Comparisons in Labor Law: Sweden and The United States





In comparative labor law, as in all comparative law, we may begin with noting the comparisons and contrasts. This, however, is but the beginning. If we do no more than this, we are little more than school children visiting the zoo, looking in wonderment at the giraffe which seems to be all neck supported by spindly legs of unequal length. How strange to the American is the Swedish rule of priority of interpretation! And how unbelievable to the Swede, the American beast of secondary boycott! Or we visit the monkey cage and are intrigued by how similar they are to ourselves. Both Sweden and United States, we note, impose on unions and employers a duty to negotiate, and make collective agreements legally enforcible; and we go on to remark on our other common traits. Everyone enjoys visiting a zoo; it is something every child, young and old, should do. But we should enjoy it for what it is—more entertainment than useful inquiry. And so it is with observing comparisons and contrasts in labor law.
    Beyond the bare comparisons and contrasts is the important question of why they exist. The process of comparison should press us to search for the reasons for the legal rules in each system, where they came from and what function they serve. When we ask the question of "why," we often realize that similar legal rules have quite different roots and serve quite different functions. We may also come to recognize that visibly different legal rules, when seen in context, perform equivalent functions.
    The value of this inquiry is less in what we learn of the other system than what we learn of our own system, for the comparative viewpoint gives us a different perspective. We are jolted into awareness that particular rules which we have unquestioningly accepted as fundamental and inevitable are neither fundamental nor inevitable. We realize that the reasons which gave birth to those rules were never good reasons or no longer exist, and that the system could function quite well with different rules. We may suddenly recognize that rules which we have viewed as subsidiary are critical in shaping the unique character of the system, or beyond the specific rules we may recognize basic characteristics of the system which give special content to those legal rules.


39-33-169 Sv Juristtidning


590 Clyde W. Summers    When we ask the question "why", comparative labor law enables us to know ourselves better, to dispel myths and question our assumptions, and to recognize the relevance of particular rules in shaping our system. This greater self-understanding can have constructive values in helping us see where changes are necessary or possible—changes which were previously unthought of or considered unthinkable. Indeed, the study of another system stirs our imagination to see possibilities for change, and the forms they may take. Although we may not be able to transplant, we may have cross-fertilization of ideas and develop hybrids adapted to our own institutional soil.
    It is not my purpose here to make a broad over-arching comparison of Swedish and American labor law. That could lead to pointless and pretentious generalities. Comparisons must confront concrete details. My purpose here is primarily to illustrate the importance of asking the question "why" in making comparisons in labor law and to suggest where such an inquiry may lead. For that purpose I want to focus on three points. First, the similarity of basic collective bargaining statutes of the two countries. Second, the contrast in the legal status of the union as representative of the employees. Third, the contrast in legal Controls on the internal processes of unions.
    Comparison in labor law presents nearly insuperable problems because it ultimately reaches into comparison of social struetures and attitudes. Comparison of bare legal rules—statutes, court decisions, and administrative regulations—is obviously incomplete, for in both Sweden and the United States one of the fundamental legal principles is reliance on self-regulation by the collective parties in the labormarket. A large body of the governing law is found in collective agreements and the courts' interpretation of those agreements. Indeed, the Swedish Act on Codetermination in Working Life of 1976 (MBL) expressly provides for its completion by codetermination agreements to be negotiated by the parties. A less visible body of law is found in the constitution or statutes of unions and employers' associations. All of these must be seen in context of the institutional structures, procedures and practices in the labor market.
    The comparisons here are limited in scope, examining only particular rules and principles. Full comparison, taking all considerations into account, is not possible. Inevitably, the comparisons are less definitive than suggestive, but hopefully they will be illuminating or provocative.


There are striking similarities between the National Labor Relations Act (NLRA) enacted in the United States in 19351 and the Act on Right of Association and Negotiations (FFL) enacted in Sweden in 1936.2 Section 7 of the NLRA declared: "Employees shall have the right to form, join or assist labor organizations, [and] to bargain through representatives of their own choosing ..."


Paralleling this, Section 3 of the Swedish statute declared: 

"The right of association shall not be infringed" and the statute defined that right as "the right ... to belong to an association ... to exercise the rights as members of the association, and to work for an association or the formation of an association."


Again, Section 8(3) of the NLRA imposed on an employer the duty to bargain collectively with representatives of his employees concerning wages, hours and other terms and conditions of employment. Paralleling this, Section 4 of the FFL established the "right of negotiation, "which was defined as "the right to institute negotiations respecting the adjustment of conditions of employment and respecting the relations between employers and employees generally". In both countries, the elaboration and enforcement of these rights was placed in a special agency or tribunal—in the United States, the National Labor Relations Board; in Sweden, the Labor Court.
    We might also be struck by the fact that both countries have, by statute, made collective agreements legally enforcible,3 but this parallel flags a difference. The Swedish statute making collective agreements enforcible was enacted in 1928, eight years before the FFL; the American provision was added in 1947, twelve years after enactment of the NLRA. This alone warns that the NLRA and the FFL might, for all their similarities, serve quite different purposes.
    When we examine and compare why the collective bargaining statutes in each country were enacted, we begin to recognize fundamental differences in the two systems. The National Labor Relations Act was enacted for the purpose of establishing a system of collective bargaining. Employers had historically engaged in a variety of oppressive tactics to prevent workers from joining unions and to frustrate collective bargaining. Employers who showed any sympathy for unions were discharged and blacklisted; new employees were required to sign "yellow dog" contracts agreeing not to join a union; union


1 Act of July 5, 1935, 49 Stat. 449.

2 Lag den 11 september 1936 om förenings- och förhandlingsrätt, replaced by MBL in 1976 which retained the provisions discussed here substantially unchanged.

3 In the United States, by Section 301 of the Labor Management Relations Act of 1947, 61 Stat. 136 (1947). In Sweden, by Lag den 22 juni 1928 om kollektivavtal. 

592 Clyde W. Summersofficers were assaulted; and unions were infiltrated with spies and agents provocateurs.4 Employers created controlled "company" unions and refused to recognize or negotiate with legitimate unions.Between 1918 and 1934, unions lost nearly 40 % of their membership, and in 1934 only 10 % of the non-agricultural work force was organized.6 The NLRA provided legal protection for unionization against employer opposition and required hostile employers to recognize unions and bargain collectively. It was a legal foundation on which to build a labor movement.
    In contrast, the Swedish statute was enacted when the collective bargaining system was fully accepted and firmly established. Unions had enjoyed an unbroken fourteen year period of growth7 and more than 75 % of the workers in major industries were covered by collective agreements.8 The purpose of the statute was to extend to farm and forest workers, office workers, foremen and other salaried workers legal protection of rights which had been generally recognized in practice for industrial workers for thirty years.9 For the Swedish unions, the statute played no major role, as they had long since secured and enjoyed these rights.
    This contrast in the purposes of the two seemingly similar statutes leads us to a more significant difference. In the United States, the system of collective bargaining rests heavily on the law; in Sweden it rests largely on agreement. The so-called December Compromise of 1906, an agreement between the Swedish Employers Confederation (SAF) and the Confederation of Trade Unions (LO), provided that "the right of association shall be inviolate". By the agreement, the organized employers explicitly accepted unions and collective bargaining.10 SAF refused to support employers who would not allow their employees to join unions and extended collective bargaining by


4 For description and documentation of these practices, see Adamic, Dynamite (1931); Brooks, When Labor Organizes (1937): Huberman, The Labor Spy Racket (1937); and the so-called La Follette Committee Hearings, "Violation of Free Speech and Labor", Senate Report No. 46, 75th Cong. 1 st Sess. (1937).

5 Twentieth Century Fund, Labor and Government (1935), p. 65— 113.

6 Troy, Trade Union Membership, 1897-1962 (1965).

7 Westerståhl, Svensk fackföreningsrörelse (1945) 31.

8 Robbins, The Government of Labor Relations in Sweden (1942) 25; Bain & Price, Profiles of Union Growth; A Comparative Statistical Portrait of Eight Countries (1980) 142-3.

9 Westerståhl, supra note 7 at 408.

10 The central purpose of SAF when it was formed in 1902 was to represent employers collectively in negotiating with the union. A mutual defense fund was created to strengthen employers in economic conflict. It was implicitly accepted that this would not be used to attack unions as such but only to support employers in collective bargaining. Hallendorff. Svenska arbetsgivareföreningen (1902—1927) (1927) 31. 41, 57. 

Comparisons in Labor Law 593negotiating industry-wide agreements even though the industry was only partially organized.11 Swedish employers, in 1906, recognized by agreement rights which American employers, in 1935, had to be required by law to recognize.
    We now must ask the potentially revealing question—why this difference? From the American viewpoint — why did Swedish employers place their stamp of approval on unions and promote the extension of collective bargaining? From the Swedish viewpoint—why did American employers so bitterly combat unionization and resist collective bargaining?
    The reasons undoubtedly reach into differences of corporate ownership and management, and into differences in social structurcs and attitudes. However, an important part of the answer may be found in the other half of the December Compromise. In return for the employers' acceptance of collective bargaining, unions agreed to inclusion in all collective agreements of Paragraph 23 (later Paragraph 32) of the SAF statutes. That paragraph provided:
"the employer is entitled to direct and distribute the work, to hire and dismiss employees at will, and to employ workers whether organized or not."

Unions thereby agreed to limit collective bargaining to wages, hours and other economic terms, and accepted employer control over decisions as to who should be promoted or transferred, who should be laid off or discharged, and how work was to be assigned or performed. Swedish employers agreed to accept unions when unions agreed not to interfere with management of the enterprise.
    American unions have never conceded to management such prerogatives, but have used the collective agreement to limit management's control.12 Historically, the closed shop and union hiring halls have been used in industries characterized by short term employment to control who will be hired.13 Seniority provisions in collective agreements regulate who is to be promoted, who is to be transferred and who is to be laid off. Negotiated job descriptions may define the duties of particular jobs, and "just cause" clauses may limit the employer's ability to discipline or discharge employees. Work scheduling, assignment to shifts, requirements of overtime, speed of production and size of work groups may all be regulated by the collective agreement.14


11 Bresky, Scherman, Schmid, Med SAF vid rodret: Granskning av en kamporganisation (1981) 34 ff: Hallendorff supra note 10 at 74. See Bain & Price, supra note 8.

12 See generally, Chamberlain, Union Challenge to Management Control (1948).

13 Slichter, Union Policies and Industrial Management (1941) Ch. 2 and 3.

14 See generally, Slichter, Livernash & Healy, The Impact of Collective Bargaining on Management (1960); Stone, Managerial Freedom and Job Security (1964). 

594 Clyde W. SummersAmerican unions can and do strike to obtain these provisions.15
    The most bitter area of dispute between unions and employers in the United States is what matters are to be regulated by the collective agreement and what matters are for employer unilateral control. Because of this dispute, the law must prescribe those subjects about which the employer is required to bargain with the union and for which the union can strike.16 Although the law limits those subjects, they include almost all of those reserved for management control by the December Compromise.17
    The important point is that for an American employer, unionization and collective bargaining means, in practical and legal terms, that the union can exercise control over decisions of the enterprise which Swedish unions agreed were solely for management. Swedish employers might not have so warmly embraced collective bargaining if Swedish unions had rejected Paragraph 32 and insisted on American type collective agreements; and American employers might have been less hostile to unions if unions had not challenged management's authority to manage.
    This perspective helps illuminate recent development in Sweden. Historically, limitation on the union's role and recognition of management prerogatives have been dominant characteristics of the "Swedish model" of labor relations. It was the precondition and unspoken assumption of the Basic Agreement between SAF and LO in 1938.18Article III of that Agreement required the employer to notify the


15 Railroad workers have repeatedly struck over the size of work crews, and legislation was necessary to resolve a strike over the number of employees needed to operate diesel locomotives. Dockworkers struck all East Coast ports to prevent or regulate the use of containers on ships, and Steelworkers closed down the steel industry for over two months to preserve limitations on management's assigning work and determining the number of employees on work crews. Collective agreements in the auto industry give the union the right to strike over the speed of assembly lines and other production standards, and this right is exercised.

16 Employers and unions are required to bargain in good faith concerning "mandatory"subjects, defined by Section 8(d) of the NLRA as "wages, hours, and other terms and conditions of emplovment." There is no legal requirement to bargain about "nonmandatory" subjects which fall outside these words, First National Maintenance v. NLRB, 452 U.S. 666 (1981). And it is an unfair labor practice for either the employer or union to use economic measures to compel the other to bargain about "non-mandatory" subjects, NLRB v. Wooster Division of Borg-Warner Corp., 356 U.S. 342 (1958).

17 "Mandatory" terms include hiring, promotions, dismissals, disciplinary rules, description of job duties, assignment or transfer of employees to particular jobs, number of employees to be used on particular jobs, scheduling of work or production. Mandatory subjects do not include, in the words of the Supreme Court, "managerial decisions which lie at the core of entrepreneurial control." Fibreboard Paper Products Corp., NLRB, 379 U.S. 203, 223 (1979), such as closing down a part of the enterprise, adding a new facility, selling the business, or merging with another. First National Maintenance v. NLRB, supra note 16.

18 Schiller, LO, paragraf 32 och företagsdemokratin (1974) Ch. 8.


Comparisons in Labor Law 595union of planned dismissals and allowed the union to call for negotiations, but the ultimate right of decision remained with the employer.19 The Works Council Agreements of 1946 and 1966 gave works councils broad rights of information and consultation, but again the right ofdecision remained with the employer. The single significant encroachment on Paragraph 32, prior to the 1970's, was the amendment to the Basic Agreement in 1964, which limited the employer's right to discharge individual employees to cases where the employer had "material grounds,"20 and this agreement was superseded by the Employment Security Act of 1974.21
    Although Swedish unions repeatedly made demands to limit orabolish Paragraph 32, there was no willingness to use industrial action to accomplish this.22 Finally, in 1976, the Act on Codetermination in Working Life (MBL) was pushed through with the political slogan "Away with Paragraph 32." The declared purpose was to give employees a voice in the decisions of the enterprise and, symbolically, Section 32 of the statute provided that "there should, if the union side so requests, also be concluded a collective agreement on codetermination for employees" on a nearly unlimited range of management decisions.23 The statute opened up for collective negotiations, with ultimate resolution by a strike, if necessary, matters reaching much more deeply into management decision-making than collective agreements in the United States.
    The climate of Swedish labor relations changed from one of cooperation to one of confrontation. Negotiation of a central codetermination agreement for the private sector came to repeated impasses. The tone of public debate between unions and employers hardened, and what was said in private was even harsher. Wage negotiations become more difficult, culminating in a nationwide strike in 1980, and fears were expressed for the future of the "Swedish model."24 The union's threat to management ontrol generated employer hostility; an American visitor in Sweden began to feel at home.
    The statute, however, has not worked the fundamental changes which unions had hoped and employers had feared. The statute did


19 Bresky, Scherman, Schmid, supra note 11 at 73 ff; Schiller, supra note 18 at Ch. 9.

20 Huvudavtalet, Kap. III, § 4.

21 Lag 1 juli 1974 om anställningsskydd.

22 Although every collective agreement included a Paragraph 32 clause, no attempt was made to join a demand for eliminating this clause with a demand for wages until 1964, and then the negotiation was only to limit the employer's freedom to dismiss.

23 See generally, Victorin, Codetermination in Sweden: The Union Way, 2 J. Comp. Securities Law 111 (1979).

24 See Storkonflikten (1980) (Broström, Ed.).


596 Clyde W. Summersnot, of its own force, change the locus of final power to decide, except for the special matter of subcontracting. On other matters, the employer is required to negotiate before making a decision, but if an impasse is reached he has the power to decide.
    Section 32 of the Act calls for negotiation of codetermination agreements, and unions could insist, to the point of striking, on contractual rules limiting the exercise of management prerogatives, or for the power to make or veto certain decisions.25 The unions, however, have been willing to accept codetermination agreements which provide structures and procedures for negotiation but leave the ultimate right to decide with the employer. Thus, the agreement negotiated with SAF for the private sector in April, 1982, reinforces the employees' right under MBL to information concerning the operation of the enterprise, and the right to discuss and express opinions concerning decisions, but it gives the employees no power to make or veto decisions.26 The unions have secured the right to discuss, but have accepted the employer's right to decide. It is not surprising that the chief negotiator for SAF welcomed this as facilitating a return to the old spirit of Saltsjöbaden.27 The union's threat to management control has been contained; management's prerogatives are subject only to the duty to discuss.
    We might at this point attempt to peel another layer off the onion by asking: Why have Swedish unions not insisted on a greater share in the decision making power on matters which daily affect the most vital interests of workers? Why have these socialist unions, whose gospel is industrial democracy, been less aggressive in challenging the employer's authority than the business oriented American unions? My purpose here, however, is not to pursue successive layers of questions in the hopes of finding a central core. Rather, my purpose is to suggest how pursuing questions of "why" may lead us to new perspectives and better understanding of our own system. That purpose may be sufficiently served by asking the next question without attempting to answer it.


25 The unions' initial proposal for a central codetermination agreement would have given employee organizations the initial power to decide on the power to veto management's decision on a wide range of subjects.

26 Utvecklings- och medbestämmandeavtal, SAF-LO/PTK, April 15, 1982. This follows the pattern set by earlier codetermination agreements for banks, cooperatives, the state sector, state enterprises, and municipal and county governments.27 Dagens Nyheter, 16 april 1982.


The basic Swedish rule is that the union, in negotiating with the employer, represents only its members, and the collective agreement binds only its members. Indeed, Section 10 of the Codetermination Act of 1976 (MBL), restating Section 4 of the 1936 statute, requires the employer to negotiate with any union which has any members among the employees, and the employer can legally have collective agreements with two or more unions for the same category of employees. The underlying principle is that the union's authority to represent is based on the individual's voluntary act of joining the union.28
    The American rule is quite the opposite. Section 9(a) of the NLRA provides that the union designated by the majority of the employees in a "bargaining unit" is the exclusive representative of all employees in the bargaining unit. The collective agreement negotiated by the majority union regulates the terms and conditions of employment of all employees, whether union members or not. If there is a majority representative, the employer is prohibited from discussing terms and conditions of employment with any other union, or to make any agreement with an individual employee which deviates from the collective agreement.29 The underlying principle is that the union's authority to represent is based on majority rule.30
    This contrast in the basic premise on which the union's representative status is based raises the question of why there should be this difference. The initial burden of explanation is on the American side, for the Swedish rule parallels that commonly accepted in other systems while the American rule is unique. Historically, the American rule has two main sources — employer hostility to collective bargaining and competing unionism. In resisting unions and collective bargaining, employers would refuse to negotiate with unions which did not have majority support on the grounds that the union did not represent the employees. When employers did deal with the union, they would undermine the union by dealing directly with employees or with "company unions" controlled by the employer, often granting to them better terms than to the unions.31 The presence of competing unions,


28 Schmidt, Law and Industrial Relations in Sweden (1977) 139—40. A member who withdraws from the union after the collective agreement is concluded continues to be bound, and an emplovee who joins after the agreement is made becomes bound. MBL. Sec. 26.

29 J. I. Case Co v. NLRB, 321 U.S. 332 (1944); Order of R.R. Telegraphers v. Railway Express Agency, Inc. 321 U.S. 342 (1944).

30 Weyand, Majority Rule in Collective Bargaining, 45 Col. L. Rev. 556 (1945).

31 Twentieth Century Fund, Labor and Government (1935) Ch. 10: Houde Engineering Corp. 1 NLRB (old) Case No. 12. 

598 Clyde W. Summerseach claiming recognition and each attempting to best the other, added confusion and confliet to the negotiation process. The employer could be caught helpless between two contending unions, or be able to frustrate bargaining by divide and conquer tactics. Exclusive representation by the majority union was considered necessary to establish an orderly and workable system of collective bargaining.32
    Neither of these two conditions which led to the exclusive representation rules in the United States had been present in Sweden. As already pointed out, since 1906 Swedish employers have not tried to combat unions, but embraced them. Nor have Swedish unions been troubled by competing unionism. The labor movement has been unified, with the Swedish Confederation of Trade Unions (LO) and the Salaried Employees Confederation (TCO) almost completely dominating the labor market.33 Each confederation has internal machinery for resolving demarcation or boundary line disputes between member unions, and the two federations have a joint committee to mediate such disputes.34 The only competing organization has been the Syndicalists (SAC), which has less than 20,000 members scattered mainly in construction and forestry.35
    This leads to the question of why there have been no competing unions in Sweden. Undoubtedly there are various political and social factors, but a major contributing cause has been the organized employers' favoring of LO and TCO unions and refusing to make collective agreements with competing unions.
    The "measuring monopoly" cases in the Labor Court indicate the preference given by employers to the established unions. In the building industry, workers are paid on a group piece rate system, with"measuring men" appointed by the employers and employees to apply the complex and uncertain formulas. The measuring men are paid by fees deducted from wages of the group. Prior to 1946, each work group chose its own measuring man, usually from the union representing a majority of workers in the group. The Syndicalists had their own work groups and their own measuring men. In 1946, the Swedish Building Contractors Association made a national agreement


32 Schreiber, The Origin of the Majority Rule and Simultaneous Developments of Institutions to Protect the Minority: A Chapter in Early American Labor Law, 25 Rutgers L. Rev. 237 (1971).

33 During the 1970's there were some instances of groups of employees breaking out of the LO unions to form independent unions. Thus, a majority of stevedores left the Transport Workers to form the Harbour Workers Union, and a group of engine drivers on the state railways formed their own union.

34 Johnson, Collective Bargaining in Sweden (1962) 62—66, 104.

35 Schmidt, supra note 28 at 17.


Comparisons in Labor Law 599with the LO unions that all measuring would be done by LO measuring men and all fees deducted from employees wages would be paid to the LO measuring office. The intended effect was to eliminate the Syndicalists' representation by their own measuring men, and to give the LO unions a measuring monopoly—in effect, exclusive representation. This was held by the Labor Court not to be a violation of the right of association, even though it compelled Syndicalists to make contributions to LO unions.36 The important point here is that employers actively aided the LO unions in attempting to eliminate competing unions.37
    Under Section 4 of the 1936 Act (FFL) and Section 10 of the 1976 Act (MBL), the duty to negotiate does not include any duty to sign an agreement.38 Swedish employers in both the private and public sector, relying on this limited duty, have refused to sign collective agreements with any but the established LO and TCO unions, even though the competing unions represented a majority of the employees. Early inthe 1970's, branch unions of dockworkers in the Transport Workers refused to consolidate with other transport workers into large branches where they would lose their identity. When they were expelled, they formed their own union, the Harbour Workers. The employers association has negotiated with this independent union but has refused to make a collective agreement with it. Instead, the employers association has made the effective collective agreement with the LO Transport Workers Union even though it now represents only a minority of the dockworkers.39 Similarly, in 1969, the State Bargaining Office announced that it would not make collective agreements with any free standing organizations, but only with the established unions. It refused to make contracts with unions of locomotive engineers, chief cooks, midwives, and harbour pilots, with the purpose and effect of driving these groups into LO and TCO unions.40 Swed-


36 AD 1947:50; AD 1948: 78; AD 1954: 19 and 20; AD 1960: 16. These decisions were sharply criticized by Professor Schmidt. See Schmidt, The Law of Labor Relations in Sweden (1962) 140-142.

37 In these cases, and in earlier cases involving claims by the Syndicalists that their freedom of association was being infringed, the SAF members of the Labor Court joined the union members to deny the Syndicalists' claims. See AD 1945: 35 and 36; AD 1945: 77; and AD 1946: 41, 59 and 64. The pattern of employer members of the court making common cause with the union members of the court to eliminate the Syndicalists is traced in Geijer & Schmidt, Arbetsgivare och fackföreningsledare i domarsäte (1958) 90—111; Summers, Freedom of Association and Compulsory Unionism in Sweden and the United States, 112 U. of Pa. L. Rev. 647, 680-4 (1964).

38 AD 1946: 68; AD 1969: 14; AD 1972: 5.

39 Hadenius, Facklig organisationsutveckling (1976) 164.

40 Schmidt, supra note 28 at 104; Fredriksson, De fristående fackförbundens svårigheter (1972). 

600 Clyde W. Summersish employers have thus given the established unions an effective monopoly in making collective agreements.
    Although the basic Swedish legal theory is that the collective agreement binds only union members, the collective agreement in fact regulates the terms and conditions of employment of non-members and limits their right to strike. The union making the collective agreement thus obtains for practical purposes the special status as exclusive representative. The Labor Court has held that unless the collective agreement expressly provides otherwise, the employer is legally bound to give non-members of the contracting union the same benefits as members, and the unions have insisted on this in order to prevent employers from hiring unorganized workers at lower rates.41 This obligation of the employer can be enforced directly only by the union; the non-union employee has no rights under the collective agreement. However, the provisions of the collective agreement constitute a custom and usage which is read into the non-union employee's individual contract of employment and he can sue on that contract in the ordinary courts. Non-union employees are also bound by the statutory peace obligation during the term of the collective agreement. Under Section 42 of the MBL they are legally liable, along with union members, if they participate in a prohibited wildcat strike, and under Section 62, are subject to any special sanctions provided by the collective agreement.42
    Legislation during the 1970's has explicitly given the established unions—those with collective agreements — the legal status of representative of all employees, whether union members or not. For example, the Act on Board Representation for Employees43 provides thatthe employee representatives on corporate boards are not elected by the employees but named by the unions having collective agreements covering the larger number of employees. The Act of Trade Union Representation at the Workplace44 gives local unions which are bound by collective agreements the right to name union representatives at the workplace. These representatives, who may act on any matter affecting employees as employees, are entitled to paid time off to perform their union mandate, to a room or other space at the workplace, and the priority to continued work in case of a lay-off. Similarly, the Work Environment Act45 provides that the safety representatives


41 Fahlbeck, Praktisk arbetsrätt (1981) 101; Schmidt, Löntagarrätt (1978) 36-9.

42 Prop. 1975/76: 105, Bil. 1, 272-3; Fahlbeck, supra note 41 at 146-7.

43 Lag den 3 juni 1976 om styrelserepresentation för de anställda i aktiebolag och ekonomiska föreningar.

44 Lag den 31 maj 1974 om facklig förtroendemans ställning på arbetsplatsen.

45 Lag den 19 december 1977 om arbetsmiljö.


Comparisons in Labor Law 601and safety committees are to be appointed by local unions bound by collective agreements. These representatives have broad functions to participate in planning new and changed locations, work processes and work methods, safety training, and general safety oversight, andthey have the power to order cessation of work where there is immediate and serious danger to life or health.
    The Codetermination Act gave an even more dominating role to the established union. Under Section 19, the right to information concerning the employer's economic situation and the guidelines of his personnel policy runs only to the union with which he has a collective agreement; and only that union has a right to examine his books and accounts. The employer's duty under Section 11 to call for negotiations before making any important changes in his business activity or employment conditions is owed to the union with which he has a collective agreement; the duty to negotiate with any other union is limited under Section 13 to changes in the employment conditions which particularly affect members of the other union. Under Section 32, a codetermination agreement can be requested and negotiated only by a union with which the employer has a collective agreement; and under Section 39 only an established union can veto an employer's decision to subcontract.
    The power of the established union to bind all employees, members and non-members alike is most explicit in the Act on Employment Security.46 In cases of layoff, because of shortage of work, it is to a high degree only the established union which is entitled to a warning notice by the employer of his planned action, and it is then only this union which has the right of discussion to determine the order of layoff. Although the statute provides rules of seniority, Section 2 allows the established union to supplant those rules by collective agreement. The union can prescribe different seniority units, and can establish other criteria such as ability to find other work, family responsibilities or the employer's need for special skills or experience. These rules, negotiated by the established union, are equally binding on its members and members of other unions or non-union employees.
    A number of other statutes give the established unions the legal status of representative of all employees with the power to supplant statutory rights by collective agreements. All of these provisions have been fully developed by Hemström in his comprehensive study of the power of the union over the individual in the Swedish system, and


46 Lag den 24 februari 1982 om anställningsskydd.


602 Clyde W. Summersneed not be restated in detail here.47 Enough has been said to make plain the central point: the organized employers, both public and private, have refused to make collective agreements with any but the established unions. The law has given special status and rights to unions which have collective agreements, and have given the agreements made by those unions binding effect on all employees, members and non-members alike. In practical and legal terms, the established unions are, in many respects, the exclusive representative of all employees, exercising a control over all employees not unlike that exercised by American unions under Section 9(a) of the NLRA.
    Although unions in the United States and Sweden have comparable powers to regulate terms and conditions of employment of all employees, they have not been held to comparable responsibilities in exercising that power. In the United States, the courts have read into the statute the duty of the union to represent fairly all employees affected by its contract. Thus, when a union negotiated a seniority provision which had the effect of putting all Negro employees at the bottom of the seniority list, the Supreme Court voided the provision, saying:
    "Congress ... did not intend to confer plenary power upon the union to sacrifice, for the benefit of its members, rights of the minority of the craft without imposing on it any duty to protect the minority ... The fair interpretation of the statutory language is that the organization chosen to represent the craft is to represent all its members, the majority as well as the minority, and is to act for and not against those whom it represents."48
    Relying on a broader principle of agency, the Court declared: "It is a principle of general application that the exercise of a granted power to act in behalf of others involves the assumption toward them of the duty to exercise the power in their interest and behalf."49
    The duty of fair representation has become a broad and fundamental principle in American law.50 It is not limited to racial discrimination, but extends to all forms of unfairness. For example, when an employer consolidated a large plant with a small plant it became necessary to combine the seniority lists. The union submitted it to avote, and the employees from the larger plant voted to place the employees from the smaller plant at the bottom of the seniority list. The court held that this overreaching by majority was unfair and enjoined the result.51


47 Hemström, Fackföreningarna och de anställda (1981).

48 Steele v. Louisville & Nashville R. Co., 323 U.S. 192, 201-2 (1944).

49 Id.

50 See generally Boyce, The Duty of Fair Representation (1978); Clark, The Duty of Fair Representation: A Theoretical Structure, 51 Tex L. Rev. 1119 (1973).

51 Barton Brands Ltd. v. NLRB, 529 F.2d 793 (7th Cir. 1976).


Comparisons in Labor Law 603    The duty has had its greatest impact, not in the negotiation of collective agreements, but in the administration and enforcement of collective agreements. Thus, the union violates its duty of fair representation if it refuses to process the grievances of employees because they are not union members,52 because of their sex,53 their political activities inside the union,54 or because of personal hostility.55 The union's duty to represent fairly includes the duty to use care and good faith in investigating, processing and presenting grievances.56 In the words of the Supreme Court, "a union may not arbitrarily ignore a meritorious grievance or process it in a perfunctory manner."57
    It should be noted that although the duty of fair representation was first used to protect non-members, the union's duty does not rest on the element that its authority to represent comes solely from the statute. The union owes the same duty to members who voluntarily joined the union and thereby authorized the union to represent them.58 Indeed, the duty is owed even to those whom the union does not represent but whose jobs the union effectively controls.59
    Although Swedish unions exercise substantial control over the terms and conditions of employment of both members and non-members and have the status, for many purposes, of exclusive representation, Sweden has not developed any analogue to the duty of fair representation. This raises the question - why not?
    The most obvious reason for this difference is the historical difference in collective agreements. In the United States the most visible and inexcusable discrimination has been in writing and applying seniority provisions, for these provide the greatest opportunity for one group of employees to profit at another group's expense. Because of Paragraph 32, Swedish agreements contained no seniority clauses; the ability to be arbitrary or discriminatory was left in the employer. Indeed, national agreements, limited as they were to economic terms, provided much less opportunity for unfairness than the locally negoti-


52 Jones v. Trans World Airlines, 495 F.2d 790 (2d Cir. 1974); Hughes Tool Co.. 104 NLRB 181 (1962).

53 Peterson v. Rath Packing Co., 461 F.2d 312 (8th Cir. 1972).

54 Hughes v. International Brotherhood of Teamsters, Local 683. 554 F.2d 365 (9th Cir. 1977); Holodnak v. Avco-Corp., 381 F. Supp. 191 (D. Conn. 1974).

55 Local 13 International Longshoremen's and Warehousemen's Union Pacific Maritime Ass'n. 441 F.2d 1061 (9th Cir. 1971).

56 Hines v. Anchor Motor Freight. Inc., 424 U.S. 554 (1976); Foust v. International Brotherhood of Electrical Workers, 572 F.2d 710 (10th Cir. 1978); Ruzika v. General Motors Corp. 523 F.2d 306 (6th Cir. 1975).

57 Vaca v. Sipes. 386 U.S. 171. 191 (1967). See generally, Summers. The Industrial Employees Rights Under the Collective Agreement: What Constitutes Fair Representation. 126 U. of Pa. L. Rev. (1977).

58 Syres v. Oil Workers International Union. 350 U.S. 892 (1955).

59 Brotherhood of Railroad Trainmen v. Howard. 343 U.S. 768 (1952).


604 Clyde W. Summersated comprehensive agreements in the United States. Collective agreements implementing LO's solidarity wage policy might be viewed by some as not being fair and equal representation because it favors low wage employees at the expense of high wage employees, but this is the kind of policy judgement which American courts have held that unions are entitled to make.
    The character of Swedish collective agreements also reduces the danger of unfairness in the administration of the collective agreement. The great majority of unfair representation cases in the United States are ones in which the union refused to process or settled grievances involving seniority, promotions or discharge, none of which were regulated by collective agreements in Sweden. There is an additional legal reason for unfairness in administration of the agreement being agreater problem in the United States than in Sweden. In the United States, rights under the agreement are normally determined and enforced through the union's grievance procedure, which may end with the union appealing to arbitration. This procedure is binding onall employees, union members and non-members alike. Under American agreements, the individual has normally no legal right to insist onarbitration, so if the union refuses to appeal, the individual is blocked, and the only recourse is to claim violation of the duty of fair representation for the union's refusal to appeal.60 In Sweden, an employee claiming violation of his rights under the collective agreement confronts no such blockage. If the union refuses to carry a member's case to the Labor Court, the member can sue in the common courts.61non-member can sue directly on his individual contract of employment without any recourse to the union.
    The position of Swedish unions, however, has been dramatically changed by the legislation of the 1970's, for the union now has significant legal status beyond its role as a party to the collective agreement. Under the Employment Security Act, unions can now create seniority rules and have a voice in the order of layoff; and under the Work Environment Act, union safety representatives can significantly affect the working conditions of individual employees. Most importantly, the Codetermination Act gives the union a voice in a nearly unlimited range of decisions at the plant and enterprise level,


60 Vaca v. Sipes, supra note 57.

61 Hemström, supra note 47 at 48. Prior to amendments to the Labor Court Act in 1976, the union member could bring his suit in the Labor Court if the union refused to proceed. Lag den 22 juni 1928 om arbetsdomstolen, Sec. 13. For a study of cases under this section, see Summers, Collective Power and Individual Rights in the Collective Agreement-A Comparison of Swedish and American Law, 72 Yale L. J. 421 (1963). 

Comparisons in Labor Law 605and those decisions affect the most immediate and important concerns of the individual employee's working life. Swedish unions, even more than American unions, now exercise a broad and pervasive influence which provides an opportunity for arbitrary or indifferent treatment of those it represents. This new role of unions at the local level creates much the same conditions in Sweden which gave rise to the duty of fair representation in the United States.
    I do not mean to suggest that Sweden will or should copy the American rule; a Swedish solution must be tailored to fit the Swedish situation. Hemström, in proposing that Swedish law should recognize that unions owe a duty to represent fairly all employees, members and non-members, has shown the kinds of considerations which must be taken into account in developing a Swedish solution.62
    The important point here is not whether or how Sweden should respond to the expanded role of unions. My purpose is to illustrate that by persisting in asking the right questions we may see more clearly significant characteristics of both systems and come to realize that surface differences in the legal status of unions obscure underlying similarities. This, in turn, may lead to recognition of lurking problems and suggest constructive proposals, such as Hemström's of how to meet developing needs.


In the preceding sections we have examined, first, an area in which the basic legal rules are, in form, quite similar, but when seen in their context reach quite different practical results; and second, an area in which the basic legal rules are in sharp contrast but other legal rules and institutions lead to quite similar results. We now examine an area in which the contrast is one of both legal form and living fact.
    In the United States, legal intervention in internal union processes is substantial. The Landrum-Griffin Act of 195963 requires unions to maintain certain minimum standards of democratic decision-making, due process and financial accountability. The underlying premise of the statute is that under the National Labor Relations Act, the union which is the bargaining representative determines employees' terms and conditions of employment, binds them by the collective agreement, and controls the grievance procedure through which their con-


62 Hemström, supra note 47.

63 Labor Management Reporting and Disclosure Act of 1959, 73 Stat. 519, 29 U.S.C. 401. See generally, Bellace & Berkowitz, The Landrum-Griffin Act (1979). 

40-33-169 Sv Juristtidning


606 Clyde W. Summerstract rights are enforced. As the Report of the Senate Labor Committee reasoned: "The Government which gives unions this power has an obligation to insure that the officials who wield it are responsive to the desires of the men and women they represent."64


    Legal intervention is limited and narrowly focused. The statute does not regulate the union's collective bargaining, political or social policies; those are to be decided by the union members. The statute requires only that in making these decisions the union must follow democratic procedures to insure that union policies are responsive tothe desires of the members. The logic was that because unions exercise powers akin to those of government, they should observe the standards expected of democratic government. The required procedures are, therefore, conceived and articulated as parallel to those required of government itself.
    Following this parallel, Title I of the Act is entitled, "Bill of Rights for Union Members." The first right is that "Every member ... shall have equal rights and privileges ... to nominate candidates, to vote in elections or referendums ... to attend membership meetings, and to participate in the deliberations and voting upon business of such meeting." The courts have held that this right is violated if union officers refuse to allow discussion and a vote at a meeting,65 or if the officers sign a contract after it has been rejected by a referendum.66
    The second right is entitled, "Freedom of Speech and Assembly," and protects "the right to meet and assemble freely with other members; and to express any views, arguments or opinions; and to expressat meetings ... his views upon candidates in an election . . . and on any business properly before the meeting." This right includes a broad right to criticize union officers and oppose union policies. A member can not be expelled from the union for claiming that an officer had misused funds, even though the accusation is not proven.67 Nor can a member be punished because he argued against an increase in dues,68 urged that a strike be ended,69 or advocated legislation opposed by the


64 S. Rep. No. 187, 86th Cong. Ist sess., p. 20 (1959).

65 Parker v. Local Union No. 1466, United Steelworkers of America, 642 F.2d 104 (5th Cir. 1981). The union officers failed to keep order, refused to recognize those advocating a particular view, and adjourned without a vote.

66 Aguirre v. Automotive Teamsters, 633 F.2d 168 (9th Cir. 1980). It is also a violation of this right for officers to circulate inadequate or misleading information on the matters to be voted upon. Blanchard v. Johnson, 388 F. Supp. 208 (N.D. Ohio, 1975), aff'd inrelevant part, 532 F.2d 1074 (6th Cir. 1976).

67 Salzhandler v. Caputo, 316 F.2d 445 (2d Cir. 1963).

68 Farowitz v. Assoc. Musicians of Greater New York, 330 F.2d 999 (2d Cir. 1964).

69 Kuebler v. Cleveland Lithographers and Photo-engravers Union Local 24-P, 473 F.2d 359 (6th Cir. 1973). 

Comparisons in Labor Law 607union.70 Debate concerning officers and policies is to be open, vigorous and unrestrained, with members free to speak both inside and outside union meetings.
    Other provisions of the Bill of Rights give union members the right to sue the union, free from any sanction or discipline, 71 and require that before a member can be fined, expelled or otherwise disciplined, the member must be given written notice of the specific charges made against him, allowed a reasonable time to prepare his defense, and provided a full and fair hearing.72
    Title IV of the statute regulates the process of electing union officers and is designed to make elections an open competition, with the campaign serving as a focus for debating union policies. Periodic elections of union officers at all levels of the union is required,73 and the union is limited in the qualifications it may require for candidates for union office.74 Every member is guaranteed the right to make nominations and to campaign for a candidate. Candidates are entitled to have their campaign literature distributed to all members; and no union money can be used to help elect a candidate.75 Thus, the union newspaper or magazine can not be used to favor a candidate. "Adequate safeguards to insure a fair election" are required, including the right of every candidate to have observers at the voting and the counting of the ballots. If there is a violation of the statute which may have affected the outcome of the election, the Secretary of Labor can obtain a court order setting aside the election and ordering that a new election be held under the supervision of the Secretary.76
    In addition to these provisions which directly relate to the decision making process, Titles II and V of the statute make union officers accountable to the members. Title II requires unions to file annual financial statements with the Secretary of Labor and to make this information available to all their members. Any member is entitled, upon showing reasonable cause, to examine the union's books andrecords to verify these reports. The members may thereby discover in


70 Mitchell v. International Ass'n of Machinists, 196 Cal. App. 796 (CA 2d, 1961).

71 Sec. 101 (a) (4).

72 Sec. 101(a) (5).

73 Elections for national officers must be held at least every five years, and elections of local officers must be held at least every three years. Sec. 401 (a) and (b).

74 The union can not require that candidates for president have held lower offices or have attended union meetings regularly where the effect is to disqualify 95 % of the members from being candidates. Wirtz v. Hotel. Motel and Club Employees Union, Local 6, 391 U.S. 492 (1968); Local 3489. United Steelworkers of America v. Usery, 429 U.S. 305 (1977).

75 Yablonski v. U.M.W.A., 305 F. Supp. 868 (D.D.C. 1969).

76 Sec. 402.


608 Clyde W. Summersdetail how the union's funds have been used and its business conducted.77 Title V declares that all union officers and representatives "occupy positions of trust" in relation to the union and its members, and are therefore under a duty "to hold its money and property solely for the benefit of the organization and its members," and to refrain from any dealings which would create a conflict of interest with the union. Union officers have been held by the courts to violate their trust for expending union funds not properly authorized by the members78 and for failing to expend money for purposes voted by the members.79 They also violate their trust if they submit a question to referendum vote without fully informing the members of the proposal and its consequences.80
    In contrast with this multi-titled American legislation, there is almost a total void of legal rules concerning the internal process of unions in Sweden. There is no legislation and only a handful of nearly forgotten court decisions which ambiguously expressed two or three elementary principles. In 1936, the Supreme Court awarded damages to a member who had been expelled from his union for refusing to take part in a boycott which had been declared illegal. The court held there were no legal grounds for the expulsion.81 During World War II, the Metal Workers expelled a member for "disloyalty" because he had urged his branch to disregard a national union circular about keeping communists out of positions of trust in the union;82 and the Printers Union expelled a member who refused to resign from the National Socialist Party.83 In both cases, the Supreme Court invalidated the expulsions. In neither case did the court articulate any basic principle of freedom of political beliefs or freedom of political activity in the union. The court went no further than to say that the expulsions were not authorized under the union rules. In 1948, the Supreme Court ordered the Bricklayers Union to accept a Norwegian bricklayer who had been denied admission. The court based its decision on twogrounds —that because of the union's position in the industry, union membership was essential for one in the bricklaying trade; and that


77 Fruit and Vegetable Packers and Warehousemen Local 760 v. Morely, 378 F.2d 738 (9th Cir. 1967).

78 Morrissey v. Curran, 650 F.2d 1267 (2d Cir. 1981).

79 Johnson v. Nelson, 325 F.2d 646 (8th Cir. 1963).

80 Blanchard v. Johnson, 388 F. Supp. 204 (N.D. Ohio 1975), modified 532 F.2d 1074 (6th Cir. 1976); Cefalo v. Moffett, 333 F. Supp. 1283 (D.D.C. 1971) modified 449 F.2d 1193 (D.C. Cir. 1971).

81 1936 NJA 672.

82 1945 NJA 290.

83 1946 NJA 83.


Comparisons in Labor Law 609the union rules provided that it was to be open to all working in the trade.84
    The court decisions establish, at most, only rudimentary principles.85 Unions, because of their importance to the worker's economic welfare, are governed by stricter standards than other "ideell" or voluntary associations; the courts will examine both the procedural and substantive basis of an association; and the unions will be required to comply with their own rules and by-laws, which are regarded as a contract between the member and his union. The courts, however, have not asserted that unions will be required to provide procedures which are fundamentally fair, or that basic democratic rights and processes in the union will be protected.
    In principle, there would seem to be as much or more reason in Sweden for the law to require unions to observe minimum democratic standards as in the United States. Swedish unions in fact play a much larger role than American unions in regulating terms and conditions of employment because of their completeness of organization, and they have much greater influence on shaping social and economic policy because of centralized bargaining. Swedish political theory has been much more explicit than American theory in relying on organizations, such as unions, to regulate economic and social problems,and viewing them as integral parts of the structure for governing society.86 Unions and employers associations are recognized in Sweden as the primary governing bodies of the labor market. The premises and logic of Landrum-Griffin would seem to be even more applicable in Sweden — the law should insure that their decisions were democratically determined, responsive to the desires of their members. Contrary to this logic, there have not been even serious suggestions that such legislation should be considered.
    This raises the question: Why is there this sharp contrast in Swedish and American law? The most obvious explanation is that there has not been a history of corrupt and oppressive practices in Swedish unions. The Landrum-Griffin Act responded to a congressional investigation which paraded across newspaper headlines and television screens a series of examples of systematic corruption and dictatorial


84 1948 NJA 543.

85 All of the cases involving unions and other voluntary associations are analyzed in Hemström, Uteslutning ur ideell förening (1972), Ch. 4. See also Schmidt, Law and Industrial Relations in Sweden (1977) 45 — 54.

86 See Heckscher, Staten och organisationerna (1951); Johansson, Folkrörelserna och det demokratiska statsskicket i Sverige (1952). 

610 Clyde W. Summerscontrol by union officers. Although these were limited to a small number of unions, some of those unions were among the most important unions, and the practices were extensive and outrageous.87 In Sweden, such corruption and misuse of union office is unknown; the recent public charges relating to the leadership in the Transport Workers is a rare exception and insignificant compared with endemic practices in the Teamsters Union in the United States. In addition,the Standard By-Laws recommended by LO, and adopted by most unions, makes available appeal to an outside arbitrator in adjudication of internal union disputes—a device shunned by all but two or three American unions.88 One obvious reason for no legislation in Sweden is that there has been no felt need for legal controls; union self-control has proven itself adequate. Another reason is that unionshave had the political power to discourage such legislation.
    This explanation, however, only raises a further question: Why have these abusive practices occurred in American unions but not in Swedish unions? There are undoubtedly a variety of reasons whichreach deep into social and cultural differences, including law-abidingness and responsiveness to social pressures, but the dominant factor is probably the difference of union ideology.
    Swedish unions historically had deep socialist roots, both in early leadership and articulate philosophy. LO was, from its beginning, closely tied to the Social Democrat party, with the unions and the party considered as two wings of the socialist movement, each working toward the same goals.89 The purpose of unions, so conceived, is not to seek the narrow self-interest of their members, but to serve broad social needs of the working class, and to change society. Because of this ideology, union leadership has been viewed more as a calling with social responsibilities, than a prize to be rewarded, or a position to be exploited. American unions, in contrast, have been permeated with "business unionism," with the dominant purpose of the union to serve the self-interest of its members. This ideology dulls idealism and fosters the attitude that union leaders should also serve their own self-interest, and the attitude tends to become accepted by the union members. The crucial test of the union leader then becomes


87 See Summers, American Legislation for Union Democracy, 25 Mod. L. Rev. 273 (1962).

88 Klein, Lynn & Feller, Public Review Boards: Their Place in the Process of Dispute Resolution, Proceedings, National Academy of Arbitrators 27 Annual Meeting (1975) 189, 205, 221.

89 Lindbom, Den svenska fackföreningsrörelsens uppkomst och tidigare historia (1938), Ch. 4, 6; Westerståhl, Svensk fackföreningsrörelse (1945), Ch. 8. 

Comparisons in Labor Law 611whether he "delivers" in terms of economic benefits to the members. If he does, many members accept that he should be well paid and not condemned for taking for himself what he can. Indicative of the difference in attitudes is the contrast in the salaries of union officers. In Sweden the salaries of top officers in the union are in principle not to be more than earned by the higher paid members of the union.90 In the United States, the salaries of union officers may be more than five or even ten times that earned by their members.91
    The two ideologies inevitably develop different types of union leaders with different conceptions of their responsibility to the union. It is not surprising that corruption in American unions has been greater in those unions where the philosophy of business unionism is the strongest; and corruption has been least where the socialist or social welfare concern has been the strongest. Centralization in the Swedish labor movement, and the effectiveness of social pressures within the movement, discourage deviant behavior in individual unions; the dominant ideology dominates. In contrast, the individual autonomy of unions in the United States enables a minority of corrupt union leaders to prosper in a predominately honest and socially conscious union movement.
    Corruption, however, was only a secondary concern in American legislation. The primary concern was the democratic process within unions. The lack of Swedish law on this requires a different explanation, for according to American standards Swedish unions are not democratic. In Sweden the principal officers are elected for life—"tills vidare"92 —in violation of the basic American principle that elected officials should be required periodically to stand for reelection. Even before the Landrum-Griffin Act, which now requires periodic elections, the action of the dockworkers union in electing their president


90 The salaries of the principal officers in the Metalworkers Union and the Textile Union were a little more than £5,300 a year when the average earnings in the manufacturing industry were £3,140. Carew. Democracy and Government in European Trade Unions (1976) 151.

91 In 1974, the President of the Teamsters Union was paid a salary of $125,000 plus $8.300 allowances and expenses. Ten other officers of the international union received total salaries and allowances above $100,000. Teamster Democracy and Financial Responsibility, A Prod Report (1976), Ch. 13. At this time. the average annual earnings of Teamsters members was substantially less than $15.000. In 1980. a survey of the salaries of 54 top union leaders showed eleven receiving over $100,000 a year. This included officers of unions with low wage workers such as Hotel and Restaurant Employees and the Food and Commercial Workers. The average salary for the 54 union officers was $77,500. Business Week, May 12, 1980. p. 86 ff.

92 Carew, supra note 90, at 139. See, for example. Stadgar. Landsorganisationen i Sverige. 1977, § 8. Mom. 1: Stadgar. Svenska Kommunalarbetareförbundet. 1959, § 21. Mom. 1: Svenska Träindustriarbetareförbundet, 1977. § 8, Mom. 1. 

612 Clyde W. Summersfor life was seen as symbol of the unmatched corruption and oppression in that union, and he was derogatorily described as "King Ryan."
    The LO constitution requires that all national unions provide in their statutes that the union's executive board have the authority to conclude a collective agreement even though it has been rejected in a referendum vote.93 Such a provision would be almost unthinkable inan American union constitution. Many American unions provide for referenda votes, union members regularly reject proposed collective agreements,94 and if union officers signed an agreement rejected by the members, they would run serious risk of being replaced in the next election.
    In Swedish unions, elections are rarely contested, except perhaps at the plant level; election campaigns in which union policy is debated are almost unknown; and there seems to be an almost total lack of organized oppostion groups within the union criticizing union officersor union policies. Swedish unions are a model of one-party bureaucracy, described by Michels,95 where the incumbent officers and staff have a monopoly of channels of communications, information and organizational resources which suppress open debate and make organized opposition impossible. Members rise to positions of influence and control, not through election by the members, but through selection by the incumbent bureaucracy. It was this tendency in American unions which the Landrum-Griffin Act sought to counteract.
    We are now forced to return to the question—why has there been legislation regulating democratic processes in the United States and no such legislation even seriously suggested in Sweden? The explanation may be found in different basic conceptions of the democratic process. The Swedish conception is described as one of representative democracy with heavy emphasis on the representative exercising his knowledge and judgment on behalf of the members. His function is not to express their views, but to act for their benefit and promote their interests; his responsibility is not to do what they tell him they think is best, but to know and do what is in fact best. Members are not


93 Stadgar 1977, Landsorganisationen i Sverige, § 11, Mom. 2. See Hadenius, Facklig organisationsutveckling (1976) 126—130.

94 Summers, Ratification of Agreements, in Dunlop & Chamberlain, Frontier of Collective Bargaining (1967) 75-102; BNA, Labor Relations Yearbook, 1975, 39-40.

95 Michels, Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy (Paul trans., 1962). 

Comparisons in Labor Law 613expected to make decisions; decision-making is delegated to the representative and his judgment is relied upon.96
    The American conception is much more one of direct democracy. The local union meeting is the forum for decision-making, and referenda are commonly used to call strikes, ratify contracts, increase dues and amend union statutes. National union officers are often elected by referenda voting which may involve several hundred thousand ballots. Representative democracy has a quite different meaning, for the elected representative is expected to reflect the desires of his constituents. He is elected because the members believe he shares their views, and he will be replaced if they disagree with his decisions. He is the members' spokesman, not their decision-maker.
    Historically, the constitutions and governmental structures of American unions reflected the membership participation and control conception of democracy, and the public expected them to follow this form of democratic process. Legislation came because some unions engaged in practices which were destructive of the democratic process, so conceived. The Landrum-Griffin Act imposed on unions legal standards to insure the minimum requirements for this form of democratic process. In contrast, the union statutes and practices of Swedish unions have historically reflected the delegated representative conception of democracy; this was what the public expected of them and what they expected of themselves. Legislation has not been considered because unions have not violated their conception of democracy.
    This explanation, of course, requires further explanation—why is there such a difference in the conception of union democracy in the


96 This conception of union democracy may now be in the process of change. The Report of the 1981 LO Congress disavows the traditional conception of representative democracy as sufficient to provide the worker participation which the MBL contemplates. The Report declares:

"So, it is not enough either to say that democracy functions if the elected representative reflects the electors' or members' views ... democracy functions fully first when the sharp boundary between elector/members and elected ceases to exist, when the whole collectivity conceives itself as participants in the democratic process." (p. 39).


"A well functioning representative democracy assumes, however, that the member most directly involved has found some forum, for example, a club meeting, where he can go and make his voice heard and take part and choose representatives, otherwise many are excluded from involvement and influence." (p 128). For a similar emphasis on a new conception of representative democracy by the TCO, see Som vi ser det (1981) 127:

"The union representative may not fall into the role to himself be the union, to beone who knows and understands and who is not specially interested or subservient to what the members want. Instead, the member-representative relation mustreflect that one who is elected to a representative position is directly subservient to those who have elected him ... The actor shall be guided by the members." 

614 Clyde W. Summerstwo countries? In part, the explanation may be found in different traditions of political and governmental institutions and practices which assume and require different conceptions of representative government and the democratic process. Union governmental structures and processes in each country tend to parallel those of the political system of which they are a part. It is not surprising that the process in LO unions of consolidating small branch unions into large regional branches was matched by the movement in government to consolidate small local governmental units into larger units.
    Parallelism between union processes and governmental processes, however, ultimately provides an unsatisfying answer, for in Sweden the political processes of unions do not parallel the political processes of government. In unions there are no competing political parties oreven any organized opposition; there are no periodic elections which provide a referendum on the performance and policies of the leadership; the incumbent officers and bureaucracy are never displaced. The union is a one-party state. The same is true of American unions. The difference is that the American legislation has, perhaps more by coincidence than by design, provided legal protections which canbring a measure of democratic process to a one-party state. The essential elements for achieving this in unions are protection of the process of criticism and dissent by those who disagree with established policies or doubt the established leadership, and requiring union officers to stand periodically for election where the acceptability of their performance and policies can be contested and measured.
    We come back to the explanation that there has been no consideration of legislation in Sweden because there has been no felt need, and there has been no felt need because, regardless of the decision-making process, union policies have been generally accepted by the members. This is true even though some policies, such as wage solidarity, are at the expense of substantial minorities, or even majorities. These policies have been accepted, partly because unions, through their educational programs, study circles and union publications, have persuaded the majority that those policies are desirable, and partly because members rely on the judgment of their representatives as to what is desirable.97 Once a policy has been decided and is perceived as being supported by the majority, there seems to be a reluctance to criticize or challenge, and a sense of obligation to accept the established view.
    The policies of American unions probably are as responsive to the


97 See Lewin, Governing Trade Unions in Sweden (1981), Ch. 5; Hadenius, supra note 39, 171, 189-191. 

Comparisons in Labor Law 615desires of their members, of even more so than the policies of Swedish unions. But union members in the Unites States who disagree with union policies seem much more ready to express their dissent. Union members have much more willingness to doubt their representatives and to criticize their officers. There is much less sense of obligation to accept the established majority view. This reflects the general political values which encourage criticism and debate, and which protect and cherish dissent. These political values make the test of democracy the process through which decisions are made. The fact that the decision reached is accepted or sought by the majority is not enough; the process of decision-making must be democratic.
    American legislation came not so much because unions made wrong or unwanted decisions, but because some failed to follow democratic procedures in making those decisions. It seems, at least to an outsider, that in Sweden there is much more concern with the substance of the decision and less concern with the procedure by which it is reached. The policy decisions of Swedish unions have been accepted and supported by the members and acceptable to the larger society. No more is expected or required.
    Pressing the inquiry as to why there is this contrast in Swedish and American law concern for substance and procedure would carry us deeper into inarticulate assumptions and attitudes as to the essentials of a democratic society. That is beyond the reach of this paper and the capacities of its author. It is enough for our purposes here to demonstrate that the differences in the legal rules have sources in basic political conceptions and values. By asking why the rules are different we come to see more clearly those sources and understand better those conception and values. We are then better able to examine those conceptions and values, either to reaffirm or question them.


The primary focus here has been on the usefulness of comparative labor law in obtaining a different perspective and better understanding of our own system. Similarities and contrasts should provoke us to ask "Why," and each question "Why" opens a door into a new and often unexpected room, which itself has other doors to be opened by further questions. Though there is seemingly no end of doors, each opened one gives us fuller knowledge of the combination mansion and madhouse in which we live. We see overlooked needs, decaying remnants and unconsidered possibilities. We might even be moved to remodel or reconstruct.
    The path of comparative law seems an unduly long and tortuous


616 Clyde W. Summersone to reach self-awareness. Do we really need to study another labor law system to ask searching questions of our own? Could we not more profitably ask directly the questions why we follow certain legal rules and institutional practices? Could we not see our needs and possibilities for change without looking at another system? The answer is that we should be able to, and do, but only in part. Most of us are bound by our unconscious premises and have difficulty envisioning what we have not seen. When we have known only one labor law system we are captives of its purported premises and their claimed consequences.We can not easily imagine that essential parts might be otherwise; we do not see many of the questions most worth asking. Studying another system is particularly useful for those of us whose imagination is limited and whose mind shrinks from being bold.