Modelling Union Democracy
— J. David Edelstein
WHILE SOCIALISTS FAVOR democracy in working-class organizations, there has understandably been widespread pessimism regarding the prospects for democracy in national unions under capitalism. The bleak American scene certainly provides little encouragement. My purpose here is to provide two important examples of large, fairly democratic unions, both in Britain, and to show the organizational basis for their democracy.
There are also implications here for democracy under socialism. Neither union has any appointed full-time officials, and both elect all of their full-time officials by direct votes of the membership. The unions are the Amalgamated Engineering Union and the National Union of Mineworkers. A widespread knowledge of these two unions' democratic features, among socialists and rank-and-file activists, could offer useful examples and inspiration.
American revolutionary socialists have rightfully emphasized rank-and-file struggles within the labor movement, within or through the unions and sometimes bypassing oligarchically dominated union structures. Within the unions the struggle has often been for democracy. But the democratic reform of a national union under capitalism is a major undertaking and, some would argue, an almost impossible task.
Nevertheless there may occasionally be opportunities for rank-and-file movements to democratize national unions by forcing through and institutionalizing changes. Miners for Democracy accomplished some of this, and Teamsters for a Democratic Union may now have such an opportunity. But even lost struggles for workers' democracy can develop the ideal of democracy and keep it alive.
Electoral competition is not the only criterion for democracy in large organizations, but its long-term absence in elections to top posts may reasonably be taken as a sign of oligarchic control, especially when incumbents are not running for re-election. Uncontested elections in filling top vacancies are extremely common in American unions, but rare in Britain, with the Amalgamated Engineering Union and the National Union of Mineworkers towards the extreme in competitiveness.
I'll first describe the unions briefly, present a model of democracy which conveys their conveys their essence, show how the organization of the two unions approached the model, and then present the evidence regarding electoral opposition.
The Amalgamated Engineering Union (AEU), formed as a craft union in 1851, is the oldest national union in Britain. It became the organizational model for other craft unions, many of which still exist. The AEU's industrial jurisdiction and the occupations among its membership, which reached over a million in the 1960s, are similar to those of the International Association of Machinists in the United States. It is no longer a craft union.
The basic structure of the AEU remained stable between 1921, when a full-time presidency was added, and 1980. The AEU formed the keystone of the Confederation of Shipbuilding and Engineering Unions, which bargained with the Engineering Employers' Federation for a basic agreement covering 3.5 million workers. In most of British industry no single union has exclusive bargaining rights, so the AEU controlled few closed or union shops.
The National Union of Mineworkers (NUM) was formed as a federation in 1889. With the nationalization of the coal mines in 1947, the NUM achieved a virtual closed shop for almost all categories of non-supervisory blue-collar workers, and it organized most white-collar workers in competition with another union. Its membership declined from over 600,000 in 1960 to 260,000 in 1973, as a result of mechanization and mine closures. With these processes continuing, and with a split during the defeated strike in 1984, the union declined to about 80,000 members by 1989.The basic organization of the Mineworkers has changed little since 1889, except that referendum elections for the top posts were added in 1918.
The Model of Union Democracy
In this idealized model there are two or more full-time officers of equal rank at each level of the hierarchy, including the top. In seeking higher office such officials are often forced to compete against each other in referendum elections, under a voting system which permits the opposition to coalesce when there are three or more candidates. Power at the very top is divided between at least two top officers. It is also limited by the regional election of second-ranking and lower officers, by substantial regional autonomy, and by independent power centers such as the executive council and the convention, with rank and filers paying the most important role in the latter. The Engineers and the Mineworkers approached this model in many respects.
Another model without any full-time officials will be discussed later.
The Democracy in the AEU and NUM
The selection of officers: There are no appointed full-time officials—all are elected, by referendum, in a multi-stage process which permits a pooling of the votes of anti-administration tendencies while assuring an absolute majority for the winner. The AEU used run-off elections, with the two leading candidates going on to a second ballot. The NUM used a transferable votes system, with voters listing their preferences 1, 2,3, etc., and the counters eliminating candidates one by one until a candidate has a majority. This avoids a run-off election.
In the AEU any member could nominate a candidate for any full-time post, with no seconding or local branch approval required, although in practice more than one nomination for a post from any one of the 2,000 branches is rare. In the NUM nominations could be made by any one of the fourteen geographic areas or six occupational or industrial subdivisions. Nominations were made differently in the various areas, by area councils with rank-and-file delegates from each local, by rank-and-file area executive committees, or by a direct vote of all area locals.
The full-time officials: The structures of full-time officialdom plus the voting systems facilitated and indeed generated competition for office throughout each union. Furthermore the overwhelming majority of full-time officials had little to fear from their superiors because they were elected within, and were responsible to, their regions. The power of assignment (and hence the potential for harassment or favoritism) by the top officers was extremely limited. Five officials could have been assigned their duties in the AEU, and none in the NUM.
Each union had a general secretary and a president, both full-time and of about equal power. The same pattern of equal, multiple full-time officials existed at each rank in both unions. Thus there was no heir-apparent below either top office or below any other full-time office. In the AEU there were seven executive council members, and two assistant general secretaries, all at the second-ranking level. This situation was repeated at the four next lower levels, each level having multiple potential full-time competitors for higher office. It is significant that only the very lowest ranking full-time office, district secretary, had a large percent of uncontested elections.
In the NUM there was no full-time second-ranking national officer. The vice-president was part-time, and usually an Area official who, in practice, had no special advantage in competition for a top post Thus the second-ranking full-time officers were the heads of the Areas, the larger of which themselves had two coequal top officers and ward-elected miners' agents below them. Thus in neither the NUM nor the AEU was there a single logical successor for a top post, or for any other full-time office.
Regional autonomy and decentralization: The regional election of the vast majority of officials is a reflection of the administrative decentralization of the two unions. This is reinforced by the absence in the union constitutions of any provision for trusteeship over unruly locals or other subdivisions. (None of the thirty-one British unions whose constitutions were studied had a provision for trusteeship.)
National conferences and committees: The annual conferences (conventions) of both unions were well shielded from manipulation by the top officers. Resolutions submitted could not be removed from the agendas or altered without the consent of the bodies submitting them. The conferences were small fifty-two s declined in membership to 80,000 today. voting delegates in the AEU (currently 121) and 135 in the NUM as was made possible by their indirect election. The AEU's conference consisted of rank-and-filers only, and every fifth year considered constitutional revisions. No full-time officials other than the president and general secretary were entitled to attend. The AEU also had an important twelve-member rank-and-file appeal court, each member elected by referendum within one of the twelve regions. The NUM's conference did not exclude full-time officials, but was elected by well-organized rank-and-file Area conferences also met, as councils, throughout the year.
Opposition in Elections
Hard evidence on the extent and effectiveness of electoral opposition is necessary to overcome the skepticism and pessimism which the American union scene has fostered, and because apologists have often labeled union oligarchies as essentially democratic.
Number of candidates: The AEU had no uncontested elections for vacancies in the full-time top posts between 1875 and 1975, and each election had three or more candidates. The average (median) since 1945 was eight candidates. All elections were by referendum. The NUM instituted referendum elections for top posts in 1918, and through 1975 it also had no uncontested elections. Of the twelve top vacancy referenda in the NUM, eight had three or more candidates. Since 1945 the average number of candidates was four.
Previous defeats for top office: More important, the winners often had to fight their way into the top office. In half of all such elections in each union since 1945, the winners had been defeated in earlier elections for a top post vacancy. For example, the winner of the NUM's presidency in 1971 had been narrowly defeated for the co-equal secretaryship in 1968.
Close elections: The votes for the runner-up in filling top vacancies have a1-most always been more than nominal, although only occasionally approaching the standard of a close governmental election under a two-party system.
In half of the AEU elections between 1945 and 1975, the runner-up received at least 77 votes per 100 for the winner. In the NUM the comparable figure was 79 per 100. In two of the AEU's six top vacancy elections the ratios were 91 to 100, and 99 to 100. Probably only the International Typographical Union here could have come anywhere near to these figures.
The competitive situation described above is exceeded at second-ranking and intermediate levels of full-time office in both unions, with the occasional defeat of second-ranking officers in the AEU. Although incumbent top officers are almost invariably opposed in the AEU, there was only one defeat, in 1913. Crop officers were removed by the executive council in 1895 and 1932.) The NUM's full-time officers were permanent (not subject to periodic elections, but removable for just cause). The 1989 Rules provide for five-year terms for full-time officials until they reach the age of 55, with retirement generally required at 65.
Other Indications of Democracy
Both unions invariably had oppositionists well represented among their second-ranking full-time officers and in their executive councils. These were generally left-Labor or Communist. In the NUM there was a sharing of power at the top level between 1946 and the early 1970s, with the three successive presidents being orthodox Labor and the first two general secretaries being Communist and the third left-Labor.
Power at the very top has been similarly shared at times in the AEU, but more consistently so in the full-time executive council, with each member elected by referendum from a separate district. The union also used referenda to elect its delegations to the annual conferences of the Trades Union Congress and the Labor Party. Voting for various posts was so frequent as to be almost a normal part of local meetings perhaps not an entirely favorable situation.
As of 1992, the major democratic features of the NUM and the AEU seem to be intact, although the NUM has fallen into hard times, and the AEU's expansion through mergers with less democratic unions may present problems.
The AEU's run-off elections between the two leading candidates gave an advantage to the best-organized but often small minorities which, between them, often had less than half of the vote. The arbitrary result was too often competition between the candidates of right-wing Labor anti-Communist groups and official Communists. The NUM's preferential ballot avoided this artificial problem.
The competition for top vacant posts in the AEU was on the basis of a low voter turn-out, between 8-12% from 1943 through 1967. Voting was at the meetings of geographically-based locals with no necessary connection to workplaces—a holdover from the union's craft years. The AEU shifted to postal ballots in 1972 over the objections of most left-wingers, who feared a loss of support Participation increased to 30% in 1975, and a left-winger was elected as an assistant general secretary in 1976. (The NUM had voting at the mines, with over 60% participation.) The NUM's avoidance of periodic elections to full-time office was intrinsically undemocratic, but it allowed independent lower-level full-time officers to compete for higher office without fear of retribution. They had independent regional power bases, and secure tenure.
Could the greater politicization in Britain and the unions affiliation to the Labor Party explain the electoral opposition in the AEU and NUM? I believe that left-right divisions provided the content for controversy but were not the root cause of electoral opposition.
Openly operating factions (caucuses) and freely circulating campaign literature were considered illegitimate in the AEU and NUM, and, with one partial exception, in other British unions. The generally accepted view was that organized internal factions are unnecessary, unfair, and probably conspiratorial (as they were, under the circumstances). The AEU banned any but hand-written campaign literature, but distributed booklets with the election addresses of each candidate through its locals. Only enough for from one-tenth to one-fifth of the membership were provided, and there was no guarantee of a fair or complete distribution. (The 1989 rules provided for mailing 500-word election addresses from each candidate with the postal ballots.) Some of the statements made were definitely misleading, and only the give-and-take of unrestricted campaigning could have smoked out their authors. The restrictions on campaigning were only partially compensated for by occasional unsigned illegal local factional flyers, and by comments in the periodicals of outside political organizations. The latter partially but ineffectively filled the void in the NUM.
In the NUM there was the rare instance of the subterfuge of a position-taking pamphlet by a candidate or his area. (There were never any full-time female officials.) The NUM did not provide election addresses and indeed prohibited campaign literature. While the restrictions against open campaigning may have been breaking down, even today it remains far from legitimate. (Recently the national executive committee has authorized election addresses.)
The NUM had only sporadic factional activity between national elections, while in the AEU factions functioned continuously but covertly. The ambiguous place in the AEU's structure of shop stewards, and of interplant stewards committees, may have contributed somewhat to factional activity, but I think not substantially. The NUM had no such situation.
On the whole, the electoral success and sharing of posts by factions, or incipient factions, were due to the same factors which made for close competition between individual candidates: It was the achievement of office by individual oppositionists and faction members, under a system conducive to this, which gave the anti-administration factions a foothold, and furthered their growth and continuity.
A Purely Rank-and-File Democracy?
It has been shown that the 8,000-member Spanish longshore union, Coordinadora, gets along democratically and well without a single full-time official and only two paid employees. (See Don Fitz, La Coordinadora: A Union Without Bureaucrats, St. Louis: Workers Democracy Press, 1990.) Working dockers run the union on a rotating part-time basis, with time off to do this guaranteed under Spanish law. To what extent is this preferable model of union democracy generalizable to other national unions?
The basis for Coordinadora's democracy is a single-occupation membership which keeps in good contact with each other on a daily basis at the hiring hall. Rotating officers can get feedback, and be controlled at quickie meetings before work The Mineworkers near their peak were eighty times larger, and more diverse occupationally, but they were a single-industry union and most members lived in mining communities. It is conceivable that all full-time officials could have been dispensed with, if ways were found to direct and monitor various specialist departments for example finances and technical and economic research.
But I think the problems of communication and coordination in the multi-industry, million-member Engineers would preclude a Coordindoratype organization. And a part-time executive council with a full-time top officer, common in Britain, has not seemed to be a prescription for democracy. The British Railwaymen elected only rank-and-filers to their national executive council, then made them in effect full-time for a three-year term, after which they had to step down. This sounds attractive, but the general secretary was the focus of power and usually had his way. Former councilors were not likely, on the basis of this experience, to become viable candidates for a top post.
Relevance for Socialist Society
I picture unions under socialism as existing alongside of more comprehensive workers' self-managed industrial/ economic institutions. Unions will remain more specialized organizations, even if their functions expand somewhat, representing the concerns of occupational groups and workers in specific work organizations. With opportunities for more rounded participation greater elsewhere in a self-managed society, and injustice at the workplace moderated, one can't count on a continuously high level of workers' involvement in union affairs. Or, worse, some degree of routinization, bureaucratization and/ or disillusionment and passivity might pervade the society, affecting not only the unions but the larger institutions of governance.
During such unfavorable intervals democracy could be sustained by the kind of competition-producing system described above, with relatively equal officers opposing each other at all levels under a fair and permissive voting system. This self-sustaining process could continue until rank-and-file movements and meaningful issues again began to superimpose themselves on it. Such a system is obviously not all that it required for a high level of democracy, but it would seem to make a necessary contribution in large organizations and supply the elements of a fail-safe mechanism there and possibly within the larger political system.
March/April 1994, ATC 49