The Harris Poll^® #59, October 15, 2003

While Most Americans Believe in God, Only 36% Attend a Religious Service Once a Month or More Often

Belief and attendance vary greatly among different segments of the population

_____________________________________

by Humphrey Taylor

Americans are far more likely to believe in God and to attend religious services than people in most other developed countries – particularly countries in Europe where philosophers have written that "God is dead." However this new Harris Poll finds that underneath a broad consensus, belief in God varies quite widely among different segments of the American public. And most people attend a religious service less often than once a month.

These are some of the results of a survey by Harris Interactive^® based on a nationwide sample of 2,306 adults surveyed online between September 16 and 23, 2003.

Belief in God and Attendance of Religious Services

This survey found that 79% of Americans believe there is a God, and that 66% are absolutely certain this is true. Only 9% do not believe in God, while a further 12% are not sure.

While most people (55%) attend a religious service a few times a year or more often, only a minority of the public (36%) attends a religious service once a month or more often, with about a quarter (26%) attending every week.

Reducing "Social Desirability" Bias

These numbers – for belief in God and for attendance at churches, synagogues and mosques – are lower than those reported in many other surveys, we believe, because of the methods we use to measure them more accurately.

One of the problems with surveys where people are interviewed by people, whether face-to-face or by telephone, is that they may not tell the truth to an interviewer, if the truth is embarrassing or if another answer is more "socially desirable." This "social desirability" bias means that many surveys underreport the number of people who are homosexual, who don’t bathe or clean their teeth, who drink alcohol, or whose children are not immunized, for example. Socially desirable behavior, such as giving to charity, voting in elections and going to church are usually over reported.

Because our online surveys, such as this one, do not involve talking to interviewers, we regularly record lower levels of behavior (and belief) on topics where there is a "socially desirable" answer. We believe that the lower levels of belief in God, and the lower levels of church-going found in this survey are more accurate than the higher levels reported in telephone and in-person surveys.

Differences in the Replies of Catholics, Protestants and Jews

Protestants (90%) are more likely than Roman Catholics (79%) and much more likely than Jews (48%) to believe in God. Religious affiliation here includes many people raised as members of a religion or religious group, regardless of what they practice or believe now.

Protestants (47%) are also more likely than Catholics (35%) to attend church once a month or more often. Only 16% of Jews go to synagogues once a month or more often.

Other differences

Belief in God is higher in the Midwest (82%) and in the South (82%) than in the East (75%) and the West (75%). It tends to increase with age from 71% of those aged 25 to 29 to more than 80% for the three age groups of people over 40, including 83% of those aged 65 and over.

Women are more likely than men to believe in God (84% versus 73%). African Americans (91%) are more likely to believe in God than Hispanics (81%) and whites (78%). Republicans (87%) are more likely to believe in God than Democrats (78%) and Independents (75%). Those with no college education (82%) are more likely to believe in God than those with postgraduate education (73%).

Church attendance (every month or more often) is higher in the Midwest (45%) and the South (40%) than in the East (30%) and the West (27%). It is lowest among people aged 25 to 29 (24%) and highest among those aged 65 and over (43%). And it is higher among women (41%) than among men (31%).

Humphrey Taylor is the chairman of The Harris Poll^®, Harris Interactive.

TABLE 1

BELIEF IN GOD AND CERTAINTY OF BELIEF

"Are you . . . ?"

Base: All Adults

	Total	RELIGIOUS AFFILIATION
	Total	Catholic	Protestant	Jewish	Atheist/ Agnostic
	%	%	%	%	%
Believe in God (NET)	79	79	90	48	15
Absolutely certain that there is a God	66	63	81	24	4
Somewhat certain that there is a God	12	16	9	24	11
Believe there is no God (NET)	9	8	4	19	52
Somewhat certain that there is no God	5	4	2	13	28
Absolutely certain that there is no God	4	4	2	5	23
Not sure whether or not there is a God	12	13	6	33	33

TABLE 2

DEMOGRAPHIC VARIATION IN BELIEF IN GOD

Base: All Adults

		Believe in God (Absolutely or Somewhat Certain)	Absolutely Certain
All Adults	%	79	66
Region
East	%	75	59
Midwest	%	82	71
South	%	82	72
West	%	75	61
Age
18 – 24	%	73	57
25 – 29	%	71	55
30 – 39	%	76	62
40 – 49	%	81	69
50 – 64	%	82	74
65 +	%	83	72
Sex
Male	%	73	60
Female	%	84	72
Race/Ethnicity
White	%	78	64
African American	%	91	82
Hispanic	%	81	66
Party I.D.
Republican	%	87	76
Democrat	%	78	65
Independent	%	75	63
Education
High school or less	%	82	72
Some college	%	77	64
College graduate	%	78	63
Post graduate	%	73	53

TABLE 3

FREQUENCY OF ATTENDING RELIGIOUS SERVICES – BY RELIGION AND RACE

"Do you attend religious services?"

Base: All Adults

				Religious Affiliation					Race/Ethnicity
		Total		Catholic		Protestant		Jewish		White		African American		Hispanic
		%		%		%		%		%		%		%
Every week or more often/Once or twice a month (NET)	36		35		47		16		35		41		37
Every week or more often	26		26		33		5		25		30		23
Once or twice a month	11		9		15		11		10		11		14
A few times a year/Once a year (NET)	23		29		24		42		23		25		28
A few times a year	19		26		20		34		19		24		24
Once a year	4		4		4		8		4		1		4
Less often/Never (NET)	28		35		28		42		29		29		25
Less often	16		19		17		13		16		15		15
Never	13		17		11		29		13		13		10
Not sure	*		*		1		*		*		*		1
Not a member of a religion	12		-		-		-		13		4		10

TABLE 4

FREQUENCY OF ATTENDING RELIGIOUS SERVICES – BY REGION AND SEX

"Do you attend religious services?"

Base: All Adults

		Region				Sex
	Total	East	Midwest	South	West	Male	Female
	%	%	%	%	%	%	%
Every week or more often/Once or twice a month (NET)	36	30	45	40	27	31	41
Every week or more often	26	22	29	30	19	22	29
Once or twice a month	11	7	16	10	8	9	12
A few times a year/Once a year (NET)	23	28	19	20	25	24	22
A few times a year	19	23	16	18	22	20	19
Once a year	4	5	3	3	3	4	3
Less often/Never (NET)	28	34	21	27	31	30	26
Less often	16	17	11	18	15	16	15
Never	13	16	10	9	17	14	12
Not sure	*	*	1	*	*	1	*
Not a member of a religion	12	8	12	13	16	14	11

TABLE 5

FREQUENCY OF ATTENDING RELIGIOUS SERVICES – BY AGE

"Do you attend religious services?"

	Total	Age
	Total	18 – 24	25 – 29	30 – 39	40 – 49	50 – 64	65 +
	%	%	%	%	%	%	%
Every week or more often/Once or twice a month (NET)	36	37	24	36	33	38	43
Every week or more often	26	29	16	24	19	28	35
Once or twice a month	11	8	8	12	15	10	7
A few times a year/Once a year (NET)	23	19	28	23	22	24	25
A few times a year	19	15	24	20	19	21	20
Once a year	4	5	5	3	3	3	4
Less often/Never (NET)	28	24	32	25	33	29	27
Less often	16	15	17	14	15	18	15
Never	13	9	16	11	19	11	12
Not sure	*	*	*	*	1	*	*
Not a member of a religion	12	19	16	16	10	9	6

Methodology

The Harris Poll^® was conducted online within the United States between September 16 and 23, 2003 among a nationwide cross section of 2,306 adults. Figures for age, sex, race, education and number of adults in the household were weighted where necessary to bring them into line with their actual proportions in the population. "Propensity score" weighting was also used to adjust for respondents’ propensity to be online.

In theory, with probability samples of this size, one could say with 95 percent certainty that the results have a statistical precision of plus or minus three percentage points of what they would be if the entire adult population had been polled with complete accuracy. Unfortunately, there are several other possible sources of error in all polls or surveys that are probably more serious than theoretical calculations of sampling error. They include refusals to be interviewed (non-response), question wording and question order, and weighting. It is impossible to quantify the errors that may result from these factors. This online survey is not a probability sample.

These statements conform to the principles of disclosure of the National Council on Public Polls.

____________________________________________

W19566
Q705, Q735, Q740