PRINCETON NJ -- Let’s take a look at a question Gallup has been asking for a number of years:
At the present time, do you think religion as a whole is increasing its influence on American life or losing its influence?
The percentage of Americans answering "increasing" has skyrocketed since Sept. 11. Last February, only 39% said religion was increasing its influence, a number typical of the response to this question over the years. In our mid-December poll, that number jumped to 71% -- the highest percentage giving the "increasing" answer in Gallup’s history of asking the question.
Our initial temptation is to use this trend to support the hypothesis that the Sept. 11 attacks caused an essential great awakening of American religion, a fundamental change in the way Americans relate to religion in their lives.
But there are several reasons why we should be cautious about this conclusion, reasons that illustrate some of the tricky issues often confronted in analyzing survey data.
For one thing, we don’t know that the change in the long-standing trend question was specifically related to Sept. 11. The time span between the two polls we are using for comparison, February to December 2001, includes Sept. 11, but the change, in theory, could have occurred at some point either before or after Sept. 11. This isn’t likely, of course -- the current change is a very large disruption in what has been a traditionally steady trend over the years, and there is a logical connection between the attacks and the focus of the question. It seems reasonable to assume a causal relationship between the attacks and the change in this question. We don’t (and can’t) know that for sure, however.
But even if we assume a direct relationship between Sept. 11 and the sharp rise in the percentage of Americans saying religion’s influence is increasing, we are left with other important and intriguing issues.
Note that the question doesn’t ask respondents if there has been a change in their own personal religion, but rather if there has been a change in the influence of religion among Americans more generally. So, this question most directly addresses the hypothesis of perceptions of change in religion across the country. If we want to determine whether or not Americans are more likely now than pre-Sept. 11 to think that religion is increasing its influence in society, then the responses give us a direct basis for accepting that hypothesis (or, looked at in a more scientific way, no basis for rejecting the hypothesis).
But relatively few people have been interested in a hypothesis about generalized perceptions of change in religion since Sept. 11. Observers have been much more likely to hypothesize that the actual influence of religion has become greater for individual Americans since the attacks. The "increasing" or "decreasing" question we have been looking at is of less value in addressing that hypothesis. It would be a mistake to make the logical leap that a larger percentage of Americans saying religion’s influence has increased necessarily means that at some individual level Americans have in fact become more religious.
Why? Most Americans can’t empirically assess the amount of change across the country, so they make assumptions. These assumptions are based on what they read and hear, primarily in the news media. Responses to the perception question, in other words, depend on the reports of others, not personal experiential data.
That’s not to say that big picture perceptions aren’t valuable in some situations. Americans’ perceptions about the crime rate, for example, may have some important implications, whatever the reality of the situation (as measured by crime statistics). If the public perceives that it is dangerous to go out at night, or that one’s household is likely to be burglarized, then there will be real consequences, regardless of changes in the actual crime rate. Similarly, if the public perceives that the economy is heading into a recession, then the consequences on retail spending behavior can be very real, whether or not the "hard" economic statistics bear that out.
In the same vein, it may be valuable to know whether or not the public thinks that religion is increasing or decreasing its influence in American society -- for a variety of reasons. The public might make more religious contributions, for example, if it is widely perceived that religion has lost its influence in society. Sales of religious books, church attendance, and volunteerism may vary depending on the perceptions of the influence of religion in society. The responses to a "perception" question can also be valuable when studying the processes of the development of social consensus in a society. Many social entities, from small towns to nations, develop shared perceptions of reality that take on a life of their own, and that can make a considerable difference in how people live their lives. Questions about perceptions help social scientists measure these processes.
But if we are trying to test a direct hypothesis about individual change in religiosity, the measurement of perception of change in religion’s influence in America is certainly not, by any means, our best measure to use.
The better way of discovering individual religious change is simple -- asking Americans to report directly on their own religious behavior and attitudes.
Luckily, we have the capability of doing just that. Gallup, for many years, has been asking the following questions about individuals’ religion:
- Did you, yourself, happen to attend church or synagogue in the last seven days, or not?
- How important would you say religion is in your own life -- very important, fairly important, or not very important?
Unlike the dramatic changes we have found in the "perceptions" question, our trend lines on the responses to the two questions listed above show very little, if any, change. Over the years, between 40% and 45% of Americans have generally answered in the affirmative to the first question, and although there was a slight increase just after Sept. 11, the percentage saying yes in two subsequent surveys has settled right back down to where it was. In similar fashion, the percentage saying "very important" in response to the second question went up slightly just after Sept. 11, but is now right in line with the pattern of responses we obtained in the months and years prior to the terrorist attacks.
So, what do we know? Our best available data, in summary, suggest that Americans are currently much more likely to perceive that religion is increasing its influence in society. However, asking Americans about their own personal religious behavior and attitudes shows little direct evidence to suggest that this is in fact the case. Additionally, as is always true, we see again how important it is to focus on the precise concept being measured when we examine the implications of specific survey questions.