keywords: Dublin Irish Ireland politics political clientelism broker brokerage clientelist urban

Personalism and Brokerage in Dublin Politics

published as : "Personalism and brokerage in Dublin politics", Irish Urban Cultures. Chris Curtin, Hastings Donnan and Thomas Wilson (eds). Belfast: Institute of Irish Studies, The Queen’s University of Belfast, 1993. pp. 79-98.

© Lee Komito; University College, Dublin¹
http://www.ucd.ie/lkomito

Introduction

The Republic of Ireland is often portrayed, by residents and outsiders alike, as a society where everybody knows everyone else, and, especially, knows everybody else's business. In the political arena, this has been phrased as 'personalism'. In academic discourse, this has often been translated into the terms clientelism and brokerage: voters going to politicians in order to obtain benefits from the government. There has been a wealth of discussion about clientelism and brokerage in Irish politics; indeed, it is one of the major topics for ethnographic research in the Republic of Ireland (see, for instance, Gibbon and Higgins 1974; Bax 1976; Sacks 1976; Higgins 1982; Komito 1984, 1989a, 1989b; and Wilson 1989). Although 'personalism' is often used as a coded allusion to political clientelism, it can be used simply to emphasize the personal and informal dimension of Irish politics (e.g., Schmitt 1973). This is especially true in the relationships between politicians and both partisan supporters and also voters. It is this personal dimension, as evidenced in Dublin politics, which this article addresses.

Most descriptions of Irish politics provided by anthropologists have derived from rural ethnographies, and the extent to which similar behaviours would exist in urban settings has not been clear. Research in Dublin shows that urban politics is both similar and different from rural politics. The nature of party politics appears to be the same; a rural party activist would feel quite at home at a urban constituency meeting. However, urban politicians and activists have less contact with constituents than their rural counterparts. The converse is also true: urban constituents have less contact with politicians and activists than their rural counterparts. This raises an interesting question: if urban constituents are less likely to have personal links with politicians, what are the consequences for the clientelist behaviour that has been characteristic of rural politics? Do constituents learn that they can go directly to state bureaucrats, or do they continue to use politicians as brokers? The interesting answer is that they do neither. Instead, they continue to look for personal mediators or brokers, if those brokers are no longer politicians. The search for trusted intermediaries remains important, although the strategies of that search are different. This raises the general issue of personalism and trust being a feature of Dublin politics, for both middle-class and working-class individuals. Why should that be the case?

In order to develop these points, this paper will first examine the daily activities of Dublin politicians, to illustrate the amount of constituency work that politicians engage in. Then, a short description of Dublin party politics provides evidence of the similarity between urban and rural party activities. However, there is a decrease in the amount of contact between politicians and voters in Dublin, as compared with rural Ireland. In this context, urban voters, like their rural counterparts, continue to look for advocates, but use different strategies and find different intermediaries or brokers. The class dimension of urban brokerage will be discussed and, finally, some general issues about Irish and urban politics will be addressed.

A politician's diary

Government in the Republic of Ireland operates at both a local and a national level, with independent elections for each tier. Unlike many such democracies, most national politicians also hold local office at the same time. The only exceptions are Ministers and Ministers of State, who are expected to resign from local office once appointed to their Ministerial position. Local office is usually a stepping stone to national office but, because of the nature of competition in Irish politics, most national politicians prefer to retain their local office even after election. Thus, many politicians were first elected as councillors to Dublin Corporation or County Council, and have also been elected to one of the houses of the Oireachtas (parliament): either Dáil Éireann and Seanad Éireann. The 'lower' house, Dáil Éireann, is the more powerful, and successful politicians are likely to be Teachtai Dála (commonly abbreviated as TD, and the equivalent of a British Member of Parliament) as well as local councillors (for a fuller description, see Coakley and Gallagher 1992 or Chubb 1992).

Partly as a result of wearing both a local and a national hat, the daily diary of a national politician makes impressive reading. He or she will attend the national parliament three days a week, and also have at least one day of local government meetings as well. Until the 1980's, salaries for national politicians were not sufficient to enable someone to be a full-time politician, and so many politicians continued in their occupations outside politics (this is less likely to be the case now, as most national politicians give up, or take leave from, their full-time occupations). Earning a living and attending local and national government meetings accounts for their weekday activities, and parliamentary sittings may last until early or even late evening.

Then, on top of these activities, one must add other meetings in the evenings and weekends. These fall into three categories: party, community, and individual meetings. The political parties all have a three tier structure of national, constituency, and branch organisation. A politician, whether local or national, is dependent on the support of the fifteen to twenty-five branches in his or her constituency, and these branches meet, individually, about once a month. Politicians try to attend the meetings of branches, which means that politicians could have fifteen to twenty-five evening meetings a month.²

In addition to party meetings, there are community meetings: at any well publicized local meeting, one finds politicians, and indeed, at all community events (such as the opening of a new church or new community hall), one finds politicians. If a local benefit, virtually anywhere in the constituency, is taking place, politicians will make an appearance. Community groups and residents' groups all expect politicians to attend their meetings and respond to their problems. Thus, during any week, politicians could easily have up to seven or eight local events or meetings to attend.

In addition, politicians also hold advice `clinics' one to four times a week. These are well publicized: they are advertised in local papers, information sheets are put into houses, and notices are posted in community centres. Anyone in the area can consult about their problems or difficulties, and politicians could see twenty or thirty people in each clinic session. Even at home politicians receive phone calls, letters, and personal visits regarding constituents' cases.

Combining all these obligations together means, to say the least, a very hectic political life. Many a politician leaves a local government council meeting in order to briefly attend a community or residents' meeting (where the information he provides is often based on material obtained from the earlier council meeting). After the residents meeting, he might have a party branch meeting where he wants to have a drink and chat with branch activists. Clearly he or she is not a full participant at any of the meetings, but they are seen to be there, and are able to keep in touch with all the various happenings in their constituency.

There is no reason to think that rural politicians have any less hectic a life; ethnographic reports (e.g., Higgins 1982) and interviews with rural politicians indicate all politicians devote an enormous amount of time to constituency activities. In a review of government salaries, the Review Body on Higher Remuneration in the Public Sector (1972: 211) asked TDs to estimate how much time they spent on political business other than attending the Dáil. Forty-four percent said 31-50 hours a week, and thirty-seven percent said over 50 hours. In addition, the TDs were asked to list the kinds of expenses they incurred. This partial list illustrates the cost of maintaining a social presence: subscriptions to organisations (mentioned by 52% of TDs), travel in constituency (52%), entertainment (46%), telephone (18%), attending clinics (17%), funeral offerings (7%), maintaining constituency office (7%), telegrams (5%), and wedding presents (4%).

Politicians would say that keeping in touch with government bureaucracy, political party, and community, even if hectic, has one major advantage: they keep their hand on the public pulse, and know what concerns people. This image corresponds with the ones presented of rural politics in Ireland by a number of ethnographic studies (Bax 1976, Sacks 1976, Carty 1981). These rural studies described politicians who are well connected with all the happenings in their electoral constituencies. In these ethnographies, it is clear that close attention is paid to the constituency in order to monitor political support. They are constantly 'nursing' their constituency to maintain their electoral support, and are keenly aware, at any given time, where their support comes from. A quintessential symbol of Irish politics was an interview broadcast over Irish television in the early 1970's. It pictured two political activists, one from Fianna Fáil and one from Fine Gael, standing on a hill, looking out over a valley in Donegal. The two them discussed each of the houses visible from the hill, and were able to identify how members of each household voted. This is personalism exemplified.

Party politics

The description of politics is one of personalism, but also competition and factionalism. Rural ethnographies are replete with stories of politicians trying to do down one's party rivals, trying to control the local party structure, trying to make bigger public claims that rivals, getting jobs for `special' friends, and so on. Political life in rural Ireland seemed to be continual party in-fighting, as well as maintaining personal links with the voters (see Bax 1976 for the most vivid description). The picture of urban party politics that emerged from a study of politics in Dublin in the late 1970's and early 1980's (Komito 1985) is little different. Political competition was largely intra-party, rather than inter-party. This was the result of structural factors. The Irish electoral system combines proportional representation with multi-seat constituencies. Thus, in a five seat constituency, voters can rank order their preferences, choosing amongst all candidates. Since there has been, until recently (Laver and Marsh 1992), a tradition of loyal party voting, politicians are in competition with other politicians from the same party for the votes of loyal party supporters. Between elections, politicians are always aware that, when the next election is called, it is their party colleagues that will pose a threat and so it is those colleagues who must be feared and watched. The local party structure is one arena for these conflicts, as politicians try to prevent their rivals from gaining control of the local party machinery.

While this conflict is latent much of the time, tensions manifest themselves during nominating conventions. If politicians cannot control which of the party candidates the voters chose, they can control who is nominated in the first place. Candidates are nominated by representatives of the constituency branches; if a politician can control the branches and the nominating convention, he or she can ensure that potential rivals do not receive a nomination. Preparations for nominating conventions are made years in advance. Politicians do their best to make sure that supporters hold important positions (such as secretary or president) in the branches; these are the individuals who can persuade other branch members to support particular candidates and are usually the branch representatives at selection conventions. One politician, for instance, planted three of his close relatives in three different branches, rather than waste them all in one branch. Politicians are often suspected by rivals, and party headquarters, of maintaining 'paper branches' -- these are branches that exist as far as the party organisation is concerned but whose only active members are the branch officials who vote as the politician wishes.³

While 'paper branches' and loyal kin in control of other branches are preferred means of safeguarding one's position in the constituency party, most branches in a constituency are either controlled by no single politician or are themselves factionalized into different camps. The monthly branch meetings are very complex; beneath the surface of conviviality and commitment to party ideals, there exist plots and counter-plots. Participants spend most of their time trying to deduce the significance of every minor event: does it imply that someone's support is shifting? Will that action somehow enhance councillor `L's' position? For example, at a constituency-wide meeting, the constituency secretary placed a local councillor in the front row of the meeting, while a local TD and also a visiting party dignitary (a TD from another constituency) were at the table facing the audience. This was an exception to the normal practice of seating all elected politicians at the head table. The TD thus received more public attention than the councillor. The councillor felt he could not make a fuss, as this would appear mean-minded, but he was annoyed and also worried: had the secretary done this deliberately (thus aligning himself with the TD whose position was increasingly threatened by the councillor)? Since no direct question could be asked to settle the matter, the councillor could only resolve to not let himself lose the limelight again and also watch the secretary more closely in the future. At selection conventions, the conflicts manifested themselves more transparently. In one convention I observed, attempts to control votes involved late night phone calls, sometimes using outright lies about rivals, to change the votes of delegates. The convention ended with almost overt accusations of betrayals, as the deceptive phone calls made the previous night were revealed in an angry post-mortum.

During the selection convention, various procedural and rhetorical strategies can be used to advance one's position. All this conflict takes place beneath the surface; many of those attending such meetings accept a superficial interpretation of harmony and shared goals. Indeed, such shared goals are often used as a weapon in party rivalries, following from Bailey's (1969) distinction between moral rules and pragmatic rules. In one case, an aspiring activist tried to obtain a nomination, and threatened to supplant the established councillor. Although the aspirant had more votes, the established councillor was able to paint a convincing picture of factionalism and divided loyalties. For the sake of party unity, he said, they should ask the national executive to permit both to run as candidates. If the aspirant had fought this, he would have seemed concerned only with personal ambition (in contrast with the other's concern for the party). He had no choice but to agree. The established politician was easily able to use his national contacts to influence the national executive; as he had expected, they decided to permit only one candidate, which was the established politician himself.

Competition between members of the same party sometimes becomes so intense that politicians prefer that members of opposition parties get elected, rather than rivals within one's own party. In one case, this preference led to a sitting TD keeping the opposition informed regarding the campaign schedule of an up and coming party rival. In this way, any publicity tactics could be countered by the politician in the other party; as far as the TD was concerned, far better that a politician from another party get elected than a rival within his own party have a chance to build his electoral base. Needless to say, the party leadership would have taken quite a dim view of this, had they known, but the sitting TD was more concerned with his career than with the overall success of his party.

These descriptions of complex party machinations would be familiar to anyone studying politics in the Republic of Ireland. Party activists are clients of patrons, or rivals of other clients, but never simply neutral participants. Party politics appears to be much the same anywhere in Ireland, whether one examines a rural or an urban constituency. Indeed, this is a picture that would find echoes in factional politics elsewhere as well, as first evidenced by Barth's model of bloc alliances composed of opposing groups at each level of the political hierarchy (Barth 1959). An important feature of factional politics in the Irish context is that their existence in the context of party solidarity and collective identification. Activists in all political parties feel as strongly about party membership as others, in different societies, would appear to feel about ethnic membership (see Cohen 1979, 1981 for further discussion of these issues). Individuals 'inherit' their party membership from their parents and few activists that I talked to would ever think about voting for a different party. The differences between parties focus on historical principles (pro or anti-Treaty) rather than party policies, but the differences even extend to personality attribution. Fianna Fáil activists see Fine Gael as less than fully Irish, too closely linked with British interests, and not really close to "the people". Fianna Fáil activists were completely serious when they would talk of "a Fianna Fáil person" or "family"; and the label carried with it specific obligations of behaviour towards other members of the same moral community. Fine Gael activists, on the other hand, saw themselves as morally upright and honest, while Fianna Fáil was seen as immoral and capable of doing or saying anything for the sake of a single increased vote. It is only within the context of these strongly held loyalties that internal party factions exist.

Politicians and the community

Party politics and the constituency activities of politicians in Dublin seems similar to politics in rural Ireland. However, the picture changes considerably when one looks at the relationship between politicians and voters. In rural ethnographies, politicians are described as having close personal contacts with voters. These contacts are developed through business and community activities, and also arise out of extensive networks of blood, marriage, and friendship relations. In Dublin politics, it would appear that politicians are unable to cultivate and depend on such an extensive network of political support. Politicians and activists tend, in Dublin, to exist in their own private world, somewhat apart and isolated from the broader society of their constituents.

One indication of this is the process, already discussed, by which party candidates are selected. As already indicated, selection conventions are occasions for overt factionalism. The basis on which candidates are chosen is also interesting; in the selection conventions which I observed, there was often some effort to chose candidates who would bring in votes from particular geographical areas of the constituency (similar to the process observed by Sacks 1970). The presumption was that people would vote for their local candidate and, having voted for him or her, would vote for other candidates of the same party. It must said, however, that this public rhetoric did not match the reasons given by politicians in private. They were happy to select candidates representing a geographical spread because they did not want their own local power base threatened by giving publicity to an aspiring candidate in the same locality. The number of electoral votes that potential candidates might obtain is rarely the basis for selection; in a number of cases, individuals selected perform very poorly in the election. This became such a problem in the 1980's that the party headquarters, for all parties, began to impose candidates on the local organization, if the local selection procedures did not appear to maximize the party vote. It would appear that this problem has not diminished over time. As recently as April 1993, a prominent member of Fianna Fáil, who had been given the task of improving party performance in urban areas, noted that "Too many candidates spend two or three years trying to win a nomination rather than out canvassing and dealing with the public" (Irish Times: 20 April 1993). At least in one party, activists and politicians in Dublin are still out of touch with voters.

Another indication of this was a change in election rules. Until 1965, only the names, but not the party affiliation of candidates, appeared on ballot papers. After 1965, however, the election rules were changed so that party affiliation could also be displayed alongside each candidate's name. The implication is clear: up to that point, political parties were confident that voters knew which parties the candidates were associated with. At that point, however, the parties recognized that social change, including increasing urbanization, meant that such a link would now have to be explicit. No longer could parties assume that voters knew, personally, the politicians who were looking for electoral support.

The social basis for party activity differs between urban and rural constituencies, with implications for close links between party and communities.⁴ Branch numbers and membership varies between rural to urban areas; in 1979, Fianna Fail had about 50-150 branches per rural constituency, and 20-30 branches per city constituency (Chubb 1982:112). One informant suggested that, in Fianna Fáil, fifteen would about the smallest number of branches for an urban constituency, with an average of twenty-five, whereas a rural constituency could have as many as seventy, subdivided into four sub-groups. Although some rural branches might have a small membership, another Fianna Fáil activist estimated that a rural cumain (branch) could have 100 members to an urban branch's 15, with the same catchment area (in terms of voting population). Furthermore, as another activist noted, that 100 members would probably include entire families, whereas urban families do not participate in the same manner. Party activists, from all parties in Dublin (particularly those from middle-class constituencies) emphasized again and again how isolated they were from the community, as contrasted with their rural counterparts. Urban activists might not see one another in the normal course of daily life, and formal branch meetings are necessary to sort out problems and discuss tactics. For rural branches, however, the formal meeting is a formality; members would have already seen each other, naturally, in the course of their non-political lives.

This apparent gulf between the political parties and the urban social environment in which they are situated arises for a number of reasons. One factor is the structure of political representation, in which the number of politicians per head of population varies widely. At national level, the constitution requires that representation, numerically, "shall so far as it is practicable, be the same throughout the country" (Article 16.2.3, 1937 Constitution). However, there is no rule as to the number of councillors for a particular area, and wide variation is possible. Urban areas tend to have fewer councillors than rural ones; the ratio in County Dublin was, in 1980, 1:7,151 voters and in Dublin City 1:8,088 voters. At the same time, there was a ratio of 1:942 in Leitrim and 1:1,875 in Clare (Roche 1982: 312-320). This means there is closer contact between councillors and voters in rural areas, simply by force of numbers. This also alters the relationship between councillors and TDs, since rural counties have a large number of councillors relative to TDs while urban ones have a small number. In 1980, there were 806 city and county councillors and 166 TDs, or a ratio of 4.86:1; yet, in the Greater Dublin area the ratio was less than 1.7:1. In rural politics, the TD must tolerate councillors who are simultaneously his helpers and also potential rivals. In urban Ireland, since most TDs remain councillors, urban rivals of TDs can be denied the stepping stone of local elected office, which helps explain the degree of factionalism evidenced at selection conventions.

There are also significant social differences between urban and rural communities. Urban communities tend to be more amorphous and anonymous than rural communities, due to geographical mobility and lessening of the extended family.⁵ This has important political consequences because of the issue of personalism and political support. A feature of rural politics is the ability of politicians to develop personal links with voters, so that politicians have their own blocs of votes as well as the support of intermediate brokers who can deliver blocs of votes. Bonds of consanguinity and affinity alone can, in a stable community, deliver hundreds of votes and constitute a strong core of electoral support. Other personal links, once developed, can also be relied on. Clearly, however, in urban communities, the resources, if not the rules of the game, change. As individuals move from one house to another, or one job to another, any existing links to politicians would become irrelevant, as the voter moves to a new constituency. This is a transformation that Dublin politicians are keenly aware of.

The lack of a personal link with voters is evidence by Dublin politicians at election time. They know the areas where they are strongest and most likely to obtain votes, but this bears no necessary relation to the areas in which voters visited clinics. People clearly do not vote on the basis of favours owed as a result of personal efforts on behalf of individuals. As has been noted elsewhere (Komito 1989a), much of a politician's time is spent trying to transform an instrumental exchange relationship (from the voter's perspective) into a personal and moral relationship. While this may be the politician's goal, the starting point is the voter who feels no particular moral obligation or personal connection with any particular politician.

Interestingly, one Dublin politician who successfully built up a network of personal contacts, did so on the basis of pre-existing links in the west of Ireland. The politician, as a candidate new in the constituency, organised his campaign through a pyramid of personal contacts. Contacting people who had moved to Dublin from his home county in the west, and whom he knew through friends, secondary school, and university, he asked some to organize small meetings and he asked others to name ten friends who might help him. He held numerous `coffee mornings' where he met small groups of housewives. Thus, he literally organised `friends of friends' to quickly penetrate the constituency. The result in the election was that he was the first person to reach the election quota.

In this context, the previous description of a politician's daily diary takes on a new meaning. Rather than evidence of how well in touch politicians are with individuals throughout their constituency, it demonstrates how out of touch politicians know themselves to be. They are trying, with the busy schedule of activities, to create the bonds of personalism that their rural compatriots can largely take for granted. Politicians are also clear that many of their activities are a waste of time. Clinics, for instance, appear to epitomize personalistic politics -- individuals are asking politicians for personal assistance, and, as they receive it, they become clients of the politician and vote accordingly. The evidence is otherwise. Voters make the rounds of all the politicians, trying to play one off against the other. Even if they are helped, there is no certainty that they will vote for the politician at the next election. Politicians are all well aware that clinics are a very mixed blessing indeed; the major reason given for holding clinics is their publicity value. It is important that voters in an area feel they are getting some attention from the politician; without it, they might decide to transfer their votes to a politician who demonstrates greater concern for the area. Thus, the clinics are part of the general strategy of maximizing a reputation in the local community, rather than a means of obtaining the support of specific individuals.

The weekly round of meetings can also be given a different interpretation. Politicians do not believe that contacts made at such meetings create personal bonds of electoral obligation; instead, past participation at these meetings become, at election time, a basis for attracting support from anonymous voters. The constituency of Dublin North provides an example of local connections which a politician creates, and depends on, for recognition. Thomas Wright was an aspiring Fianna Fáil candidate in the 1982 election. In addition to owning a fish and poultry business in the centre of Malahide, he was Malahide `Community Personality' for 1981. He was chairman of Malahide Festival for the previous six years, a current member of Bord Iascaigh Mhara (the National Fisheries Board), Portmarnock Community Centre, Malahide Chamber of Commerce, and the Malahide Tennis and Cricket Club. He further noted that he was the coach of the Irish schoolgirl's international basketball team, as well as member and former player in the local Gaelic Athletic Association (GAA) club and remained active in GAA affairs. His Fine Gael rival for the marginal seat was Nora Owens. She advertised that she had previously been editor of the local newsletter, as well as having been involved with resident's associations, community councils, school boards, the Old Malahide Society, and the Tidy Towns Committee. Both candidates hoped to create an identification with the locale, in hopes of gaining extra votes. Neither presumed that previous activities in these groups were sufficient to ensure votes at election time.

Voters, trust and personalism

Descriptions of rural politics emphasize the personal link with politicians, and, even though politicians may cultivate such a link to ensure electoral support, it seems clear that the voters themselves want that personal link. In ethnographic explanations, this personal link is demanded because people do not trust anonymous civil servants in Dublin or anonymous local offices in the county council. They can only imagine extending trust to individuals that are known to them personally (Sacks 1976: 47-48; Bax 1970: 184). Part of the mythology of modernization has been that people would realize that a personal relationship is not a prerequisite for fair treatment; thus, as modernization proceeds, one should see a move from politicians as personal mediators to direct approaches to bureaucrats. What is the situation in Dublin? Ethnographic evidence certainly suggests that a personal link between voters and politicians is difficult to sustain in Dublin. If voters are faced with a choice between going to anonymous bureaucrats or somewhat less anonymous politicians, do they turn their attention to bureaucrats?

The evidence suggests that urban voters are not likely to approach agents of the state directly; they, like rural voters, want personal relations of trust. However, since the structure of social and political relations is different in Dublin, the strategies by which those personal relations are attained is also different. Part of the evidence for this comes from a survey designed to compare Irish `civic culture' with other countries' civic culture, as studied by Almond and Verba (1965). The survey was jointly carried out by the Economic and Social Research Institute (ESRI) and Stein Larsen, then at the Institute of Sociology at the University of Bergen in the early 1970's (and reported in Raven and Whelan 1976). Amongst other questions, the survey asked: "If you should have any problems with the authorities, which of the people here in the street (area) would you think of consulting? (We do not assume that you are personally acquainted with the person)." Only one person could be named; those who indicated they did have someone in mind were then asked to name the person's occupation. It is a particularly interesting survey because respondents were free to choose anyone; most comparable surveys only permit respondents to choose between politicians and civil servants. The survey did not specify Dublin versus other respondents, and the categories used to classify respondents were somewhat problematic, so it is only possible to distinguish, in a general way, between urban and rural respondents.⁶

In accordance the expectations of modernization theory, individuals in urban areas are less likely to chose politicians as the preferred broker. However, in contradiction to modernization theory, there is, however, only a marginal increase in preferences for bureaucrats. Respondents who do not prefer to go to politicians choose, instead, someone else in the local community, who, by definition, is neither a politicians nor an employee of the state. This individuals have been referred to, in the following survey, simply as 'community figures':

Table 1: Percentage distribution of respondents,
by brokerage preferences and area of residence
(adapted from Komito 1992).

Thus, out of a sample size of 848, just under sixty percent of respondents were classified as rural, and just over forty percent were classified as urban. This is roughly similar to the 1971 Census figures.⁷ While sixty-two percent of rural respondents would go to a politician, only forty-two percent of urban respondents would make the same choice. However, in the urban context, those going to officials (local authority officials, civil servants, social welfare workers, community health nurses, and so on) only increased to just over six percent. It is the residual category of `community figure' that shows a marked increase; in other words, urban dwellers may no longer go to politicians, but they still want go to some mediator rather than deal with the state directly.

The actual occupations which constitute this residual category of `community figure' is as interesting as the size of the category. Firstly, if urban and rural respondents are taken together, the occupations grouped together as 'community figure' include solicitor (chosen by 8.3% of respondents), priest (7.1%), farmer (4.7%), businessman (2.9%), and shopkeeper (2.1%). These high-status occupations are the ones most likely to be consulted when individuals need assistance. Looking only at the urban respondents, the high-status individuals remain important: solicitor (10.9%), priest (7.3%), businessmen (4.2%), farmer (2.0%)⁸, shopkeeper (2.8%). But, in addition, new figures emerge as significant choices: guard (2.5%), printer (2.0%), housewife (3.4%), chairman of resident's association (2.0%), teacher (1.4%), and publican (1.4%). There is a further 10 percent comprising occupations too diverse to summarize. Overall, the figures indicate a greater diversity of brokerage choices in urban areas than rural areas.

There is a relatively clear consensus about appropriate brokers in rural contexts: politician, solicitor, priest, and farmer account for 82 percent of respondent choices. However, no such consensus exists in the urban context, where a similar selection of occupations accounts for only 62 percent of choices. Urban dwellers utilize a wider range of brokerage choices than rural respondents. It is also notable that, in many cases, there are no obvious links of economic or political dependency between the individual and the broker, which would explain why that person was approached. Why, for instance, should an individual choose printer or housewife, except that they are choosing `friends? Thus, the decrease in personal contacts with politicians has not led voters to turn to officials; in their search for trusted mediators, they simply cast their brokerage net more widely.

There is some indication from the survey that middle class respondents were more likely than working class respondents to chose `friends' rather than politicians as brokers, although the number of respondents is too small for statistical significance. This would be in line with another survey (Komito 1989), and interview data also suggests that middle-class urbanites are likely to have friends who can act as brokers or mediators. Individuals who had obtained a third-level education (for instance, from University College Dublin) have a pool of potential brokers to draw on. In many cases, they know someone in the civil service or, failing that, someone in one of the professions (such as law or medicine) who can act as intermediary. This also happens in working class areas (one political activist worked for the telephone company and could provide contacts), but there are fewer instances of this, and the contacts provided are less likely to have any significant resources at their disposal. Thus, in working-class areas, voters are less likely to have any effective alternative to politicians.

One common feature of brokerage, whether middle-class or working-class, is its conspicuous absence of class rhetoric. This lack of class politics is unremarkable in the Irish context, since its absence is a feature of rural and urban, working-class and middle-class Ireland. It is remarkable, however, if one compares Irish urban politics with urban politics elsewhere in Europe, where class, ethnicity or religion are often the basis of collective political action (Mair 1992a, 1992b). Most of the brokerage that occurs (based on ethnographic research) is either individual or community oriented. That is, people usually want something for themselves as individuals (social welfare assistance, medical assistance, or public housing) or as a member of a resident's group (community park, improved street lighting, or provision of public buses). The only extent to which class differences manifest themselves is that working-class voters are more likely to want individual benefits for themselves, while middle-class voters are more likely to be concerned with community amenities.

Although the number of urbanites who approach politicians is less than in rural Ireland, it still remains the major occupational category (42 percent). Voters tend to approach politicians with suspicion rather than trust: in politicians' clinics, the rhetoric of exchange between politician and voter is imbued with the sense of a bargain being struck. People do not overtly say "I voted for you, so you should now come through", but this is the covert implication. Certainly they do say, explicitly, when explaining why they expect politicians to fill out forms that they, the constituent, could fill out themselves: "Sure, isn't that what he is elected for?" The voters are uncertain how much they can trust politicians, and hope that they have some hold over politicians by virtue of the vote. This helps explain the relative lack of interest in Community Information Centres (CICs). These Centres are staffed by volunteers, who have received training from the National Social Services Board. There are 77 such Centres around the country, and they provide information about administrative services on a confidential basis. Both local and national officials recognize that inquiries from such Centres are legitimate -- that is, they will respond as they would to a politician's inquiry (although not always as quickly). These Centres are used to some extent (in 1990 they dealt with 130,987 queries), but not to the same extent as politicians' clinics. People do not believe CICs are as effective as politicians, and, more significantly, they also do not believe they will work as hard as politicians. At least with a politician, one has the threat of the vote as a bargaining stick; one has no moral or personal hold over a stranger in a CIC. Thus, the preferred basis for assistance in Dublin continues to be some bond of trust, rather than any alternative institutionalized assistance; when that is not available, the next best alternative is an implicit transactional bargain between voter and politician.

Conclusion

To summarize, Irish party politics remains the same whether one speaks of urban or rural areas. Party politics, throughout the Republic of Ireland, is built on a pyramid of patron-client links. At each level (branch, constituency, national), individual activists and politicians compete with others, use the support of their clients in that competition, and are, in turn, clients of higher level patrons. These factional conflicts exist within each political party, membership of which is similar to the moral community of ethnicity. These maneuvers conform to the transactional models of politics discussed by Bailey and Barth.

The major difference between Dublin and the rural communities described by Sacks and Bax is the relationship between political activists and the voters. In rural communities, politicians provide a relationship of trust for voters who distrust the state. Although politicians do so in order to advance their own political careers, that does not alter the perception of individuals that such a trusted link is both needed and provided. In Dublin, however, relations of trust do not involve politicians; due to a widening gulf between politicians and voters, politicians are unable to provide that personal link. However, in the absence of such a link, voters continue to look for individuals who can act as personal links to the state. These individuals are not necessarily going to hold positions of power or prestige in the local community; the first prerequisite is that they be trusted.

To some extent, this differs from the expectations implicit in earlier rural ethnographies. In these ethnographies, brokerage was a consequence of beliefs and behaviours that are stereotypical 'peasant', such as a distrust of impersonal bureaucracy and a distance from central authority. In such a context, one would have expected Dubliners to be less likely to be dependent on personal contacts with politicians, and more likely to deal directly with state functionaries. Although Dubliners are indeed less likely to depend on personal contacts with politicians, they did not directly approach state functionaries. Instead, they continue to prefer personal links, but with other potential mediators.

This preference for personal contacts seems to operate independently of class. Although middle-class Dubliners tend to avoid politicians by using other brokers, they still prefer the mediation of brokers rather than directly approaching representatives of the state. Thus, in Irish politics, personal relations is not a phenomenon of rural, as opposed to urban politics; nor is it necessarily a phenomenon of underdeveloped or post-colonial politics. The urbanization or modernization process has not encouraged a shift away from personal contacts.

Regardless of class, personalism and trust are features of Dublin politics. This iIllustrates the importance of the personal dimension in the way individuals relate to state, following on from Eisenstadt and Roniger (1984). Is there any reason, unique to Ireland, that would explain this preference for personal contacts? The Irish civil service is criticized as secretive, authoritarian, and unresponsive; not only is little attempt made to release information, it seems there are positive attempts to conceal even trivial bits of information. As one expert of Irish public administration said,

"the operations within government departments and within state-sponsored bodies are almost entirely closed to public scrutiny. We know about what goes on within them only in so far as a conscious decision is taken to publish a decision or report, often presented as a remarkable act of magnanimity on the part of the body concerned." (Barrington 1980:191-2)

One politician observed that there was an "almost paranoid refusal by public agencies to divulge even the minimum information about projects or concerns of public interest" (Chubb 1982:379; see also Roche 1982, Komito, 1985). Perhaps this lack of access, which excludes middle-class as much as working-class Dubliners, explains the dependence of personal contacts. Irish bureaucracy has been made more accountable and more accessible in recent decades; will this alter the desire for personal ties?

The preference by Irish people for personal contacts with the bureaucracy has been noted before (Pyne 1974). Even though few actually need brokerage assistance, people still want politicians to act as brokers (Komito 1989). That is, despite people's own experiences that the formal system actually works reasonably well, they still distrust it and believe instead in the need for personal contacts. Furthermore, such personal contacts are perfectly possible in Ireland. As one famous Irish politician remarked in the 1960's,

"there is hardly anyone without a direct personal link with someone, be he Minister, TD, clergyman, county or borough councillor or trade union official, who will interest himself in helping a citizen to have a grievance examined and, if possible, remedied." (Chubb 1982:316)

In the 1980's, people could, if they wanted, still find a personal advocate to act of their behalf (even if that advocate was only a politician). Is this preference for personal contacts universally true, but only manifested in a small-scale society like Ireland, or is this part of the legacy often attributed to Ireland's "post-colonial" heritage?

1. The data for this article derives from ethnographic research carried out continuously from 1978-1981 in Dublin, intermittent research in Dublin since 1981. The research focused on two areas: a working class constituency in Dublin Corporation, and a largely middle class constituency in County Dublin. It also included interviews with politicians and local officials in the Dublin region.

2. A politician would make sure to attend any meetings in his or her local 'bailiwick' (see Sacks 1970 for a discussion of the bailiwick system), and might miss meetings in other parts of the constituency where his or her support was weaker.

3. Evidence of this came from interviews with party activists and politicians in all three political parties: Fianna Fail, Fine Gael, and the Labour Party.

4. For further detail on party organisation, see Gallagher (1985), Laver and Marsh (1992), or Mair (1987).

5. It is, of course, well known that urban communities are not always either amorphous or anonymous. The classic studies of working class communities in Dublin (Humphreys 1966) and London (Young and Willmott 1957) found dense networks commonly associated with rural communities. For a discussion of the urban versus rural distinction from a network perspective, see Southall (1973), Hannerz (1980), and Wellman and Berkowitz (1988).

6. Although this question was not extensively analysed in the original publication, the data was made available to the author for further analysis. The survey divided respondents into seven categories: 1) residential area in city centre, 2) main industrial area in town, 3) mixed residential and industrial/business area in town, 4) suburban area/council estate in large town, 5) large village (500+) in rural area, 6) small village, 7) sparsely populated area. The categories are neither clear cut nor unambiguous, and the safest strategy was to simply dichotomize the variable area of residence into rural (areas five through seven) and urban (all remaining areas).

7. The Census figures are 40 percent of residents in towns with a population of ten thousand or more, and 44 percent if residents of towns with a population between five thousand and ten thousand are included (Central Statistics Office 1972).

8. The choice of farmers by urban respondents appears puzzling. Within the general category of 'urban', it was those classifed as living in a "residential area in city centre" who chose farmers or agricultural labourers. There are a few possible reasons for this choice: these 'farmers' are urban dwellers who own land elsewhere, the respondents chose someone who they knew but who did not actually live nearby, or some of the respondents actually lived in rural areas and were classified incorrectly. It is not clear which of these is the correct explanation.

References