United States General Accounting Office GAO Chairman, Committee on Commerce, Science, and Transportation, U.S. Senate June 2003 AIRLINE LABOR RELATIONS Information on Trends and Impact of Labor Actions GAO-03-652 June 2003 AIRLINE LABOR RELATIONS Information on Trends and Impact of Highlights of GAO-03-652, a report to the Labor Actions Chairman, Committee on Commerce, Science, and Transportation, U.S. Senate Labor negotiations in the airline Since the airline industry was deregulated in 1978, the average length of industry fall under the Railway Labor Act. Under this act, airline negotiations has increased, strikes have declined, and nonstrike work actions labor contracts do not expire, but (e.g., sickouts) have increased. After 1990, the median length of time needed instead, become amendable. To for labor and management at U.S. major airlines to reach agreement on help labor and management reach contracts increased from 9 to 15 months. Of the 16 strikes that occurred at agreement before a strike occurs, those airlines since 1978, 12 occurred prior to 1990, and 4 occurred the act also provides a process— subsequently. All ten court-recognized, nonstrike work actions and all six including possible intervention by presidential interventions occurred since 1993. the President—that is designed to reduce the incidence of strikes. Summary of Negotiation Trends Since Deregulation Despite these provisions, negotiations between airlines and their unions have sometimes been contentious, and strikes have occurred. Because air transportation is such a vital link in the nation’s economic infrastructure, a strike at a major U.S. airline may exert a significant economic impact on affected communities. Additionally, if an airline’s labor and management were to engage in contentious and prolonged negotiations, the airline’s operations—and customer service—could suffer. GAO was asked to examine trends in airline labor negotiations in the Airline strikes have had obvious negative impacts on communities, including 25 years since the industry was lost income for striking and laid off workers, disrupted travel plans, and deregulated in 1978, the impact of decreased spending by travelers and the struck airline. However, such airline strikes on communities, and the impact of lengthy contract impacts have yet to be thoroughly and systematically analyzed. The potential negotiations and nonstrike work net impacts of a strike on a community would depend on a number of factors, actions (such as “sickouts”) on such as availability of service from competing (nonstriking) airlines and the passengers. length of the strike. For example, of two recent strikes, one lasted 15 days and one lasted 24 minutes. GAO’s analysis indicates that passenger service has been affected more adversely by nonstrike work actions than by an increase in the length of negotiations. Generally, but not always, as negotiation periods increased, there has been a slight decline in on-time flights. However, the impact of these negotiations has been unclear because the decline may also have been www.gao.gov/cgi-bin/getrpt?GAO-03-652. affected by other factors such as poor weather. By comparison, the 10 court- recognized, nonstrike work actions more clearly resulted in negative impacts To view the full product, including the scope and methodology, click on the link above. on passengers, as shown through such measures as a decrease in the number For more information, contact JayEtta Z. of on-time flights, an increase in the number of flight problem complaints, and Hecker at (202) 512-2834 or a decrease in passenger traffic. HeckerJ@gao.gov. Contents Letter 1 Results in Brief 3 Background 5 Length of Negotiations and Number of Nonstrike Work Actions Have Increased, While Number of Strikes Has Declined 9 Airline Strikes Adversely Affect Communities, but Impacts Have Not Been Fully Analyzed and Vary from Place to Place 17 Nonstrike Work Actions Have Greater Impacts on Passengers than Lengthy Negotiations 24 Agency Comments 31 Appendix I Additional Questions 33 Appendix II Objectives, Scope, and Methodology 35 Appendix III Additional Background Information on the Railway Labor Act 38 Key Provisions of the RLA 38 Collective Bargaining Process under the RLA 39 Appendix IV Contracts Negotiated and Ratified or Settled by the Amendable Date 41 Appendix V Airline Strikes That Have Occurred Since Deregulation 42 Appendix VI Court-recognized, Nonstrike Work Actions Since Deregulation 43 Page i GAO-03-652 Airline Labor Relations Appendix VII Number of Presidential Interventions Since Deregulation 44 Appendix VIII Comments from the National Mediation Board 45 Appendix IX GAO Contacts and Staff Acknowledgments 51 GAO Contacts 51 Staff Acknowledgments 51 Tables Table 1: Unions Representing Selected Crafts or Classes at Major Passenger Airlines as of February 1, 2003 5 Table 2: States That Include Binding Arbitration or Last, Best Offer Arbitration as a Dispute Resolution Option 33 Table 3: Congressional Interventions in Railroad Negotiations 34 Figures Figure 1: Summary of Negotiation Trends Since Deregulation 3 Figure 2: Collective Bargaining Process under the Railway Labor Act 8 Figure 3: Length of Time Taken to Negotiate Contracts, 1978 to 2002 11 Figure 4: Median Negotiation Lengths by Carrier 12 Figure 5: Frequency of Strikes, Presidential Interventions, and Court-recognized, Nonstrike Work Actions by Year 15 Figure 6: 2001 Market Shares for Major Airlines at Their Hubs 21 Figure 7: Spoke Communities Served from Cincinnati and Minneapolis-St. Paul Retain Service from Competing Airlines 23 Figure 8: On-Time Flight Statistics for American and Delta at Dallas/Fort Worth International Airport, August 1998 to December 1999 26 Figure 9: Flight Problem Complaints for American and Delta from August 1998 to December 1999 27 Page ii GAO-03-652 Airline Labor Relations Figure 10: Passengers Carried on American and Delta at Dallas/Fort Worth International Airport, February 1998 and 1999 28 Figure 11: On-Time Flight Statistics for Delta and Continental at Atlanta Hartsfield International Airport, August 2000 to August 2001 29 Figure 12: Flight Problem Complaints for Delta and Continental, August 2000 to August 2001 30 Figure 13: Change in Passenger Traffic on Delta and Continental at Atlanta Hartsfield International Airport in December 1999 and 1 Year Later during the Nonstrike Work Action in December 2000 31 Abbreviations AFA Association of Flight Attendants AIRCon Airline Industrial Relations Conference ALPA Air Line Pilots Association AMFA Aircraft Mechanics Fraternal Association APA Allied Pilots Association ATA Air Transport Association CAPA Coalition of Airline Pilots Associations CESTA Communities for Economic Strength Through Aviation CWA Communications Workers of America DOT Department of Transportation IAM International Association of Machinists and Aerospace Workers IBT International Brotherhood of Teamsters NMB National Mediation Board PAFCA Professional Airline Flight Control Association PEB Presidential Emergency Board RLA Railway Labor Act SAEA Southwest Airlines Employee Association SWAPA Southwest Airlines Pilots Association TWA Trans World Airlines TWU Transport Workers Unions UFA Union of Flight Attendants Page iii GAO-03-652 Airline Labor Relations This is a work of the U.S. Government and is not subject to copyright protection in the United States. It may be reproduced and distributed in its entirety without further permission from GAO. It may contain copyrighted graphics, images or other materials. Permission from the copyright holder may be necessary should you wish to reproduce copyrighted materials separately from GAO’s product. Page iv GAO-03-652 Airline Labor Relations United States General Accounting Office Washington, DC 20548 June 13, 2003 The Honorable John McCain Chairman, Committee on Commerce, Science, and Transportation United States Senate Dear Senator McCain: Observers of the interactions between airline management and labor have long characterized these labor relations as contentious and adversarial. Negotiations between unions and airlines, for example, have taken up to 4 years to complete. Unions and airlines have each prolonged negotiations for their financial benefit. The importance of labor relations has recently been magnified by the financial crisis facing many airlines. Since September 11, 2001, US Airways, United Airlines, and Hawaiian Airlines have entered Chapter 11 bankruptcy, and American Airlines is fighting to avoid bankruptcy. US Airways, United, and American have all had to obtain the consent of their unions for contract concessions to substantially cut labor costs. At stake, according to industry observers and financial officials, has been the continued existence of at least two airlines. The process under which labor negotiations in the airline industry take place is substantially different than the process of most other industries. Airline labor contracts do not expire, and their negotiations can include a series of steps—including mediation, arbitration, and presidential interventions—specifically designed to avoid an impasse that would interrupt the flow of essential commerce. Since 1936, airline labor negotiations have been conducted in accordance with the requirements of the Railway Labor Act, which contains an established framework to reduce the incidence of strikes. Although the act is designed to bring about settlements without unions resorting to a strike, negotiations between the airlines and their unions have sometimes been contentious, and strikes have occurred. Recently, negotiations have at times been marked by nonstrike work actions, such as sickouts and work slowdowns. These actions are designed to place economic pressure on an airline. Because of ongoing concerns about the scope and impact of airline labor negotiations, you asked us to examine a number of issues concerning airline negotiations, strikes, and nonstrike work actions. As agreed with your staff, we examined the following three questions: Page 1 GAO-03-652 Airline Labor Relations • What have been the major trends of labor negotiations in the airline industry since the industry was deregulated in 1978, including the number and length of negotiations and the number of strikes, presidential interventions to avoid or end strikes, and nonstrike work actions? • What has been the impact of airline strikes on communities? • What have been the impacts of the length of negotiations and the occurrence of nonstrike work actions on passengers? In addition to these questions, you also requested information on states using binding and last offer arbitration1 for essential employees and the number of times Congress has intervened in railroad labor negotiations in the last 25 years. See appendix I for data on states using binding arbitration and congressional interventions in railroad negotiations. To determine the trends of labor negotiations, we analyzed data on negotiations, strikes, and nonstrike work actions from airlines, labor unions, the National Mediation Board (NMB), industry groups, and academic experts. We also interviewed officials with major U.S. airlines, labor unions, the NMB, industry groups, and academic experts. To determine the impact of strikes on communities, we reviewed available published studies from academics and other experts, and we analyzed data on airline schedules. To determine the impact of nonstrike work actions, we defined such actions as those in which airlines obtained either temporary restraining orders or injunctions against unions to prevent various actions. We then analyzed data from the U.S. Department of Transportation (DOT) on airline operational performance and passenger service. Except where noted, all data collected were current as of December 2002. Because of various data limitations, our analyses are restricted to major U.S. passenger airlines.2 We did not evaluate the efficacy or effectiveness of the Railway Labor Act or the impact of any 1 Last offer arbitration is a form of arbitration in which the dispute resolution procedure limits an arbitrator to choosing the final offer made by one of the parties. 2 DOT defines a “major” airline as one with annual total operating revenues in excess of $1 billion. For purposes of this report, we include only those major carriers that were in business as of 2001 for which we could obtain data (Alaska, America West, American, Continental, Delta, Northwest, Southwest, TWA, United, and US Airways). We excluded American Eagle, American Trans Air, cargo, and regional airlines, including American Eagle, some of which fit the $1 billion criteria, because data for these airlines were not available. Page 2 GAO-03-652 Airline Labor Relations possible changes to the act. Appendix II contains a more complete description of our scope and methodology. Since the airline industry was deregulated in 1978, labor negotiations have Results in Brief taken increasing amounts of time and have been marked less by strikes and more by nonstrike work actions. For the contracts we reviewed that had been negotiated between major carriers and labor unions since deregulation, the overall median length for contracts negotiated between 1978 and 1989 was 9 months, while the median negotiation time from 1990 to 2002 increased to 15 months. (See fig. 1.) However, some carriers such as Continental Airlines, Southwest Airlines, Alaska Airlines, and United Airlines have been more successful than others at reaching agreement with their labor unions in much less time. Of the 16 strikes that occurred, 12 took place from 1978 to 1989, and 4 took place since 1990. Various presidential interventions that have prevented or halted 6 strikes have all occurred since 1990, and all 10 court-recognized, nonstrike work actions have also taken place since 1990. Although the complete number of nonstrike work actions is not known because they are difficult to document, our evidence suggests that their use has increased in the past 12 years. Figure 1: Summary of Negotiation Trends Since Deregulation Page 3 GAO-03-652 Airline Labor Relations While strikes cause obvious negative impacts on affected communities, we could identify no published studies that comprehensively analyzed the full impacts of a past strike. Negative impacts of strikes include the lost income of striking and laid off workers, disrupted travel plans due to cancelled flights, decreased spending by the struck airline, and less spending by travelers. However, the overall economic impact of any past strike, including direct and indirect effects (and the offsetting effect of various mitigating factors such as the presence of service from competing airlines) has not been quantified. Our analysis of past strikes and other information indicates, however, that a strike’s potential impacts could vary greatly from community to community. For example, a community with substantial amounts of service provided by competing airlines is less likely to be affected than a community that is heavily or entirely dependent on the service provided by the striking airline, because passengers have continued access to air service. As a result, a thorough assessment of a strike’s impact on one community would be difficult to generalize to other locations. Our analysis indicates that passenger service has been affected more adversely by nonstrike work actions than by an increase in the length of negotiations. Generally, but not always, as negotiation periods increased, there has been a decline in on-time flights.3 However, the impact of these negotiations has been unclear, because the decline may also have been affected by other factors such as poor weather, aircraft maintenance, runway closures, air traffic control system decisions, or equipment failures. By comparison, the 10 court-recognized, nonstrike work actions more clearly resulted in negative impacts on passengers, as shown through such measures as a decrease in the number of on-time flights, an increase in the number of flight problem complaints, and a decrease in passenger traffic. For example, during an American pilot slowdown in 1999, the percentage of flights that arrived or departed on time declined by 11.6 percentage points. Also, customer flight complaints with DOT about American nearly quadrupled during the period of the nonstrike work action, rising from 53 to 203 complaints, while passenger traffic fell by 15 percent as compared to the year before. 3 DOT defines an on-time flight as one that is less than 15 minutes after the scheduled gate arrival or gate departure time. Page 4 GAO-03-652 Airline Labor Relations All the major airlines have some union representation of at least part of Background their labor force. The various crafts or classes4 that unions typically represent include pilots, flight attendants, mechanics, and dispatchers. Sometimes unions also represent customer service agents and clerical workers, aircraft and baggage handling personnel, and flight instructors. The extent of unionization among the major carriers varies significantly. At Delta, unions represent the pilots and two small employee groups; at Southwest, on the other hand, unions represent 10 different employee groups. Different unions may represent a given employee craft or class at different airlines. For example, the Air Line Pilots Association (ALPA) represents pilots at United, but the Allied Pilots Association represents American pilots. Table 1 summarizes the representation of different crafts or classes at the major airlines. Table 1: Unions Representing Selected Crafts or Classes at Major Passenger Airlines as of February 1, 2003 Flight Mechanics Fleet Airline Pilots attendants and related Dispatchers service/ramp Alaska ALPA AFA AMFA TWU IAM America West ALPA AFA IBT TWU TWU American APA APFA TWU TWU TWU Continental ALPA IAM IBT TWU (none) Delta ALPA (none) (none) PAFCA (none) Northwest ALPA IBT AMFA TWU IAM Southwest SWAPA TWU AMFA SAEA TWU United ALPA AFA IAM PAFCA IAM US Airways ALPA AFA IAM TWU IAM Legend AFA = Association of Flight Attendants ALPA = Air Line Pilots Association AMFA = Aircraft Mechanics Fraternal Association APA = Allied Pilots Association APFA = Association of Professional Flight Attendants 4 NMB defines a craft or class as a group of employees seeking representation grouped around factors such as their functions, duties, responsibilities, and the general nature of the work performed. Page 5 GAO-03-652 Airline Labor Relations IAM = International Association of Machinists and Aerospace Workers IBT = International Brotherhood of Teamsters PAFCA = Professional Airline Flight Control Association SAEA = Southwest Airlines Employee Association SWAPA = Southwest Airlines Pilot Association TWU = Transport Workers Unions Source: International Association of Machinists and Aerospace Workers. Note: American completed its purchase of Trans World Airlines (TWA) in April 2001, and this table lists union representation as of February 1, 2003; hence TWA is not included in this table. In general, airline labor contracts include three major elements: wages, benefits, and work rules. Work rules generally refer to those sections of a contract that define issues such as hours to be worked and what work is to be done by what employees. Negotiations between airlines and their labor unions on these contracts are conducted in accordance with the requirements of the Railway Labor Act (RLA). This act was passed in 1926 after the railroads and their unions agreed to set in place a legal framework that would avoid disruptions in rail service. The act was amended in 1936, after discussions with airline labor and management, to include the airline industry and its labor unions. See appendix III for a summary of the history and key provisions of the RLA. Airline labor contracts do not expire; rather, they reach an amendable date—the first day that the parties can be required to negotiate the terms of a new contract. Labor negotiations may begin before or after the amendable date, however. While a new contract is being negotiated, the terms of the existing contract remain in effect. Under the RLA, labor negotiations undergo a specific process that must be followed before a union can engage in any kind of work action, including a strike, or before a carrier can change work rules, wages, and benefits.5 After exchanging proposed changes to contract provisions, the airline and the union engage in direct bargaining. If they cannot come to an agreement, the parties must request mediation assistance from the NMB. By statute, if the NMB receives a properly completed application for 5 This limitation on a union’s legal authority to engage in work actions or a carrier’s legal authority to change work rules is known as “maintaining the status quo.” Page 6 GAO-03-652 Airline Labor Relations mediation, it must make its best effort to mediate an amicable settlement. If negotiations are deadlocked after mediation, the NMB must then offer arbitration to both parties. If either party declines arbitration, the NMB releases the parties into a 30-day cooling-off period. While this process is set by law, the decision about when the negotiations are deadlocked is left to the NMB. If the NMB concludes that a labor dispute threatens to interrupt essential transportation service to any part of the country, the act directs the NMB to notify the President of this possibility. The President then can, at his discretion, convene a Presidential Emergency Board (PEB), which issues a nonbinding, fact-finding report.6 If the President does not call a PEB, after the 30-day cooling-off period ends the union is allowed to strike, and the airline is allowed to alter working conditions unilaterally. These actions are known as self-help. If the President does convene a PEB, it is given 30 days to hold hearings and recommend contract terms for a settlement to the parties. The union and the airline then have an additional 30-day cooling-off period, after the PEB makes its recommendations to the President, before either can engage in self-help. After a PEB, Congress may also intervene in the contract dispute by legislating terms of a contract between a carrier and a union. Congress, however, has never intervened in airline negotiations since deregulation. Figure 2 summarizes the key steps in the negotiation process under the RLA. 6 The NMB notifies the President that a potential strike would result in a “possible substantial interference with interstate commerce.” At the President’s discretion, a board can be established to investigate the dispute between the union and the company and make recommendations for settlement. The recommendations are not binding on management or labor. These boards are known as PEBs. Page 7 GAO-03-652 Airline Labor Relations Figure 2: Collective Bargaining Process under the Railway Labor Act Besides negotiations on contracts that are nearing or have passed the amendable dates, airline management and labor may also engage in other negotiations. For example, if an airline introduces a new type of aircraft into its fleet, management and labor will negotiate “side agreements” to Page 8 GAO-03-652 Airline Labor Relations the contract that set pay rates and work rules governing the operation of that aircraft. An example of this situation was when Delta and its pilots settled on pay rates for flying Delta’s newly introduced Boeing 777s in 1999. This agreement was an amendment to a contract that was ratified in 1996. Conversely, during financially difficult times, an airline’s management and labor may negotiate concessionary agreements before contracts reach the amendable date. For example, since 2001, several airlines have requested pay cuts from their unions due to the precarious financial condition of the airlines. In April 2003, American employees agreed to $1.8 billion in wage, benefit, and work rules concessions to help the airline avoid bankruptcy. In April, United employees represented by ALPA, Association of Flight Attendants (AFA), the International Association of Machinists and Aerospace Workers (IAM), the Transport Workers Union (TWU), and the Professional Airline Flight Control Association (PAFCA) agreed to $2.2 billion in average yearly savings to avoid liquidation or having all labor contracts abrogated by the bankruptcy court. Through January 2003, US Airways employees, including unionized, nonunionized, and management personnel, agreed to over $1 billion in cuts to avoid liquidation. In the 25 years since deregulation, airline contract negotiation lengths Length of have increased while the frequency of strikes has declined, but the number Negotiations and of nonstrike work actions have increased. For the 236 contracts that the major passenger airlines negotiated since 1978, available data suggest that Number of Nonstrike the median time taken to negotiate contracts has risen substantially since Work Actions Have 1990, although this varies among the different carriers. In addition, 75 percent of strikes occurred prior to 1990. By comparison, all presidential Increased, While interventions and all identified nonstrike work actions (such as sickouts Number of Strikes or refusals to work overtime) occurred after 1990. Has Declined Page 9 GAO-03-652 Airline Labor Relations Airline Contract The length of time to negotiate airline contracts has increased since Negotiation Lengths Have deregulation. From 1978 to 1989, the median contract negotiation was 9 Increased Since 1978 months while the median negotiation length from 1990 to 2002 increased to 15 months.7 In other words, from 1978 to 1989, half of the contracts were negotiated in more than 9 months while from 1990 to 2002, half of the contracts took more than 15 months to reach an agreement. However, in 1978–1989, 6 contracts were ratified or settled by the amendable date where as from 1990–2002, 9 contracts were ratified or settled by the amendable date.8 (In all, during the two time periods from 1978–1989 and 1990–2002, the number of negotiations that began before the amendable date were 65 and 51, respectively.) Conversely, the number of contracts that required more than 24 months to negotiate more than doubled between the two periods. Figure 3 summarizes changes in the length of time taken for airline labor negotiations from 1978 to 2002.9 7 While we measured negotiation length from the date the carriers reported as the start of negotiations through the ratification/settlement date, the NMB measures negotiation lengths from when it first receives a request for mediation services. While NMB’s measure accurately reflects the period of time they are involved in negotiations, our measure was designed to portray the total period involved in negotiations. 8 See appendix IV for a list of contracts ratified by their amendable date. 9 We were not able to calculate negotiation lengths for all 236 contracts because key dates were unavailable for 89 contracts: 83 had unknown start dates or ratification/settlement dates, and 6 were listed as first-time contracts. Page 10 GAO-03-652 Airline Labor Relations Figure 3: Length of Time Taken to Negotiate Contracts, 1978 to 2002 1978-89 1990-2002 71 contracts 75 contracts Median: 9 months Median: 15 months Negotiated before amendable date 12% 54% Less than 1 year 8% 28% 29% 24% 1-2 years 14% 31% More than 2 years Carriers differed in the degree to which their median negotiation lengths increased—if they increased at all. Negotiation lengths increased at six carriers that were measured, in some cases more than doubling. On the other hand, negotiation lengths decreased or remained constant at three: Continental, United, and Trans World Airlines (TWA). Figure 4 shows the change in median negotiation lengths at the major U.S. passenger airlines before and after 1990. Page 11 GAO-03-652 Airline Labor Relations Figure 4: Median Negotiation Lengths by Carrier Note: America West Airlines was excluded from the measurement because negotiations for this airline were not listed prior to 1990. a Continental had only one contract pre-1990 that had both a known negotiation start date and ratification/settlement date. Contract complexity may play a role in lengthening negotiations. In the 1980s, for example, scope clauses (provisions in labor contracts of the major airlines and their unions that limit the number of routes that can be transferred to smaller, regional jets) could be very short—sometimes only one paragraph. Now, however, such scope clauses can be 60 or more pages. Also, contracts negotiated during the 1980s tended to consist mainly of wages and benefits, while those negotiated in the 1990s included Page 12 GAO-03-652 Airline Labor Relations corporate governance issues such as code sharing,10 regionals,11 and furloughs. Another factor in the length of negotiations is the relationship between labor and management. According to industry experts who examined labor relations in the industry, the quality of labor relationships is defined by the parties’ level of trust, their level of communication, and their ability to problem solve.12 Those carriers that industry officials and labor- management experts13 regard as having positive labor relations tended to have shorter negotiation periods than carriers with acrimonious relationships. Industry officials noted increased tension within labor- management relationships during the 1990s, when the industry recovered from economic hardship to enjoy the biggest boom in its history. An industry official explained that during the recessionary economic period of the early 1990s, unions tended to stall negotiations to avoid making concessions. Conversely, during the peak economic period in the mid to late 1990s, some airlines’ management tried to further improve their profits by prolonging negotiations. Carriers described by industry officials and labor-management experts as having had positive labor relationships include Continental (following 1993) and Southwest. In the 1990s, their median negotiation periods were 7 and 13 months, respectively. Labor-management experts credit Continental’s current CEO for creating relationships of trust, and re- establishing Continental as a profitable carrier after its bankruptcy in the early 1990s. Industry officials also credit Southwest’s labor relationships to 10 The Regional Airline Association defines code sharing as when one airline uses the two- letter designator code of another airline to designate its flights, for example, Comair using Delta’s designator code (DL) to designate one of its flights. 11 The Federal Aviation Administration (FAA) Aerospace Forecasts defines regional airlines as those carriers that provide regularly scheduled passenger service and whose fleets are composed primarily of aircraft having 60 seats or less. 12 Jody Gittell, Andrew von Nordenflycht, and Thomas Kochan, “Mutual Gains or Zero Sum? Labor Relations and Firm Performance in the Airline Industry,” Industrial and Labor Relations Review, (forthcoming), and James Schultz and Marian Schultz, “Northwest Airlines Strike and Labor Negotiations,” American Association of Behavior Social Sciences Journal 2 (1999) 254- . 13 Industry officials represent airline management, airline interest groups and/or industry sponsored research organizations. Labor-management experts include academics who study airline labor relations, authors of studies regarding labor relations in the airline industry and lawyers practicing airline industry labor and employment law. Page 13 GAO-03-652 Airline Labor Relations 30 years of profitability while maintaining its original leadership. Both companies have been recognized for extended periods of low conflict in labor negotiations, underpinned by high-trust workplace cultures.14 Carriers that have been described by labor-management experts as having had contentious relations with their unions include American, Northwest Airlines, TWA, and US Airways. Also, all have a history of strikes and/or court-recognized, nonstrike work actions. Furthermore, in the 1990s, many of these airlines had negotiations that tended to take much longer than Continental’s and Southwest’s. For example, the median length of time to negotiate contracts at US Airways in the 1990s was 34 months. By contrast, the length of time to negotiate contracts at Southwest was 13 months. Strikes Have Decreased The incidence of strikes in the airline industry has decreased over time. Of and Nonstrike Work the 16 strikes that occurred since 1978, 12 occurred prior to 1990, and 4 Actions Have Increased occurred subsequently.15 These strikes ranged from as short as 24 minutes to more than 2 years. Figure 5 summarizes the incidence of strikes, during the 1990s presidential interventions, and court-recognized, nonstrike work actions16 between 1978 and 2002. 14 Jody Gittel, Andrew von Nordenflycht, and Thomas Kochan, “Mutual Gains or Zero Sum? Labor Relations and Firm Performance in the Airline Industry,” Industrial and Labor Relations Review, (forthcoming). 15 See appendix V for a description of airline strikes that have occurred since deregulation. 16 For purposes of this report, we define court-recognized, nonstrike work actions as any labor actions, other than a strike, performed outside of the self-help period and judged necessary by a court of law to warrant a temporary restraining order (TRO) or an injunction. See appendix VI for details on each of these nonstrike work actions. Page 14 GAO-03-652 Airline Labor Relations Figure 5: Frequency of Strikes, Presidential Interventions, and Court-recognized, Nonstrike Work Actions by Year Note: There was one presidential intervention and one court-recognized, nonstrike work action in 2002. Six presidential interventions have been used to prevent strikes since deregulation. All six occurred since 1990.17 Not all presidential interventions were PEBs. In 1993, the President recommended binding interest arbitration for American’s flight attendant negotiation. In 1998, and again in 2001, two PEB warnings occurred; one occurred during 17 In the most recent airline strike—the 89-day strike at Delta Connection carrier Comair in 2001—NMB regarded the impact as not significant enough to warrant a presidential intervention, as they did not believe the strike would substantially threaten to interrupt interstate commerce to such a degree as to deprive any section of the country essential transportation service. (The Comair strike was not included in the list of strikes as it is not a major carrier. See appendix II for additional limitations to the scope of this report.) Page 15 GAO-03-652 Airline Labor Relations Northwest’s pilot strike and the second for American flight attendants.18 Still, PEBs have been used three times in the airline industry since 1978: during a 1994 American pilot negotiation, a 1996 Northwest mechanic negotiation, and a 2000 United mechanic negotiation. Compared to strikes, the pattern for nonstrike work actions has been the opposite: their incidence has increased over time. In all, 10 court- recognized, nonstrike work actions have occurred, each since 1998. Such actions included various forms of slowdowns such as sickouts, work-to- rule, and refusals to work overtime.19 According to a labor-management expert, carriers believe there have been many more nonstrike work actions than the 10 recognized by the courts, but their existence is difficult to prove. Airline management has been unable to produce the evidence needed to prove the actions are taking place.20 Those nonstrike work actions that were not identified by the court 18 The President can take various measures short of a PEB to put pressure on the parties to settle. Such measures include sending a presidential representative to meet with the parties. For example, in the ALPA pilots’ negotiation with Northwest in 1998, the President sent his Senior Counsel and Transportation Secretary to meet with the federal mediator to help the parties resolve their differences. The President can also publicly announce his readiness to empanel a PEB. For example, during the 2001 APFA negotiations with American, the Transportation Secretary announced publicly that the President would use all tools necessary to ensure there was no disruption in service. See appendix VII for a list of presidential interventions. 19 Slowdowns are an organized effort by workers to decrease production in order to pressure the employer to take some desired action. Slowdowns can include refusing to work overtime, sickouts, and work-to-rule. A union official shared that, typically, minor FAA rules that do not concern safety may be overlooked in order to maintain flight schedules. According to two labor-management experts, during a work-to-rule action, airline labor strictly follows such minor rules in order to slow the flight schedule. For example, pilots may refuse to fly a plane if a tray table does not stay in the upright position. 20 According to an International Brotherhood of Teamsters (IBT) union official, unions do not participate in nonstrike work actions, although some might admit to performing work- to-rule actions to put pressure on the carriers. The union official also stated that individual employees have taken actions into their own hands. For example, IBT reported that individual employees promoted a Northwest flight attendant sickout in 2000 by using the Internet. After IBT leadership accessed the Web site, they told their members and the airlines that the union was not in favor of the members’ actions. In separate actions, the Seventh and Eleventh Circuit Courts of Appeal, ruling on behalf of United and Delta, respectively, declared that unions are responsible for controlling labor actions. Specifically, the Eleventh Circuit Court of Appeals held that, when Delta’s pilots engaged in concerted activity in violation of the RLA, “ALPA…ha[d] a duty to end…unlawful action” by its members. According to the court, that duty is not met by mere statements of policy and exhortations to refrain from unlawful activity, but must be backed with action, including union-imposed sanctions (Airline Management Publication). Page 16 GAO-03-652 Airline Labor Relations include a number of highly publicized labor disruptions. For example, the reported, but unconfirmed, nonstrike work action taken by United’s pilots in the summer of 2000 was widely publicized by the media,21 yet the airline never brought the issue before a court of law. Additionally, it has been reported that the reason why these actions are difficult to detect is because a concern for safety often masks the source of such actions. Airline labor strikes have exerted adverse impacts on communities, but we Airline Strikes identified no published studies that systematically and comprehensively Adversely Affect analyzed a strike’s net impact at the community level. For some strikes, we were able to identify evidence of individual impacts, such as reduced air Communities, but service to and from the community, lost salaries or wages by striking or Impacts Have Not laid-off airline workers, or lower airport revenues. However, no studies have yet synthesized such information for a thorough picture of a strike’s Been Fully Analyzed impact on a community. Our analysis indicates that a strike’s potential and Vary from Place impacts would likely vary greatly from community to community, because to Place of differences in factors such as the amount of service available from other airlines. Thus, even if the impact of a strike were to be thoroughly studied at a particular community, it would be difficult to generalize these results to other locations. Airline Strikes Have Had With the reduction of air service stemming from an airline strike, Negative Economic communities have experienced economic disruptions from a number of Impacts on Communities sources. Lost income of airline employees, fewer travelers and less spending in travel related businesses, and less spending by the airline are just some of the ways that local economies have been affected by a strike. For example, canceled flights have lead to the layoff of nonstriking employees, fewer travelers in the airport spending money in concessions, and reduced landing fees for airports. Because passenger traffic dropped, spending at hotels suffered. Local reports illustrated some of a strike’s economic impacts on a community during the 2001 Comair pilot strike. Comair, a regional carrier for Delta, has its main hub at the Cincinnati/Northern Kentucky 21 See, for example, “United Pilots’ Slow Taxiing Causes Delays at O’Hare,” Chicago O’Hare Air Traffic Control, TheTracon.com, July 26, 2000, http://www.thetracon.com/news/times072600.htm, or “United Airlines Scraps Nearly 2,000 Flights,” CNN.com, August 8, 2000, http://www.cnn.com/2000/TRAVEL/NEWS/08/08/united.cancellations.ap. Page 17 GAO-03-652 Airline Labor Relations International Airport. Over the course of the strike, which lasted 89 days, Comair did not operate its 815 daily flights, causing the 25,000 passengers who would normally have been on those flights in an average day to curtail their travel or make arrangements on other airlines. The airline’s 1,350 striking pilots, many of whom are based in the area, lost an estimated $14 million in salaries, and the airline reported laying off an additional 1,600 nonstriking employees in the greater Cincinnati area as well. A concourse at the Cincinnati/Northern Kentucky International Airport closed during the strike. Reports stated that the concourse’s 16 stores and restaurants lost more than $3 million in sales, and that 152 of 193 workers were laid off. The airport also lost $1.2 million in landing fees from Comair during the strike. Impacts can be felt not only at hub communities like Cincinnati, but also at smaller spoke communities that may be served only by the striking airline. When Northwest Airlines pilots struck in 1998, for example, Mesaba Airlines, a regional affiliate, suspended operations as well. At least 12 of the communities served by Mesaba during the Northwest strike had no other air service. One of these locations was Houghton, Michigan. According to local reports, travelers to and from Houghton had to drive as far as Green Bay (213 miles from Hancock, Michigan, location of Houghton’s airport) or Wausau, Wisconsin, (192 miles away) to find alternative flights. DOT also recognized the possible impacts of halting all airline service. The department ordered Mesaba to return service to 12 communities served from Minneapolis under the terms of Mesaba’s Essential Air Service contract. However, before the order was implemented, the strike ended, and service was restored to these communities. Full Impacts at the While the available information indicates that airline strikes can and do Community Level Are have adverse impacts on communities, we identified no published studies Largely Unknown that attempt to comprehensively measure these impacts at the community level. The kinds of impacts cited above, for example, may have mitigating factors that need to be taken into account. In the Comair strike, for example, union strike funds replaced some of the lost income of strikers. ALPA approved payments of $1,400 per month to striking Comair pilots during the strike period, allowing them to spend at a reduced rate in the community. A study that reliably estimated the impact of a strike at the community level would need to take factors such as these into account. No such study has been done. Page 18 GAO-03-652 Airline Labor Relations Another reason for uncertainty about the full impacts of a strike on a community is that the impact of a strike on passengers’ travel decisions is often unknown. For example, while more than 100 communities lost Comair service to and from Cincinnati during the strike, all of these communities had service to Cincinnati from another airline. Thus, although hotel occupancy reportedly fell by more than 18 percent in Northern Kentucky in the strike’s first month, the degree to which this drop was attributable solely to the strike is unknown. Apart from community-level analysis of strikes, some studies have examined the overall economic impacts of aviation on regions or states. For example, the Campbell-Hill Aviation Group, on behalf of an industry interest group, published a report examining the state-level impact of a potential loss of aviation service, but this study did not evaluate the impact of any particular strikes on local or regional economies.22 For example, the study stated that, in the year ending in March 2002, Delta had 10 percent of the passenger traffic in Texas and projected that a 10 percent reduction in aviation benefits would cause a daily reduction of $17.7 million in one measure of the Texas economy, its gross domestic product (GDP).23 DOT also has on occasion produced wide-ranging assessments of the impacts of potential airline strikes, but these studies have never addressed the impacts of strikes that actually occurred. These studies are conducted at the request of the NMB, which uses them in evaluating whether the labor dispute threatens to interrupt essential transportation services in any part of the country. Once the NMB makes this assessment, it notifies the President, who may, at his discretion, empanel a PEB. If the NMB believes an airline strike is probable, it may request the department to examine the possible economic consequences of that strike. The department reports the extent of potentially lost air service to hub and spoke cities of the affected carrier, the number of passengers that would have no service if a 22 We found studies published by Wilbur Smith Associates, Wisconsin Department of Transportation, and the University of Cincinnati that reported on the total economic impact of aviation on the United States, the state of Wisconsin, and the greater Cincinnati/Northern Kentucky region. None examined any strike impacts. 23 In the above example, one should not equate a strike against Delta with a 10 percent reduction in Texas’ aviation benefits. Aviation benefits stem from other aviation related activity such as general and military aviation as well as scheduled airline service. Also, one would have to assume that all airline flights in Texas were completely filled with paying passengers so none of Delta’s 10 percent of Texas fliers could be accommodated by other airlines. Page 19 GAO-03-652 Airline Labor Relations strike were to occur, possible financial impacts on the carrier, indirect impacts on the national economy, and the mitigating and aggravating factors on the impacts of a strike. While DOT’s reviews may examine many areas that could be affected by a strike, they examine only potential strikes and are not conducted after actual strikes. Community-Level Impact While comprehensive studies of community-level impacts of past strikes of Any Future Strike are not available, one thing that emerges from our analysis is that any Would Depend Partly on future strike’s impact on a given community is likely to be affected by the level of service available from other airlines. If alternative service is Service Available from greatly limited, travelers may have to take alternative—and less direct— Other Airlines routes offered by other airlines, or, in extreme cases, travel great distances to other airports in order to fly at all. Those impacts on travelers and businesses will vary depending on whether the community is a hub or spoke destination and even among an airline’s hubs and spoke destinations. The impact of a future strike at an airline’s hub locations would depend in part on which airline is involved in the strike and its market share at the hub. Some airlines dominate air traffic at their hubs to a much greater extent than other airlines do, and a strike involving an airline with a dominant position at most of its hubs would likely have more impact than a strike involving an airline that is hubbing out of locations where competition is greater. In 2001, the airlines with the most and least dominated hubs (based on the percentage of total available seats controlled by the hubbing airline) were US Airways and America West. (See fig. 6.) US Airways averaged 81 percent of the seats offered at its hubs, while America West averaged 32 percent. Thus, based on the loss of seating capacity at its hubs, a strike at US Airways that halted service would likely have substantially more impact on its hub communities than a strike at America West that halted service. Page 20 GAO-03-652 Airline Labor Relations Figure 6: 2001 Market Shares for Major Airlines at Their Hubs Among a single airline’s hub cities, the impact of a strike would also likely vary depending on service available from alternate carriers at those cities. Again, the impact of a strike at the hubbing carrier or its regional partners would be more substantial at more highly dominated hubs. For example, in 2001, Delta and its regional partners accounted for 91 percent of the seats available in Cincinnati, but only 19 percent of available seats at the Dallas/Fort Worth International Airport, which has the lowest market share among Delta’s hubs. Consequently, a strike against Delta would likely have caused much greater disruption in Cincinnati than in Dallas. In contrast to the differences among Delta’s hubs, the impact of a strike at Northwest would likely be felt equally at its Minneapolis/St. Paul, Detroit, and Memphis hubs. At each of its hubs, Northwest offered between 77 and 80 percent of available seats. As at hubs, the impacts of strikes on available air service at spoke cities would also depend on the amount and type of available alternative service. Those communities with air service from other carriers have a greater opportunity to mitigate the potential impact of a strike by enabling travelers to access the national air system using competing airlines. For Page 21 GAO-03-652 Airline Labor Relations example, figure 7 shows available air service, as of April 2003, at spoke communities served by Delta’s regional partner, Comair, from Cincinnati, and by Northwest’s regional carrier, Mesaba, from Minneapolis-St. Paul. Comair provided nonstop service to a total of 101 U.S. communities from Cincinnati. All but one of these communities had alternative service to Cincinnati from another airline—64 with nonstop service, 36 with one-stop service.24 Thus, if Comair’s operations were to be disrupted by a strike, passengers at these communities would still have the opportunity for service to and from Cincinnati. The picture at Minneapolis-St. Paul is somewhat different. There, 10 of the 47 spoke cities served by Mesaba would have no alternative service to Minneapolis-St. Paul. 24 The one exception (Melbourne, Florida) also had one-stop service to Cincinnati, but not from a competing airline. Melbourne passengers could still have one-stop service to Cincinnati from a combination of another Delta regional carrier (to Atlanta) and then a Delta mainline flight to Cincinnati. Page 22 GAO-03-652 Airline Labor Relations Figure 7: Spoke Communities Served from Cincinnati and Minneapolis-St. Paul Retain Service from Competing Airlines Note: Data are from airline schedules for the week of April 21-25, 2003. Other Factors Also Several other factors could also influence the impact of a future strike on a Influence the Total Impact community. The length of the strike is one such factor; longer strikes are of Airline Strikes more likely to have an adverse impact. Since deregulation, strikes have varied from 24 minutes for an American pilot strike in 1997 to almost 2 years for a Continental mechanics strike (1983–1985). Another likely factor is financial preparation; as already mentioned, the local impact of the Comair strike was likely mitigated somewhat by the union’s payments to striking pilots. Similarly, the ability of airlines to operate through a strike—whether by hiring replacement workers or having union members cross picket lines—could also influence a strike’s impact. For example, during a strike by Continental mechanics lasting almost 2 years, some Continental workers crossed the picket line and continued working. This Page 23 GAO-03-652 Airline Labor Relations allowed Continental to continue operation after a shutdown of only 3 days. Tactics used by the striking union can also reduce the overall impact. Alaska flight attendants used a technique called “CHAOS” (Creating Havoc Around Our System) that involved intermittent walkouts of certain crews on certain days.25 This tactic kept certain flights from operating, but did not shut down the entire airline. Our analysis indicates that passenger service has been affected more Nonstrike Work adversely by nonstrike work actions than by an increase in the length of Actions Have Greater negotiations. Generally, but not always, as negotiation periods increased, there has been a slight decline in on-time flights.26 However, the impact of Impacts on these negotiations has been unclear, because the decline may also have Passengers than been affected by other factors such as poor weather, aircraft maintenance, runway closures, air traffic control system decisions, or equipment Lengthy Negotiations failures. By comparison, the 10 court-recognized, nonstrike work actions more clearly resulted in negative impacts on passengers, as shown through such measures as a decrease in the number of on-time flights, an increase in the number of flight problem complaints,27 and a decrease in passenger traffic. Impact of Negotiation Our analyses found a slight correlation between the length of negotiations Lengths on Passengers Is and adverse impacts on passengers.28 We analyzed 23 negotiations between Unclear airlines and pilot unions from 1987 to 2002.29 As negotiations lengthened, the frequency of on-time arrivals declined slightly. However, it is not clear 25 CHAOS, as practiced by the Alaska flight attendants, was found by the federal courts to be a legal form of self-help and not an illegal work action. The Alaska flight attendants did not engage in this activity until after release by the NMB and the 30-day cooling-off period. 26 DOT defines an on-time flight as one that is less than 15 minutes after the scheduled gate arrival or gate departure time. 27 Flight problem complaints include complaints regarding cancellations, delays, or any other deviations from the schedule, whether planned or unplanned. 28 As a measure of adverse impact on passengers, we analyzed the number of on-time arrivals in relationship to the length of contract negotiations. The correlation between these two items was -.25. 29 These 23 pilot contracts were chosen because they had the most complete information, including amendable dates and ratification dates. Concessionary agreements ratified before the amendable date, first time pilot contracts, or pilot contracts with missing information were not used in this analysis. Page 24 GAO-03-652 Airline Labor Relations if the change in on-time flights is attributable solely to negotiation lengths, or if other factors may also have contributed to the on-time performance. DOT’s data on flight arrival and departure timeliness indicate whether a flight is delayed, but not what caused the delay. Common factors for delays include severe weather, aircraft maintenance, runway closures, customer service issues (e.g., baggage and accommodating passengers with special needs, such as those in wheelchairs or youths requiring escorts), air traffic control system decisions, and equipment failures. Thus, despite the apparent relation between lengthening negotiations and a deterioration of service quality, other exogenous factors may explain the change in flight delays. Nonstrike Work Actions Available data indicates that nonstrike work actions have had adverse Have Clearer Adverse impacts on passengers. While DOT data do not specifically identify these Impacts on Passengers actions as the causes for the delays or the reasons for the complaints, increases in the number of late flights, passenger complaints, and decreases in passenger traffic during the period of the actions suggest a clearer relationship than is apparent with these same measures and lengthy negotiations. The periods in which nonstrike work actions occur show decreases in on-time flights, increases in passenger complaints, and decreases in passenger traffic. Two examples of such actions, the American pilot sickout and the Delta pilot slowdown, are described in the next two sections. American Pilot Sickout American experienced decreases in on-time flights, increases in customer complaints, and drops in passenger traffic during a pilot sickout. (Under FAA regulations, any airline pilot can take himself out of the cockpit if he is sick, overly stressed, or does not feel “fit to fly.” During a sickout, pilots utilize these regulations to excuse themselves from work in order to put economic pressure on the airline during the negotiation.) In December 1998, AMR Corp, the parent company of American, purchased Reno Air, whose pilots were then to be integrated into a single workforce. In early 1999, American pilots began a sickout over a dispute involving a side agreement that would integrate Reno Air operations. On February 10, 1999, a federal judge ordered the pilots to return to work. Subsequently, the number of flights cancelled increased. On February 13, 1999, the judge found the pilots’ union in contempt of court.30 By February 16, the airline 30 The judge fined the American pilots’ union, the Allied Pilots Association, $45.5 million for contempt. Page 25 GAO-03-652 Airline Labor Relations reported a return to its normal schedule but, reportedly, pilots were still refusing to work overtime and were adhering to work-to-rule practices, meaning that they would follow every regulation stipulated by the FAA in order to slow the airline. Figure 8 illustrates on-time arrival and departure rates at Dallas/Fort Worth International Airport for the period of August 1998 through December 1999 for American and Delta, which also operates a hub at that airport. The on-time flight statistics for the two airlines are relatively equal prior to the sickout period. During the next several months, American’s on-time record fell below that of Delta. Both carrier’s on-time rates declined somewhat, suggesting that other factors such as weather might also influence flight operations. However, the difference between the two airlines during this period is greater than in other periods. In August 1999, when Reno Air’s operations were officially integrated—even though no agreement was made—the two airlines’ records resumed a more closely parallel path. Figure 8: On-Time Flight Statistics for American and Delta at Dallas/Fort Worth International Airport, August 1998 to December 1999 Page 26 GAO-03-652 Airline Labor Relations The American sickout also caused increases in passenger flight problem complaints. Figure 9 compares the change in complaints against American and Delta. The complaints began to rise in February of 1999 and, generally, continued to increase into the summer, when American reached an agreement with its pilots. Figure 9: Flight Problem Complaints for American and Delta from August 1998 to December 1999 A comparison of passenger traffic between American and Delta at Dallas/Fort Worth International Airport indicates that passenger traffic declined either to avoid the carrier experiencing the nonstrike work action or due to grounded flights. (American grounded up to 2,250 flights per day during the sickout period.) (See fig. 10.) During the American pilot sickout in February 1999, there was a drop in American’s passenger traffic. Compared to the year before, American’s passenger traffic declined by 15 percent while Delta’s passenger traffic rose by 5 percent. Page 27 GAO-03-652 Airline Labor Relations Figure 10: Passengers Carried on American and Delta at Dallas/Fort Worth International Airport, February 1998 and 1999 Delta Pilot Slowdown Another example of the impact of nonstrike work actions on passengers is the Delta slowdown in 2000–2001. In September 1999, Delta began negotiations with its pilots and submitted a contract proposal, which sought to tie future raises to the company’s financial performance. As a result, Delta pilots began refusing to fly overtime in the winter of 2000. When compared to Continental’s operations at Atlanta Hartsfield International Airport,31 Delta experienced substantial declines in on-time flights and increases in flight problem complaints while also experiencing declines in passenger traffic. Delta first went to court on December 5, 2000, and was denied an injunction. The airline then took the suit to the Eleventh Circuit on January 18, 2001, and the denial was overturned and remanded for injunction. 31 AirTran operates more flights than Continental at Atlanta Hartsfield International Airport. However, because DOT does not classify AirTran as a major airline, we compared Delta’s operations with Continental. Page 28 GAO-03-652 Airline Labor Relations Figure 11 shows the percent of on-time flights for both Delta and Continental at Atlanta’s Hartsfield International Airport for the period of August 2000 to August 2001. During the slowdown period from December to January, there is a decline in Delta’s on-time flights relative to Continental’s. Once the court issued an injunction against the union, the two airlines resumed a more similar pattern. Figure 11: On-Time Flight Statistics for Delta and Continental at Atlanta Hartsfield International Airport, August 2000 to August 2001 Delta’s pilot slowdown also showed an increase in passenger complaints during this period. Figure 12 compares the change in passenger flight problem complaints about Delta and Continental during Delta’s slowdown. Flight complaints rose sharply in December and January, peaking at 185 in January 2001, and immediately declining after the union was enjoined on January 18, 2001. Page 29 GAO-03-652 Airline Labor Relations Figure 12: Flight Problem Complaints for Delta and Continental, August 2000 to August 2001 Finally, Delta’s passenger traffic at Atlanta Hartsfield International Airport also declined during the slowdown, but the pattern was less pronounced than for the American sickout discussed earlier. (See fig. 13.) In December 2000, when Delta first pursued an injunction in court, Delta’s and Continental’s passenger traffic dropped by 9 and 4 percent, respectively. Unlike the American sickout (when up to 2,250 flights were grounded per day), Delta pilots’ refusal to fly overtime grounded far fewer flights—about 100 to 125 per day—which means less passengers were affected by cancelled flights as compared to American. Page 30 GAO-03-652 Airline Labor Relations Figure 13: Change in Passenger Traffic on Delta and Continental at Atlanta Hartsfield International Airport in December 1999 and 1 Year Later during the Nonstrike Work Action in December 2000 We provided copies of a draft of this report to NMB for review and Agency Comments comment. NMB indicated it generally agreed with the accuracy of our report, and it provided technical clarifications, which were incorporated into the report as appropriate. The NMB also provided an additional statement, which is included in appendix VIII. We also provided selected portions of a draft of this report to the major airlines and unions to verify the presentation of factual material. We incorporated their technical clarifications as appropriate. As agreed with your office, unless you publicly announce the contents of this report earlier, we plan no further distribution until 30 days from the report date. At that time, we will provide copies to the Honorable Francis J. Duggan, Chairman of the National Mediation Board; the Honorable Norman Y. Mineta, Secretary of Transportation; and other interested parties. We also will make copies available to others upon request. In Page 31 GAO-03-652 Airline Labor Relations addition, the report will be available at no charge on the GAO Web site at http://www.gao.gov. If you or your staff have any questions about this report, please call me at (202) 512-2834, HeckerJ@gao.gov or Steve Martin at (202) 512-2834, MartinS@gao.gov. Appendix VIII lists key contacts and key contributors to this report. Sincerely yours, JayEtta Z. Hecker Director, Physical Infrastructure Issues Page 32 GAO-03-652 Airline Labor Relations Appendix I: Additional Questions Appendix I: Additional Questions In addition to the three primary questions, you asked us how many states use a system of binding arbitration and last offer arbitration with their essential service personnel. You also asked how many times in the last 25 years has Congress had to intervene in a dispute with railroads and what were the outcomes. As of November 2002, according to information from officials of Harvard University, 23 states—including the District of Columbia—use binding arbitration and/or last offer arbitration as arbitration options. (See table 2.) Of those, none use last offer arbitration as their sole arbitration option. Table 2: States That Include Binding Arbitration or Last, Best Offer Arbitration as a Dispute Resolution Option State Type of arbitration included in resolution options 1 California Binding arbitration 2 Maine Binding arbitration 3 Illinois Last offer arbitration 4 Michigan Last offer arbitration 5 Oklahoma Last offer arbitration 6 Tennessee Last offer arbitration 7 Colorado Binding arbitration and last offer arbitration 8 Conneticut Binding arbitration and last offer arbitration 9 Delaware Binding arbitration and last offer arbitration 10 District of Columbia Binding arbitration and last offer arbitration 11 Hawaii Binding arbitration and last offer arbitration 12 Iowa Binding arbitration and last offer arbitration 13 Maryland Binding arbitration and last offer arbitration 14 Minnesota Binding arbitration and last offer arbitration 15 Montana Binding arbitration and last offer arbitration 16 Nevada Binding arbitration and last offer arbitration 17 Ohio Binding arbitration and last offer arbitration 18 Oregon Binding arbitration and last offer arbitration 19 Pennsylvania Binding arbitration and last offer arbitration 20 Rhode Island Binding arbitration and last offer arbitration 21 Texas Binding arbitration and last offer arbitration 22 Washington Binding arbitration and last offer arbitration 23 Wisconsin Binding arbitration and last offer arbitration Source: Harvard University. According to information from the National Mediation Board, in the last 25 years Congress intervened in railroad negotiations eight times. These Page 33 GAO-03-652 Airline Labor Relations Appendix I: Additional Questions interventions occurred between 1982 and 1992. (See table 3.) Congressional interventions do not involve the airlines. Table 3: Congressional Interventions in Railroad Negotiations Date Remark 1 6/26/92 Binding arbitration imposed by Congress 2 6/26/92 Binding arbitration imposed by Congress; parties reached voluntary agreement 3 6/26/92 Binding arbitration imposed by Congress in three cases; parties reached voluntary agreement in all others 4 4/18/91 Terms imposed by Congress 5 8/4/88 Status quo extended by Congress 9/9/88 Terms imposed by Congress 6 1/28/87 Status quo extended by Congress; parties reached voluntary agreement 7 8/21/86 Status quo extended by Congress 9/30/86 Terms imposed by Congress 8 9/22/82 Terms imposed by Congress Source: NMB. Page 34 GAO-03-652 Airline Labor Relations Appendix II: Objectives, Scope, and Appendix II: Objectives, Scope, and Methodology Methodology This report examines the following three questions: • What have been the major trends of labor negotiations in the airline industry since the industry was deregulated in 1978, including the number and length of negotiations and the number of strikes, presidential interventions to avoid or end strikes, and nonstrike work actions? • What has been the impact of airline strikes on communities? • What have been the impacts of the length of negotiations and the occurrence of nonstrike work actions on passengers? To determine the trends of airline labor negotiations, including the length of negotiations, the number of strikes, the number of presidential interventions, and the number of nonstrike work actions, we analyzed data from multiple sources. We obtained our data from major U.S. airlines and various labor organizations. The labor groups included the Air Line Pilots Association (ALPA), the Coalition of Airline Pilots Associations (CAPA), the Association of Flight Attendants (AFA), the International Association of Machinists and Aerospace Workers (IAM), and the International Brotherhood of Teamsters (IBT). We also received substantial negotiation and contract data from the U.S. National Mediation Board (NMB) and the Airline Industrial Relations Conference (AIRCon), a group funded by major U.S. airlines to facilitate the exchange of contract negotiation information and other labor relations matters among carriers. Because data were not available for commuter (regional) and all-cargo carriers, we originally limited our analysis to passenger airlines that are considered majors by the U.S. Department of Transportation (DOT) that were in operation during 2001. These airlines were Alaska, America West, American, American Eagle, American Trans Air (recently renamed as ATA Airlines), Continental, Delta, Northwest, Southwest, TWA1, United, and US Airways. We later were not able to include American Eagle or American Trans Air, which met the DOT criteria, in our analysis because we were not able to obtain information on these airlines. Dates listed as negotiation start dates differ between the airlines, AIRCon, and NMB, therefore, limiting the accuracy of the data collected. A negotiation’s “start date” can be when the carrier’s management or union 1 American completed its purchase of TWA in April 2001, and TWA no longer exists as a separate entity. Analysis of activity from deregulation through April 2001 is included in this report. Page 35 GAO-03-652 Airline Labor Relations Appendix II: Objectives, Scope, and Methodology exchange a written notice stating that one of the parties desires a change in rates of pay, work rules, or working conditions or when face-to-face negotiations actually begin (i.e., when the two parties sit at a table and verbally negotiate the contract). By contrast, the NMB defines a “start date” only when it is called for mediation. For the purposes of our data collection, we first used dates provided by the airlines to AIRCon at the time the contract was being negotiated. If those were not available, we turned to the dates provided directly to us by the airlines from their files when available. We were supplied different dates, including ratification dates and settlement dates, for the end point of negotiations. We know of at least one union that did not have its members vote to ratify contract changes until after 1982. Again, we first used AIRCon provided ratification or settlement dates, if possible, and, in cases where these were not available, we used airline provided dates, or dates provided by NMB. We were unable to calculate a negotiation length for 83 of the 236 contracts because we could not identify either a start date or a ratification or settlement date for them. In addition, we did not calculate negotiation lengths for 6 initial contracts, the first contract a union signs after a craft or class becomes recognized at an airline. To obtain information on nonstrike work actions, we also examined media sources and also reviewed federal court records. Based on the information we were able to review, we defined court-recognized, nonstrike work actions as those work actions for which airlines obtained either temporary restraining orders or injunctions against unions. Officials from the airlines we spoke with stated that there have been many more nonstrike work actions than the 10 judged by the courts. Even some union officials stated that union members have taken actions that they considered legal under their contract or Federal Aviation Administration (FAA) regulations. These same actions, on other occasions, have been found to be violations of the status quo by the courts. Additional cases of nonstrike work actions, however, have been difficult to prove. Airline management has either been unable to produce the needed evidence in court or airlines never took unions to court. Union officials also strenuously deny illegal activity on the part of their unions. We interviewed officials from airlines, labor unions, the NMB, and industry groups. The airlines we spoke with included American, American Trans Air, Continental, Delta, Northwest, Southwest, Comair, Atlantic Coast Airlines, Federal Express, United Parcel Service, and Airborne Express. We only analyzed data from airlines where we could obtain full data. The labor groups we interviewed included ALPA, CAPA, AFA, IAM, and IBT. We also held discussions with officials from NMB, the Air Transport Page 36 GAO-03-652 Airline Labor Relations Appendix II: Objectives, Scope, and Methodology Association (ATA), Communities for Economic Strength Through Aviation (CESTA), and AIRCon. To determine the impact of airline strikes on communities, we searched for studies of these impacts from airlines, industry groups, and academic institutions. Specifically, we talked with United, Delta, Comair, ATA, and CESTA. Based on suggestions from airlines, unions, interest groups, and our own research we also talked with faculty at Harvard, the Massachusetts Institute of Technology, the University of Cincinnati, and the University of Kentucky. None of these sources knew of any published studies on specific impacts of past strikes on any community. In discussions with NMB, we learned that DOT produces studies, solely at the request of NMB, on the likely impacts of probable airline strikes on the airline and local and national economies. We obtained a copy of one of these studies from DOT. We also analyzed data on airline schedules and market share from Sabre, Inc.; BACK Aviation Solutions; and the Campbell-Hill Aviation Group. We also reviewed local media reports from communities affected by strikes. Due to the lack of published studies or generally accepted methodology to determine the impact of strikes, we cannot discount other possible causes for these impacts. To determine the impact of the length of negotiations and court- recognized, nonstrike work actions on passengers, we analyzed data on airline operational performance from DOT’s Air Travel Consumer Report and passenger traffic information from BACK Aviation Solutions. To determine the impact of negotiation lengths, we compared on-time performance throughout the course of 23 negotiations between airlines and pilot unions. To determine the impact of nonstrike work actions, we compared airlines’ on-time performance and flight complaints between airlines before, during, and after the 10 court-recognized, nonstrike work actions. We also analyzed changes in passenger traffic among airlines during these actions. Though our analysis included performing a correlation between on-time arrivals and the length of airline labor contract negotiations, we did not perform any multivariate analysis, and thus, cannot rule out possible alternative causes. We conducted our review between August 2002 and May 2003 in accordance with generally acceptable government accounting principles. Page 37 GAO-03-652 Airline Labor Relations Appendix III: Additional Background Appendix III: Additional Background Information on the Railway Labor Act Information on the Railway Labor Act The Railway Labor Act, 45 U.S.C. § 151, et. seq., (RLA) was passed by Congress in May 1926 to improve labor-management relations in the railroad industry. In January 1926, a committee of railway executives and union representatives jointly presented a draft bill to Congress that was universally supported by those in the industry. Congress did not make any changes of substance to the bill, and the RLA was signed by the President on May 20, 1926.1 Congress has not altered the basic structure of the act that labor and management use to resolve what are known as “major disputes,” i.e., disputes over the creation of, or change of, agreements concerning rates of pay, rules, or working conditions. After discussions with airline management and labor the act was applied to air carriers in 1936.2 As a method to keep labor disputes from interrupting commerce, the new law represented a significant departure from past labor practices by requiring both sides to preserve the status quo during collective bargaining and preventing either side from taking unilateral action. When labor and management representatives drafted the legislation, they agreed that both sides of a labor dispute should negotiate the dispute and not make any change in the working conditions in dispute until all issues were worked out under the deliberate process outlined in the act. The RLA is not a detailed statute. The main purposes of the act are Key Provisions of the threefold. First, Congress intended to establish a system that resolves RLA labor disputes without interrupting commerce in the airline and railroad industries. The statute requires both labor and management “to exert every reasonable effort to make and maintain agreements ... and to settle all disputes ....”3 The Supreme Court has described that duty as being the “heart” of the act.4 Second, the act imposes on the parties an obligation to preserve and to maintain unchanged during the collective bargaining process “those actual, objective working conditions and practices, broadly conceived, which were in effect prior to the time the pending dispute arose and which 1 P.L. No. 257, 69th Cong., 1st Sess., 44 Stat. 577 (1926). 2 45 U.S.C. § 181. 3 45 U.S.C. § 152. 4 Chicago & North Western Ry. v. UTU, 402 U.S. 570, 574 (1971). Page 38 GAO-03-652 Airline Labor Relations Appendix III: Additional Background Information on the Railway Labor Act are involved in or related to that dispute.”5 This is generally known as “maintaining the status quo.” Finally, the act requires that: “Representatives, for the purposes of this Act, shall be designated by the respective parties ... without interference, influence, or coercion exercised by either party over the self-organization or designation of representatives by the other.”6 That obligation was strengthened in 1934 so as to prohibit either party from interfering with, influencing, or coercing “the other in its choice of representatives.”7 The collective bargaining process established by the RLA is designed to Collective Bargaining preserve labor relations peace. The carrier is required to maintain the Process under the status quo before, during, and for some time after the period of formal negotiations. The union and the employees have the reciprocal obligation RLA to refrain from engaging in actions that are designed to economically harm the company, such as strikes during the same period. These actions are termed economic self-help in the act. Airline labor and management periodically engage in negotiations to reach a comprehensive collective bargaining agreement that will remain in effect for a defined period, usually 2 or 3 years. The parties are required to submit written notices (“Section 6 notices”) of proposed changes in rates of pay, rules, and working conditions. In some cases, parties may agree that collective bargaining is required to proceed according to a particular time schedule. If those direct discussions do not result in an agreement resolving a dispute, either party or the National Mediation Board (NMB) can initiate mediation. The RLA requires both parties to maintain collectively bargained rates of pay, rules, and working conditions while they negotiate amendments to the agreement. This requirement extends the status quo after an existing agreement becomes amendable if no agreement is reached by that time. If mediation proves unsuccessful, the NMB appeals to the parties to submit the dispute to binding interest arbitration. If that is unsuccessful, the statute provides for a 30-day cooling-off period. There can be no lawful 5 45 U.S.C. §§ 155, 156, and 160. 6 45 U.S.C. § 152. 7 45 U.S.C. § 145. Page 39 GAO-03-652 Airline Labor Relations Appendix III: Additional Background Information on the Railway Labor Act self-help by either side during this period. Even after the termination of the 30-day period, the self-help option is contingent. If a dispute threatens “substantially to interrupt interstate commerce to a degree such as to deprive any section of the country of essential transportation services,” the President, upon notification by the NMB, is empowered to create an emergency board to investigate the dispute and issue a report that is followed by an additional 30-day period for final negotiations. After this process, the parties are left to self-help and further negotiation to reach a settlement. The only alternative is congressional action, which has never been used in an airline labor dispute. Page 40 GAO-03-652 Airline Labor Relations Appendix IV: Contracts Negotiated and Appendix IV: Contracts Negotiated and Ratified or Settled by the Amendable Date Ratified or Settled by the Amendable Date Amendable Ratification or Carrier Union Craft date settlement date 1 Alaska ALPA Pilots 4/30/80 4/15/80 2 Alaska ALPA Pilots 4/30/83 3/2/83 3 Alaska ALPA Pilots 5/1/91 4/29/91 4 Alaska ALPA Pilots 4/30/93 2/16/93 5 Alaska ALPA Pilots 12/1/97 10/15/97 6 American TWU Mechanics 3/1/93 10/7/91 7 Continental TWU Dispatchers 4/1/99 6/1/98 8 Delta PAFCA Flight control 1/1/82 1/22/81 9 Delta PAFCA Flight control 1/1/86 10/25/85 10 Northwest ALPA Pilots 7/1/80 6/28/80 11 Northwest ALPA Pilots 3/1/94 7/6/93 12 Northwest IAM Mechanics 7/1/88 6/6/88 13 TWA ALPA Pilots 9/1/95 10/3/94 14 United ALPA Pilots 11/30/94 7/12/94 15 United IAM Dispatchers 11/30/94 7/12/94 Legend ALPA = Air Line Pilots Association TWU = Transport Workers Union PAFCA = Professional Airline Flight Control Association IAM = International Association of Machinists and Aerospace Workers Sources: National Mediation Board, airlines, and labor unions. Note: At least one union notified us that they did not have their members ratify agreements before 1982. Page 41 GAO-03-652 Airline Labor Relations Appendix V: Airline Strikes That Have Appendix V: Airline Strikes That Have Occurred Since Deregulation Occurred Since Deregulation Duration of Carrier Union Craft or class negotiations Dates of strike Duration of strike 1 Alaska IAM Mechanics 2/17/84-6/3/85 3/4/85 - 5/4/85 2 months 2 American APA Pilots 6/30/94-5/5/97 2/15/97 24 minutes 3 American APFA Flight attendants 11/18/92-10/10/95 11/18/93 - 11/22/93 5 days 4 American TWU Flight instructors Not available 11/4/79 1 day 5 Continental ALPA Pilots Not available 10/1/83 - 10/31/85 2 years 6 Continental IAM Mechanics 1981-1985 8/13/83 - 4/16/85 1 1/2 years 7 Continental IBT Flight engineers Not available 9/23/79 - 10/6/79 13 days 8 Continental UFA Flight attendants Not available 12/5/80 - 12/21/80 16 days 9 Continental UFA Flight attendants Not available 10/1/83 - 4/17/85 1 1/2 years 10 Continental IAM Flight attendants 1985-1989 3/15/89 - 12/15/89 9 months 11 Northwest ALPA Pilots 8/27/96-9/12/98 8/29/98 - 9/12/98 15 days 12 Northwest IAM Mechanics 9/29/81-6/16/82 5/22/82 - 6/25/82 1 month Flight kitchen 13 Southwest IAM Mechanics Not available 1/13/80 - 2/1/80 19 days 14 United ALPA Pilots 1/30/84-6/17/85 5/17/85 - 6/14/85 29 days 15 United IAM Mechanics 10/1/78-5/24/79 3/31/79 - 5/27/79 2 months Ramp and stores Food services Dispatchers Security officers 16 USAir IAM Mechanics 2/14/90-10/13/92 10/5/92 - 10/8/92 3 days Legend IAM = International Association of Machinists and Aerospace Workers APA = Allied Pilots Association APFA = Association of Professional Flight Attendants TWU = Transport Workers Union of America – AFL-CIO IBT = International Brotherhood of Teamsters UFA = Union of Flight Attendants ALPA = Air Line Pilots Association Sources: NMB, airlines, and labor unions. Page 42 GAO-03-652 Airline Labor Relations Appendix VI: Court-recognized, Nonstrike Appendix VI: Court-recognized, Nonstrike Work Actions Since Deregulation Work Actions Since Deregulation Date of court Carrier Union Craft Work action Plaintiff request decision Outcome 1 American APA Pilots Sickout TRO sought 2/10/1999 Awarded 2 American TWU Mechanics Slowdown TRO sought 2001 Awarded 3 American TWU Mechanics Slowdown TRO sought 1998 Awarded 4 American TWU Mechanics Slowdown Injunction sought 1999 Granted 5 Delta ALPA Pilots Refuse overtime Injunction sought 2001 Granted 6 Northwest AMFA Mechanics Refuse overtime Injunction sought 5/11/2001 Granted 7 Northwest IAM Clerical Slowdown Injunction sought 2/25/1999 Granted Flight kitchen Stock 8 Northwest IBT Flight attendants Sickout Injunction sought 1/5/2000 Granted 9 TWA IAM Mechanics Sickout and work TRO sought 1998 Awarded stoppage 10 United IAM Mechanics Slowdown Injunction sought 7/1/2002 Granted Legend APA = Allied Pilots Association TWU = Transport Workers Union of America – AFL-CIO ALPA = Air Line Pilots Association AMFA = Aircraft Mechanics Fraternal Association IAM = International Association of Machinists and Aerospace Workers IBT = International Brotherhood of Teamsters Sources: NMB, airlines, and courts. Page 43 GAO-03-652 Airline Labor Relations Appendix VII: Number of Presidential Appendix VII: Number of Presidential Interventions Since Deregulation Interventions Since Deregulation Presidential Carrier Union Craft Amendable date intervention date Actions taken 1 American APA Pilots 8/31/94 2/15/97 Presidential Emergency Board 2 American APFA Flight attendants 11/1/98 2001 Presidential Emergency Board warning 3 American APFA Flight attendants 12/31/92 1993 President recommends binding interest arbitration 4 Northwest ALPA Pilots 11/2/96 September 1998 Presidential Emergency Board warning 5 Northwest AMFA Mechanics 9/30/96 3/12/01 Presidential Emergency Board 6 United IAM Mechanics 7/12/00 1/19/02 Presidential Emergency Board Legend APA = Allied Pilots Association APFA = Association of Professional Flight Attendants ALPA = Air Line Pilots Association IAM = International Association of Machinists and Aerospace Workers AMFA = Aircraft Mechanics Fraternal Association Sources: NMB and airlines. Page 44 GAO-03-652 Airline Labor Relations Appendix VIII: Comments from the National Mediation Board Appendix VIII: Comments from the National Mediation Board Page 45 GAO-03-652 Airline Labor Relations Appendix VIII: Comments from the National Mediation Board Page 46 GAO-03-652 Airline Labor Relations Appendix VIII: Comments from the National Mediation Board Page 47 GAO-03-652 Airline Labor Relations Appendix VIII: Comments from the National Mediation Board Page 48 GAO-03-652 Airline Labor Relations Appendix VIII: Comments from the National Mediation Board Page 49 GAO-03-652 Airline Labor Relations Appendix VIII: Comments from the National Mediation Board Page 50 GAO-03-652 Airline Labor Relations Appendix IX: GAO Contacts and Staff Acknowledgments Appendix IX: GAO Contacts and Staff Acknowledgments JayEtta Z. Hecker (202) 512-2834 GAO Contacts Steven C. Martin (202) 512-2834 In addition to those individuals named above, Jonathan Bachman, Staff Brandon Haller, David Hooper, Terence Lam, Dawn Locke, Sara Ann Acknowledgments Moessbauer, Stan Stenersen, and Stacey Thompson made key contributions to this report. Page 51 GAO-03-652 Airline Labor Relations (544050) The General Accounting Office, the audit, evaluation and investigative arm of GAO’s Mission Congress, exists to support Congress in meeting its constitutional responsibilities and to help improve the performance and accountability of the federal government for the American people. GAO examines the use of public funds; evaluates federal programs and policies; and provides analyses, recommendations, and other assistance to help Congress make informed oversight, policy, and funding decisions. GAO’s commitment to good government is reflected in its core values of accountability, integrity, and reliability. The fastest and easiest way to obtain copies of GAO documents at no cost is Obtaining Copies of through the Internet. GAO’s Web site (www.gao.gov) contains abstracts and full- GAO Reports and text files of current reports and testimony and an expanding archive of older products. The Web site features a search engine to help you locate documents Testimony using key words and phrases. You can print these documents in their entirety, including charts and other graphics. Each day, GAO issues a list of newly released reports, testimony, and correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site daily. The list contains links to the full-text document files. To have GAO e-mail this list to you every afternoon, go to www.gao.gov and select “Subscribe to daily E-mail alert for newly released products” under the GAO Reports heading. Order by Mail or Phone The first copy of each printed report is free. Additional copies are $2 each. A check or money order should be made out to the Superintendent of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more copies mailed to a single address are discounted 25 percent. Orders should be sent to: U.S. General Accounting Office 441 G Street NW, Room LM Washington, D.C. 20548 To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202) 512-6061 Contact: To Report Fraud, Web site: www.gao.gov/fraudnet/fraudnet.htm Waste, and Abuse in E-mail: email@example.com Federal Programs Automated answering system: (800) 424-5454 or (202) 512-7470 Jeff Nelligan, Managing Director, NelliganJ@gao.gov (202) 512-4800 Public Affairs U.S. General Accounting Office, 441 G Street NW, Room 7149 Washington, D.C. 20548
Airline Labor Relations: Information on Trends and Impact of Labor Actions
Published by the Government Accountability Office on 2003-06-13.
Below is a raw (and likely hideous) rendition of the original report. (PDF)