately be most obvious, that the self-reports aren’t

actually accurate. You should trust the behavioral measures over the self-reports. What’s happening

here is that religious people are inflating reports of their actual prosocial behavior.

Everybody tends to inflate these. The expectation is that maybe religious people are fudging

a little bit more than nonreligious people. There’s another explanation, which is that

the behavioral task should not be tested. These are not actually reflective of any real

differences that might exist outside of the lab. There may actually be a pro-social religious

advantage that we’re not tapping into with these artificial tasks that we’re looking

at. I think both of these are true to some degree. Let’s start with the first one.

There has been a bunch of work showing that religious people do score higher on measures

of social desirable responding. They tend to indicate that they’re better than they

actually are, more than do nonreligious people. There have been a couple meta-analyses recently.

This one demonstrated that strongly. It makes the case that religious people tend to be

self-enhancing. This paper actually makes it really interesting argument. If you could

think of the intuitive explanation, the one that may have occurred to you guys is something

like religion might make people more self-enhance. Or there could be some sort of personality

factor that makes you both religious or self-enhancing. These guys argue something separate. They

argue something more like this, that it’s actually self-enhancement, the tendency to

enhance yourself, that makes you more religious. They say that there is something called the

self-enhancing personality and that, and I quote, “people strategically use an ingenious

array of means to satisfy the self-enhancement motive…Religion is a prevalent and important

such aspect. People…will be likely to capitalize on it for satisfying the self-enhancement

motive.” So people who are going to want to aggrandize themselves find a way to do

so in religion. They present a bunch of compelling evidence for that. I’m convinced about this

to some degree. Another point that needs to be taken into account—this is not incompatible

with that—is if you ask people if they’re religious and if you ask them how pro-social

they are, it’s not that they may only be inflating how much they’re pro-social. They

may also be inflating how much they say they are religious. We can’t get super accurate,

implicit measures of what people actually believe, but we can see if people inflate

their religious attendance compared to how much they actually do attend religious services.

So this is a recent study. What they did was they asked people how much they attend, but

they asked people in 2 separate ways. They randomly assigned people to be actually asked

in person; face-to-face to somebody to whom they’re going to have to respond, or online

where you’re not actually talking to anybody. The assumption is that people are actually

going to be more honest when they’re revealing this online, but when they’re confronted

with a person they’re going to hedge a bit. That’s what they found. So when you look

at what people say when asked online, you have about 31% saying that they go weekly

or more. When they’re asked by telephone, that number increases to 36% and the percentage

of people who say they never go actually falls from 43% down to 30%. So they’re less willing

to admit when they’re actually directly talking to somebody “Oh, I don’t go to

church.” There have been a number of studies like this. One of my favorites is from the

90s, where they actually just did a head count of everybody who said they went to church

last Sunday and compared that to actually the number of people who were there. They

did this across the United States. This is for a few of the cities. When you actually

ask people how many of you went to church last Sunday, these are the rates you get.

When you actually count how many people showed up, it’s much lower. In some places, it’s

less than half as much. So people do seem to be inflating their actual religious attendance.

What that means is that when we’re measuring what might be a relationship between religiosity

and prosociality, we might actually be measuring a relationship between your tendency to inflate

how much you go to church and your tendency to inflate how much prosocial behavior you

enact in. So really what you’re doing is you’re just finding that people who fudge

on one thing fudge on the other. I guess that’s interesting. Let’s tackle the other possibility,

that our behavioral tasks are not actually telling us what we want them to be telling

us. Here, I think it’s good to consider the situations in which religious people might

actually in the real world have a prosocial advantage. It’s not in these cold, sterile

lab rooms when that would emerge. It’s when they’re actually enmeshed in the community,

or thinking about their religious beliefs, that are actually getting them to engage in

prosocial behavior. So it doesn’t look so much like this, a cold, sterile lab room,