Selecting a sample of unions to study proved to be quite challenging because there exist no aggregate data on transnational labor relationships. Unions do not maintain records of every contact and interaction they have with unions in other countries, and only large events and campaigns appear in union publications and documents. Moreover, because many unions do not have departments dedicated to managing international work, the reporting of that work in formal archived sources is not routinized and tends to be sporadic. Finally, international work is usually conducted by key actors in unions through informal mechanisms – there is little institutional memory regarding how relations emerged and developed. Determining the universe of unions involved in transnational relationships is therefore extremely problematic.

To mitigate these challenges, I employed a three-prong strategy to identify industrial unions involved in transnational relationships. I began by selecting an initial sample of the ten largest U.S. industrial unions in 1991, prior to NAFTA's negotiation. Table A.1 presents them from largest to smallest. It also provides membership data for 1993 – before NAFTA went into effect – and net and percentage losses in membership between 1991 and 1993.