1. Introduction
In the past several decades, the Catholic Church has been rocked by multiple sexual abuse scandals. While there are many dimensions to this, one key aspect that has emerged from the growing numbers of reports, investigations, and academic studies is the special role of bishops in creating and maintaining the ecclesial cultures and contexts in which abusive practices fester and thrive (or else find it harder to do so). Some bishops have been found to be sexual abusers themselves. Others, through inaction, buck-passing, the seeking of ‘quiet’, ‘pastoral’ solutions to problems (e.g., moving abusive priests from parish to parish), ignoring or silencing accusers, and/or active cover-ups have been found to be more-or-less culpable in ‘enabling’ it.
The prospect of successful interventions or reforms in this area is, however, hampered by a general paucity of studies investigating episcopal culture and dynamics in general, let alone with a particular focus on the abuse crisis. In this paper, we introduce one particular means of doing this which has proven successful in exploring multiple other areas of organizational and group dynamics: Social Network Analysis. For readers unfamiliar with this particular methodological and theoretical toolkit, we explain some of its main elements, elucidate several features of recent episcopal scandals in the United States (i.e., McCarrick, Bransfield) from a social networks perspective, and present the results of some preliminary, exploratory studies, mapping the network structures of the Bishops’ Conferences of England and Wales, and the United States.
The collection of tools, theories, and methods commonly described as Social Network Analysis (hereafter “SNA”) has helped shed light on diverse areas of social, cultural, political, intellectual, and religious life.
1 These include the epidemiology of contagious diseases (
Keeling and Eames 2005), the formation and spread of rival philosophical schools in medieval Europe (
Collins 1998, pp. 451–522), and collaborations among twentieth-century British classical composers (
McAndrew and Everett 2015). Undergirding much of this empirical work is an avowedly
relational understanding both of individual subjects, and of the various, overlapping social worlds which they collectively inhabit, create, maintain, and/or change. Accordingly, if—as argued by one of relational sociology’s leading proponents—“The most appropriate analytic unit for the scientific study of social life is the network of social relations and interactions between actors (both human and corporate)”, then SNA has established itself as one of the principal means by which “we can identify
mechanisms within interaction, relations and networks which help to explain and understand events in the social world” (
Crossley 2011, pp. 1, 4).
This paper has two main purposes. Firstly, we seek to show how certain methods and theoretical insights within SNA might illumine a subject of critical and, we believe, urgent significance within the sociology of religion. While appreciably slower to “catch on” in this field than in many others, there is now a reasonably substantial body of literature applying SNA within the sociological study of religion (
Everton 2018). To the best of our knowledge, none of it has focused on episcopal networks, whether Catholic or otherwise. In fact, there does not exist a huge amount of empirical social research on Catholic bishops, either as individuals or as collective actors—albeit with some notable exceptions. Closest in spirit to the approach adopted herein is Melissa
Wilde’s (
2007) application of Social Movement Theory to episcopal politicking during the debates of Second Vatican Council. In addition, there has been a handful of psychological studies of US bishops based on mailed questionnaires (e.g.,
Sheehan and Kobler 1977;
Schroeder 1978), though these are now forty years old. The American Jesuit priest and journalist Thomas Reese, who holds a PhD in Political Science, published detailed studies of several aspects of the Church’s internal “power structure” (e.g.,
Reese 1989,
1992). Most recently, researchers at the Center for Applied Research in the Apostolate in Washington, DC, have published a new monograph based on surveys and/or interviews with a majority of current “Ordinaries”
2 (
Fichter et al. 2019). Promisingly, recent years have witnessed a small-but-growing number of studies focusing on the Catholic episcopacy vis à vis the handling of sexual abuse cases, and the abuse crisis more broadly (specifically,
O’Keefe 2019;
Ballano 2019;
Byrnes 2020;
Pogorelc 2020). Though not focused exclusively on the abuse crisis, notable recent monographs by Brian
Dive (
2020) and Adam
DeVille (
2019) each, in their different ways, concern institutional failures and pathologies within the current episcopal structures, and propose possible remedies.
Our second intention is to present initial findings from our own exploratory, “proof-of-concept” application of formal SNA to Catholic episcopal networks. This encompasses (i) a pilot study, using data from the (relatively small) Catholic Bishops’ Conference of England and Wales (we used this, primarily, to test out methodological decisions regarding the sampling, compiling, and coding of the dataset, and also to assist with hypothesis generation), and (ii) various analyses, using network data for each member of the United States Conference of Catholic Bishops as of July 2018 (i.e., the month in which the
New York Times first reported on serious allegations of sexual misconduct against then-Cardinal Theodore McCarrick; see
Goodstein and Otterman 2018). This, we argue, lends considerable support to the hypothesis, widely shared among commentators on the Catholic abuse crisis, that there is a critical
network dimension both to the problem and, perhaps, to its solution. ‘The key now is to uncover the networks within the clergy and episcopate’ (
Dreher 2018a). We conclude by suggesting various possible areas for future, deeper research.
2. Background and Rationale
The motivation for this present study comes from three different sources. The first is the recognition that SNA has proven especially transformative in exploring networks in which the influence of patronage, preferment, and indebtedness are typically important. These could include anything from the mafia in Italy or the USA (
Varese 2013;
Mastrobuoni and Patacchini 2012), Chinese political elites (
Keller 2016), the academic job market (
Kawa et al. 2019), or appointments to corporate boards of directors (
Koenig and Gogel 1981). Prima facie, episcopal appointments, culture, and therefore governance within the Catholic Church are ideal topics for serious SNA attention (see also
Ballano 2019;
Pogorelc 2020). Note, for example, the following sui generis features (cf.
Reese 1984;
O’Callaghan 2007, pp. 119–52;
Allen 2016a;
Dive 2020, pp. 104–15):
The special nature of the bishop–clergy relationship (e.g., “Clerics are bound by a special obligation to show reverence and obedience to […] their own ordinary”; Code of Canon Law, 273), which in purely practical terms extends far beyond the ordinary employer–employee dynamic within the secular world (outside of certain forms of bonded or indentured labour);
- (1)
The crucial role that the Ordinary plays in identifying, encouraging, and mentoring (including appointments to certain senior roles) potential future bishops;
- (2)
The critical role that a priest’s own Ordinary plays in championing the ‘candidacy’ of a specific priest to be considered for episcopal office;
- (3)
The role that certain other, influential bishops have in supporting or undermining a particular candidate (especially the local Metropolitan, certain prelates known to “have the Nuncio’s—or Pope’s—ear”, the previous Ordinaries of a given See, and/or members of certain national or Vatican committees (cf. Code of Canon Law, 377);
- (4)
The fact that, aside from an “intense eight-day training program” held annually in Rome for newly appointed bishops, the primary way in which “bishopcraft” is learned is through informal apprenticeship under one’s own previous bishop(s), especially through particular official roles within the diocesan administration, or (more formally) as an Ordinary’s auxiliary bishop.
Whatever the
theological justifications for, and desirability of, these features, from the perspective of SNA one might plausibly hypothesize a number of potentially negative properties emerging from this situation. For example, these might include the potential for ambitious clergy (or seminarians) to actively seek the favour and patronage of their own (and/or other influential) bishops, or indeed for bishops to use the hope or even promise of preferment as a means of incentivizing or rewarding loyalty. It could result in certain “types” (in terms of personality, class, ethnic background, theological vision, etc.) of priests being favoured and/or formed, in line with the type of their own bishop, and perhaps of a wider episcopal “mould” or “culture”. This homophilizing tendency would then be intensified by the fact that “how to be a bishop” is learned, in very great measure, through a process of imitation and socialization. It might lead to the creation of identifiable “factions” or “cliques” of bishops, bound by mutual bonds of
preferment and favour, and who act, formally or informally, in concert, and who each support and promote each other’s protégés. Furthermore, given all this, it might feasibly create shared senses of solidarity among particular groups of bishops, such that if “one falls, we all do”. To give a hypothetical example: suppose that the bishops in a given province had all served as vicars general, chancellors, and/or auxiliary bishops for each other, and had in turn (even absent direct or formal collusion; cf.
Bourdieu [1984] 1988, pp. 84–89, on similar dynamics at work within academic appointments) returned the favour by supporting the promotion of each other’s chancery favourites. Should one of the senior bishops in this group then be rumoured to have committed crimes while in office, it is not hard to imagine how others in the network might seek a “quiet” solution to the problem, to prevent either themselves or their patrons becoming implicated, even if by association, to varying degrees.
Secondly, this a priori fittedness of an SNA-informed approach to understanding episcopal governance structures receives considerable a posteriori support of various kinds. As noted above,
Wilde (
2007) demonstrates the value in adapting methods and perspectives commonly used to study political movements in order to shed light on the often-shadowy world of church politics. Furthermore, in reporting on and analysing such topics, well-informed journalists and academics naturally speak in terms of “networks”, “factions”, “cliques”, “lobbies”, “protégés”, and “patrons”—that is to say, precisely the kind of folk-descriptions of social and political life which SNA seeks to quantify and interrogate. To give a single illustration here, note these excerpts from veteran Rome correspondent Robert Mickens, writing for the
National Catholic Reporter:
The current system the church uses to seek out and appoint candidates for episcopal service is far too often based on cronyism inherent in an old boys’ network […] The apostolic nuncio plays a major role in drawing up the terna of (the top three) candidates for a particular episcopal post. [… The] roughly 30 cardinals and other ranking prelates from around the world who are members of the Congregation for Bishops […] discuss and vote on the candidates. […] However, well before this happens, bishops, in too many cases, have already begun “grooming” someone,–perhaps a star seminarian or their priest-secretary–to be a future member of their very exclusive club, the episcopal college. […] Customarily, the ordinary of [a] large diocese has a fairly good chance of pinpointing the man or men he wants as an auxiliary bishop. And if he’s well connected in Rome, especially with members of the pertinent congregation, this major hierarch can often help advance an auxiliary (or another bishop friend) to head his own diocese.
Finally, and more specifically, existing analysis of the scandals engulfing the Catholic Church in the United States (as in several other countries) highlights the role that precisely these kinds of network dynamics may have contributed, directly and indirectly, to both individual and institutional failures (and/or crimes) in adequately dealing with accusations of sexual abuse. Recognition of the need to focus on
organizational cultures and contexts in both diagnosing and treating the “sexual abuse crisis” in toto is not wholly novel (
Keenan 2011). Yet, it was thrust to the fore in 2018 primarily, though by no means exclusively, by the revelations surrounding Theodore McCarrick, formerly Cardinal-Archbishop of Washington, DC. While the full details, including allegations of grooming and sexual abuse of both boys and young men,
3 are beyond the scope of this paper, and have besides received extensive documentation in both the media (see, in detail,
Altieri 2020) and the official investigative report released by the
Holy See (
2020), a number of network-relevant aspects are worth highlighting.
3. The McCarrick Case: A Relational Perspective
First, McCarrick’s predilection for identifying select groups of “especially favoured” (
Goodstein and Otterman 2018) seminarians and young priests—i.e., “young men under his authority in the Church” (
Dreher 2018b)—whom he showered lavishly with alcohol, flattery, handwritten notes, gifts, meals, overnight stays at his personal beach house, and prophecies of great futures in the Church. These he referred to as his “nephews” and encouraged them to call him “Uncle Ted”.
Second, the unwillingness of subordinates within his diocese(s) to refuse dubious requests, or to speak out in other ways. Note here the testimony of Fr Boniface Ramsey, who was a faculty member at Newark’s archdiocesan seminary during McCarrick’s 1986–2000 tenure as Archbishop, and whose repeated, unheeded attempts at whistleblowing have now come to light. Regarding McCarrick’s widely-known practice of sharing a bed with his seminarians—he would deliberately invite more “nephews” to his Jersey Shore beach house than there were beds to accommodate—Ramsey rhetorically asks: “what member of the faculty would approach the archbishop to tell him that it just wasn’t right?” Commenting on a “sense of resignation” among the seminary faculty, Ramsey recalls his own first attempt at raising the issue with his immediate superior: “The rector knew exactly what I was talking about and promised to do what he could to stop it, after admitting that
he felt strung between his loyalty to his archbishop and his realization that what the archbishop was doing wasn’t right” (
Ramsey 2019; emphasis added). The Holy See’s own ‘McCarrick Report’ records several other such instances of his subordinates not acting on credible information, or even direct knowledge, of his predatory behaviour—up to and including two of his (former) Newark auxiliaries being present at the sexual assault of a seminarian (
Holy See 2020, pp. 87–93).
Third, McCarrick’s acknowledged status as the “the
kingmaker for appointments in the Curia and the United States” (
Viganó 2018, p. 8) among American bishops. This claim has featured prominently in reports following the 2018 revelations, along with questions concerning the extent to which other high-placed US bishops and/or cardinals in his “network” might have benefited from his championing (e.g.,
Dougherty 2018;
Dreher 2018c). Nor is this mere
post-facto speculation. Well prior to the scandals erupting in June 2018, McCarrick’s outsized pull on episcopal placings, over and above the Cardinal-Archbishop of Washington’s traditional clout, was generally acknowledged. His direct role as the “architect” of specific appointments was also regularly reported in the media (
Magister 2014;
Palmo 2016;
Allen 2016b). Importantly, however, this kind of influence is not something unique to McCarrick: the existence of powerful “bishopmakers” follows naturally from the episcopal selection process as described above. Certain bishops have reputations for exerting various types of sway, whether official or unofficial, to have their own favourites and protégés elevated to the episcopacy. Against this, the official McCarrick Report is keen to stress that he did
not have any disproportional influence over bishop making (e.g., “there is no evidence that McCarrick played a significant role with respect to episcopal appointments in the United States or elsewhere during this period”), to the point of devoting long footnotes to ‘setting the record straight’ regarding some of the cases noted above (e.g.,
Holy See 2020, p. 415, n. 1302). In a separate paper, we have investigated this very question, again applying the tools of social network analysis. We believe our analysis “provide[s] systematic evidence to the claim that McCarrick was indeed a ‘kingmaker’ within the USCCB” (
Sadewo et al. 2021, p. 256).
Fourth, the fact that serious allegations about McCarrick were widely known, and even more widely rumoured, among the highest echelons of the hierarchy for decades, but were dismissed, ignored, or in at least two cases, paid-off by his previous dioceses with five- or six-figure settlements. This retrospective “everybody knew” aspect is, of course, a common theme in the exposure of high-profile serial sex offenders (e.g., Jimmy Savile, Harvey Weinstein). A small number of (relatively junior) USCCB members have gone so far as to accuse their fellow bishops of, at best, negligence and, at worst, complicity and conspiracy. Thus, Robert Barron, who was then an auxiliary bishop of Los Angeles but has since been appointed as Ordinary in Winona-Rochester, has written:
[I]t seems numerous bishops, archbishops, and cardinals, both in this country and in the Vatican, knew all about McCarrick’s outrageous behavior and did nothing in response to it; or, rather worse, they continued to advance him up the ecclesiastical ladder, from auxiliary bishop, to bishop of a diocese, to archbishop, and finally to cardinal. Even after he resigned from his post in Washington, DC, […] McCarrick continued to be a roving ambassador for the Church and a kingmaker in the American hierarchy–again, while everyone knew about his disturbing and abusive tendencies.
Understandably, questions have been raised as to
what specific bishops close to him, including members of a so-called “McCarrick caucus” (
Allen 2016b) whose church careers he appears to have helped, knew and
when, and what they did (not)
do and
why (not). Of particular interest here are those bishops who served in various senior and/or trusted roles, and especially those who lived
with him for months or years (
Bullivant 2020;
Bullivant and Sadewo 2021).
Fifth, the Catholic Church’s male-only priesthood means that, while Catholic sexual abuse is
not exclusively homosexual in nature, sexual activity (whether abusive or not) among bishops, priests, and seminarians
ipso facto is. Furthermore, while celibacy is demanded of all Latin-rite priests
4 and seminarians, straight and gay, simply
being same-sex attracted is, in and of itself, officially problematic (see
Congregation for Catholic Education 2005). From a network-perspective, this combination of factors has important relevance for understanding the McCarrick case. Not least, there is clear potential for
mutually compromised networks of homosexually active (or once-active) priests, such as McCarrick appears to have cultivated among his “nephews” (By contrast, illicitly
heterosexually active bishops and priests, of whom there are no shortage of examples, can only be so with those
outside of the priestly networks). The existence of “homosexual subcultures” within US Catholic seminaries or diocesan power structures, while understandably a sensitive topic, is well-established in the academic literature, as too are the disproportionately high numbers of same-sex-attracted seminarians and clergy in the first place (
Greeley 2004, pp. 42–46;
Cozzens 2004, pp. 124–39;
Regnerus et al. 2021). In itself, that same-sex-attracted seminarians and priests might form friendship and support groups with those who can empathize with their trials is wholly unsurprising. However, combined with an intensely homosocial environment, a culture of secrecy and shame combined with legitimate fear for one’s vocation or ministry should one be “outed”, and a special bishop–priest/seminarian relationship which, irrespective of its other virtues and/or theological rationale, is certainly open to exploitation, then the risk of other McCarrick-esque cases is certainly a real one.
Sixth, details from subsequent episcopal scandals have shed further light on the wider culture in which McCarrick thrived for so long. Most telling here is the case of Bishop Michael Bransfield, whose late-2018 retirement from the Diocese of Wheeling-Charleston, West Virginia, was followed immediately by an official investigation into his conduct, overseen by Archbishop Lori of Baltimore (in his capacity as Bransfield’s Metropolitan). In the words of the official report, leaked by the
Washington Post in December 2019: ‘Bishop Bransfield engaged in a decades-long campaign of predatory behavior targeted at seminarians, young priests, and other young men, dating back to the period from 1982 to 1987 during Bishop Bransfield’s tenure at the National Shrine’ (
Zuckerman Spaeder 2019, p. 9). Although never a direct subordinate of McCarrick’s, Bransfield’s time in Washington coincided with the latter’s tenure as Archbishop, and they appear to have been close (
McCarrick 2011, pp. 203, 405). McCarrick was among Bransfield’s co-consecrators when he was made a bishop in 2005.
The Bransfield Report reveals many parallels to McCarrick’s own modus operandi. For example:
Bishop Bransfield subjected multiple seminarians and priests to unwanted sexual overtures, sexual harassment and sexual contact […]. Numerous witnesses reported classic ‘grooming’ behaviors by Bishop Bransfield […]. Despite witnessing multiple instances of harassing and abusive behavior over several years, none of the Vicars took action to address Bishop Bransfield’s behavior. The Vicar General and Judicial Vicar, in particular, acknowledged that they were aware of Bishop Bransfield’s tendencies and sexual harassment, but took no steps to prevent it based on a combination of fear and loyalty toward the Bishop (
Zuckerman Spaeder 2019, pp. 2–3).
Furthermore, investigators uncovered huge numbers of “gifts” made from his personal account before being routinely reimbursed from diocesan funds, amounting to hundreds of thousands of dollars (
Boorstein et al. 2019).
Bishop Bransfield frequently gave gifts to various high-ranking clergy, as well as young priests and seminarians, both in the form of cash and material goods, routinely utilizing his administrative staff to order and send material gifts […] In total, the [Diocese of Wheeling-Charleston] spent more than
$200,000 on Bishop Bransfield’s gifts to family members, clergy and others, often in recognition of celebrations such as first communion, weddings and anniversaries (
Zuckerman Spaeder 2019, pp. 47–48).
Personal payments of four- or five-figure sums were regularly made to other bishops, especially those in influential positions in America and Rome—a practice which, as it transpires, is perfectly common. As one veteran Vaticanista puts it: “the impression one gets from bishops’ public statements is that very few of them thought anything was strange about the money going around. It’s just what high churchmen do, at least in the US” (
Altieri 2019). McCarrick, too, was noted for his ‘customary gift-giving and donations’ (
Holy See 2020, p. 4). Indeed, he and Bransfield worked closely together on the Board of a major US-based Church fundraising charity, the Papal Foundation (
O’Brien 2018;
Flynn 2019).
Several bishops and cardinals who received Bransfield’s checks have since made clear that these gifts came with no strings or conditions attached, and thus were in no sense “bribes”. Maybe so. However, as generations of social scientists are all too aware, cultures in which reciprocal gift-giving is an embedded practice invariably tend to produce complex (and networked) relationships of trust, indebtedness, solidarity, obligation, and counter-obligation, even if the actors are not themselves fully conscious of them. That said, one presumes that McCarrick himself, having earned a PhD in Sociology from the Catholic University of America (
McCarrick 1963), might not be wholly unacquainted with the classic theories of
Malinowski (
1922) and
Mauss (
[1950] 2002).
4. Social Network Analysis: An Introduction
Since the precise techniques and jargon of SNA are not typically familiar to sociologists and other scholars of religion (
Everton 2018, loc. 549 n. 6), we feel it may be helpful here to provide a brief primer before proceeding.
A social network is defined as a collection of a finite set of actors (nodes) and the relation/ties between them (
Robins 2015). Actors (nodes) are defined as discrete individuals or groups and usually they have characteristics or attributes. The terms actors and nodes will be used interchangeably in this paper. As for ties, it seems that there is no formal definition in the literature except that they establish a linkage between a pair of actors (
Borgatti et al. 2013).
Network Centrality. Social selection processes would determine the nodes position in the network. Nodes in a central position tend to be regarded as having more prestige, influence, power, autonomy, and so on. Studies found that people in a central position tend to reap certain benefits such as having better grades (
Bruun and Brewe 2013), higher personal accomplishment (
Shapiro et al. 2015), better organizational citizenship behaviour that leads to better well-being (
Tsang et al. 2012), and so on. People in the central positions of the network might also determine the life and death of the network. In criminal networks, for example, it is sufficient to take out the key players to pacify the whole entire network (
Sparrow 1991).
There are several measures of centrality. In this paper, we would like to focus on degree centrality. Degree centrality refers to the number of edges (or ties) that a node (or vertex) has (
Robins 2015). The higher the number of edges/ties, the higher the degree centrality of the node/vertex. Nodes with high degree centrality would be highly visible and be considered as important in the network. Most of the information that runs through the network would pass by these nodes, giving them a clearer picture of what is happening in the network. Decision makers, leaders, and social influencers tend to be in such positions. Based on this premise, bishops/cardinals that have a high degree of centrality in a network would have more power or influence to control the network and they will have more knowledge about the network. The high number of people that they have direct contact with would mean that they have the potential to diffuse information to the network quickly.
In a directed network, degree centrality is divided into indegree and outdegree. Indegree is the number of ties that a node received, while outdegree is the number of ties that a node sent. The interpretation for indegree and outdegree is quite different. In a network diagram of which bishops have served under whom, a high indegree would mean that a bishop had a lot of others serving under him, while high outdegree would mean that this bishop had served under many bishops. Our focus is directed into indegree centrality as we theorize that when a certain bishop has high indegree, he would have more power over the network as he has more people in the network that was “moulded” by him. There are still many other measures of centrality such as betweenness, beta centrality/bonacich, closeness, and k-reach centrality, but those are beyond the scope of this paper.
5. Pilot Study: England and Wales
Before creating the USCCB dataset, we felt it important to do a test run based upon a much smaller dataset. This would allow a number of basic methodological choices regarding the scope (people, timeframe) and coding (what types of “tie”, and how differentiated) of the underlying adjacency matrix to be honed effectively: repeat “do-overs” are not so easy with a manually compiled 424 × 424 matrix, as our US dataset turned out to be. The Catholic Bishops’ Conference of England and Wales was selected for this sandboxing stage, due both to its size (with 27 dioceses and equivalents,
5 compared to the US’s 197), and to one of us being familiar with its personnel and reported dynamics.
Naturally, there are many ways in which bishops might be socially “tied” to each other: e.g., attending the same seminary or university, being involved with certain apostolates, or simply through getting to know each other at various (formal or informal) get-togethers. This is particularly true of the English and Welsh Church, which is a small world both geographically and otherwise. For reasons outlined above, however, (i) the special, sui generis relationship between an Ordinary and his literal subordinates, and (ii) the significant influence of an individual’s current and former Ordinaries over whether he becomes a bishop himself, and if so, over his subsequent episcopal “career”, appear to be of overriding importance. This Ordinary–subordinate “tie” is not a uniform one: it is reasonable to hypothesize differing “strengths” to the relationship based on different roles that a given priest/bishop has served in under a given bishop.
Given these assumptions, the basic sample was restricted to all living bishops who either were, or had been (i.e., emeriti, many of whom remain active in pastoral and administrative work), members of the Catholic Bishops’ Conference of England and Wales. This excluded any papal diplomats from England and Wales who have been ordained as (arch)bishops, but who are not and never have been CBCEW members. It included, however, one bishop (Roche) now assigned to a fulltime role in the Roman Curia, but who was previously the Bishop of Leeds. Data was originally collected in mid-October 2018, then slightly updated in October 2019 (for publication of our original working paper). Ties were defined according to which bishops the members of this original sample had served under (including any now deceased, though these are not included in the below network map), coded for three different “weights” of seniority/trustedness of role:
- (1)
Whether A has served directly under Ordinary B in some capacity (i.e., as a priest of B’s diocese, or–in the case of Bishopric of the Forces–as a priest on loan to the Ordinary).
- (2)
Whether A has served in one of a small set of especially high-trust diocesan positions (Vicar General, Episcopal Vicar, Chancellor, seminary Rector, Private Secretary within diocese) under B, or as General Secretary of the Bishops’ Conference under B as Cardinal-President.
- (3)
Whether A has served as an Auxiliary or Coadjutor Bishop under B.
6
Where A has served B at one or more different levels—e.g., a stint as an Episcopal Vicar (2), following on from many years as a regular parish priest (1)—the tie was recorded in accordance with the highest level (so 2). Given the precedence accorded to the Ordinary–subordinate relationship, cases where, e.g., C was a religious priest serving in a parish in B’s diocese, but not
under B’s canonical authority (that is, C’s Superior remained his own Abbot or Provincial), were not counted as a tie (though serving under B at levels 2 or 3 would be). These data were collected and inputted manually into Excel, creating an 80 × 80 adjacency matrix (including those deceased). One virtue of this coding methodology is that, in principle, these ties and their weights are based upon objective, publicly available data: whether or not B has served under A, and when and in what formal capacities, are facts of the kind that would feature in any “ecclesiastical CV”. A variety of web-based sources were used, including biographies given on official CBCEW and diocesan websites, individual bishops’ Wikipedia pages, and (for dates of ordination and episcopal appointments)
http://www.catholic-hierarchy.org/. (accessed on 1 November 2018).
Figure 1 shows the resulting network map of all living CBCEW members, active and retired (names in parentheses), constructed using the NodeXL software package (
Smith et al. 2010). The graph was drawn using the Harel-Koren Fast Multiscale algorithm (
Harel and Koren 2002), then lightly amended for ease of presentation. Edges are directed, with “A → B” signifying “A served under B”, and are assigned a tone/width corresponding to the three weights outlined above (with 1 being the lightest/thinnest, and 3 the darkest/heaviest). Nodes are sized according to their indegree (i.e., the total number of other bishops/nodes, irrespective of edge weights, who have served under them). Of 58 nodes, 40 form part of a single constellation, connected by 49 edges; a further 7 are one of three much smaller constellations of one or two edges each; and 11 are isolates. The main constellation has a diameter of 9, with a mean distance between nodes of 3.8 (NB: both here, and in the following two figures, there are a number of bishops with no ties to other bishops in their respective datasets. This is normally for one of two reasons: either they have been appointed to the bishops’ conference from outside of the usual Latin-rite diocesan set-up, e.g., from a religious order, or as a member of one of the Eastern Catholic eparchies; or else, they have outlived any other bishops whom they served under or who served under them. For completeness’ sake, such ‘untied bishops’ are arranged at either the top or bottom of the relevant figures).
The purpose of this pilot study was simply to test out various sampling and coding strategies, and to see whether the application of SNA to a Bishops’ Conference could produce a meaningful network map. This it certainly appears to have done. The influence of certain Metropolitan Archbishops, due both to their special role in the nomination process, and their typically having several auxiliary bishops under them, is clear: all 7 nodes with an indegree of 3 or more are, or were, Archbishops of Westminster (Nichols), Birmingham (Nichols, Longley), Southwark (Bowen, McDonald, Smith), Liverpool (Kelly, M. McMahon), and/or Cardiff (Smith). Furthermore, the graph clearly shows the overwhelming influence and centrality of Cardinal Nichols, with a quarter of all bishops (15 out of 58), and a third of all non-retired bishops (11 out of 34), having served under him, in most cases as auxiliaries. Nichols’ centrality is, moreover, probably greater than that of his recent predecessors, having led another large metropolitan archdiocese for nine years prior to his translation to Westminster (which neither Hume, 1976–1999, nor Murphy-O’Connor, 2000–2009, had).
The fact that the graph conforms, in these and other basic respects, to what any observer of English and Welsh Catholicism “could have told you anyway” is a good sign: a network map seeming to show, say, Nichols as a marginal figure within the CBCEW, or one of the long-retired auxiliaries as its kingpin, would be (correctly) suspected of having fatal methodological flaws. That said, the above graph does
not simply represent a wildly convoluted and laborious method of arriving at obvious facts. It also, arguably, sheds light on the more subtle dynamics of CBCEW politics. Two examples may be given here. Firstly, it was reported that, prior to the 2016 EU referendum, a Nichols-backed attempt to put out an official CBCEW statement in favour of Remain (i.e., anti-Brexit) was thwarted by a minority of bishops (
Thompson 2016). If one knew nothing more about the Bishops’ Conference than the information presented in
Figure 1, then a betting person might reasonably suppose this rebellion to have been fomented within the relatively dense “Southwark cluster” towards the left of the graph, most of whose members are at two removes from Nichols’ direct circle. This was in fact the case, with Peter Smith, then Southwark’s Archbishop, reportedly leading the charge. Secondly, there are two bishops whose occasional public statements have been widely perceived as being out of step with the CBCEW consensus on certain “hot button” issues (e.g., on the reception of communion by the divorced-and-remarried, or by politicians supporting abortion or same-sex marriage). Devotees of SNA will not be surprised to learn that these two, Egan and Davies, are a) themselves closely tied, with both having served under Noble, and Egan having been Davies’ Vicar General; and b) out on a limb, in both senses, in terms of the wider CBCEW network (Egan, at 6 removes, is the joint-furthest vertex from Nichols in the main constellation).
7. Discussion
These exploratory studies offer an original contribution to several live issues regarding episcopal structures and governance. As can be seen from our maps of bishops of England and Wales, and United States bishops, they support the view that it is meaningful to talk in terms of there being defined “cliques” of bishops. While previous discussions of this phenomenon have tended to be limited to considering the direct relationships between a single bishop and his various “proteges” (e.g.,
Dreher 2018a), our network visualizations reveal more complex kinds of clustering. That is to say, whether bishops B, C, and D have both served under A is not the only factor to take into account. Important too is whether B, C, or D have served under each other (e.g., in our England and Wales network, Wilson and Stock have both served under Nichols, but Wilson has also served under Stock), or under a bishop Z of whom bishop A was himself once a protégé (and so on). As is clear from
Figure 3, the US bishops do indeed tend to cluster into various more-or-less dense “flocks”.
It is important to state that network proximity to a McCarrick (or whomever) is not itself any indication of another bishop’s knowledge of, complicity in, or imitation of a McCarrick’s misdeeds. Indeed, we know from other cases that sexual offenders are often adept at hiding their crimes, including from those closest to them. Nevertheless, given the nature of both the clerical Ordinary–subordinate relationship, the outsized influence that a priest’s own (former) Ordinary/ies has/have in advancing his own ecclesiastical career, and the great extent to which “bishopcraft” is passed on through a form of apprenticeship, then it would be naive in the extreme to ignore this “network dimension” entirely. We wish to highlight two significant issues where we think this kind of network-thinking might make a direct, positive contribution.
The first relates to several inquiries into “historical”
8 sexual abuse and its handling by church authorities, not least those commissioned by multiple states’ Attorneys General. Apparently without exception, these reveal a “standard response to reports of abuse by church leadership” (
Missouri Attorney General 2019, p. 2) practised with remarkable consistency across different dioceses. As one scholar has recently summarized it:
A more-or-less standard pattern emerged: credible allegations against an abusive priest being kept quiet, with assurances made to the victims and their families; the priest in question being quietly reassigned, perhaps after a period of ‘successful’ counselling at one of a small group of Church-run treatment centres specializing in precisely this, or sent to another diocese with a glowing letter of recommendation; no thought whatsoever being given to this new set of young people being put into very serious harm’s way; and this process being repeated, multiple times, for years if not decades (
Bullivant 2019, p. 225).
As the Grand Jury tasked with investigating six dioceses in Pennsylvania arrestingly puts it: “Special agents testified before us that they had identified a series of practices that regularly appeared, in various configurations, in the diocesan files they had analysed. It’s like
a playbook for concealing the truth” (
Pennsylvania Attorney General 2018, p. 3; emphasis added).
This begs the question of precisely
how this unwritten, secret “playbook” came both to be and to spread. There is no evidence of any explicit conspiracy between Catholic dioceses to create a set of norms or procedures to be followed in such circumstances, and yet it also seems a stretch simply to suppose that almost exactly the same “solutions” arose complete independently, by spontaneous generation, in each chancery. Much more plausible, we contend, is to view this metaphorical “playbook” as a set of routinized practices and norms, or
habitus (see
Bourdieu [1972] 2013), emerging and diffusing “organically” within and through ecclesial networks. SNA studies have explored the ways in which the creation and spread of social norms, of various kinds, are affected by differences in network structure (e.g.,
Sen and Sen 2010). Feasibly, the combination of (i) socially learned “bishopcraft”, with senior chancery roles serving as a quasi-apprenticeship for future bishops; (ii) the existence of both strong, densely clustered groups of bishops in particular regions; and (iii) weaker, though still significant, links
between these local groups, due to rarer cross-country promotions of bishops (cf.
Granovetter 1973); provides the ideal conditions for this unwritten “playbook” to become informally codified. Also amenable to serious SNA explanation, at least in principle, is the “underground railroad” shunting problem priests between dioceses: a rigorous mapping of which bishops loaned abusive priests to which other bishops could be a fascinating study in its own right.
The second issue we wish to highlight is the scope for very serious
conflicts of interest, especially among bishops with close ties to each other. The way episcopal appointments currently happen, it is frequently the case that bishops in a certain region are relatively densely clustered. This follows from several features (
not bugs) of the system, not least the tradition of auxiliary bishops being created locally (with the Ordinary having a fairly free hand in nominating candidates),
9 and the role that a province’s Ordinaries have in recommending candidates to nearby vacant sees. Undoubtedly, this system brings many benefits: for example, it is far more efficient than an unwieldy national system and increases the chances of bishops actually knowing the people they are putting forward. One of several side-effects, however, is that if complaints are made against the former bishop of a diocese, then there is a strong likelihood of the
current bishop being quite closely networked with him: even if neither has previously served under the other, the odds are good that they have mutual ties with other bishops who have. Indeed, this is precisely what happened with McCarrick. When the (arch)dioceses of Metuchen, Trenton, and Newark all made large cash settlements to McCarrick accusers in the mid-2000s, Metuchen’s Bootkoski and Trenton’s Smith (now deceased) had both been appointed by McCarrick as his auxiliaries. Meanwhile, Newark’s Myers, while not directly tied to McCarrick,
is tied to eight other bishops who served under McCarrick, including Bootkoski (see
Figure 4). Since new Ordinaries obviously “inherit” the auxiliaries and other clergy of their predecessors, these kinds of overlapping connections are impossible to avoid fully. It is not hard to think up scenarios in which bishops X and Y “sharing” subordinates B, C, and D, even in the absence of a direct tie between X and Y, might create conflicted loyalties between the two. After all, if Y is found to have been engaged in serious wrongdoing, then questions will naturally be asked as to what knowledge or involvement B, C, or D might have had. Since B, C, and D may now be X’s own closest collaborators, friends, and protégés, the temptations of a quiet, out of court, and/or (to use a favoured euphemism) “pastoral” solution to the problem may be irresistible. More invidious still, consider that B, C, and D may now have intimate knowledge of any “dirt” on X, which, if X goes public with accusations against Y, might itself be leaked.
This recognition has significant implications regarding new canonical norms issued by Pope Francis in 2019, specifically to address the question of how serious complaints against bishops are properly to be dealt with and investigated. One noteworthy feature is that primary jurisdiction is granted to a province’s Metropolitan Archbishop to investigate allegations against bishops, past or present, in that province. (If the Metropolitan is himself the subject of the allegations, then this role would normally devolve to the “most senior suffragan Bishop by promotion, to whom, if such is the case, the following provisions regarding the Metropolitan apply”;
Vos Estis, art. 8 §2). These canonical norms owe much of their genesis to the US crisis. Indeed, the fundamentals of this so-called “Metropolitan model” were widely reported to have been drafted by Wuerl and Cupich in the run-up to the USCCB’s own plenary meeting held in November 2018, something they themselves denied (
White 2018).
The new norms are, explicitly, alive to the potential for conflicts of interest:
The Metropolitan is required to act impartially and free of conflicts of interest. If he considers himself to be in a conflict of interest or is unable to maintain the necessary impartiality to guarantee the integrity of the investigation, he is obliged to recuse himself and report the circumstance to the competent Dicastery (Vos Estis, art. 12 §6; see also 13 §3).
The practical difficulty, however, is this: given the structure of episcopal networks, there would seem to be a very low probability of any Metropolitan being “free of conflicts of interest” if asked to investigate other bishops, past or present, of his own diocese or province. Even where no direct ties exist, there will frequently be other close ties between mutual subordinates or superiors, protégés, or mentors. Additionally, this is quite apart from other proprietary concerns that a Metropolitan might have about the moral, financial, and reputational liabilities of his own diocese and its personnel. Even apart from the kinds of Ordinary–subordinate ties which we have particularly privileged in this paper, given other kinds of ties between US bishops, one wonders how any senior bishop can ever be, and/or be seen to be, truly “free of conflicts of interest” when investigating “brother bishops”. Indeed, this much is clear from the very first “Metropolitan model” investigation carried out in the US (
before the model was enshrined in canon law): the above-discussed investigation of Bransfield by William Lori, Archbishop of Baltimore. As it turned out, Lori was himself one of the high-ranking prelates who received cash gifts from Bransfield: USD 7500 in total, including USD 5000 to mark his
becoming Bransfield’s Metropolitan. At Lori’s request, his own name was redacted out of the version of the Report sent to the Vatican, along with the names of ten other bishops (
Ferrone 2019;
Pogorelc 2020). Furthermore, our own network data show that while Lori and Bransfield are not closely tied, Lori served, if briefly, as an auxiliary of McCarrick’s in Washington (indeed, as ‘an extraordinary help, guide, and friend’;
McCarrick 2011, p. 5). McCarrick too received cash from Bransfield. In what other area of life would oneself and one’s close associates having received thousands of dollars in “gifts” from the person whose case one is investigating not obviously and automatically be an objectionable “conflict of interest”?
8. Conclusions
During this period of intense (deserved) scrutiny of various aspects of church governance and culture from many quarters, it has been common for commentators to think instinctively in relational terms. Even when the focus of attention has been on specific, named individuals, discussion has often looked beyond personal misdeeds and culpability to
also consider the extent to which individuals are enmeshed within, and thus both influence and influenced by, wider relationships and networks—and, therefore, also to the cultures, practices, and norms which they co-produce and co-maintain. Various examples of this from the reporting of the McCarrick case have been given above. Here, let us add the earlier and more general observation of two Catholic legal scholars: “Bishops in the United States are part of a particular ecclesiastical culture that has its own influences, attitudes, and beliefs. This culture significantly influenced their decisions regarding clergy sexual abuse” (
O’Reilly and Chalmers 2014, p. 215). If this is so, then tools which might help scholars, church officials, or indeed criminal investigators better understand the “social architecture” undergirding this culture are surely worth serious consideration.
In fact, our basic contention goes much deeper than this. An understanding of Catholic episcopal governance structures as “network[s] of social relations and interactions between actors” (
Crossley 2011, p. 1) is not
just a critical lens with which to investigate church scandals. Rather, social networks are integral to how such “ecclesiastical capital” (
pace Bourdieu) is organized and maintained at all times, whether functioning well or not. Saintly and heroic bishops are, that is, just as much a “product” of networks as are defrocked former bishops. True, cases of abuse and cover-up serve as particularly apposite case-studies—both due to their clear and present “public interest” nature (i.e., in exploring how they have arisen, and how future ones might be prevented) and to the level of investigative journalism that has gone into uncovering the full details (on which we have drawn, gratefully and liberally, in offering our own interpretations). However, they remain case studies supporting our much more wide-ranging claim. That is, the significance of SNA as one valuable, and hitherto untried, tool for understanding and/or improving Church governance. (As we hope is implicit in every paragraph of this paper, we do
not think that SNA alone—uninformed by the insights of other forms of theological or social-scientific inquiry—is a kind of methodological “silver bullet” for this subject, or indeed any others.).
Given all this, we wish to conclude by suggesting seven specific areas where we think a relational, network-informed approach has the potential to yield interesting fruit, several of which we are already working on/towards.
While we have privileged the Ordinary–subordinate tie here, there are of course others which may also be significant (
Pogorelc 2020). Among those that are, at least in theory, quantifiable, are: (co-)consecrators at episcopal ordinations; which seminaries/universities and years studied at; service together on certain committees or Boards of Directors; which other bishops are ‘honoured’ in a bishop’s own coat-of-arms. The degree to which these types of network overlap or correlate with each other would be a very interesting topic for enquiry. In addition, if we have some measure of the ‘ideological leaning’ of each bishop, we can input that information to the network(s) mentioned above to see if there is some similarity in ideology between bishops who were connected to each other. Gauging the ideological leaning can be quite difficult though because the difficulty in operationalizing it and there are many bishops who remain silent in terms of their ideological position.
Simply relying on the visualization, the amount of information that can be extracted is quite limited, especially if we are dealing with a large map. For that reason, we should use a more advance method of SNA to further understand the network pattern behind the visualization. One of our studies (
Sadewo et al. 2021) had applied core-periphery analysis from the Ordinary–subordinate tie to investigate which bishops were in the core of the network (i.e., key players). Further study can also use a more cutting-edge SNA such as Exponential Random Graph Model (ERGM;
Robins et al. 2007) to understand the network structure in more detail.
Extending Ordinary–subordinate (or other types of) ties backwards through time. To what extent are “kingmakers” among the previous generations of bishops still significant over current episcopal politics? Is it useful to think in terms of episcopal “family trees”, or even dynasties? (On this theme, see
Bullivant and Sadewo 2021 regarding Cardinals Spellman and Cooke, McCarrick’s own former Ordinaries in New York.) Does the influence of different bishops or dioceses wax and wane over time (perhaps with changes in the pope or nuncio)?
Given the specific policies made by some bishops’ conferences (USCCB included) for affirmative action in promoting ethnic or linguistic minority candidates, for example in the form of national lists, to what extent do these policies actually work?
Thinking practically and positively, if episcopal networks currently do exhibit tendencies towards certain “network pathologies”, how might these same methods aid in reforming them? Would a policy of appointing “qualified outsiders” (i.e., suitable bishops not already tied into regional clusters) to major dioceses help in mitigating the kinds of conflicts of interest suggested above, especially in helping to “clean up” scandal hit dioceses?
Moving beyond national episcopal politics, what might be gained from applying SNA methods to studying the Roman curia (itself a major object of scrutiny under the current pontificate). Which are really the powerful dicasteries? Which bishops, from which countries, sit on which especially influential Congregations (especially the Congregation for Bishops)?