Here is the amicus brief filed in Harris v. Quinn on behalf of Respondents by a group of labor law professors (and authored by Charlotte Garden and Matthew Bodie).  The brief makes a number of arguments, among them the following (from the Summary):

I. Unlike in the private sector, state employers choose for themselves whether to permit their employees to bargain collectively, and can also set—within established First Amendment limits— countless other parameters, such as the subjects over which bargaining will occur and the permissibility of strikes. In other words, each state may structure its labor law in the manner that it concludes best promotes sound workforce management. While some states accordingly reject bargaining with their workforces altogether, others reasonably conclude—and indeed, research suggests—that collective bargaining can be an effective way to aggregate, clarify, and channel workers’ preferences, and that collective bargaining can promote the state’s own efficiency interests by promoting workforce engagement and longevity.

States that permit collective bargaining almost universally adopt the exclusive representation model. Under that model, once a union has been democratically selected by a bargaining unit, the state bargains with only that representative. Further, that union represents all the workers within the unit, including those who do not join the union. The alternative—in which a union represents only those workers within a bargaining unit who choose representation—would allow a potential multitude of unions to demand separate negotiations and separate contracts with an employer. States that adopt collective bargaining almost always reject this system, which threatens to raise bargaining costs and increase intra-workforce conflict.

Likewise, states may reasonably conclude that, within a system of exclusive representation, employees should be required to pay their share of the costs of representation. In particular, states may reasonably conclude that allowing bargaining unit members to choose not to pay their share towards the costs of representation would lead to free riding that would undermine the benefits sought from the exclusive representation system.

Illinois’s particular interests in allowing its Medicaid-funded personal assistants to select a union to represent them in bargaining with the state are evident. A union can serve an important quasi- human-resources role in aggregating and communicating information about this geographically dispersed workforce. This information can aid the state in determining how best to attract and retain a qualified and professional workforce, potentially improving service delivery and decreasing program costs. Further, unions can help improve workforce health and safety through cooperative partnerships with public managers, as in this case, where the union has negotiated training and equipment programs.