Pray at the Pump

The L.A. Times reports on a group claiming that the recent reduction in gas prices was caused by prayer. “If the whole country keeps on praying, we can bring down prices even more — to even less than $2,” says Rocky Twyman, founder of Pray at the Pump.

If prayer did cause the price to drop, did it do so through a shift in the demand curve or was it through a shift in the supply curve?

My sense is that the article is newsworthy because some people do not believe it is appropriate to pray for reduced prices. But praying for relief from crisis is a classic use of prayer. Georgia’s governor, for example, recently prayed for rain; some people believe that it worked.

Others do not believe prayer to be effective. But several studies have run randomized trials to determine whether prayer (or what is sometimes called “distance healing”) can be effective. For example, a small randomized study in a coronary care unit found:

The therapeutic effects of intercessory prayer (IP) to the Judeo-Christian God, one of the oldest forms of therapy, has had little attention in the medical literature. To evaluate the effects of IP in a coronary care unit (CCU) population, a prospective randomized double-blind protocol was followed. Over ten months, 393 patients admitted to the CCU were randomized, after signing informed consent, to an intercessory prayer group (192 patients) or to a control group (201 patients).

While hospitalized, the first group received IP by participating Christians praying outside the hospital; the control group did not. At entry, chi-square and stepwise logistic analysis revealed no statistical difference between the groups. After entry, all patients had follow-up for the remainder of the admission. The IP group subsequently had a significantly lower severity score based on the hospital course after entry (P< .01). Multivariant analysis separated the groups on the basis of the outcome variables (P<.0001). The control patients required ventilatory assistance, antibiotics, and diuretics more frequently than patients in the IP group. These data suggest that intercessory prayer to the Judeo-Christian God has a beneficial therapeutic effect in patients admitted to a CCU.

There is even a Cochrane Review of single-blind studies (where patients do not know whether others are praying for them). The results are mixed. But across six studies and more than 6,000 patients, there was a statistically significant reduced risk of death. (Good Bayesians are more likely to accept this result if it accords with their prior beliefs — and more likely to reject this result if it conflicts with their prior beliefs. Bad Bayesians who disagree will not even allow it to move their prior an iota.)

The idea of testing for the impact of prayer goes all the way back to 1883, when Francis Galton, the father of regression, sought to test the efficacy of prayer.

Some people have claimed that it is wrong to even try to empirically test the power of prayer. The Bible includes several passages admonishing us: “You shall not test the Lord thy God.”

But I also worry that this is a question that may not be fully resolvable by resorting to randomized study. If there is an omniscient God, then it is impossible to run a truly double-blind study. This creates the possibility for a transcendent kind of Hawthorne effect.

If God behaves differently in response to testing prayers than to non-testing prayers, then we will not learn whether non-testing prayers help (or hurt). On the other hand, if the null hypothesis is that prayer should have no impact, and we find one in patient-blind randomized control trials, then the atheists have some explaining to do.


Compare Europe with the US in regards to medical health and religion. In general far less Europeans are religious (and thus pray less for other people) but in general are far more healthy and live longer lives than Americans.

Now we're talking a big number of test subjects in otherwise similar environments.

Oh just a minor question about the CCU test mentioned: they say the praying patients went outside of the hospital to pray (in groups one assumes). Perhaps getting out of the hospital and talking to other people (in a friendly environment, such as at a church) is helpful? ;)

Daniel Reeves

While hospitalized, the first group received IP by participating Christians praying outside the hospital; the control group did not.

The problem I've always had with this study is that there may be others praying for the people who were hospitalized that weren't assigned to do so, e.g. names placed in church bulletins.

The effects of prayer-- if there are any-- aren't easy to quantify.

How about comparing the life expectancy of Christians and atheists? That could say something.

jim gundlach

Georgia's governor was smart enough to check the weather forecasts before he prayed for rain in public. Alabama's governor was not so smart so his prayer did not work.

Johnny Appleseed

"Where revelation comes into its own is where reason cannot reach. Where we have few or no ideas for reason to contradict or confirm, this is the proper matters for faith...that Part of the Angels rebelled against GOD, and thereby lost their first happy state: and that the dead shall rise, and live again: These and the like, being Beyond the Discovery of Reason, are purely matters of Faith; with which Reason has nothing to do." John Locke

"the doctrine of a personal G-d interfering with natural events could never be refuted... by science, for [it] can always take refuge in those domains in which scientific knowledge has not yet been able to set foot." Albert Einstein


Science is a methodology. Emotionally-driven subjects such as religion, passion, art, or love cannot be measured. You might as well try to express Fermat's Last Theorem through interpretive dance.
This is not to say that religion cannot be challenged by science: where it spills out of churches and into the real world, it must be subjected to the same stern scrutiny as any philosophy, without special treatment or coddling. But religion as an ascetic, experiential state is immune to measurement or testing.

Matt K


Was just about to mention the amputee angle, glad I read all the comments first.

Just Another God

21. said: "Naturally, nobody bothers conducting a study into whether prayer can heal an amputee, because deep down in our hearts we know the truth: there is no God."

So, we know in our hearts that God does not exist because if God existed he grow new legs for amputees at the whim of anyone who decided to pray for it? Sorry, does not compute. I don't believe in God the 'toy' or God the 'puppet' of humanity.

How about: you are God and I am God and everything surrounding us (both seen and unseen) is God because everything is made 'of' God.

Instead of praying for some God 'out there' separate from ourselves to heal someone; maybe sending the infinite and ubiquitous love and light 'of' God to those in need would give them the strength and courage to heal and believe in themselves and the value of their own lives.


Where to begin?

On the subject of praying for rain: Freakonomics blog itself posted an article ( discussing how accurate - or not - weather forecasts are. According to that article, rain forecasts, at one day out, are ~85% accurate. Let's use that result to estimate the plausibility of the light rain on November 15th being a miracle...

For simplicity, I will define a 30 day drought as "30 days during which the one-day forecast is always for no rain". For a 30 day drought, and assuming a binomial distribution (reasonable for the options of rain / no rain) anything from 0 to 8 days with rain would be consistent, at a 95% level, with the 85% accuracy in forecasts (meaning that 95% of 30 day periods with all one-day forecasts of no-rain would have between 0 and 8 days of rain. If you would like to check that result, read up on the binomial distribution, and the gaussian distribution with which I approximated it for ease of calculation) In short, the handful of days in November 2007 in which Georgia received rain provide no basis for believing this is a miracle.

On to the medical effects of intercessory prayer... The abstract copy-pasted in this blog article certainly makes a strong claim. However, in this 1988 article (1988? If the best example is 20 years old, that doesn't speak well for the field of research), a key error is made in the significance estimates. Of 26 variables tracked, 6 show a difference at the 5-7% level. One or two variables showing such a difference would be considered insignificant, meaning it could be chalked up to chance. But 6 variables? If these variables were independent, this could allow for a strong claim. However, the need for diuretics, antibiotics, or intubation and the development of pneumonia, congestive heart failure, and cardiac arrest are not independent variables. Congestive heart failure is a risk factor for cardiac arrest... treatments for all three of these diseases can include diuretics or intubation/ventilation, and antibiotics are commonly called for in treating pneumonia. These sorts of inter-related effects are exactly why we require strong evidence to accept fantastic claims - it's too easy to accidentally get magnified results. And, to boot, an attempt was made to replicate this study, with no success (Harris, Gowda, Kolb 1999)

As for the linked review of research on intercessory prayer studies (Roberts, Ahmed, and Hall 2006), the numbers stated in the abstract linked to don't paint as strong a picture as Ian makes it sound. The NNT, or Number Needed to Treat, indicates the number of people who would need to receive the treatment in order to expect 1 to actually benefit. An NNT of 42 for prayer preventing death is about a 3% effect. From the numbers in this abstract, telling someone they are receiving prayer (Number Needed to Harm of 14) is more likely to hurt them than your prayer is to save them! In fact, Ian makes a stronger claim than do the author's themselves. The authors of this meta-analysis, in their summary, state:

"Most of the studies show no real differences. Prayer was found to be helpful in one study of women receiving fertility treatment whereas another trial found those aware of prayer had more complications following an operation."

Highlighting the need to analyze large numbers of studies as a step toward drawing a conclusion, another meta-analysis, this one with 14 studies (Masters, Spielman, and Goodson 2006) found no statistically significant effect for prayer, either good or bad (btw, the MSG study includes the ten papers in the RAH study, plus four others).

Ian also misunderstands the meaning of double-blind. In a double blind experiment, neither the patient nor the caregiver knows if the patient is receiving the intercession or not. Nothing more, nothing less... the presence of a deity would not change this. His sly suggestion that maybe a deity would just refuse to play along, and not answer any prayers tied to a study... well, who could find such a childish god worthy of worship anyway?

However, I couldn't agree with Ian more when he says that if we find an influence of prayer in double blind experiments, "the atheists" have some explaining to do... if, by "the atheists," he also includes the theistic doctors who practice evidence based medicine. However, the results on intercessory prayer, so far, are that it has no significant effect (when used in this way, significant has a more nuanced meaning than in day to day speech. It means that if there is an effect, it's so small we haven't been able to reliably measure it yet). I'll pit my antibiotics against prayers anyday!

Unfortunately, the skill to interpret statistics is not generally taught in grade school or college, the skepticism to go beyond the abstract and actually look at the work done is not generally taught, access to the science journals is not generally available, it's not easy to find explanations of field-specific jargon (i.e. NNT; prospective randomized double-blind protocol), and who has the time anyway? This is why we have journalists who distill the big picture to us.

Well, that's something I'll keep praying for...



You could really mess the market here. Praying for increases in conflicting commodities or shares would lead to confusion.

Or it could just be utterly stupid and the real economic question is why has the world's leading economy produced people who are so openly willing to waste their time when the economy needs all hands to the pump?

Joe Glenn

@18 : "But if prayer to God works, then the way it works is certainly outside of human understanding"

I disagree. Think of "prayer to God" as only a name that we assigned to the current act & result of the human thought process we call "prayer". From a biological/psychological point of view, the act of prayer is simply brain activity directed to a specified purpose. Granted, our current means of measuring all this and the far-reaching implications it may or may not have are currently inadequate. However, I think relatively soon, we'll have the tools to accurately measure it in a very scientific way and we'll be able to determine if the brain activity we call "prayer" really has the power to help or harm others in a significant way. This is nothing out of reach of human understanding.

The biggest problem here is the religious fervor that gets in the way of science because people associate the brain activity called "prayer" with some omniscient God that may or may not exist and actually has very little to do with the brain activity called "prayer". My personal theory is that a prayer is only as strong as the faith and belief one puts in it. A prayer recited with no faith nor emotion will be powerless but one imbued with extreme faith and positive emotion just might have the power to affect patients in a positive way. I only hope I live long enough to see enough scientific proof and effort in that direction.


Luke Burton

The efficacy of prayer could easily be tested by asking for God to heal an amputee.

For something like petrol prices going down, someone's cancer disappearing, there are innumerable other reasons why these things might have taken place. However, if an amputee who was being prayed for spontaneously grew back his or her leg, that would be unequivocal evidence of a miracle.

Naturally, nobody bothers conducting a study into whether prayer can heal an amputee, because deep down in our hearts we know the truth: there is no God.

We conveniently choose ambiguous cases and claim that God had something to do with whatever change was being prayed for, but when faced with a very clear cut example, nobody dares test it.

There are plenty of very unfortunate amputees out there - Cambodian children, war veterans, honest workers who suffered accidents. Yet God apparently prefers to tweak petrol pump prices and make cancers come and go, rather than giving these people what they truly need: their limbs back.

Believers will turn themselves inside out trying to explain this, but no explanation will be satisfactory for a Cambodian child who no longer has legs.


Chris H.

God answering a prayer would indicate fallibility. When God created the world, he knew how everyone's life would turn out (otherwise, he wouldn't be very powerful, would he?) He knew your aunt would get cancer, so why are you praying? Her fate was determined at the moment of creation.

People really should pray for the strength to accept the future, whatever it is. Asking God for favors is ridiculous.

Not so Fast

@15: But if the rise of Christianity does correlate with disease, pestilence, natural disasters, and military successes what does it really prove? Whose side God is on? Or that Christians are better equiped to survive such events?

Alas, we're stuck with the old "correlation does not imply causation" rule.


Most who read this blog will probably agree that a belief in God, the power of prayer, or that there is truth in religion is decidedly unscientific. Yet here we are discussing the scientific ways of testing prayer? This is laughable. If God does exist or if praying to Him does work, how can we even dream of expecting a scientific test to prove or disprove it? The scientific method and statistical analysis are tools developed by humans. But if prayer to God works, then the way it works is certainly outside of human understanding, and likely to be outside the realm of what our tools and methods can really test for. Why is this concept so difficult for both the religious and non-religious to deal with?

S Clinton

It would be interesting to find out what were the wider reasons for the above experiment. Presumably it was hoped that it would prove/disprove the effectiveness of prayer as a form of medical treatment and thus certain implications about a God could be made.

If they went straight for the wider question of "Is there a God?" then the answer is fairly obvious from a scientific point of view. There is no credible evidence in favour of the theory and most of the surrounding claims cannot be proved or disproved in a satisfactory manner. If it were not for the fact that we as a species have believed in God for so long then there would be no question whatsoever about his existence today.

Mr Kid

@#30 Peter: Problem with that theory is that believers believe that heaven is a better place to be than earth, so maybe they are dying younger because they want to?


I don't have the time to find it right now, but there have been similar blind studies on praying for patients which have showed worse outcomes in the prayed for group. What have we learned from this? There's a high degree of variability in outcomes for patients in the study.

The results aren't reproducible. It's as simple as that. Some studies can be well set up, and will still result in a great deal of noise, some indicating that prayer helps, some with the opposite results.

Additionally, it's impossible to correct for people outside the study contributing their own prayers.


What is the mechanism through which prayer is expected to work? It seems like this is the question being asked in the comments in different ways. Has any work been done on this?


I think the fumes are making them lightheaded, not enlightened.

E Olson

Or no shift in any curves - just a stronger dollar. Can you pray for that too?