(Note: Part three of six.)

A steady decline in union membership has led union organizers and sympathetic politicians to introduce "labor reform" legislation designed to make it easier for unions to gain representation rights over more workers without becoming more accountable to those workers.

The main labor reform bill before Congress, the "Employee Free Choice Act," has two particularly problematic provisions. The first is "card check" recognition, which as discussed in part two of this series last week, would hurt workers by doing away with the use of secret-ballot elections to determine worker preferences for union representation. The second troubling aspect of the labor reform proposal is its misuse of binding arbitration.

Arbitration is a valuable method for resolving disputes and is frequently used in labor relations to resolve grievances that arise under existing contracts. But the Employee Free Choice Act’s process of using arbitration to create a contract where parties are unable to agree, referred to as "interest arbitration," is a clumsy approach that is seldom seen outside of government.

Currently, negotiations on an initial contract after a union is recognized are treated much the same as any other contract: The parties negotiate in good faith until they settle on terms. If they fail to do so the union may call a strike, and the employer may either implement its last offer or even lock workers out. But EFCA would short-circuit negotiations, allowing either party to call for a government-appointed mediator after 90 days. Either party could unilaterally have the matter submitted to arbitration after 30 days of mediation, and the results of arbitration would be binding on both parties for two years.

In its current form the bill does not specify many details about how arbitration will be carried out, but Michigan has been using arbitration in a similar way for police officers and firefighters since 1969. Michigan’s process is fairly typical among states that use interest arbitration to resolve bargaining impasses with government employees. The results are not encouraging for those who hold binding arbitration out as a solution for labor strife.

When negotiations break down to the point where arbitration is needed, Michigan law calls for a three-member panel to determine wages and other terms of employment. The government employer and union each appoint a panelist, while the third, a neutral arbitrator who serves as chairman, is chosen from a list provided by the state, with union and employer striking off names until only one is left. Because the members appointed by the union and the employer can be counted on to support their own side, the arbitration process ultimately hinges on the opinions of this neutral member.

Based on Michigan statute, the process of arbitration is supposed to go quickly. Assembling the arbitration panel should take less than three weeks. Once the panel is named, the first hearing should be held within 15 days, and hearings are supposed to be wrapped up 30 days after they commence.

In reality, the process is much longer. In the early 1990s, only one out of every six arbitration cases was resolved within 300 days of a petition being filed. Since then the pace of arbitration has improved, but not dramatically. A review of 29 arbitration cases resolved in 2005 and 2006 showed that only seven were resolved within 300 days, fewer than one out of four. On average, arbitration takes almost 15 months from the date that a request is filed to the date that a decision is reached. This delay ties up government resources because arbitrators’ awards are retroactive, meaning that pay raises awarded by arbitration often involve back pay that local officials must prepare for in advance.

This uncertainty about both future pay and back pay is a serious burden for local governments. In the private sector these delays could do even more damage, as uncertainty about future wages and working conditions make it more difficult for companies to recruit personnel or respond to changes in the marketplace.

On the critical economic terms, in particular base wages, Michigan law calls for the arbitration panel to select either the employer’s terms or the union’s terms. The panel is not allowed to put together a compromise package of its own. Not all arbitration laws have this sort of provision, but there are disadvantages either way. Michigan’s approach makes arbitration riskier, with unions and employers entering arbitration knowing it is likely to be an all-or-nothing affair. Allowing arbitrators to write their own economic terms, by contrast, creates the temptation for them to take the safe path and avoid dealing with the merits of the parties’ arguments by splitting differences down the middle. (The union wants a 4 percent raise and the employer wants to give 2 percent? Easy, we’ll just say its 3 percent).

Michigan law lists a number of criteria that the panel is to consider in making a decision, such as the ability of the government employer to pay, comparisons with similar communities, trends in private-sector employment and the local cost of living. But in the end the arbitrator decides what weight to put on any of these factors, with little oversight by state or local government and virtually no risk that his or her ruling will be overturned by the courts. Using interest arbitration on initial contracts might make the process even less predictable, because arbitrators will not have prior collective bargaining agreements to look to for guidance.

Many states rely on interest arbitration to settle public-sector labor disputes, especially those involving public safety employees, but according to Frank Zotto, vice president for case management at the American Arbitration Association, it is "a rare thing for this to happen in the private sector." The few private-sector cases where interest arbitration was used to resolve a bargaining impasse mostly involve airlines, a troubled industry where many major carriers have been through bankruptcy in the last few years, an example that is especially discouraging for business owners.

The proposed law does not specify how arbitration panels will be set up, and federal regulators will have considerable leeway to establish their own procedures. Federal interest arbitration under EFCA is likely to be similar to Michigan’s process, but it could be different in important ways. The federal government might assemble panels where all three participants are neutrals, reducing the importance of any single panelist. It might do a better job of keeping panels on schedule, reducing the wait for a decision and the size of back-pay awards.

But one thing cannot be changed about the arbitration process — binding contracts will not be drawn up by management and labor working out their differences and reaching compromises that they both can live with. Instead, both sides will be bound by the decision of arbitrators who will be unaccountable and insulated from the results of their handiwork. The arbitration process may, or may not, have the confidence of both sides. The decision makers may be seen as fair and judicious, or they may be distrusted by one side — or both. It will not matter: The arbitrators’ decisions will almost always be final. And the arbitrators themselves will not have to live on the low wages that they might force workers to labor under, nor will they suffer layoffs or bankruptcy if their awards prove to be too generous.

This makes interest arbitration something of a gamble, one that unions and private-sector employers seldom agree to take on their own. It would be unwise for Congress to force employers and unions into taking that gamble today.

#####

Paul Kersey is senior labor policy analyst at the Mackinac Center for Public Policy, a research and educational institute headquartered in Midland, Mich. Permission to reprint in whole or in part is hereby granted, provided that the author and the Center are properly cited.