CESNUR - Centro Studi sulle Nuove Religioni diretto da Massimo Introvigne


The End of the Mass? Over-Reporting and Catholic Mass Attendance: An Empirical Study

by PierLuigi Zoccatelli
A paper presented at the CESNUR 2010 conference in Torino.© PierLuigi Zoccatelli, 2010. Please do not quote or reproduce without the consent of the author

During this presentation, I will try, in the short time at my disposal, to sum up the great quantity of data and the tables which we have compiled over the course of a year-long study. The findings have been published in detail in a volume, a copy of which has been given out to all the participants of this conference, so that you may have a chance to study and refer to them as needed.

If we need to expand into any of the aspects touched on in this presentation, we will be assisted, not only by referring to the volume I just mentioned, but also by Massimo Introvigne, with whom I collaborated in conducting this research, and who has offered to co-operate in answering questions from the audience. I’d like to acknowledge his assistance in preparing this paper.

In a CESNUR research project carried out in 2009 regarding the frequency of attendance at religious services in Central Sicily, we focused on twelve townships covering the area located within the boundaries of the diocese of Piazza Armerina, one of the eighteen Episcopal sees of Sicily, comprising towns belonging to two provinces: Enna and Caltanissetta. The surface area of this territory of 2,003 square kilometers has a population of 220,308, according to the 2009 census estimate. The largest city is Gela.

Townships Males Females Total
Aidone 2,474 2,719 5,193 (2.4%)
Barrafranca 6,202 6,896 13,098 (5.9%)
Butera 2,394 2,626 5,020 (2.3%)
Enna 13,352 14,725 28,077 (12.7%)
Gela 37,866 39,251 77,117 (35.0%)
Mazzarino 5,735 6,328 12,063 (5.5%)
Niscemi 13,012 13,511 26,523 (12.0%)
Piazza Armerina 9,941 10,900 20,841 (9.5%)
Pietraperzia 3,392 3,902 7,294 (3.3%)
Riesi 5,339 6,009 11,348 (5.1%)
Valguarnera Caropepe 3,959 4,377 8,336 (3.8%)
Villarosa 2,544 2,854 5,398 (2.5%)
Total 106,210 (48.2%) 114,098 (51.8%) 220,308 (100.0%)

Previously in 2008, in this same territory, CESNUR carried out a research project regarding non-Catholic religions and movements. In this first investigation, we found a total of 28 religious minorities present in the area.

Denomination Members
Assemblies of God in Italy 790
Italian Christian Pentecostal Church 1,710
Christian Pentecostal Congregations 105
Kurion Iesoun 124
Apostolic Church In Italy 250
Old Apostolic Church 40
United Pentecostal Church International 5
Free Popular Mission 20
Christian Community 10
The Church on the Rock (Elim) 300
Ekklesia (Elim) 80
International Evangelical Church 200
Evangelical Church of the Reconciliation 220
Christian Mission 60
Charismatic Christian Mission 300
Waldensian Church 100
Brethren Assemblies 180
New Apostolic Church 83
Seventh-day Adventist Church 400
Jehovah’s Witnesses 1,610
Church of Jesus Christ of Latter-day Saints 64
Romanian Orthodox Diocese 40
Atisha Serlingpa Center 20
Soka Gakkai 30
Islam 585
Bahá’í 200
Reiki Universal Love Association 50
Church of Scientology 50
Total 7,626

Overall, the 7,626 members of the 28 minorities represent 3.5% of the resident population, contrasting with the 2.1.% of Italians belonging to religious minorities on the national scale. This is a unique case in Italy: a sufficiently large territory within which numerous minority religions operate, representing a percentage of the resident population which is almost double compared to the rest of the nation.

The research carried out by CESNUR in 2009 is the logical continuation of our previous investigation into religious pluralism. The aim of our research in 2008 was to measure, describe, and analyze the overall situation regarding religious practice through a study of minority religions other than the Catholic church. In the 2009 research we focused on Catholics.

The practice of a religion may be measured according to three main levels: believing, belonging, and behaving. While the study of religious practice is a separate enquiry with respect to the study of beliefs and behaviors, our 2009 project focused on what is considered to be the main indicator of “belonging,” i.e. the frequency with which subjects attend religious services.

This is both a fundamental and controversial field of enquiry. In fact, research into the frequency of attendance at religious services is generally conducted using the sampling method of questionaires and interviews, thereby obtaining a report which only measures how frequently the respondents claim to attend religious services. This data is important, but it may suffer from over-reporting, or, in some cases, under-reporting, i.e. subjects claim to attend more or less frequently than they actually do. Some research has focused on measuring actual attendance at religious services, and the project carried out by CESNUR in 2009 in Central Sicily attempted to do just that.

One of the aims of our research, conducted through a survey, was to measure declared participation in the area under study through computer-assisted telephone interviews. The other aim of our research was to consider all the Catholic churches existing in the territory, and measure church attendance in a systematic and thorough way. Thus, we counted the numbers of people actually present, so as to compare them with the frequency of attendance declared by respondents in the telephone interviews. It must be immediately emphasized that it would be wrong to consider the actual attendance as “true” and the declared attendance as “false.” A long international debate on over-reporting has at least clarified that any interpretation of data in terms of true or false is a gross over-simplification.

On the basis of a representative sampling in terms of territorial distribution, sex, and age, during October 2009, 1000 questionnaires were administered through the CATI (Computer-Assisted Telephone Interviewing) system. The 1000 subjects interviewed were asked five questions: (a) At the present time, what religion or faith do you feel you belong to? (b) Excluding weddings and funerals, how frequently do you attend Mass, or if you are not Catholic, other religious services? (c) Last weekend did you go to Mass, or if you are not Catholic, did you attend another religious service? (d) How frequently do you take communion? (e) How often do you go to confession? Here are the results of our questionnaires.

(a) At the present time, what religion or faith do you feel you belong to?

Catholic 92.2%
Protestant Evangelical 1.5%
Orthodox 0.9%
Islamic 0.5%
Jewish 0.0%
Buddhist 0.1%
Jehovah’s Witnesses 1.1%
Hindu 0.0%
Other religion 0.4%
No religion 3.3%
Total 100.0%

(b) Excluding weddings and funerals, how frequently do you attend Mass, or if you are not Catholic, other religious services?

Never 17.6%
Once or twice a year 10.1%
Several times a year 20.9%
Once a month 10.7%
Two or three times a month 7.1%
Weekly 27.2%
Several times a week 6.4%
Total 100.0%

(c) Last week-end did you go to Mass, or if you are not Catholic, did you attend another religious service?

Yes 33.9%
No 66.1%
Total 100.0%

(d) How frequently do you take communion?

Never 23.7%
Every few years 5.2%
Only at Christmas, Easter, or other religious holidays or occasions 18.4%
Several times a year 17.4%
About once a month 7.6%
Two to three times a month 6.8%
Every week 17.2%
Several times a week 3.7%
Total 100.0%

(e) How often do you go to confession?

Never 32.4%
Every few years 10.9%
Once or twice a year 27.1%
Several times a year 16.1%
Monthly, or almost monthly 11.8%
More frequently 1.7%
Total 100.0%

Differently from the questionnaire, which was based on a sampling of persons over 15 years of age, our measurement of church attendance on Sunday took into account the population over the age of 10, i.e. 196,473 individuals, or 89.2% of the resident population.

A group of two hundred volunteers was created, instructed and trained to perform the practical task of measuring attendance. Thus we were able not only to measure attendance at all of the 320 Masses celebrated between Saturday and Sunday evening of the week end of November 21-22, 2009, but also count the number of people taking communion. Our results obtained from two distinct groups of volunteers also included the fourteen Eucharistic celebrations of the Neocatechumenal Way, which are often excluded from surveys since they are not announced to non-members. Those being unable to attend Mass who asked to receive communion at home or in a hospital were counted as if they had participate to a Mass. The following table shows a synthesis of the data collected, covering all Catholic services held in the area on that week end, without exception.

Townships Presence (M / F) % of population (M / F) N° Masses Communion (M / F)
Aidone 837 (274 / 563) 17.5% (12.0% / 22.6%) 10 577 (182 / 395)
Barrafranca 2,198 (541 / 1,657) 18.7% (9.8% / 26.7%) 20 1,292 (234 / 1,058)
Butera 1,678 (617 / 1,061) 36.4% (28.3% / 43.8%) 12 1,090 (327 / 763)
Enna 5,971 (2,048 / 3,923) 23.2% (16.8% / 28.9%) 64 3,645 (989 / 2,656)
Gela 13.315 (5,059 / 8,256) 19.6% (15.3% / 23.7%) 68 10,674 (3,800 / 6,874)
Mazzarino 2,065 (693 / 1,372) 19.0% (13.6% / 23.9%) 26 1,476 (445 / 1,031)
Niscemi 2,189 (687 / 1,502) 9.4% (6.1% / 12.7%) 19 1,542 (426 / 1,116)
Piazza Armerina 2,773 (932 / 1,841) 14.9% (10.5% / 18.8%) 39 1,992 (577 / 1,415)
Pietraperzia 1,390 (418 / 972) 21.1% (13.8% / 27.3%) 18 868 (181 / 687)
Riesi 1,155 (318 / 837) 11.4% (6.7% / 15.5%) 17 967 (232 / 735)
Valguarnera Caropepe 1,969 (586 / 1,383) 26.5% (16.7% / 35.2%) 16 1,034 (221 / 813)
Villarosa 885 (228 / 657) 18.2% (10.1% / 25.2%) 11 619 (128 / 491)
Total 36,426 (12,401 / 24,025) 18.5% (13.2% / 23.5%) 320 25,776 (7,742 / 18,034)

So how many “practicing Catholics” are there in our area? Our previous analysis revealed that 33.6% of those living in the area (not only Catholics) declared to attend a religious service once a week or more often, whereas only 18.5 % actually attended Mass during the week end selected for study.

Adding to this the 3.5% of non-Catholics (and considering that members of religious minorities in Italy are rarely “non-practicing), we obtain 22.0 % of the entire population under study as a likely measure of those who actively practice their religion in a sample week end. In our telephone survey, 51.4% (Catholics and non-Catholics) claimed that they practiced their religion once a month. 92.2% of subjects interviewed declared they were Catholics, and 96.7% (Catholics and non-Catholics) declared they were religious, leaving us with a 3.3% of nonbelievers.

In interpreting this data, we must avoid certain misunderstandings. In the past, studies which have attempted to measure over-reporting by comparing the results of questionnaires and interviews with the actual count of attendance in places of worship have been sharply criticized, particularly in the United States. This should not keep us from conducting research which, if carried out with caution and methodological humility, may yield interesting results.

Far from the spotlights of American discussions, since 1980, the Episcopal Conference of Poland has organized an annual “Statistics Sunday” during which data is collected regarding church attendance and communion in every place of worship in Poland where Mass is celebrated. This data, which is not kept secret, but which is also not excessively publicized outside Poland, does not include non-Catholic religious minorities.

A comparison of the counting in these “Statistics Sundays” in Poland with telephone surveys must be made with caution. The Polish telephone surveys do not share the same temporal series and are not carried out by the same institutes assigned the task of physically counting participants at Mass. With due caution, we have created the following table from the Polish data available.

Year CATI Survey Direct Count
1998 57.0% 47.5%
1999 57.0% 46.9%
2000 56.0% 47.5%
2001 56.0% 46.8%
2002 57.0% 45.2%
2003 58.0% 46.0%
2004 57.0% 43.2%
2005 57.0% 45.0%
2006 58.0% 45.8%
2007 56.0% 44.2%
2008 57.0% 40.4%
2009 58.0% 46.9%

Apart from an anomalous drop registered in 2008, attendance at Mass in Poland remains consistent, and is much higher than in all the other European countries, with the exclusion of Malta. There is also an almost stable over-reporting of 10-12 percentage points. In Italy the existence of over-reporting has been hypothesized and even hypothetically quantified through models that have attempted to measure the “survey effect,” and the truthfulness of respondents’ answers to questionnaires. A pioneering study in this sense was carried out in 2004-2005 in the territory of the Patriarchate of Venice by sociologist Alessandro Castegnaro.

The Polish bishops distinguish between what they call the dominicantes, i.e. people counted as actually attending Mass on their “Statistics Sundays”, from “declared practicing Catholics,” those who defined themselves as practicing Catholics during the telephone surveys. Then there are also “irregular practitioners,” those who claim they attend Mass at least once a month. And fourthly, there are those who claim to be Catholics, i.e. they have been baptized, and do not claim to belong to any other religion, or to be atheists or agnostics, but do not attend Mass with any regularity. In Italy, the Italian Conference of Catholic Bishops, following research by leading sociologists, has stressed for some time that whenever we ask “How many Catholics are there?” we must differentiate between at least two concentric circles, one including a much larger “baptismal community” and the other a smaller “Eucharistic community.”

Our research allows us to make a few further distinctions. First we must repeat that the Diocese of Piazza Armerina has a large non-Catholic presence (3.5%) notably higher than the national average. Thus, assessing the religious participation and practice will differ according to whether we take into account only the Catholic Mass or also the religious practice of other faiths. If we limit ourselves to Catholics, the image of the concentric circles is of use. But the dominicantes (18.5%) are not “true” practitioners compared to the hypothetical “false” practitioners declaring themselves as practicing Catholics in the telephone survey. Statistical data cannot be made to say more than it actually says.

We must avoid the temptation to mechanically compare the data, because as a more detailed analysis would show, the age ranges considered in the survey and in the actual counting of people present were not identical, and in the survey questions were also asked concerning participation in non-Catholic services. Removing the data referring to non-Catholics (collected through our research on religious minorities) and taking into consideration this uncertainty, the circle of “declared”, or self-identified practicing Catholics is about 30% of the whole population.

This figure is important and is not challenged or denied by the actual headcount at Masses. It indicates the respondents’ intention and aspiration to participate in the Mass, which is extremely important in any discussion of Catholic identity and identification with the Church. Next come the larger circles of occasional practitioners (51.4%), and of Catholics who feel they belong to the Church as a “baptismal community” (92.2%), a percentage which does not necessarily correspond to the number of people baptized, since the latter also includes people who, after Catholic baptism, have converted to another religion or become nonbelievers. Although the newspapers would revel in such a conclusion, it would be wrong to deduce from our research that only 18.5 % of the population in the diocese is Catholic.

From another point of view, in fact, one could say that Catholics make up 92.2% of that population. The discrepancy between the two figures shows that “Catholic” is no longer a term with a single meaning. At least in theory, the meaning may be one with respect to doctrinal identity. But certainly “Catholic” has multiple meanings with regards to sociological identification. So, in order to receive a sensible answer to the question “Who is Catholic?”, we need to be aware of nuances. And it would also be wrong to say that “practicing Catholics” represent 18.5% because the notion of “practice” is also complex and has multiple meanings.

How can we explain the discrepancy between the dominicantes and the self-identified practicing Catholics in the telephone survey?

Previous studies of over-reporting have attempted to answer this question, suggesting that surveys themselves are responsible for this phenomenon, because, for example, proper procedure is not followed in conducting interviews, the population selected for sampling is not homogeneous, or the questions themselves are misunderstood. In other cases, over-reporting has been attributed to the subjects’ wish to present themselves to the interviewers in a sociably more favorable light, affirming positive behaviors which correspond to their own personal religious aspirations and identities.

Even though these explanations suggest interesting directions for further research, they are not totally convincing. The former, concerning the sociologists who design the surveys, underline the intrinsic limits of this type of enquiry but rarely explain what the “right question to ask” would be. It would be quite remarkable if, in over twenty years of research, the right question had never been hit upon. The latter, regarding the subjects interviewed, formulate hypotheses with which we might agree when they distinguish between various levels of identification. But they go on to speculate on the state of mind and the wishes of the respondents, which sociologists have no way of measuring and which lie outside their field of enquiry.

We must ask ourselves whether it is truly necessary to seek “explanations” for this anomaly, or even more radically, if an anomaly really exists. What the studies of over-reporting conducted in the USA, Poland, and Italy actually demonstrate is that declared practice is just that: “declared.” It measures identity and also perhaps identification, but not acts and behaviors. To think otherwise would be like expecting opinion polls regarding consumer approval of products to measure sales in the stores. We are dealing with two diverse realities here, not the same one measured differently, and certainly not a reality measured in terms of “true” and “false.”

The nucleus of the dominicantes appears to be characterized by a strong identification with the Church and the sacraments. In the survey 20.9% of the total sample claim to take communion once a week. Our research confirmed that 70.8% of those attending Mass in our own “Statistics Sunday” also took communion. There is a significant difference with the situation in Poland, where on “Statistics Sundays” the number of people taking communion remains constant, roughly one third of the people attending Mass. This figure probably indicates a different approach of the Polish priesthood, depending in part on a historical tradition, to the theme of communion.

13.5% of the total sample in our survey also claim to go to confession at least once a month. With regards to confession, however, the data is uncertain and every attempt to measure it runs up against limitations imposed by the privacy laws.

In any case, despite what is usually affirmed with regards to the crisis of confession in Italy and elsewhere, the telephone survey may indicate that this sacrament is still holding strong. This is a sign of the high quality of the “hard core” of the dominicantes in the diocese. Although a minority among those who perceive themselves a Catholics, the number of actual communions and what they declare about confession seem to indicate a high level of commitment to the sacramental life of the church.

This circle of dominicantes should be evaluated together with and not “against” those self-identified as practicing Catholic but not actually counted as Mass attendants. This “hard core” of dominicantes remains a consistent presence in the spiraling of the concentric circles that define the complex relations of Sicilians and Italians with religious practice.