THE EFFECT OF PRAYER ON GOD’S ATTITUDE TOWARD MANKIND
JAMES HECKMAN*
This paper uses data available from the National Opinion Research Center’s
(NORC) survey on religious attitudes and powerful statistical methods to
evaluate the effect of prayer on the attitude of God toward human beings. (JEL
C21, Z12)
I. INTRODUCTION where f#(x) is the derivative of f(X ) at X 5 x.
This article uses data available from the Thus, from the population distribution of
National Opinion Research Center’s survey prayer, we can estimate the population regres-
on religious attitudes and powerful statistical sion function of God’s attitude as a function of
methods to evaluate the effect of prayer on the prayer. For a derivation, see Singh (1977).
attitude of God toward human beings.
The technique—due to Singh (1977)—is II. EMPIRICAL APPLICATION
briefly described here. Let Y be God’s attitude
arrayed on a scale ranging from 0 to 1. This is Greeley estimates that f(X) is bimodal for X
an unobserved variable. Let X be the intensity scaled between 0 and 1. Many people almost
of prayer in the population. It too is scaled never pray and many others pray a lot. Using
between 0 and 1. The population density of Greeley’s numbers, we reach the following
prayer is summarized by a univariate density important conclusions reported in Table 1,
f(X ), which has been estimated by Father where ‘‘t’’ statistics for the Parzen (1962)
Greeley (1972). kernel estimator of f(X) are presented in
Accept on faith that the conditional density parentheses. See Figure 1, which graphs the
of X given Y is of the form: estimated relationship at points of evaluation.
The empirical conclusion from this analysis is
ð1Þ gðX jY Þ 5 aðY ÞexpðXY Þ; important. A little prayer does no good and may
make things worse. Much prayer helps a lot.
where a(Y) is an unknown, continuous, posi-
tive, and differentiable function. Singh dem-
III. DISCUSSION
onstrates that under his conditions:
Using a powerful method due to Singh, we
f #ðxÞ have established a relationship between God’s
EðY jX 5 xÞ 5 ;
f ðxÞ attitude toward man and the amount of prayer
TABLE 1
*The first draft of this article circulated in July 1980. It
evoked a response from Father Andrew Greeley, the noted
Estimated Regression
Catholic priest and sociologist, which I attach. Ivan E(Y |X = x) t statistic SE
Werning made helpful comments on the current draft.
Heckman: Henry Schultz Distinguished Service Professor of x5 0.0 0.4 (3.2) 0.1
Economics, Department of Economics, University of x5 0.2 1.3 (4.6) 0.3
Chicago, 1126 East 59th Street, Chicago, IL 60637. Phone x5 0.4 2.3 (6.7) 0.3
(773) 702-0634, Fax (773) 702-8490, Email jjh@uchicago.
edu; Professor of Science and Society, UCD Geary Insti- x5 0.6 3.0 (9.3) 0.3
tute,UniversityCollegeDublin,Belfield,Dublin4,Ireland; x5 0.8 0.0 (2.1) 0.0
ResearchProfessor,AmericanBarFoundation,750North x5 1.0 2.0 (1.8) 1.1
Lake Shore Drive, Chicago, IL 60611.
234
Economic Inquiry doi:10.1111/j.1465-7295.2009.00224.x
(ISSN 0095-2583) Online Early publication June 2, 2009
Vol. 48, No. 1, January 2010, 234–235 © 2009 Western Economic Association International
HECKMAN: THE EFFECT OF PRAYER ON GOD’S ATTITUDE TOWARD MANKIND 235
FIGURE 1
Plot of Estimated Regression
transmitted to God. The method presented time but the reaction, thus far, has always been that She
here is applicable to a number of important was not home. Econometric models are much neater!
I’m enclosing a copy of a brief memo of mine about the
problems. Provided conditional density (1) is persistence of prayer. My colleague, Phil Morgan, has
assumed, we do not need to observe a variable developed some log linear models which show that prayer
to compute its conditional expectation with increases with age because, the older people are, the more
respect to another variable whose density likely they are to believe that prayer works. This may be
wishful thinking, but the fact that it’s wishful thinking
can be estimated. For example, one can extend doesn’t necessarily mean that it’s not true!
current empirical work in a variety of areas of And the attitude that prayer works seems to be a life
economics to estimate the effect of income on cycle rather than a generational or cohort effect.
happiness or the effect of income inequality on The next thing to inquire about is whether the saints
democracy. We conjecture that this powerful are listening!
AMG/jk
method can be extended to the more general cc: Richard Robb.
case when X is not observed either.
APPENDIX
Appendix from Father Andrew Greeley (personal
communication, 1983) in response to this article REFERENCES
DATE October 10, 1983 Greeley, A. M. Unsecular Man: The Persistence of
Religion. New York: Schocken Books, 1972.
TO James Heckman Parzen, E. On Estimation of a Probability Density Func-
FROM Andrew M. Greeley tion and Mode. Annals of Mathematical Statistics,
SUBJECT 33, 1962, 1065–76.
Richard Robb showed me your interesting essay about Singh, R. S. Applications of Estimators of a Density
the effect of prayer on God’s attitude. I envy you your and Its Derivatives to Certain Statistical Problems.
ingenuity of economic models. Our survey interviewers Journal of the Royal Statistical Society: Series B
have been trying to get an interview with God for a long (Methodological), 39, 1977, 357–63.
