The use of social data in the evangelization of Europe: methodological issues.

Paas, Stefan

From the very beginning of the modern missionary movement, extensive data have been collected and published regarding the numbers of new church plants, workers in the field, demographics, and the like. With the West increasingly being seen as a mission field, it has become an area for collecting mission statistics as well. This is particularly true for Europe. Today the websites of virtually all organizations and denominations concerned with the evangelization of Europe contain a large amount of quantitative information about European countries, with varying degrees of accuracy. We read, for example, that "in many European countries less than five percent of the population attend any church," (1) that Portugal had 3.0 percent evangelicals in June 2010, (2) or that 40.4 percent of Ukraine's "People Groups" are "unreached." (3)

Most of these statements are presented without clearly indicating the sources of the information, delineating the research methodology followed, or defining the core concepts employed. They present supposedly "hard" data but without explanation or theoretical framework. Obviously, most of these surveys, with their maps, descriptions of unreached people groups, statistics, and definitions of "felt needs," are not primarily meant as social research data but as "mobilization rhetoric." (4) Although they point us in the right direction (collecting social data is imperative in any missiological analysis), they are virtually worthless as serious research.

Fortunately, there are reliable data collections that missiologists can use. In fact, Europe (especially western Europe) is among the best-researched areas of the world in terms of religion. These collections contain a large amount of accessible data, collected according to the highest social-scientific standards. This does not mean, however, that they have no problems. There are important methodological issues to be considered in using these databases in the context of efforts to evangelize Europe, some of which I discuss here.

Resources

Several social databases address religion, especially Christianity, in Europe. Probably the most complete and easiest to access are the combined data of the European Values Study (EVS) and the World Values Survey (WVS). (5) These studies consist of a longitudinal series of quantitative surveys, from 1981 to 2008. They have been severely criticized, however, because of their lack of clear theoretical orientation and poor framing of questions. Moreover, the wording of some questions in consecutive EVS surveys has been changed, rendering it sometimes difficult to compare answers from different years.

A second source of information is the database of the International Social Survey Programme (ISSP), starting in 1984 with four founding members. (6) Currently forty-five member countries are involved (twenty-seven of them European). ISSP has conducted three surveys on religion in Europe: in 1991, 1998, and 2008. Although this survey has stricter social science controls, as well as concern for continuity of questionnaire design, it is limited to a fifteen-minute module in a larger survey.

Another important and recent database is produced by the European Social Survey (ESS). This biennial effort started in 2002 and has continued in 2004, 2006, 2008, and 2010. (7) The ESS claims to have higher scientific standards and to use more flexible questionnaires than other surveys.

Besides these international surveys, most countries have their own social research institutes. Census data and, in some cases, denominational records provide additional information.

How Hard Are "Hard" Data?

Collections of quantitative data may appear to be "neutral" pools of data, needing only to be interpreted by sociologists or missiologists. This is, of course, not true. The databases mentioned above have been composed and funded with specific aims, such as supervision of European integration. Their selection of questions is not meant to further the cause of Christian mission. When observed carefully, they also appear to be less "factual" and obvious than may at first be apparent. In fact, there are considerable methodological problems, giving us good reason to be reluctant to trust these data blindly, regardless of how responsibly and transparently they have been collected.

For example, data from different surveys are often inconsistent or even contradictory. For cross-national surveys a specific problem is the different way in which the same question may be understood in various countries because of cultural differences. (8) It is also important to remember that most surveys have relatively small sample sizes (around 1,000 people per country). Conclusions about minority groups or certain age cohorts therefore tend to be based on very small response groups, which affects their reliability? (9)

Here I consider three methodological issues that are relevant for Christian mission: (1) the definition of core concepts, (2) the comparison of data from different periods, and (3) the phrasing of questions.

Definition of core concepts. Most questionnaires cluster their questions under headings such as "Perceptions of Life," "Family," "Religion and Morale," and "National Identity" (EVS/WVS). This reflects the common assumption that the realm of religion can somehow be isolated from the rest of human social life. People are defined as religious when they say they believe in God, attend church services, pray, and so forth. However, singing the national anthem with tears in your eyes or paying high entrance fees and painting your face in order to watch your favorite soccer team play is not counted as religious activity.

This is a notoriously difficult issue, generating much discussion among scholars who study religion. From a social-scientific perspective, the advantage of having a clear concept of religion is obvious. It helps us to formulate scientific problems, and it allows us to register social change as, for example, the secularizing of society. There are also political interests involved, such as the question of whether certain groups and organizations may be considered religions and be protected by the laws of religious freedom. (10)

Social theorists typically distinguish substantive and functional definitions of religion. According to the former, religions are unique in having certain contents (usually the belief in extraordinary, or "supernatural," phenomena). Functional definitions describe certain effects that religion has for individuals or society. (11) Substantive definitions are often narrower than functional definitions, although this depends on the number of functions a scholar would categorize as typically religious. In any case, the questionnaires mentioned seem to assume an implicitly substantive definition of religion, which is distinct from the nonreligious domains "national identity" and "family." Such distinctions will always be debated, as will the juxtaposition of "religion" and "morale."

From a missiological point of view, it may be more fruitful to employ a functional definition of religion, which allows us to do justice to the traditional Christian belief--at least since Augustine--that religion is not just a matter of having certain convictions and habits but, more broadly, is also an orientation toward anything that ultimately concerns us. According to Christian philosopher James Smith, religious institutions are those "that command our allegiance, that vie for our passion, and that aim to capture our heart with a particular vision of the good life." (12) They seek to provide more than merely entertainment or an education; they desire to make us into certain kinds of people. A specifically Christian critique of culture or of idolatry may develop from such a definition.

Functional definitions suggest that changes in religious orientation among populations pertain not so much to the disappearance of religion as to its transformation. These sweeping definitions, however, tend to blur our focus on real differences, not only between separate social domains (such as the church and the soccer stadium), but also between "then" and "now." Even if we would admit that modern Europeans are not altogether "irreligious," still there are huge differences between the general practice of religion today and that of, say, eight hundred--or eighty--years ago.

People who do not believe in God or go to church are usually unhappy being labeled (functionally) religious. Some social scientists would say that using "religious" for such people is a form of "concept-imperialism," since it does not respect the self-understanding of agnostics and atheists. (13) While this criticism may be valid in the context of social science, which operates with a social perspective, it is a different story in the context of theology, whose practitioners must be granted the privilege of using a "religious" lens in their research. After all, most sociological definitions of religion also would not match the self-understanding of believers. Concepts like "worship," "idolatry," and "liturgy" belong to the toolkit of theology, and no one can blame theologians for applying these terms to any group or institution, regardless of whether it considers itself to be religious. Just as social scientists would counter theological objections to their analysis of religion by referring to the undeniably social dimensions of religion, so too theologians can respond to nonbelievers who protest against their theological analysis by saying that no one should be blind to the specifically religious aspects of human behavior--even if participants themselves fail to acknowledge them.

Another issue is the definition of a "Christian." For evangelicals, this is a most crucial matter. The problem here is not that sociologists are less capable than theologians in determining whether someone is a Christian. It is true that a sociological definition of the word "Christian" is based on external criteria, such as church attendance or affirmation of certain beliefs. Compared with theological or confessional definitions of a Christian, these sociological approaches may seem superficial. However, as soon as our theological definitions must be applied in real life, theologians likewise find that they need to focus on observed behavior. According to one's definition, a Christian may be someone who shows the fruits of the Spirit in his or her life. But how should one determine whether a specific person--for example, one who asks to be admitted to the sacrament of baptism--is a Christian? In this case the church needs to define what these fruits are (e.g., love, patience, gentleness) and whether these fruits are actually being displayed. Most theologians would say that only God knows the quality of one's heart, even as we need some external criteria by which we can establish whether someone may receive the sacrament or be elected to church leadership. Which brings us back to observable criteria of beliefs and behavior. In short, there is no privileged inner perspective on Christianity that is accessible only to theologians. Although both theologians and sociologists may formulate abstract doctrines and theories, both also must deal with data that are epistemologically equal--that is, equally observable.

This means that we should not take issue with the mere fact that sociologists look at people and their religion "from the outside." We all do. It is important, however, to discuss the kind of criteria we use and the kind of questions we ask. The available databases contain only the most general questions, such as "Do you believe in God?.... Do you attend religious services on a regular basis?.... Do you pray?" or "Were you raised with religious values?" Evangelicals would also be interested in more specific questions, such as "Do you believe that you are justified by grace alone?" (14)--provided, of course, that people understand the terms of such a question.

So it is important to note that the available surveys of religion in Europe (such as EVS and ISSP) show roughly how many people are Christians (in a nonconfessional sense), and also how seriously they take their own Christianity, but they do not tell us about their stance toward theological differences, such as the one between Roman Catholicism and the churches of the Reformation. A decision whether Catholic or Orthodox countries are legitimate "mission fields" for evangelical missionaries, for instance, cannot be made solely on the basis of these data. In the end, this is a theological decision, although it can and should be informed by reliable social information.

Comparing data from different periods. Some theologians are frustrated at sociologists for wording their survey questions exactly the same year by year. Social scientists defend this practice by pointing at the necessity of comparing data from different periods. If in 1980, for example, 45 percent of a certain population answered "yes" to the question "Do you believe in God?" and in 2010 only 35 percent of the population did so, we conclude that this belief has dropped 10 percent in thirty years. Or has it? The conclusion assumes that respondents in 2010 understood this question in exactly the same way as respondents in 1980. Some scholars would reject this conclusion, because it does not take into account that religious traditions are continually changing.

The same applies for cultural interpretations of the word "God." People now may perhaps feel more hesitation in answering "yes" to a question whether they believe in God, not because they have become atheists or agnostics, but because they no longer feel that the word "God" is adequate to describe whom or what they believe in. These surveys perhaps record a shift from belief in the traditional Western (theistic) concept of God as a "supercause" or a heavenly guardian toward a more diffuse and vague awareness of God, an awareness people find difficult to express in words. (15)

An analogous issue is belief in heaven. In evangelical circles traditional concepts of heaven and the afterlife have recently been criticized as unbiblical and Platonic. (16) If these views influence thinking Christians to such an extent that they begin to answer "no" on questions like "Do you believe in heaven?" outsiders may see this development as evidence of the demise of orthodoxy. The opposite, however, may be true--at least from an insider's point of view.

Seen from this perspective, what is measured is the extent to which people affirm or deny traditional statements of church doctrines, but not the possible changes in their own religious life or their corrections of doctrine. In this way, there may be a hidden similarity between conservative theology and the sociology of religion as it is usually conducted. Both record only what is disappearing, measured against the standards of the past. Newer theological reforms fly under their radar or are dismissed as changing the rules of the game. Sociologists would counter that they do in fact conduct research into contemporary religious attitudes and "new spirituality." Such a response may not be adequate, for the new religious expressions are usually very different from the church-bound Christianity of the past.

A related problem is the change in the meaning or status of certain (religious) institutions over the years. For example, in comparing baptism figures across both time and place in Europe, we must be aware of the changing perceptions of this practice. (17) In the more secular parts of Europe, baptism is increasingly seen as initiation into a voluntarist organization rather than as a badge of national identity. (18) Over the generations, people's understandings of rites and institutions like baptism, confirmation, marriage, citizenship, and nation change, which complicates considerably the task of interpreting data from different time periods.

The phrasing of questions. A third methodological fault line is the question of what exactly is being measured. This question has several aspects.

Conclusions as to the religiosity of Europeans are obviously based on their own answers to questionnaires and interviews. But how should we understand these answers? For example, since these surveys usually do not qualify terms like "God," "heaven," or "pray," it is impossible to establish exactly what a person means if he or she answers affirmatively the question "Do you believe in heaven?" Is this person an orthodox Christian (who furthermore believes that it is, indeed, orthodox to believe in heaven)? someone who has never seen a church from the inside? or perhaps someone who has just read The Discovery of Heaven by novelist Harry Mulisch? (19) In other words, these databases need backup by qualitative research, allowing us to probe further what people mean when they say they believe in "heaven."

Also, it is well known that people's answers tend to be influenced by the cultural status of Christianity (or any other religion) in their country and by general expectations regarding religious behavior as related to their national or class identity. Do high rates of religious beliefs and behavior tell us more about the status of religion in a particular country than about the actual beliefs of people in this country? For example, research among North Americans indicates that self-reporting of church attendance may be as much as twice as high as one's actual church attendance. (20) This discrepancy occurs most among those who consider themselves regular churchgoers. It seems, therefore, that the traditionally high church attendance rates in the United States may tell us more about how Americans want to see themselves than about what they actually do. (21)

This phenomenon raises questions also about religious data from the past. Could historic recordings of church attendance and Christian beliefs have been exaggerated because of a possibly greater tendency in those days to conform? What is being measured now may still be a genuine decrease in church membership and a diminishing of traditional Christian beliefs, but the numbers are perhaps influenced by the extent to which large groups of church members nowadays feel free to admit that they do not really belong or believe as they might feel they are supposed to--feelings that may have been more disguised or suppressed in the past. And could the reverse be true in countries where religious life stands under cultural suspicion or is even persecuted? Such reversals of context likely help to explain the sudden surge of religious activity reported in most post-Communist nations after 1989. Also it is conceivable that many people in thoroughly post-Christian countries are reluctant to call their activities religious because of the unpopular associations of the term, although an external observer might be able to see clearly religious dimensions in their behavior. (22)

A final issue of measuring involves the influence that the wording of questions has on people's responses. Here are two examples. One relates to the religious behavior of young people. When asked whether they pray to "God," only 11 percent of European teenagers surveyed answered affirmatively. (23) Asked in another survey whether and how they pray, without pushing them in a certain direction, 30 percent of those questioned mentioned "God" spontaneously. (24) Apparently, using the word "God" in the question influenced (or intimidated) the respondents considerably.

Another example: in the Netherlands two large statistical agencies track the religious behavior of the Dutch. One, the Centraal Bureau voor de Statistiek (CBS), reports that 60 percent of the Dutch are members of a church, whereas the other, the Sociaal-Cultureel Planbureau (SCP), puts the figure at 40 percent. This huge difference (more than 3 million people!) can be explained by the questions used. The CBS asks a single question: "To which of the following denominations do you belong?" with "none" being one of the options. The SCP uses two questions: "Do you belong to a denomination?" and "If yes, to which one?" Merely to change the query into two separate questions causes a reduction in church membership in the Netherlands by one-third! This difference probably reflects the large number of nominal church members who do not value their membership very much and who do little to practice it. (25) The fact that a simple variation in the presentation of questions exerts such a great influence on respondents' answers gives much food for thought. (26) This factor likely explains the frequently contradictory results produced by different surveys in the same country.

Conclusion

Within a missiological framework, these remarks may suffice to create at least some restraint in taking seemingly "hard" sociological data at face value. The data can help give us an impression of long-term trends, but they are less reliable when taken as giving us an exact picture of the contemporary religious condition of a given country. Although collections of quantitative data may seem impressive, it is clear that they must be complemented by qualitative research.

Notes

(1.) See www.geocities.ws/demosvalera/about_cai.htm. Put in this imprecise way, the statement conveys very little useful information.

(2.) See www.operationworld.org/port; also Jason Mandryk, Operation World: The Definitive Prayer Guide to Every Nation, 7th ed. (Colorado Springs, Colo.: Biblica Publishing, 2010), 692 (for discussion of the statistical base point used, see xxviii).

(3.) See www.joshuaproject.net/countries.php?rog3=UP; click on "People Progress." Joshua Project, a ministry of the U.S. Center for World Mission, is a "research initiative seeking to highlight the ethnic people groups of the world with the fewest followers of Christ" (www.joshuaproject.net/joshua-project.php).

(4.) "The danger in this new age of the 'infomercial' is that the lines between reliable research and mobilization rhetoric are sometimes blurred," Gary R. Corwin, "Sociology and Missiology: Reflections on Mission Research," in Missiology and the Social Sciences: Contributions, Cautions, and Conclusions, ed. Edward Rommen and Gary Corwin (Pasadena, Calif.: William Carey Library, 1996), 27.

(5.) The two surveys can be found at www.europeanvaluesstudy.eu and www.worldvaluessurvey.org.

(6.) See www.issp.org.

(7.) See www.europeansocialsurvey.org.

(8.) For an elaborate discussion of these methodological problems, see Frane Adam, "Social Capital across Europe: Findings, Trends, and Methodological Shortcomings of Cross-National Surveys" (paper prepared for the Wissenschaftszentrum Berlin fur Sozialforschung, 2006).

(9.) See Grace Davie, Religion in Modern Europe: A Memory Mutates (Oxford: Oxford Univ. Press, 2000), 12.

(10.) See, for example, the dispute in Germany over the legal status of the Church of Scientology.

(11.) For a discussion of substantive and functional definitions, see Inger Furseth and Pal Repstad, An Introduction to the Sociology of Religion: Classical and Contemporary Perspectives (Aldershot: Ashgate, 2006), 16-22.

(12.) James K. A. Smith, Desiring the Kingdom: Worship, Worldview, and Cultural Formation (Grand Rapids: Eerdmans, 2009), 90. Smith mentions the shopping mall, the nation, and the university as such religious institutions.

(13.) Furseth and Repstad, Introduction, 22.

(14.) Corwin, "Sociology and Missiology," 23, makes this point in discussing the question whether supposedly Christian countries like France, Greece, or Argentina are legitimate foci for evangelical missions. See also J. Andrew Kirk, Mission under Scrutiny: Confronting Current Challenges (London: Darton, Longman & Todd, 2006), 50: "When a question is phrased in terms of 'do you believe in God?' it still usually evokes a positive response .... Were the question to be, 'What difference does God make to the way you bring up your children?' Or, 'how does God influence the way you do your job?' I suspect that an honest answer in most cases would be, 'He doesn't.'"

(15.) This shift has been defended by, among others, the Dutch Roman Catholic theologian Anton Houtepen. See his God: An Open Question (London: Continuum, 2002) and "De secularisatie in het westen en de 'Terugkeer van God': Hoe God verdween uit de sociologie," Religie & Samenleving I (2006): 65-85.

(16.) See, for example, Paul Marshall, Heaven Is Not My Home: Living in the Now of God's Creation (Nashville: Word Publishing, 1998), and N. T. Wright, Surprised by Hope: Rethinking Heaven, the Resurrection, and the Mission of the Church (New York: HarperCollins, 2008).

(17.) Bernice Martin, "Beyond Measurement: The Non-quantifiable Religious Dimension in Social Life," in Public Faith: The State of Religious Belief and Practice in Britain, ed. Paul D. L. Avis (London: SPCK, 2003), 1-18.

(18.) For this discussion, see Peter Berger, Grace Davie, and Effie Fokas, Religious America, Secular Europe? A Theme and Variations (Aldershot: Ashgate, 2008), 115-16.

(19.) Harry Mulisch, The Discovery of Heaven: A Novel (London: Penguin Books, 1998).

(20.) C. Hadaway, Marler Kirk, Penny Long, and Mark Chaves, "Over-Reporting Church Attendance in America: Evidence That Demands the Same Verdict," American Sociological Review 63 (1998): 122-30; and Tom W. Smith, "A Review of Church Attendance Measures," American Sociological Review 63 (1998): 131-36.

(21.) This topic is in itself a very interesting qualitative research subject. Even if from this overreporting it would appear that Americans do not attend church much more than Europeans do, it is clearly considered "normal behavior" to do so in the United States (see Berger, Davie, and Fokas, Religious America, Secular Europe? 42-43).

(22.) On these conceptual difficulties, see Nancy T. Ammerman, "Studying Everyday Religion: Challenges for the Future," in Everyday Religion: Observing Modern Religious Lives, ed. Ammerman (Oxford: Oxford Univ. Press, 2007), 219-38.

(23.) Roland J. Campiche, ed., Cultures des jeunes et religions en Europe (Paris: Cerf, 1997).

(24.) Jacques Janssen and Maerten Prins, "The Abstract Image of God: The Case of the Dutch Youth," Archives de Sciences Sociales des Religions 109 (2000): 31-48.

(25.) Jos Becker, De vaststelling van de kerkelijke gezindte in enquetes: 40% of 60% buitenkerkelijken? (The Hague: SCP, 2003); and "Church Membership Investigated (1950-2000)," in The Dutch and Their Gods: Secularization and Transformation of Religion in the Netherlands since 1950, ed. Erik Sengers (Hilversum: Verloren, 2005), 61-66. Similar problems are described in the United States and in the United Kingdom (see Hadaway, Kirk, Long, and Chaves, "Over-Reporting Church Attendance").

(26.) For another example, see Andrew M. Greeley, Religion in Europe: A Sociological Profile (New Brunswick, N.J.: Transaction Publishers, 2003), 77, with severe criticism of the "loaded questions" in the 1998 ISSP survey on negative attitudes toward religion (the questions were "designed to draw a stereotyped answer which could be hostile").