POLARIZING CATHOLICS?: DON’T BELIEVE YOUR MAIL! ANDREW GREELEY "Our main contribution to
society is to debunk the cherished myths of popular discussion and so-called
"informed debate." |
|
|
| .INTRODUCTION | ||
| Is polarization increasing among American Catholics? The position paper prepared for the Common Ground project argues that it is. Much of the reaction to the project assumes the serious nature of the polarization problem and argues about the cause of polarization. It would seem that Catholic polarization is but a part of the more general polarization which besets the country. More than four out of five Americans believe that there is more polarization than there used to be. Many observers of American religion believe with Professor James Davidson Hunter that there are culture wars going on between orthodox and liberal religious groups that might tear American society apart. | _ |
|
 |
. Yet until recently there has been very little serious consideration of available data which might specify and explain the nature of American polarization. However, in the November 1996 issue of the American Journal of Sociology, three scholars from Princeton University – Paul DiMaggio, John Evans, and Bethany Bryson – have examined data from two ongoing surveys (Michigan’s National Election Study and NORC’s General Social Survey) to examine the extent of a possible increase in polarization in this country from the early nineteen seventies to the middle nineteen eighties. They report that on only one major issue – abortion – has there been an increase in polarization. On other issues – race, crime, sexual morality, women’s public roles, family gender roles – there has been no increase in abortion and in fact an increase in consensus. Americans are less polarized than they were in 1972. CATHOLIC POLARIZATION While Professor DiMagio and his colleagues considered many subgroups (where they also found little polarization) they did not consider possible polarization among Catholics. (Why bother looking at one quarter of the American population?) I propose in this article to apply the Princeton model and methods to that question. To summarize my findings: it was unfortunate that the Princeton team did not look at Catholics; for on issues of sexual morality, the role of women, divorce, and especially abortion there is even less evidence of polarization than there is in the rest of American society. These findings contradict sharply the Common Ground position paper. I support the general intent of the project. However, I am convinced on the basis of this analysis that the basic assumption of those who have proposed it is inaccurate and misleading. I will not apologize for the complexity of this paper. Polarization is a complex issue that must be discussed rigorously and carefully. One does not find evidence for it by listening to dinner table conversation, reading the entrails of dead chickens or wetting one’s finger and holding it up to the wind. The data for this paper are taken from the General Social Survey and can be found on the G.S.S web page (http://www.icpsr.umich.edu/gss). For a fuller discussion of the methods used the reader should consult the article by Professor DiMaggio and his colleagues. There are three issues which must be considered when one wants to investigate change in attitudes over time – average (mean), dispersion (variance) and bimodality (kurtosis). By way of comparison let us look at NBA basketball scores. The mean has clearly declined (to say nothing of free throw averages). However, one must ask whether teams (and players in the matter of free throws) still cluster around the mean as they have clustered around means in previous years. Are the (hated) Knicks and the Bulls, for example, about the same distance from the present mean as they were from the former? Or have teams spread out along the distribution, so that the Knicks and the Bulls, for example, are now further from one another than they used to be? If the latter is what has actually happened, then there has been an increase in dispersion and thus an increase in polarization in team scoring. Moreover it might also be the case that the middle may weaken and the teams may cluster together at the top and the bottom of points scored, with one group of teams near the Bulls at the top and another near, let us say, the Nets, at the bottom. If that is the case, then bimodality is occurring and the teams are polarizing at opposite ends of the distribution of scores. Note that this phenomenon can occur even if the average score remains the same. The similarity in averages masks a clustering at the top and at the bottom of scores. In fact, the Princeton scholars discovered that, while there has been little change in average scores on the NORC abortion scale in the last two decades, there has been an increase in both dispersion and bimodality. Polarization has occurred without a change in the average score because the polarizing elements have cancelled each other out. In a longer and more technical version of this article (on my home page – www.agreeley.com ) I present correlations with time for mean scores, dispersion, and bimodality on six items –abortion, sexual morality, divorce laws, premarital sex, family gender roles and public roles of women. A positive correlation with means indicates that Catholics have move to more "conservative" positions between 1972 and 1994), a negative indicates that they have moved in a more "liberal" direction. A positive correlation with dispersion and a negative correlation with bimodality indicates an increase in polarization, a negative correlation with dispersion and a positive correlation with bimodality indicates that there has been a decline in polarization. A statistically insignificant correlation indicates that there is no change in polarization. On five of the six measures the mean score has declined for Catholics over time, meaning that they have moved to the more "liberal" position over the last two decades, ranging from a correlation of -.04 between time and abortion to a correlation of -.26 between time and family gender roles. Only on the issue of divorce laws has there been no change over time in average Catholic attitudes. I repeat here what I have often said before: the fact of such a change does not mean that the change is in the proper direction. Descriptions do not imply norms: one does not arrive at sound morality by counting noses. Nor is the point in this paper to report changes in Catholic attitudes which have been reported often in the past but to ask whether these changes have led to polarization. Does this shift in Catholic attitudes mean that Catholics are "polarizing" on these issues? Has dispersion increased as a result of the changing average: Are Catholics now more likely to be spread out along the response continuum than they once were? Has bimodality increased: Are Catholics now more likely to cluster at opposite ends of the continuum than they used to be? This is not the case. In fact, while the Princeton study shows an increase in both dispersion and bimodality on the abortion issues for the American population, there is a decrease in polarization among Catholics: Americans in general are polarizing about abortion, but Catholics are not polarizing on this or any other of the issues under analysis. With a drift to the "left" and no increase in polarization, it follows mathematically that the "right wing" segment of the Catholic population is declining. This does not mean that they are wrong, only that they are less numerous. Take for example two issues – the immorality of pre-marital sex and legal abortion when the woman’s health is in danger. 7.7% of the Catholic population said that premarital sex was always wrong AND opposed legal abortion even when the mother’s health was in danger in the first five year period. This declined to 6.3% in the final five year period. Since both of these positions would be considered essential in any measurement of a Catholic "hard" right, it would follow that this group represents somewhere in the neighborhood of three million Catholic adults. This is not a trivial number. Moreover it may well be a saving remnant. However it is not a major component of American Catholicism. If we ask what proportion supports either one OR the other of these two positions but NOT BOTH, we see in the accompanying table that the rate has declined from more than 38% to less than 26%. The Catholic "right" would hardly count these among their numbers. Or to put the matter the other way around the proportion of Catholics rejecting both positions has increased from 62% to 74%. I don’t rejoice in this change; I merely report it. Indeed if one adds such non-doctrinal matters as married priests, the popular election of bishops, and decentralization of power to national hierarchies as I have been able to do in other research, the proportion of American Catholics who are "hard" right declines to less than two percent, less than a million people. Not very many for all the noise they make – and all the letters they rignt. Are there any flaws in this analysis? 1) It would be much more satisfactory if one had available a wider battery of questions designed on specifically Catholic subjects which measured Catholic attitudes over the last twenty years. In the absence of such data one must make do with the items that are available in existing longitudinal data sets. There are no measures of attitudes either towards the ordination of women or a lifting of the ban on married priests. Yet attitudes on these issues in other surveys correlate strongly with the measures that are used in this analysis. It would be remarkable if polarization on abortion has decreased while polarization on the ordination of women has increased. 2) It might also be said that this analysis is sociological gobbledygook. Everyone knows that there is polarization; any sociology which purports to question that obvious fact should be dismissed as irrelevant. In other words, once you have made up your mind, facts to the contrary must be dismissed. As Professor Michael Hout observed on reading this paper "What’s dangerous is that overly simplistic summaries invite polarization. The polarizing metaphor itself come from the photographers technique of eliminating gray from an image. Points which were gray are forced to black or white. It produces a dramatic image -- much more dramatic than the gray dominated but more realistic original. That’s fine in art which aspires to be provocative. But it is hurtful in social analysis and totally unacceptable in social science which should describe rather than provoke." 3) Some might object that such arcane sociological terminology as "dispersion" and "bimodality" (to say nothing of "variance" and "kurtosis") serve only to obscure a relatively simple question: are American Catholics more polarized now than they were in the decade after the Second Vatican Council? As someone remarked recently, "I’m sick of surveys which tell us what we believe." Perhaps, but the alternative is to judge "what we believe" from an examination of conscience or from talking to one’s friends and relatives. Reality is rarely simple. Polarization is a complex issue not because the Princeton team and I have made it complex but because it is inherently complicated. It cannot be understood with the rigor and the precision that is required for Church policy decisions by the mere citation of what everyone knows to be true. Those who prefer simple and clear generalizations on this subject to precise and nuanced – if admittedly gray-tinged -- analysis are dangerous. The complaint that sociologists use their own terms and methods is valid only if one is prepared to deny all disciplines (such as theology or scripture study) their own terms and methods. 4) Presumably it will also be said that characteristically I find what I want to find in the data. To that I reply the data are available on the General Social Survey home page. Anyone who wishes to do their own analysis is welcome to do so. If they don’t like the General Social Survey then they should use their own money to collect their own data. Ad hominem arguments are not an adequate response save for the ignorant and the foolish. How can one explain the phenomenon of declining polarization in an environment in which everyone thinks there is increasing polarization? Among the reasonable answers that one might provide are the following:
If Common Ground serves only as a pretext for elite leaders to continue to shout at each other (or to posture as "moderates") it will accomplish nothing positive and may well have a negative impact on church policy because it will confirm the false image of a polarized population. It is possible that as Professor Hout proposed to me that if the elites talk to one another, they be less likely to demonize one another and may lower the tone of their rhetoric. Candidly I doubt that this would happen; the reaction among the hard "Right" to very idea of Common Ground indicates that they are not likely to stop demonizing. The magazine Crisis implicitly called for Cardinal Bernardin’s resignation because of Common Ground when he was already dying of cancer. In fact, the serious problem facing American Catholicism is illustrated in this paper not by the measures of dispersion and bimodality but by measures of the average (or mean). They show that the alienation of the body of the Catholic population from their leadership has increased over the past two decades. The laity as a body are less likely to take seriously what the Pope or the bishops say. Moreover this alienation has affected the laity in such a way that they move in the direction of the "Left" as a group and with no increase in polarization. The hard line "Right" is increasingly a smaller minority no matter how loudly it shouts. This is not a matter of personal opinion, but of statistical fact. Alas, mail is more powerful than statistics – in Rome, in Washington, in chancery offices, and in rectories. For the Common Ground dialogue to make any sense it must be a dialogue between the leaders and the members, a highly desirable dialogue but one that does not seem very likely in the present structure and culture of Catholicism. The present philosophy of church leadership seems to be that if one issues commands, gives orders, demands assent, invokes authority, than one has discharged one’s obligations. Moreover "loyalty" demands that one pretend that these techniques have been successful. To listen and to try to persuade is somehow to be disloyal. It is worth observing that most of the time period covered in this analysis has occurred during the administration of the present pontiff. The data in this analysis suggest that the strategy of teaching based on authority without the need to listen and to persuade has not been successful in preventing the continued drift of the laity away from its leadership. I shall be accused of being the cause of the drift or of approving it because I describe it. In fact, I deplore the increasing alienation between laity and leadership as a tragedy. What might be called the "teaching Church" and the "believing Church" (or the "Church as institution" and the "church as community") are at odds with one another; this is a profound loss for Catholicism. It will not do to say that the believing Church must swallow its proud disobedience and obey. As Cardinal Bernardin said in his final press interview (with the New York Times), that has been tried and it doesn’t work. Moreover, I trust that it is clear that my criticism of some elements of Common Ground comes from the opposite direction as that of the East Coast Cardinals who think there is no need to dialogue about anything. The assumption in the Common Ground project that the Catholic population is splitting into two polarized groups obscures this deeper and more serious problem. There is no evidence that a polarizing split is happening. To the extent that this assumption continues to dominate the discussion, the basic goal of the project will be perverted and the project itself will become mischievous. In my first parish assignment, I served under a pastor who nervously feared that every negative letter – especially if it were anonymous -- represented vast hordes of dissatisfied parishioners. Moreover a single complaint on a hot summer Sunday that the air conditioning was too cold would send him rushing towards the sacristy to turn off the cooling system. For some reason the scores, even the hundreds, of parishioners who were stifling would complain only to the curates. The alleged polarization of American Catholicism is of the same order of reality. I plead with those responsible for this mythology in the same words that I used to the Monsignor in those days, "Don’t believe that your mail is typical." I have the impression that this plea will be no more effective today than it was then.
APPENDIX Table 1 Changes in Mean, Dispersion, and Bimodality in Abortion Attitudes Means Dispersion Bimodality Table 4 Changes in Mean, Dispersion, and
Bimodality in Premarital Sex Attitudes |
|
