The D.C. Circuit once observed that “[i]t is a fact of life in NLRB lore that certain substantive provisions of the NLRA invariably fluctuate with the changing compositions of the Board.” In 2000, the Clinton Board found that teaching and research assistants at private universities are “employees” covered by the NLRA; in 2004, the Bush Board found that they were not, and in 2016, once again, the Obama Board found that they were employees. This has led to speculation that a Trump Board will deny employee status to teaching and research assistants. In acknowledging this possibility, I don’t want to suggest that this result would be reasonable – the majority opinion in the Board’s 2016 Columbia University case offers a compelling statutory analysis in support of its conclusion. By contrast, the dissent’s position largely relies on speculation about the effects of collective bargaining on universities, with a particular emphasis on the potential disruption from the use of economic weapons. Oddly, the dissent fails to acknowledge that many of these weapons – strikes, lockouts, loss of academic credit, loss of prepaid tuition – would be available even if the Board denies employee status to teaching and research assistants. In fact, Congress enacted the NLRA in the hope that encouraging collective bargaining would minimize industrial strife and unrest. But, if a Trump Board nevertheless rules that teaching and research assistants are not “employees,” what will happen at Harvard and Columbia, where teaching and research assistants have already voted on unionization? Assuming they vote in favor of unionization, their unions should be safe for at least an initial contract cycle.
The NLRB does not simply issue fiats setting forth policies. Instead, it decides particular cases. In deciding cases, the Board often sets policies that have much broader implications, but even if a majority of Board Members would like to overturn a particular precedent, they must wait until they have a case that raises the issue. You might think that Harvard or Columbia could raise the issue with the Trump Board simply by refusing to bargain with a victorious union. But, when an employer refuses to bargain with a newly certified union, since the earliest days of the NLRA the Board has adhered to a policy of refusing to allow the employer to raise issues that “were or could have been litigated in the underlying representation hearing.” This is true even where the issue raised by the employer is jurisdictional. For instance, the NLRA definition of “employee” excludes individuals employed as supervisors. But, where employers have argued that a bargaining unit improperly includes supervisors, the Board has refused to address those claims in refusal-to-bargain cases following a union election. This has been true even where the Board Members have suggested that they were sympathetic to the employer’s position on the merits. In Evergreen New Hope Health & Rehabilitation Center, a 2002 case, the employer argued that a newly certified unit improperly included statutory supervisors. Board Members Hurtgen and Bartlett both noted in a footnote that they did not necessarily endorse the decision that had been reached in the representation case, but nevertheless the issue raised by the employer was not “properly litigable” in the refusal-to-bargain case.
If a university has voluntarily agreed to recognize and bargain with a union representing teaching and research assistants, the employer may not simply change its mind in the hopes of getting the Board to rule on the question of employee status. The leading case here is a Ninth Circuit case called E.G. & H. v. NLRB. In that case, several employers had reached an oral agreement with the union, but refused to execute the agreements, defending their actions on the grounds that the bargaining units improperly included supervisors, and therefore the NLRB lacked authority to enforce the agreements. The Board argued, and the Ninth Circuit agreed, that once an employer recognizes a union, the employer may not repudiate the bargaining relationship on the ground that the bargaining unit was inappropriate.
If the Board does find a case to rule that teaching and research assistants are not employees, the existing bargaining units would not be safe forever. The employers would be able to withdraw recognition upon the expiration of a collective bargaining agreement. But, it may well be that once the universities see that the parade of horribles they fear does not transpire, they will realize that the advantages of labor peace outweigh the costs of reopening the issue on campus.