Anatomy of an FLA report

FLA’s Own Monitoring Reports Reveal Stunning Ineptitude in Factory Monitoring

The FLA must be hoping that no one will actually read its factory reports. Maybe that’s why they are showcased in a barely squintable 6-point-font, under layers of flashy graphics, in the FLA’s annual report. We have squinted our eyes and read the reports and what we found was a disturbing lack of seriousness for the enforcement of basic worker rights, especially the right to organize.

Analysis of the Fair Labor Association’s Monitoring Activities

This report analyzes the FLA’s monitoring activities based primarily upon the FLA’s own self-published data. As explained in detail below, this self-published data reveals on the part of the FLA a disturbing lack of seriousness and a haphazard methodology for investigating and correcting worker rights violations. The failure of the FLA to competently monitor for worker rights abuses is most alarming in the area of freedom of association, which we will demonstrate in our analysis below.

This analysis includes:

  • General observations of the FLA’s annual report and monitoring report data
  • A statistical overview of the FLA’s findings regarding respect for freedom of association
  • An analysis of FLA findings and actions on individual cases involving violations of freedom of association
  • An analysis of cases in which workers have sought to use the FLA’s third-party complaint system to address worker rights violations

General Observations of Annual Report

Before turning to an analysis the FLA data, we offer some general comments regarding the FLA annual report for 2005, which summarizes the FLA’s audits for 2004.

  • The FLA annual report purports to summarize “94 independent, external monitoring visits at factories producing for FLA companies”. It is important to recognize that these “independent, external” monitoring visits “conducted by the FLA,” were not, in fact, conducted by the FLA and were not independent, at least in any serious sense of the word. They were conducted by contract monitors hired by the FLA, and we are left to rely entirely on the competence and candor of these hired firms, most of whose income is derived from doing contract work for the multinational brands and retailers they monitor. The report entitled “Cal-Safety Compliance Corporation and the FLA”, published by FLA Watch, provides an in-depth look at one of these firms.
  • None of the information provided in the FLA factory monitoring reports is verifiable. This is because the FLA and its member corporations continue to refuse to publicly disclose the names of the factories being monitored, instead insisting on referring to them with nicknames such as “factory 08021540C”. This leaves workers and solidarity movements with no way to challenge the veracity of the FLA’s specific findings. It is possible, however, to get a sense of the quality of the FLA’s corporate monitoring based simply upon reading the self-published factory reports. As discussed below, the findings from this exercise are disturbing.
  • The FLA report makes heavy use of statistics and graphs to summarize “instances of non-compliance”. The FLA deserves credit for the attractive presentation. But the report’s use of statistics can only be called pseudo-science, as any trained researcher would recognize. First, of course, the statistics are only as good as the quality of monitoring that is taking place; the reports themselves indicate that the monitoring is seriously FLAwed. Second, in purporting to describe the frequency of various code of conduct violations, the report begs the question: what counts as an “instance” of non-compliance? If a supervisor has been shouting degrading verbal insults at all 40 women workers he supervises on a daily basis, and he has been doing so for the past two years, then how many instances of non-compliance are we observing? One? One hundred? One thousand? Who knows? There is no substantive basis, and no legitimate reason, to try to put any number on it. In doing so anyway and without explanation, the FLA report conveys to the public a sense of scientific rigor that is in fact absent. In the FLA’s attempts at quantitative analysis they have failed at the most important qualitative aspects of monitoring- those dealing with harassment, discrimination, freedom of association and basic respect and dignity.
  • The FLA annual report makes mention of only two cases in which the FLA responded to a third-party complaints In one of these cases, the FLA mischaracterizes a case as positively settled when in fact the FLA did not even recommend a corrective action regarding the most serious abuse identified (the massing firing of a workforce following a unionization effort). USAS is aware and has knowledge of a much larger number of other cases in which workers have submitted third-party complaints none of which receive mention in the FLA’s report. We describe several of these cases in some detail below – in each case the FLA took more than a year to respond to worker complaints and then failed to address the issues of most serious concern.

Analysis of FLA Monitoring of Freedom of Association

The FLA has recently made bold public comments about its deep commitments to the right of workers to organize and bargaining collectively. The FLA recently stated that “The FLA believes that good labor relations and the ability of workers to negotiate and determine their own needs are the foundation of sustainable improvements in workplace conditions.i ” Given these bold pronouncements, what has the FLA actually done to ensure the associational rights of workers are respected?

To answer this question, we analyzed the findings on freedom of association from a sample of 52 of the 94 factory audits that the FLA contract monitors conducted in 2004, published in the FLA’s 2005 annual report. These 52 reports represented the full universe of tracking charts published by the FLA as of November 20, 2005; additional tracking charts were released subsequently. The following conclusions can be drawn from this self-published FLA data.

  • In most factories monitored (54%), the FLA monitor made no finding whatsoever about whether the right to organize and bargain collectively is respected. The rate of factories where no finding was made is even higher in China (70%).
  • The charts do not identify a single case where worker leaders terminated for organizing won reinstatement as a result of the FLA’s monitoring, though the reports indicate that such firings had taken place.
  • In not a single case summarized in the FLA’s tracking charts did workers succeed in organizing a union while the factory was under FLA scrutiny, though it is apparent that such efforts by workers were underway during or before the FLA’s audit.
  • In a number of cases, the FLA monitors noted inappropriate involvement by management in the affairs of unions or worker organizations, but then failed inexplicably to characterize this behavior as inappropriate.
  • Overall, the tracking charts indicate a haphazard methodology and a disturbing lack of seriousness in pressing for respect for associational rights, as described in a lengthy series of examples below. The monitors repeated accept the word of factory managers or easy-to-produce written policies as evidence of problems fixed, without conducting any verification to ensure that corrective actions are actually taken.

The following chart provides a general overview of our review of the FLA tracking charts.

Summary of FLA Tracking Charts

Number Percentage
Total number of factories audits reviewed


Factories in which the FLA monitor reports nothing regarding Freedom of Association



Factories in which FLA monitor reports a finding on Freedom of Association



Factories in which a union was recognized through the FLA remediation process



Factories in which a union existed in a factory at the time of the audit



Among factories where a union existed, number of factories where the information indicates the union is run or is inappropriately influenced by management.ii



Of factories where unions are run or are inappropriately influenced by management, number at which FLA monitor nevertheless treated the union as legitimate.



Factories where there is a worker-management committee



Of factories with worker-management committees, number where election of worker representatives was organized or inappropriately influenced by management



Factories in China


Factories in China with no finding on FOA



Factories in countries known to have rampant violations of Freedom of Association iii



Of cases in these countries, number with no finding on Freedom of Association.



Analysis of Individual Factory Audits

The following are examples of the findings made by FLA monitors regarding freedom of association in a number of factories from the FLA’s own tracking charts. This self-published data indicates that FLA monitors lack basic training to identify violations of associational rights and display an alarming lack of seriousness in following through on violations that are identified.

  • At factory 08021540C in China, the FLA reports that factory management committed to “Establish clear policies and procedures to provide workers with the right to freely associate and collective bargaining. Communication channels should be available.” This commitment was then “verified” by the FLA’s monitor, who found that the “Right to associate is recognized in ‘factory rules’.” No other information is provided. This report is illustrative of many in which the existence of company documentation on an issue is sufficient for the FLA monitor to conclude that a problem is solved, even as a basic knowledge of the context – in this case the fact that in China, independent organizing and bargaining collectively by workers is illegal and harshly repressed – casts doubt on or contradicts the conclusion.
  • At factory 34032282C in El Salvador, the monitor found evidence indicating that the factory had been blacklisting trade unionists – “Workers said that people who have been members of unions are not hired by the company” – a finding corroborated by testimony from multiple workers. Yet the only actions in the company’s remedial action plan, accepted by the FLA, were: a) revising its written discrimination policy to include a clause stating that workers’ union membership will not affect hiring, and b) providing training to employees, presumably by management as no other actor is mentioned. The FLA did not require any training of management, the party in a position to actually stop the blacklisting. Nor, apparently, did the FLA attempt to verify that the blacklisting was no longer occurring at the factory – a matter that could be readily tested by follow-up research. This is another example in which the presence of documentation is accepted as proof of a problem fixed, without any attempt to verify that a company’s behavior has actually changed.
  • At factory 10025003C in Mexico, according to the tracking chart “many workers told auditors that workers were dismissed when leading efforts to form an independent union” and workers feared the company would retaliate if workers sought to organize. The company’s remediation plan, accepted by the FLA, consisted of the following statement: “Respect employees’ rights to comply with all national/local laws and regulations concerning freedom of association and collective bargaining. Employees should be free to join organizations of their own choice without being subject to intimidation or harassment.” The only specific commitments for action promised by the factory were to place a notice on the factory’s announcement board “reminding workers of their freedom of association” and to “emphasize freedom of association” in management’s course for new workers. There is no indication that the FLA made any effort to identify workers who were fired for organizing or demand that they be reinstated. Apparently the FLA monitor did not even verify that the factory carried out the limited actions promised.
  • At factory 12008477C in Vietnam, the FLA monitor identified a patently bogus company-run union, in which the “union leader is Deputy Factory Manager of Factory X…. All candidates in the election and all delegates voting in the election were line leaders, supervisors or managers. Regular line workers were not directly involved in the process…. A CBA was signed recently (October 2004)… [but] there was no negotiation process, no collecting of opinions from workers, and no worker involvement in the process.” Yet, inexplicably, the FLA then concluded that the union could become legitimate if management made certain adjustments to the arrangement. The factory agreed to “provide time budget of the union president and records on all union activities to demonstrate the involvement of the union president and other 12 committees” And – as if a company-run union could be made authentic simply by management reorganizing it – the factory then agreed to a remediation plan, involving “scheduling regular meetings in consultation with union representatives to review roles and responsibilities” of the union and submitting to the FLA, as documentation, “(a) union restructuring plan (b) agenda for scheduled meetings and trainings.” The FLA’s monitor confirmed the submission from management of management’s “union restructuring plan.”
  • At factory 07003236C in India, the FLA monitor found evidence indicating blacklisting of union members – “Worker with union affiliation is not hired.” The company denied the practice. The only remedial response of the company, accepted by the FLA, was for “management to verify that factory policies do not encourage any discrimination at the time of hiring.” The FLA monitor then confirmed that the factory added a non-discrimination clause in its hiring policy and conducted trainings for its supervisors and managers. No effort was apparently made by the FLA to verify that the company is in fact no longer blacklisting union workers. This is yet another case where the FLA monitor accepts documentation as sufficient evidence of a problem fixed, without any independent verification that practices have changed.
  • At factory 36000502C in Bangladesh, the FLA monitor found a Workers Welfare Committee headed by “the Administrative Manager and all the members are selected from management.” At the FLA’s urging, the factory added 23 worker representatives to the committee. It is not completely clear, but it appears that management selected these workers or organized an election for them, as the workers reportedly did not understand their purpose on the committee. The FLA then asked the company to provide training to these workers.
  • At factory 07007671C in Turkey, the FLA monitor found that a number of workers “quit” because “management did not support the Trade Union.” The report states that “this has been resolved”, though it is entirely unclear what this statement means. The report goes on to note that the factory then established a worker committee, consisting of two employees, which meets weekly with management to discuss grievances. The FLA monitor concluded that the new committee is working effectively. In sum, the report indicates that a union organizing drive was busted and in its place the company set up a 2-person worker committee, with no power to negotiate a contract or represent the workers legally, and this outcome was deemed positive by the FLA monitor.

Discussion of FLA Third-Party Complaints

This section provides an analysis of the FLA’s “third-party” complaint system. The FLA annual report makes mention of only two cases involving the third-party complaint procedure. However, USAS is aware of a much larger number of other cases in which workers have submitted complaints which are not reported on in the FLA’s report. In some of these cases, USAS has intimate knowledge of the cases as a result of relationships with the concerned workers and the experience of working as worker advocates on these cases. This section includes reviews of one of the two complaints reported upon in the FLA annual report and two additional cases which receive no mention in the FLA annual report.

Gildan Activewear

One of the two cases of third-party complaints summarized in the FLA Annual Report 2005 was that of a factory in Honduras owned by the Canada-based Gildan Activewear.

In early 2004, the Worker Rights Consortium and the Fair Labor Association initiated investigations of Gildan’s factory in El Progreso in response to complaints by or on behalf of workers. Both the WRC and FLA investigations concluded that Gildan had violated its workers’ rights by illegally firing dozens of workers who had attempted to organize an independent trade union. In July 2004, in the midst of joint negotiations with the WRC and FLA to correct violations, Gildan suddenly announced it was shutting down the El Progreso factory and firing all 1,800 of the plant’s workers. While Gildan claimed the closure was unrelated to the investigations and the attempt by workers to form a trade union, an analysis by the WRC found that Gildan’s justifications were pretexts and that the company had closed the factory to put an end to the inquiries and the organizing efforts by workers and that the mass firing of workers was illegal.

In October 2004, the FLA put Gildan under special review. As a condition of regaining its normal FLA member status, the FLA asked Gildan to take several actions, but these actions did not include the most central demand of the fired workers and of the WRC – to either reopen the factory or provide meaningful employment opportunities to the illegally terminated workers. Instead, the FLA asked Gildan to correct a number of misrepresentations of the FLA’s position on Gildan’s website and with the media, provide training to supervisors, meet with an NGO involved in the case, and provide back pay to only half of the workers terminated for organizing before the closure. Thus, Gildan was allowed to illegally terminate its entire workforce of some 1,800 workers to bust an organizing drive – in the middle of an FLA investigation – and yet the FLA did not require Gildan to do anything to remedy the harm done to these illegally terminated workers.

Ultimately, after Gildan was reinstated by the FLA as a full member and without any assistance from the FLA, the WRC extracted a commitment from Gildan to offer employment opportunities to illegally fired workers in other Gildan facilities in the region.

PT Jaba Garmindo

One case in which workers filed a third-party complaint to the FLA that does not receive mention in the FLA report is PT Jaba Garmindo, located in Jakarta, Indonesia.

In April 2004, workers from PT Jaba Garmindo filed a third party complaint with the FLA, alleging that the factory was violating the workers’ rights to associate freely and bargain collectively. PT Jaba Garmindo produces collegiate licensed goods for Nike. The FLA responded by email several days after the complaint was filed, promising that the FLA would review the case in the next several days and move forward with the FLA third-party complaint process.

But for the next year, from April 2004 to April 2005, the workers heard virtually nothing from the FLA or from Nike. During this period of time, the FLA never sent a representative to the factory to meet with the workers. To the knowledge of workers or their advocates, no investigation was ever carried out, either by the FLA or Nike. If an investigation was carried out by a third party on behalf of the FLA and/or Nike, it was carried out without speaking to the workers who filed the complaint and without or sharing any findings or recommendations with these workers or their union, known as GSBI.

More than one year after the complaint was filed, at the end of April 2005, a FLA investigator finally came to Indonesia in person to “investigate” the Jaba Garmindo case. But by this time, the issue was moot. During the year than had elapsed, the workers had aggressively pursued the issues through the Indonesia legal system and through public pressure and were able to reach a resolution with the factory’s management. It was only through the workers’ tenacity and management’s willingness to resolve the issue that the issues were ever addressed.

PT Victoria

The case of PT Victoria in Indonesia is another in which workers filed third-party complaints, which the FLA’s annual report does not mention.

At the end of December 2003, the PT Victoria factory shutdown and its owners fled to Hong Kong without paying workers more than million dollars in legally mandated compensation, including severance compensation and compensation for time they had worked. The factory had been producing garments for FLA corporate member Eddie Bauer and the factory had not paid workers for a substantial period of time they spent sewing Eddie Bauer garments prior to the closure. The workers pursued the issue in the Indonesia courts and, in March 2004, the court ordered the company to pay a full severance settlement, but PT Victoria refused to abide by the court’s ruling. In August 2003, having pursued all local avenues for justice in Indonesia, the PT Victoria union decided to send a letter of concern/ complaint to FLA.

From August 2004 to August 2005, the workers heard virtually nothing from the FLA. Finally, in August 2005 – fully one year after receiving the complaint –an FLA investigator finally came to Indonesia in person to “investigate” the case, which meant starting from zero in terms of understanding the issues involved. The workers were so upset that the FLA took so long to respond to their complaint that they staged a protest of the FLA when representative arrived.

In May 2005, just as Eddie Bauer was being targeted in an international campaign for its failed response to this case, the FLA chose to provide full accreditation to Eddie Bauer’s monitoring program – to the shock of workers in Indonesia and worker advocates throughout the globe. Eddie Bauer’s role in this case was viewed as particularly disturbing because following the illegal closure of PT Victoria, Eddie Bauer maintained a business relationship with the factory’s parent company (shifting the production of goods to a sister factory in Cambodia), but failed to use this influence to compel the factory to pay the money it owed the workers. Eddie Bauer flatly refused to take any responsibility for the debts owned to workers, despite the fact that the factory had failed to pay workers for time spent sewing Eddie Bauer garments. The FLA supported Eddie Bauer’s position on this.

In August 2005 – nearly two years after the factory closed without paying workers’ legally mandated compensation – the FLA participated in a series of meetings with Eddie Bauer and the worker representatives. The WRC provided the FLA with translation and arranged the logistics for the meetings. The FLA and Eddie Bauer public credited themselves for their involvement in these convenings, calling the meetings examples “of their serious efforts to resolve the case”. However, nothing substantive came from the meetings toward actually correcting the nonpayment of wages to the workers. To date, the workers continue to be owed more than a million dollars.

To our knowledge, the FLA has not indicated any intention to address the case further.


An analysis of the FLA’s annual report leaves a disturbing impression regarding the FLA’s competence and seriousness as a factory monitoring organization, particularly with respect to the protection of associational rights. This is not suprising considering the structure of the FLA.

The FLA’s improper use of statistics in reporting on its work would be derided as pseudo-science in any serious academic setting. Indeed, a graduate student at one of the FLA’s member universities employing the same techniques would be sent back to redo his or her work.

When one takes the time to read the data on which these statistics are based (the tracking charts presented in 3.5 size font), one finds monitoring practices that are haphazard, sloppy, and lacking in basic competencies. In a majority of cases, the FLA monitors make no finding whatsoever about whether the right to organize and bargain collectively are respected. The monitors repeated accept the word of factory managers or easy-to-produce written policies as evidence of problems fixed, without conducting basic verification to ensure that corrective actions are actually taken. In some cases, the FLA monitors notice violations, such as the firing of union members, but then apparently do nothing to correct the violations. In other cases, the monitors notice obviously inappropriate behavior, such organizing by management of union elections, but somehow do not identify the practices as inappropriate.

A review of the FLA’s third-party complaint procedures finds yet more disturbing practices. The FLA monitors failed to meaningfully respond to worker complaints for more than a year after the complaints were filed. When the FLA did follow-up on worker complaints, it failed to take the most basic actions to correct the abuses identified and – remarkably – went so far as to certify the companies involved as full FLA corporate members even as these companies’ failure address the violations were being denounced by workers and credible worker advocates across the globe.

Our review of the FLA’s practices gives strong reason to doubt the candor of the FLA and the effectiveness of the monitoring practices it employs. We hope this critical assessment will be helpful in stirring what is clearly much needed change in the way university anti-sweatshop policies are being monitored and enforced.

i. Fair Labor Association’s “General Points for Schools to Consider Regarding the FLA Program and the Designated Supplier Program Proposal”

ii. These include factories where union officials are members of management, where union members or leaders are appointed by management, or where management apparently organized the union’s election.

iii. These include China, Bangladesh, Vietnam, Mexico, and El Salvador.