IX. INCENTIVES IN THE PUBLIC AND NONPROFIT SECTORS: DO HIGH-PERFORMANCE WORKPLACE PRACTICES WORK?

Local Union Political Competition and Bargaining Performance

MORRIS M. KLEINER
University of Minnesota and the National Bureau of Economic Research

ADAM M. PILARSKI
AVITAS, DNV

Abstract

      How does direct union democracy, with much competition or representative democracy, with little rivalry, affect bargaining outcomes? This paper provides new data from two similar local unions negotiating in the aerospace industry. In addition, we gain insights on bargaining relationships from interviews with the participants in the labor–management events during the late 1960s through the late 1990s. Based on the wage settlements in the collective bargaining agreements and other measures of union success during the contract, the union members at the more democratic union did better than unionists with little political competition, although it comes at a price of greater strife.

Introduction

       Professor Henry Farber’s often cited review of the analysis of union behavior, he laments that no one has compared an “operating democratic union” with the “completely unfettered leadership-run union”(Farber 1986). Although there have been empirical estimates of the effect of small variations in union democracy on bargaining outcomes, these results have shown generally murky quantitative results (Fiorito and Hendricks 1986). In this paper we attempt to provide unique in-depth case evidence of the kind that Farber was alluding to in attempting to understand what might be the bargaining outcomes of unions with different levels of internal competition.

       This study provides new evidence from two similar local unions as well as insights from interviews with the participants of the events in this union and company, from the 1970s through the 1990s, of the impact of greater political competition within a union on bargaining outcomes and the impacts on the shop-floor bargaining. The in-depth analysis of the events, personalities, and outcomes in the union provides unique insights that could not be obtained by only examining the theory and data.

Incentives for Remaining a Union Leader

       From the perspective of union leaders in office, there are ample incentives to create an environment where challenges to their job are reduced. Being a union leader in most organizations gives the individual super seniority during potential layoffs, and takes the official off the mundane assembly line work for long periods of time. For those union leaders with public service ambitions, it can provide a springboard to local government work, political office, or opportunities to obtain a job with the national union. With these benefits for union leaders, the incentives for union behavior have long been a topic of debate and discussion in industrial relations.

       Conflicting perspectives in industrial relations--namely, what is the objective of unions and what is the role of political competition?--suggests that a more democratic union would be somewhat more successful not only in bargaining contracts but also in fractional bargaining on the plant floor in part because of the competition within the political system (Kuhn 1961). Recent theoretical analysis states that union dissidents typically accuse the union leaders of being too soft in their negotiations with firms. Dissidents rarely accuse the leadership of costing the union members jobs by negotiating too generous an agreement from management’s perspective. Professor Michael Kremer states that union leaders who are not seriously challenged are often prepared to sacrifice workeroriented provisions, such as wages, for union-oriented provisions such as union security, automatic checkoffs of union dues, the right of the union to participate in all grievance negotiations, and preferential seniority for union officials (Kremer 2000). Among the features that give incumbents strong advantages are indirect elections of leaders. If, at the national level, incumbents see that a local union is not supporting them, they have the power to place a local under receivership, and the national union takes over the affairs of the local. Similarly, if a local president sees a steward in a work group or team who is not supporting him, she or he has the power to remove that individual. Moreover, there are often charges of vote-stealing in local elections. Union officers are not required to give memberships lists to opposition candidates. If a local union is spread out over a large geographic area, then this becomes an even greater impediment for a potential challenger because it imposes greater time and money costs of a challenge. Finally, local staffs are known to contribute to the campaign of local union leaders, making it even more difficult for challengers when campaigns are more expensive. Furthermore, when union staffs are geographically dispersed and less visible, then opposition candidates will find it even more difficult to mount significant political challenges.

Comparing Two Local Unions

       In Figure 1 we show how union democracy at the local level is determined for our comparison of two similar unions, and how it may impact bargaining outcomes. In the far left box, the organizational structure of a union is assumed to impact the level of democracy. For example, if the structure of a local union is an amalgamated one, it would then consist of several plants, generally in the same geographic area. In contrast, if a local union represents only one plant, this would affect the level of union democracy through the level of competition for key positions. Further, within local unions, the structure of voting would also matter in the determination of the level of competition. Direct elections would have their greatest impact on equalization of candidates running for office. Consequently, if there is a great deal of competition for the top union positions, then the union leaders, which include the union stewards, would need to be more responsive to the interests of the members. In the middle box of the figure, the greater the level of democracy and competition in the union, the tougher the union leaders must be in negotiations, which results in better collective bargaining agreements from the perspective of the membership, all else equal. Of course measures of the success of bargaining outcomes would include “orbits of coercive comparison” relative to other negotiated agreements or inflation (Dunlop 1944). This model suggests that union democracy, as defined by more competition within the union, affects what union members receive at the bargaining table and the power they have vis-à-vis management at the shop floor.

       Local 148 of the United Automobile Workers (UAW) represented employees at McDonnell Douglas, now Boeing Aircraft, at its Long Beach, California facility for more than 60 years and is the focus of our analysis because of its unusually high level of democratic competition.¹ Throughout the 1960s and into the early and mid-1990s, the local had between 10,000 and 35,000 members. The production of large commercial aircraft was in one large plant and subsidiary storehouse plants surrounding the main facility. This made it easy to organize rival political parties within this union local, and there were direct elections of all union officials. The plant also comprised the entire local, an unusual arrangement for unions in the UAW. This union was highly democratic using virtually any measure. For example, there were rival political parties with many candidates, hotly contested elections, and conflicts with the international UAW over bargaining demands during collective bargaining. Union meetings were open to direct legislation from the membership at the floor of the meetings, which made direct democracy a tradition and part of the culture within the local. This union would epitomize Farber’s view of an “operating democratic union” and would be consistent with the first box in Figure 1. As a consequence, the union leadership had to be responsive to the membership or a rival leader would organize a new political party and challenge the leadership in the next election.

       Our discussions with past union presidents, union stewards, and national union officials responsible for aerospace negotiations concluded that one of the major reasons for this high level of democracy evolved from the way the plant was designed for the assembly of large commercial aircraft.² Their opinion was that the high levels of union democracy developed largely because all the employees were “under one roof.” This meant that getting political information out to the members was easy and cheap. Campaigning could be done during lunch hours, coffee breaks, or just before or after work. Getting political information to the membership in this plant involved going to a central location in the plant and handing out pamphlets that were often mimeographed on plain paper.³ There were hotly contested races for all major leadership positions and many candidates from the 1960s through the 1990s. No union president served more than two terms, and only one served two consecutive terms as the leader of the local. The spread between the winner and loser in presidential elections during the period was no greater than 5 percent. This competition ranged from union presidents who wanted cooperation with management to ones who desired conflict because it would produce a more lucrative contract for the membership.

       In contrast, Local 887, which was also a UAW affiliate in the aerospace industry, had a representative form of union democracy spread over several plants in the Los Angeles, California metropolitan area and within commuting distance of the plant that 887 represented. They had about the same number of members, roughly 24,000 depending on the period of time and demand for the product, and there was considerable growth and decline depending on government defense orders. This union was an amalgamated union that represented eight plants for the Rockwell company in the Los Angeles metropolitan area, a large defense contractor that produced mainly military aircraft. As a consequence of geographic dispersion, the international UAW organized this local as a representative form of democracy rather than a direct democracy, which was the case at Local 148. During the 1970s through the 1990s, Local 887 had no hotly contested elections. In most elections, there was either no opposition candidate or token opposition, and no union president was voted out of office during the period. There were three presidents during the 1970s through the 1990s, and the ones who left office did so either to enter local or state political life or to obtain managerial positions with the international union. Since there was no meaningful opposition to the union leader, there may have been fewer incentives to perform in the interests of the members. For example, few grievances went to arbitration, and management was able to implement work teams, with lots of employee involvement with little opposition from the local union. In contrast to Local 148, distance and indirect democracy made it difficult to organize opposition to this local’s entrenched political party.

       These differences in structure were associated not with regular union meetings where motions “from the floor” from the membership could be acted upon, but with monthly meetings that were conducted through representatives of the local in each of the eight plants. Second, union stewards were often appointed by the representatives of the local in the plant. Annual meetings were well organized and, some have suggested, orchestrated to serve the interests of the union president. The union officials argued that this system provided stability and stewards with specific human capital who knew how to handle grievances. In contrast to Local 148, which had considerable turnover of stewards and top officials, Local 887 had individuals in place who were not likely to lose their position in the union. It was costly to form any opposition to the leadership both in terms of time spent organizing and in the potential loss of political favor with the leadership. Consequently, there were few unionists who attempted to do so. Management liked the stability and certainty that this type of union organizational structure provided the company. This local union is close in Farber’s terminology to being a “completely unfettered leadership-run union.”

       The stability of the staff at Local 887 provided job-specific human capital that would allow the union to be knowledgeable about doing their job and serving their members. When it came to negotiating collective bargaining agreements, they knew the relevant information on wages for comparable employees in the region and country. In handling grievances, long tenured stewards have a longer institutional memory about previous contracts, the issues negotiated in previous contracts, and how arbitrations works. This longer experience by the leadership would give Local 887 a major advantage in negotiating better collective bargaining agreements.

       From an analytical perspective, one unique aspect of our “quasi-experiment” is that by picking a local comparison union with such similar outward characteristics as Local 148, we are able to difference out common elements between the two local unions, and any remaining differences in bargaining outcomes is a consequence of the variation in union democracy.⁴ The common elements of both union locals are the same national union, worker skills and education, metropolitan area, membership size, and industry. The major differences in the two local unions are the levels of internal competition and the related organizational structure.

Bargaining Performance of the Two Locals

       Although clearly not a controlled experiment, Locals 148 and 887 do provide what may be considered a “quasi-experiment” of the effects of union democracy. In spite of the pattern bargaining structure that dominated the policies of the international UAW, there were differences in wages for similar jobs. In Table 1 we give the differences in the maximum salary for two common occupations in the two companies--tool and die makers, a skilled occupation, and janitors, an unskilled occupation--for contracts negotiated from 1969 through 1997. This was the last year the companies were independent of Boeing. We also chose these occupations, since they appeared in all the contracts, whereas many other occupations appeared in some but not in other labor contracts over this 28-year period. Since most workers in each job were at the maximum wage in both plants, this hourly wage reflects the mode of wages for these occupations in both organizations.

       The results show that these employees at Local 148 usually had higher wages depending on the time period across these many collective bargaining contracts. There were small differences in favor of Local 148 in the 1960s of approximately 2 to 4 percent for tool and die makers, but this value grew to almost 11 percent by the late 1990s. Wage differences in 1969 for janitors favored those employed in Local 887 but, by the late 1990s, janitors in Local 148 had a more than 26 percent wage advantage. This reflects the impact of strikes, work slowdowns, and tough negotiations at Local 148 and McDonnell Douglas that resulted in major wage gains for union members in the company. Union leaders in 148 needed to be responsive to the members’ desires to obtain good contracts or be voted out of office, and this seemed to dominate the benefits of the greater experience of the union leaders in Local 887.

       All other major categories of jobs showed a similarly higher pay range for employees in Local 148. Moreover, the shift premium for working for second shift was 20 percent higher at Local 148, but the third shift was the same for employees in both local unions. Local 148 also had pension benefits that were approximately 8 percent higher per year employed by the company for employees in similar jobs with the same years of seniority for contracts negotiated in the late 1990s. Local 148 also had a more generous 401(k) provision in their collective bargaining agreement so that the workers, together with the company contribution, could save 14 percent in the more politically competitive union versus 10 percent for Local 887. Bargaining unit members in Local 148 had one seniority plan for the whole plant, but Local 887 had varying plans based on the specific Rockwell plant, and there were bumping rights across the plants. This came in spite of official publications by the national UAW that they were attempting to get the same benefits for members at both unions.⁵

       In obtaining due process at the workplace and cooperation from management, our discussions with management at McDonnell Douglas and union leaders at both locals also showed that Local 148 had greater power in fractional bargaining. The union leadership sometimes bragged that becoming a top corporate leader within labor relations at McDonnell Douglas meant having the approval of the union leadership. “Clear it with 148” was a common refrain heard from management in the industrial relations office regarding those who hoped to advance at the company in labor relations. This type of power in plant-level post-contract negotiations meant having a strong political identity and incentives to serve the membership even if it meant “crossing” management.

       There were other major differences on labor relations events. Over the same period, there were three strikes and one work-to-rule policy between Local 148 and McDonnell Douglas. However, there were no strikes, no work to rule, and an explicit TQM policy with teams and employee involvement agreed upon by the union with Rockwell. Perhaps the solidarity that links high levels of democracy also leads union leaders to show that they can deliver the best agreement possible, showing that the union acts as a political institution when it comes to wage setting. Based on the collective bargaining agreements for both unions, and virtually all other measures of union success--including voice at the workplace, wages and benefits, and having a balance of power with management at the shop floor--our results show that unionists at Local 148 did better than those at 887 in spite of the greater experience of unionists at this local.

       Certainly other factors such as differences in the structure of the locals, the economic situation of the company, the level of union leadership-specific human capital, and demand for the planes that were being produced by firms also influenced bargaining outcomes. Employees in Local 887 work primarily on military planes, whereas workers at Local 148 are primarily in commercial aviation production. The military production market has only one customer, the U.S. government, whose principal consideration is the quality of performance in the mission. On the other hand, the main criterion for sales of commercial aircraft is price, and McDonnell Douglas had to be much more sensitive to the cost of factors of production, especially labor costs, than Rockwell. Another possibility for higher wages and benefits at Local 148 is potentially higher productivity and effort relative to Local 887. This also seems unlikely since Local 887 had an employee involvement program with considerable employee participation for most of the 1970s through the 1990s, but Local 148 had only a short-term program that failed. Most of the empirical evidence supports the role of employee involvement in increasing productivity (Ichniowski et al. 2000). All of these economic factors would favor a better collective agreement for Local 887. Nevertheless, the results showed better outcomes for Local 148.

       From the company’s perspective, having this level of economic power by the union would result in Local 148 being able to put short-run pressure on management to make wage and work-rule concessions in order to meet current production schedules. However, in the long-run, management was able to contract out work and put new production facilities in neighboring states or other countries, where union power was perceived to be much weaker. In this case the Dunlop view of wage bill maximization was not realized in the long run by the members of the union, as employment declined dramatically to less than 5,800 employees 2 years after the takeover by Boeing during the late 1990s.⁶ However, Local 887 employment also declined, and there were about 1,400 union members by the year 2000 following the Boeing takeover of Rockwell.

Conclusions

       How does direct democracy, with much competition, or representative democracy, with little rivalry, affect bargaining outcomes? We provide a comparison of a highly democratic union with one that has one party rule and is consistent with a model of union behavior suggested by Henry Farber as an appropriate experiment in analyzing this issue.

       There were both political and economic factors that influence the behavior of the two unions examined in this paper. In contrast to Local 148, Local 887, which represented several plants in the same industry and location, had a substantially lower level of democracy within the plant. Local 148 had many changes in leadership that represented a variety of political parties that had differing views toward management. On the other hand Local 887 had little contested changes in leadership, which resulted in a stable group of union representatives and a long-standing employee involvement program. In contrast Local 148 was involved in three strikes and one work-to-rule action, and had major swings in the policies of the union leaders.

       The economic objectives of the national union were to obtain comparability in wages and benefits between union locals in aerospace in the Los Angeles metropolitan area. Further, the company that negotiated with Local 887 had considerably fewer economic incentives to cut costs and had almost no competitors in its specialized segment of aerospace. In spite of these constraints by the national UAW and with the competitive structure of the product market, Local 148 still had many more economic and job security benefits than Local 887. In short, the level of political competition within the union appears to have led to many more economic gains relative to a similar union local, but the contract gains also should be measured against the costs to the union members of the lost earnings during the three strikes, and the reaction by management, which resulted in subcontracting work to other vendors and the eventual selling of the business to its main rival.

Acknowledgments

       We thank William Appleby, Jeff Balfour, Sandy Donovan, Ken Erickson, Douglas Griffith, Ken Hill, Bruce Kaufman, Paul Schrade, and Shirley Underwood for their comments and assistance with this paper.

Endnotes

1. For more details on the election campaigns and election outcomes in both unions, see Kleiner and Pilarski (2001).

2. We had lengthy discussions with Doug Griffith, a past president of Local 148, and with Shirley Underwood, a former union steward in the local and now an international representative for the UAW. In addition, we discussed the politics of Local 887 with Ben Aceves, an international representative with the UAW, and a former union officer with Local 887, Paul Schrade, as well as the current president of the local. We visited the Local 148 headquarters and held discussions with members of several of the political parties in the local as well as members of the retirees association. We also visited the McDonnell Douglas plant, saw production facilities, and discussed union politics with members of the labor relations department in the company.

3. We were given access to many campaign materials, handouts, and strategies used during the elections by former president Griffith and former union steward Underwood.

4. See Arvey et al. (1991) and Freeman and Kleiner (1990) for a fuller explanation of the statistical impact of choosing a close companion/twin organization in differencing out common elements in statistical inference.

5. News releases from the UAW aerospace group for members were showing that each of the contracts were similar, calculations from the contracts obtained from the international for both unions showed the important distinctions in the contracts that are noted in the text.

6. It should also be noted that, because of the decline in defense orders, union employment at Rockwell also declined substantially during the 1990s.

References

Arvey, Richard, Thomas Bouchard, Nancy Segal, and Lauren Abraham. 1989. “Job Satisfaction: Environmental and Genetic Components.” Journal of Applied Psychology, Vol. 74, no.2, pp. 187–92.

Dunlop, John. 1944. Wage Determination Under Trade Unions. New York, NY: MacMillan.

Farber, Henry. 1986. “Analysis of Union Behavior” In Orley Ashenfelter and Richard Layard, ed., Handbook of Labor Economics, New York, NY: North-Holland, pp.1031–90.

Fioito, Jack, and Wallace E. Hendicks. 1987. “Union Characteristics and Bargaining Outcomes,” Industrial and Labor Relations Review, Vol. 40, no. 4 (July), pp. 569–84.

Freeman, Richard B., and Morris M. Kleiner. 1990. “Impact of New Unionization on Wages and Working Conditions: A Longitudinal Study of Establishments.” Journal of Labor Economics, Vol. 7, no. 3, pt. 2, pp. S-8–S-25.

Ichniowski, Casey, David Levine, Craig Olson, and George Strauss. 2000. The American Workplace. Cambridge, U.K.: Cambridge University Press.

Kleiner, Morris M., and Adam Pilarski. 2001. “Does Internal Political Competition Enhance Its Effectiveness?” In S. Estreicher, H. Katz, and B. Kaufman, eds., Internal Governance and the Organizational Effectiveness of Labor Unions. New York, NY: Kluwer Publishing, pp.103–19.

Kremer, Michael. 2000. “ An Epidemiological Model of Unions,” Working Paper, Harvard University, pp. 1–24.

Kuhn, James. 1961. Bargaining in the Grievance Settlemen., New York, NY: Columbia University Press.