The goal of the Cornell-ILR Labor Action Tracker is to provide a comprehensive database of strike and labor protest activity across the United States displayed on an interactive map. This document contains information about our methodology, specifically how we capture and collect information about strikes and labor protests. We outline our methodology through three sections below. In Section I, we distinguish between strikes and labor protests by developing separate definitions for each. In Section II, we discuss our search and verification protocols to outline how we find and add events to our tracker. In Section III, we define and explain the variables we capture for each event. We also plan to release a short summary via our Twitter account on the third Monday of each month detailing the labor actions we captured over the prior month.
Our tracker provides links to original sources that we use to compile and manually input information about each strike or labor protest. However, we often have to use our own judgment to make final determinations about certain variables, as available information can be partial or ambiguous. We meet regularly to discuss certain labor actions, especially to decide whether to label a specific event a strike or a labor protest. We will continue to be transparent about any changes we make to this document as our project grows and develops. Note that we often refer to strikes and labor protests in the below sections as "events."
On our interactive map, users can filter based on whether we label an event a strike or a labor protest, which we define below. Users also can search for both strikes and labor protests, which compiles our complete database of labor action across the United States.
Strike: “A temporary stoppage of work by a group of workers in order to express a grievance or to enforce a demand. Such a grievance or demand may or may not be workplace-related.”
--The first sentence of this definition is heavily influenced by Peterson (1937: 3),1 also
used by Hyman (1989: 17).2 The only difference we make to the first part of this definition is by changing “employees” to “workers.”
Labor Protest: “Collective action by a group of people as workers but without withdrawing their labor in order to express a grievance or enforce a demand. Such a grievance or demand may or may not be workplace-related. A labor protest may also consist of a group of people not acting in the protest as workers as long as the central demand is workplace related.”
We distinguish between strikes and labor protests as a core component of our labor action tracker. The major distinction between strikes and labor protests relates to whether a group of workers stopped work during the course of the event. We believe this definition of strikes is relatively inclusive, but we need to convincingly demonstrate that a stoppage of work led by a group of workers occurs to label an event a strike.
We collect information on strikes and labor protests from a variety of sources, building on many of the protocols that the Bureau of Labor Statistics (BLS) employ to search for work stoppages.3 These sources include, but are not limited to, the BLS strike report, the Federal Mediation and Conciliation Services (FMCS) work stoppage summary, Bloomberg Law’s work stoppage database, major media sources, organizational press releases, and social media. We are currently searching for English and Spanish-language sources.
Our verification protocols determine the number and type of sources we require to add an event to our tracker. We identify three different categories of sources with corresponding protocols:
Tier I) These are sources that have tracked strike data for a considerable period of time. We can feel confident both that the event occurred, and the nature of the event was presented accurately.
Tier I sources include the BLS strike report, the FMCS work stoppage summary, and Bloomberg Law’s work stoppage database.
Tier II) These are sources either connected to the labor movement or press coverage of the labor movement. We can feel confident that the event occurred, but we need to make sure that the event is labeled correctly according to our definitions and that we accurately capture correct information pertaining to the variables listed in Section III. To add events generated from one of these sources, we require additional verification from another Tier II source or Tier III source.
Examples of Tier II sources include, but are not limited to, union/labor organization press releases, Labor Notes, The New York Times, local affiliates of major networks such as NBC or CBS, local news sources such as The Buffalo News or The Ithaca Voice, and news sources that cover specific industries.
Tier III) These are nontraditional sources that may capture activity not covered by Tier I or Tier II sources. These sources are more difficult to rely on if we want to feel confident that the event is labeled correctly according to our definitions and that we accurately capture correct information pertaining to the variables listed in Section III. To count these events, we require additional verification from one Tier II source or one Tier III source. Every effort should be made to verify a Tier III source with at least one additional Tier II source.
Examples of Tier III sources include, but are not limited to, social media posts, such as Twitter and Facebook, and online forums, such as Reddit. An email exchange or other contact with informants, such as workers or labor organization representatives, to verify or learn more about an event also counts as a Tier III source. When using informants as a source, we will state this under the “Notes” section of an event.
For each event, we try to capture data on a set of variables to provide as much information as possible. We gather all of this information from the sources we cite for the event, occasionally looking up additional information (such as a specific address for the event if not mentioned in the sources). Not all variables are captured for each event, either because they are not applicable to the event or because we are not able to capture reliable information for particular variables. The variables are listed below in the order they appear when searching a specific event. Users are able to filter according to some of these variables (marked below by *) using our filtering mechanisms available on the main page.
Please note that the number of events listed on our map does not necessarily reflect the number of individual strikes and/or labor protests, but rather the total number of locations of strikes and/or labor protests. For example, a single strike at five locations will count as five separate “results found” on our map.
Employer: Name of the employer involved in the event, if applicable. We generally list the employer name referenced in our sources. In the event that workers are legally considered independent contractors, we try to list the company that organizes and manages their work.
*Labor organization: Name of the labor organization(s) involved in the event, if applicable. This includes unions, worker centers, and/or other labor organizations. For unions, the name of the international union will generally go here. When filtering for labor organization, please use an acronym (e.g., SEIU, IBT) for most inclusive results.
Local: Name of the local labor organization involved in the event, if applicable. This includes locals of unions or local associations of other labor organizations such as worker centers.
*Industry: Industry of employer involved in the event, if applicable. We generate our list of industries from the 2017 NAICS.4
*Bargaining Unit Size: The number of workers in the bargaining unit related to the event if we are able to determine this information. This is only applicable for events that feature unionized workers or workers in the middle of a unionization campaign.
Number of Locations: The number of locations for a given event. For example, a single strike at 12 locations would read 12 for this variable. Note that when this number is greater than one, link(s) to the other locations will appear after the last variable below. Also note that we cannot find all locations for some events, which will be acknowledged in the Notes variable below.
Address (city, *state, zip code): The address of the event, which allows us to pinpoint the event on our interactive map. Note that users can filter by state, which includes all states and territories of the United States.
*Strike or Protest: Whether we categorize the event as a strike or a labor protest according to our definitions in Section I.
*Approximate Number of Participants: The number of participants in a given event, if we are able to determine this information. Note that this category potentially applies to both union and nonunion events. We choose the word “participants” as some events may involve community members and other individuals beyond the workers directly involved in the event.
*Start Date: The start date of the event.
End Date: The end date of the event.
*Duration Amount: The number of days, hours, or minutes of an event. Note that we only capture duration for strikes and that this generally includes all calendar days that the strike covers.
Duration Unit: Days, hours, or minutes of an event. Note that we only capture duration unit for strikes. Strikes are generally captured in days unless we can demonstrate the strike lasted a certain number of minutes or hours less than one day.
*Authorized: Whether the local union or local union leadership actively supported the event. This variable only applies to unionized workplaces (after formal recognition of a union). Note that we only capture the authorized variable for strikes.
*Worker Demands: The main demands associated with an event. We have developed a list of twelve demands (see below) that users can filter by based on some of the most frequently identified worker demands from our first approximately 1,000 entries.
Note that these twelve demands do not preclude us from listing other demands for individual events.
Source: Link(s) to the source(s) we used to capture the event. The number and type of sources are guided by our verification protocols as explained in Section II. We sometimes go beyond these protocols and link to additional sources to capture more information relevant to our variables and/or to provide more information on a specific event.
Notes: Used to provide additional information about an event, if necessary.
1Peterson, F. (1937). Strikes in the United States: 1880-1936. Washington: United States Department of Labor.
2 Hyman, R. (1989). Strikes: Fourth Edition. London: Macmillan.
3See (2019). Work Stoppages. Bureau of Labor Statistics. Retrieved from https://www.bls.gov/opub/hom/wsp/pdf/wsp.pdf.
4See (2021). 2017 NAICS. United States Census Bureau. Retrieved from https://www.census.gov/naics/?58967?yearbck=2017.