California Ex Rel. Harris v. Safeway, Inc.

Chief Judge KOZINSKI,

with whom Judges TALLMAN and RAWLINSON join,

dissenting in part.

By going out of its way to rule on the non-statutory labor exemption, the majority decides an important legal question that will have absolutely no effect on anyone involved in this case. We hold that there’s no categorical antitrust violation under the quick look doctrine; nor can such a violation be established on remand, because California stipulated to dismissal if it didn’t prevail under quick look. Since no antitrust violation can ever be established in this case, we have no occasion to decide whether any exemption from antitrust liability would apply. The majority’s groundbreaking ruling on the labor exemption is thus very likely an advisory opinion and beyond the scope of our Article III jurisdiction. See Thomas v. Anchorage Equal Rights Comm’n, 220 F.3d 1134 , 1138 (9th Cir.2000) (en banc). Such dicta are particularly unwise when they are effectively insulated from Supreme Court review, as the grocers probably will have neither incentive nor standing to petition for certiorari.

Worse, I seriously doubt the majority decides the labor exemption issue correctly because it fails to grapple with the complex dynamics of this case. Had it done so, it would have realized that each and every factor the Supreme Court found relevant in Brown v. Pro Football, Inc., 518 U.S. 231 , 116 S.Ct. 2116 , 135 L.Ed.2d 521 (1996), supports finding the revenue-sharing provision (RSP) protected by the labor exemption. Contra maj. op. at 1131-32.

First, the RSP was inextricably intertwined with the collective bargaining process. See Brown, 518 U.S. at 250 , 116 S.Ct. 2116 . Contra maj. op. at 1131-32. The grocers’ agreement was a direct response to the union’s anticipated use of whipsaw tactics. As courts have long recognized, whipsaw tactics are particularly devastating for employers, because

the union strikes against one member of a multiemployer bargaining unit, but allows the other employers to continue operating in order to maximize the competitive pressure brought to bear upon the struck member ...; the idea is thereby to force each employer individually to capitulate through a series of such strikes, thus defeating their attempt to stand together.

Int’l Bhd. of Boilermakers v. NLRB, 858 F.2d 756 , 760 (D.C.Cir.1988).

The grocers here were legitimately concerned that the union would selectively strike and picket only one chain, diverting their customers to the others. The unions would thereby upset the prevailing competitive balance, crippling the target and ruining any chance of bargaining as a group. The grocers sought to blunt the disproportionate losses borne by the targeted chain by redistributing some of the windfall profits reaped by the others as a result of the union’s tactics. The agreement was limited to the duration of the strike plus two weeks and became operative only if, and only to the extent, the union succeeded in redirecting consumers from the targeted store to other stores in the bargaining group. The RSP was thus narrowly tailored to counter the union’s divide-and-conquer strategy. Its effect, moreover, was entirely pro-competitive: It helped keep all competitors in the market rather than letting one be wiped out by the strike.

As it turns out, the grocers were right to be concerned. Less than a week after *1141 their contracts expired, the unions struck, and Vons was the exclusive target. Per the grocers’ Mutual Strike Assurance Agreement, Albertson’s and Ralphs locked out their workers. The unions then picketed all three stores, but soon pulled pickets from Ralphs to zero in on the other chains. The unions’ strategy was highly effective: Vons and Albertson’s lost market share to Ralphs, as customers switched stores to avoid the picket lines. Some of these losses were offset by payments from Ralphs under the RSP, but many — such as the loss of long-time customers — were permanent. The RSP was thus a limited, defensive tool that was directly related to the collective bargaining process.

The second Brown factor- — -that the practice is “unobjectionable as a matter of labor law and policy” — also points in favor of exemption. Broum, 518 U.S. at 238 , 116 S.Ct. 2116 . Contra maj. op. at 1129. Although the NLRB has not had occasion to rule on the specific RSP at issue here, both the Supreme Court and the NLRB have generally sanctioned the use of economic weapons to combat whipsaw tactics. See, e.g., NLRB v. Brown (Brown Food), 380 U.S. 278 , 85 S.Ct. 980 , 13 L.Ed.2d 839 (1965). The Supreme Court has recognized a particularly strong interest in “preserving the integrity of the multiemployer bargaining unit” against targeted attacks. Brown, 518 U.S. at 245 , 116 S.Ct. 2116 (quoting Broum Food, 380 U.S. at 289 , 85 S.Ct. 980 ) (internal quotation mark omitted). Accordingly, the Court has approved many practices employers use to maintain a common front: They may, for example, agree to lock out all workers if any one of them is struck, see NLRB v. Truck Drivers Local Union No. 449, 353 U.S. 87, 89, 97 , 77 S.Ct. 643 , 1 L.Ed.2d 676 (1957), or hire temporary replacement workers, see Brown Food, 380 U.S. at 279-80 , 85 S.Ct. 980 . Revenue-sharing provisions that are designed to maintain cohesiveness in the teeth of union efforts to disrupt the bargaining group fit squarely within this category of accepted labor practices.

On the handful of occasions that courts have evaluated the legitimacy of revenue-sharing provisions, they have been upheld. In Kennedy v. Long Island R.R., 319 F.2d 366 (2d Cir.1963), the Second Circuit upheld the use of a strike insurance plan. Like the grocers here, the railroads were concerned that the union would engage in whipsaw tactics and wanted to spread the losses among employers. Id. at 368-69 . The railroads contributed to an insurance fund that would pay out in the event of a strike. Id. As the proceeds were distributed, the fund was supplemented by further deposits from non-struck railroads. Id. The union sued, claiming the strike insurance fund violated both antitrust law and the railroad’s duty to bargain in good faith. Id. at 368 . The Second Circuit rejected both claims, reasoning that “the strike insurance plan, far from constituting a violation of the railroad’s duty to bargain in good faith, was an instrument of self-help properly employed in the process of collective bargaining.” Id. at 371 .

The D.C. Circuit reached a similar conclusion in Air Line Pilots Ass’n Int’l v. Civil Aeronautics Bd., 502 F.2d 453 (D.C.Cir.1974). There, six airlines entered into a “Mutual Aid Pact” to “soften the impact of strikes against individual companies.” Id. at 455 . The Mutual Aid Pact contained a provision, almost identical to the RSP, under which a “strikebound company received payments from other Pact members equal to their increase in revenues resulting from the strike.” Id. As in Kennedy , the union claimed the revenue-sharing provision violated antitrust law and national labor policy. Id. The D.C. Circuit likewise rejected these claims, explaining that “[t]he national labor policy *1142 rests on the principle that parties should be free to marshall [sic] the economic resources at their disposal in the resolution of a labor dispute.” Id. at 456 . Although both of these cases were decided under the Railway Labor Act, their analysis was based on national labor policy and relied heavily on National Labor Relations Act cases. See, e.g., id. at 456 n. 3 (relying on NLRB v. Ins. Agents’ Int’l Union, 361 U.S. 477, 490 , 80 S.Ct. 419 , 4 L.Ed.2d 454 (1960)); Kennedy, 319 F.2d at 371 n. 4 (relying on Truck Drivers, 353 U.S. 87 , 77 S.Ct. 643 ). The majority dismisses these cases in a footnote, but they fatally undermine the majority’s claim that the precaution adopted by the employers here is too novel and exotic to fall within the labor exemption.

Courts have also approved other strategies that redistribute the financial pain wrought by a strike. When unions engage in selective striking, for example, striking employees are often paid benefits to compensate for lost wages. See Kennedy, 319 F.2d at 372 . Unions do this because it would be unfair to force one set of employees to bear the entire burden when the eventual benefit will inure to everyone. Strike benefits also strengthen the union’s bargaining position by ensuring all employees share the same incentives, and none will be driven by personal hardship to undermine the strike. Strike benefits are thus considered a legitimate economic weapon. See Int’l Bhd. of Boilermakers, 858 F.2d at 767. RSPs serve precisely the same purposes for employers bargaining as a group: Why should one employer alone bear the heavy cost of selective striking or picketing, when the eventual contract will bind the entire group? And why should unions be able to spread the burdens of a strike to reinforce their bargaining position, while employers can’t?

When unions pay benefits to striking workers, their collusive actions are protected by statute. See HA. Artists & Assocs., Inc. v. Actors’ Equity Ass’n, 451 U.S. 704, 713-14 , 101 S.Ct. 2102 , 68 L.Ed.2d 558 (1981). To protect the collective bargaining process, employers too must be allowed to share losses without fear of antitrust liability, which is the very point of the non-statutory labor exemption. See Brown, 518 U.S. at 237 , 116 S.Ct. 2116 .

My colleagues reach the wrong conclusion because they misread Brown . The majority looks for some affirmative approval in labor law for the RSP, maj. op. at 1129, but that is far more than Brown calls for. Brown requires only that the conduct be “unobjectionable as a matter of labor law and policy.” 518 U.S. at 238 , 116 S.Ct. 2116 (emphasis added). The very fact that the union did not challenge the RSP as an unfair labor practice, despite having raised a procedural dispute before the NLRB, itself is proof that the RSP is unobjectionable as a matter of labor policy. The second Brown factor is clearly satisfied.

The third Brown factor similarly favors exemption because the RSP concerned only parties with a direct stake in the outcome of the collective bargaining agreement. See Brown, 518 U.S. at 250 , 116 S.Ct. 2116 . Contra maj. op. at 1131. Vons, Albertson’s and Ralphs were part of the same multi-employer bargaining unit, and Food 4 Less’s own contracts were set to expire just months later. Food 4 Less, standing alone, had no hope of doing better against the union than Vons, Ralphs and Albertson’s had done by working together. In all likelihood, the contract the union wrung from them would set the bar for Food 4 Less’s own negotiations. Food 4 Less thus had a strong labor-related interest in standing shoulder-to-shoulder with the three other chains.

*1143 Equally important, Food 4 Less was required by its existing contract to contribute to employee benefits at a rate tied to that of Ralphs. As Ralphs’s Vice President of Human Resources and Labor Relations explained, “the rate ... for Food4Less employees’ health and welfare [was] directly connected to the contract we ... bargained.” If the unions managed to extract more favorable benefits from Ralphs and the other chains as a result of the strike, Food 4 Less would be forced to provide more benefits for its own workers. It thus had a direct and immediate financial stake in the contract negotiations. The majority’s claim that “the inclusion of [Food 4 Less] suggests that the conduct is not anchored in the collective-bargaining process,” id. at 1131, is belied by the record.

The fourth factor — whether the conduct involved subject matter that the parties were required to negotiate collectively— simply doesn’t apply to this case. See Brown, 518 U.S. at 250 , 116 S.Ct. 2116 . In Brown , the Court had to decide whether the NFL’s unilateral imposition of its last good faith salary offer was exempt from antitrust review. In such a case, it’s relevant whether these substantive terms relate to “wages, hours, and working conditions,” or whether the terms are unrelated to collective bargaining. Id. at 241 , 116 S.Ct. 2116 ; compare id. at 234 , 116 S.Ct. 2116 (agreement to set salaries is exempt), with United Mine Workers v. Pennington, 381 U.S. 657, 663 , 85 S.Ct. 1585 , 14 L.Ed.2d 626 (1965) (agreement to set prices of goods is not exempt). Where the employer action at issue involves a procedural bargaining tactic, rather than a substantive term of the contract, it makes no sense to ask whether the tactic relates to “wages, hours, and working conditions.” It never does. Yet procedural tactics go to the heart of the exemption’s purpose— to free from antitrust scrutiny employer actions that are “needed to make the collective-bargaining process work.” Brown, 518 U.S. at 234 , 116 S.Ct. 2116 (emphasis added); see also id. at 247 , 116 S.Ct. 2116 . A lockout, for example, is a bargaining tactic that has nothing to do with “wages, hours, and working conditions” yet is exempt from antitrust review. See id. at 245 , 116 S.Ct. 2116 ; see also id. at 254 , 116 S.Ct. 2116 (Stevens, J., dissenting); maj. op. at 1130. That the revenue-sharing provision — which is a procedural tactic designed to put pressure on the union during the course of negotiations — doesn’t relate to a mandatory subject of collective bargaining is simply beside the point and cannot count against application of the labor exemption. Contra maj. op. at 1129-30.

Fifth, and finally, the conduct indisputably “took place during and immediately after a collective-bargaining negotiation.” Brown, 518 U.S. at 250 , 116 S.Ct. 2116 . The revenue-sharing provision clearly centered on the time period of the labor dispute, as it lasted only for the duration of the strike plus an additional two weeks. In its haste to condemn the RSP, the majority omits this factor from its discussion.

A fair reading of the RSP can leave no doubt that all the relevant Brown factors weigh heavily in favor of exempting the RSP from antitrust review. We are not dealing with employers who were using a labor dispute as a pretext to engage in price-fixing; it’s perfectly clear that the employers were responding to union tactics in the course of a strike, and only to the degree the tactics were effectively deployed by the union. The majority’s contrary dicta have no basis in the record, common sense or precedent.

Worst of all, we may never be able to correct this error. Strikes are costly en *1144 deavors for everyone involved, and introducing the additional threat of antitrust liability — with its protracted litigation, unpredictable rule of reason analysis and treble damages — will no doubt force employers to think twice before entering into a revenue-sharing agreement in the future. Today’s gratuitous decision thus has the unfortunate consequence of “forcing [employers] to choose their collective-bargaining responses in light of what they predict or fear antitrust courts, not labor law administrators, will eventually decide.” Brown, 518 U.S. at 247 , 116 S.Ct. 2116 . Should this fear prevent employers from entering into an RSP, we will have effectively usurped the role of the NLRB by dictating the tools that can and cannot be used in labor disputes. Because I find this result irreconcilable with Brown , inimical to sound labor policy, completely unnecessary to the resolution of this case and outside the scope of oui* constitutional authority, I dissent from Part III of the majority opinion.