||Printer-friendly version (HTML)
What Can You Tell Me About Collective Bargaining Expirations and Work Stoppages?
Originally Posted: May 28, 2003
There are several reasons why it is difficult for BLS to identify accurately the relationship between collective bargaining agreement expirations and work stoppages; most reflect the fact that the data series were established as separate programs, were maintained for different purposes, had separate databases, and were not coordinated for research purposes.
Some of the questions that the Bureau’s collective bargaining staff receive pertain to the relationship between collective bargaining agreement expirations and work stoppages: 1) Does the work stoppage series distinguish between strikes and lockouts? 2) What percent of first agreements1 lead to work stoppages? 3) How many collective bargaining agreement expirations result in a work stoppage? This article attempts to provide some background and answers to these questions.
The major work stoppage series
When the National Labor Relations (Taft-Hartley) Act of 1947 was passed, BLS was specifically authorized to collect information about industrial disputes: "The Bureau of Labor Statistics in the Department of Labor is authorized to furnish upon request of the Service, or employers, employees, or their representatives, all available data and factual information which may aid in the settlement of any labor dispute..."2
For almost 35 years, the Bureau’s work stoppage program included all work stoppages (strikes and lockouts)3 that idled 6 or more workers for 8 hours or a full shift or longer in all private and public sector industries. Because of budget cutbacks, the work stoppage program was amended in 1982 to cover only major work stoppages--those that involve 1,000 or more workers. Like the earlier work stoppage series, major work stoppages include worker-initiated strikes and lockouts by employers. Because of the complexity of disputes, BLS has never attempted to distinguish between strikes and lockouts in its statistics; both are included in the term "work stoppages." The work stoppage series excludes slowdowns, work-to-the rules disputes, and political strikes.4 In addition, because of the limitations of the program, BLS does not currently collect data on the number of first agreements that lead to a work stoppage.
Measures of major work stoppage activity. BLS currently publishes several measures of major work stoppage activity, including the number of work stoppages, the number of workers involved in stoppages, and the number of days of idleness resulting from stoppages. Data for the number of workers involved and days of idleness include all workers made idle in establishments directly involved in a stoppage; they do not account for secondary idleness resulting from material or service shortages at establishments not directly involved in the dispute. (See table 1 for data covering 1993-2002.)
Data for the number of days of idleness are calculated by summing the number of workers idled on each day for each of the stoppages in effect in the reference period. Data for days idle as a percent of working time are calculated by dividing the days of idleness for all stoppages by available working time (computed by multiplying total employment by the number of days typically worked) for the reference period. Total employment figures are from the BLS publication Employment and Earnings, published by the Office of Employment and Unemployment Statistics.5
Detailed work stoppage information published on a monthly or annual basis includes the following:
- Employer and union names
- Location of the stoppage
- Beginning and ending dates of the stoppage
- Number of workers involved
- Number of work days lost and percent of working time lost
- Duration of stoppages, in ranges
Work stoppages data can be found at http://www.bls.gov/wsp/mwsdetail.htm.
The collective bargaining agreement file
The Taft-Hartley Act also authorizes BLS to collect collective bargaining agreements and maintain a file6 of these agreements for public inspection: "For the guidance and information of interested representatives of employers, employees, and the general public, the Bureau of Labor Statistics in the Department of Labor shall maintain a file of copies of all available collective bargaining agreements and other available agreements and actions thereunder settling or adjusting labor disputes."7
Like the work stoppage program, the collective bargaining agreement file has evolved over time in response to public demands and budgetary constraints. Until the early 1980s, for example, BLS collected all known agreements covering 500 or more workers; but due to declining resources, the number of workers covered was increased to 1,000. Thus, the current file includes all major collective bargaining agreements covering 1,000 or more workers in all private and public sector industries, except the railroad and airline industries, whose agreements are collected by the National Mediation Board.8 Currently, BLS has one or more agreements in its file for more than 2,180 collective bargaining situations.
The collective bargaining agreement (CBA) database. BLS catalogues all the agreements in its file in its CBA database. Among other things, the CBA database contains a number of contract identifying fields:
- The names of the employers and unions that are signatories to the agreement
- A bargaining unit identifier (e.g., production and maintenance workers)
- The geographic location of the bargaining unit
- The number of workers covered by the agreement
- The effective and expiration dates of the latest agreement in the file
When an agreement is renegotiated and received by Bureau staff, the CBA database is updated to reflect changes, including new effective and expiration dates.
"Disconnects" between work stoppages and collective bargaining expirations
There are several reasons why it is difficult for BLS to accurately identify the relationship between CBA expirations and work stoppages. Most of the difficulty is due to the fact that, because of budgetary reasons, work stoppages and the collective bargaining file were established as separate programs and were maintained for different purposes, had separate databases, and were not coordinated for research purposes.
The following table shows a comparison of the similarities and differences in the scope and major characteristics of the two programs:
||Collective bargaining agreements
|Includes stoppages idling 1,000 or more workers
||Includes bargaining units of 1,000 or more workers
|Covers all private and public sector industries, including railroads and airlines
||Covers all private and public sector industries, except railroads and airlines
|May cover two or more bargaining units that combined idle 1,000 or more workers in the same stoppage
||Excludes small, coordinated multiple units totaling 1,000 or more workers if covered by separate agreements
|Excludes bargaining units of 1,000 or more workers if less than 1,000 are idled
||Each bargaining unit must cover 1,000 or more workers
|May result from a reopener9, first agreement, or grievance
||Only relates to contract expirations
As can be seen, the scopes and characteristics of the two programs differ in major ways, so that a work stoppage incident may not have a corresponding entry in the CBA database. Nor is there currently a common identifier to link an incident in the work stoppage database to an existing, associated expiration in the CBA database. The CBA database was established exclusively to catalogue the collection of agreements in the Bureau’s file and was not designed to measure collective bargaining activity or work stoppages.
The expiration dates in the CBA database pose another problem. The database is updated when new agreements are received (almost daily); and new expiration dates replace old expiration dates, which are not retained. Thus, the parties’ bargaining history is lost, making it harder to link a contract expiration with a stoppage if they occur in different years.
In addition, counts of the number of expiration dates for a particular year will vary depending on when they are produced. For example, a count of the expirations for 2004
done in April 2003 will differ from the same count performed in October 2003 because a number of agreements’ expiration dates will have been updated. Also, the database "expiration date" field refers to the last, available printed agreement in the file. If there is an agreement that has not been printed or made available to Bureau staff, then the expiration date field will not be updated. In these cases, the CBA database expiration dates are "stale" and do not reflect the most recent settlements.
In addition, the CBA program also was not established to capture data on reopeners or first agreements. To further complicate matters, expirations may occur in one year and work stoppages in subsequent years; thus, making it more difficult to link the two sets of data. For instance, an agreement may expire in 2004, but the work stoppage may occur in 2005.
Historical collective bargaining activity data
BLS does have limited historical data that can shed light on the relationship between CBA expirations and work stoppages. The data come from the bargaining calendar program, a byproduct of the Bureau’s major collective bargaining settlement program, which provided costing information on collective bargaining settlements, deferred wages, and cost-of-living adjustment payments. The collective bargaining settlement program was eliminated in 1995 due to budgetary constraints; thus, the bargaining calendar program was also eliminated.
The bargaining calendar program:
- Included all bargaining situations covering 1,000 or more workers
- Included both contract expirations and reopenings
- Covered all private sector industries, including railroads and airlines
- Included two or more bargaining units in a coordinated bargaining situation if the combined number of workers covered equaled or exceeded 1,000
Because of these characteristics, it is possible to more readily link the data from the bargaining calendar to corresponding data on major work stoppages. Table 2 shows data from the last 5 years of the program. These data reveal that about 5.6 percent of contract negotiations led to a work stoppage, and about 10.4 percent of workers involved in contract negotiations were idled by a work stoppage.
A "snapshot" of 2001 private sector expirations
Does BLS have any current data that can shed light on the relationship between expirations and work stoppages? A rough approximation has been attempted. Every month, BLS produces public listings of the collective bargaining agreements in its database so visitors to the file room will have a recent "picture" of the latest available agreements in the file. One such listing was used as a snapshot to try to gauge the number of private sector agreements that expired in 2001. The listing showed that 348 (27 percent) of 1,271 private sector agreements in the database as of October 2001 had expirations in 2001. In the same year, there were 24 major work stoppages in the private sector. Simply dividing 24 by 348 will lead to the inference that 6.9 percent of major private sector agreements expiring in 2001 were associated with a major work stoppage--similar to data seen in 1991-95. Is this an accurate picture of expiration and work stoppages in 2001? The figure is probably too high because the number of expirations in 2001 is probably underestimated.
BLS is currently investigating whether it can produce annual snapshots that would shed light on the question of how many expirations lead to work stoppages. The work may allow BLS to link work stoppages and expirations by industry, thus producing industry dispute incidence rates. In addition, it may allow BLS to link duration of work stoppages to length of contracts.
1 A first agreement is the initial agreement between the company and the union, coming after a union is certified to represent the employees covered under the agreement or after the union is voluntarily recognized by the company to represent these employees.
2 Section 211(b), Title 29, Chapter 7, Subchapter II, United States Code.
3 A strike is a temporary stoppage of work by a group of employees (not necessarily members of a union) to express a grievance or enforce a contract demand. A lockout is a temporary withholding or denial of employment by a company during a labor dispute to enforce terms of employment upon a group of employees.
4 A political strike is conducted against a governmental agency, not an employer. An example would be when unionized taxicab drivers "strike" because they think taxicab rates (set by a governmental agency) are too low.
5 See "Major Work Stoppages Technical Note," Compensation and Working Conditions, Fall 2001, p. 85.
6 The Bureau maintains a public file of the agreements in its Washington, DC offices. The file room is open from 8:30 am to 4:00 pm so data users can review the agreements. If data users can’t or do not want to visit the Washington, DC office, they can have BLS staff copy agreements at a nominal cost.
7 Section 211(a), Title 29, Chapter 7, Subchapter II, United States Code.
8 Section 212, Title 29, Chapter 7, Subchapter II, United States Code specifically authorizes the National Mediation Board to collect these agreements. The Bureau, therefore, has never collected railroad and airline agreements.
9 Reopeners are clauses in collective bargaining agreements that state the time and circumstances under which negotiations can be requested prior to the expiration of the contracts. Reopeners are usually restricted to wage and benefit issues.