Bangladesh Safety Inspections

To:WRC Affiliate Universities and Colleges
From:Scott Nova
Date:August 18, 2014
Re:Bangladesh Safety Inspections

I write to provide an update on an aspect of the factory inspection program of the Bangladesh Accord: inspections at factories producing for both Alliance and Accord brands.  
As we have explained previously, the decision of a number of large US apparel companies not to join their US, European and Asian counterparts in signing the Accord – and, instead, to create their own smaller initiative – has significantly complicated safety efforts. The civil society and worker organizations involved in creating the Accord would have greatly preferred a single, unified initiative, which would have simplified safety renovations; instead, Walmart, Gap, VF and others rejected calls for unity and instead launched a second program, the Alliance. Making matters more difficult, the Alliance has weaker, non-binding requirements for brands and retailers, which cannot be enforced by third parties.  
One area where the existence of a separate Alliance program has created particular logistical challenges is with inspections of factories that produce for Alliance brands as well as Accord brands. These factories represent about 15% of Accord factories and are subject to inspection by both programs. For obvious reasons, it is preferable to avoid duplicate inspections, but only if this can be done safely. It is also important to avoid multiple, conflicting renovation plans.  
For this reason, the Accord committed last year that it would work to avoid the need for duplicating initial factory inspections by attempting to reach an agreement with the Alliance on the treatment of these factories. The Accord then scheduled inspections of the joint factories toward the end of its inspection calendar, so there would be time to resolve the issue. However, the Alliance proceeded to inspect the joint factories, despite the absence of any agreement.  
The Accord thus faced a dilemma: should it accept the Alliance inspections? On the one hand, the Accord wanted to avoid duplication. On the other hand, the Accord could not allow its inspection standards or the safety of workers to be compromised.  
In the end, despite relentless demands from factory owners and their allies to treat the Alliance inspections as equivalent to the Accord’s, the Accord determined that it could not do so and still protect the safety of workers – because of methodological shortcomings in the Alliance inspection process, because the Alliance does not require its brands to help pay for factory renovations, and because many of the Alliance inspections are directly controlled by individual Alliance brands, rather than the Alliance staff. As you know, brand-controlled inspection programs have a dismal track record in Bangladesh, over many years, with workers dying in larger numbers in factories ostensibly covered by such programs. The Accord determined that inspections directed by Alliance brands cannot be given consideration and that inspections directed by the Alliance staff can be considered only with significant qualifications. 
The Accord has therefore adopted the following policy, announced last Friday, which can be read in its entirety here:

  • At joint factories that have been inspected by Alliance brands (including those inspected by the licensee VF Corporation), the Accord will disregard the Alliance inspection reports and will carry out its normal program of initial inspections.
  • At joint factories where inspections have been done under the direction of the Alliance staff, the Accord will consider the Alliance’s initial inspection reports when formulating action plans, but will also carry out follow-up inspections during the corrective action phase, as called for under the Accord program, to ensure that any mistakes or omissions by the Alliance inspectors are caught and corrected.
  • The Accord has also reiterated to its signatory brands and retailers that their obligation to provide all needed financial assistance to factory owners will remain in place at the joint factories, regardless of whether Alliance brands choose to contribute.
  • The Accord will seek unified corrective action plans wherever possible, but these will, in all cases, be subject to the approval of the Accord’s independent Chief Safety Inspector.

The Accord Steering Committee approved this policy, concluding that it will allow duplication of initial inspections to be avoided where this can be done safely, while still ensuring that workers are protected.  
From the WRC’s perspective, this policy makes sense. It represents an effort by the Accord to find ways to ameliorate the logistical problems created by the existence of a second inspection program, while still ensuring that the weaknesses of the Alliance program are not allowed to place workers at the joint factories at risk.  
Unfortunately, no such safeguards are available to workers at the several hundred factories in Bangladesh producing only for Alliance brands. An indication of the risks those workers face are the recent events at a factory called Sinha Knitting, a large production facility inspected in July by the Accord’s structural engineers, and prior to that by inspectors working for VF Corporation.  
First, it is important to understand that all initial structural inspections (whether Accord, Alliance or local government inspections) utilize assumptions about the strength of concrete in making initial determinations as to the soundness of structures. Actual testing of concrete strength is a complicated and time-consuming process that cannot feasibly be performed in the context of large numbers of initial factory inspections. Instead, assumptions are used initially to calculate the strength of concrete columns. In all cases where the initial calculations indicate any significant risk, more extensive inspection, including testing of concrete strength, is then scheduled. This approach allows the highest-risk buildings to be identified quickly, which is the most urgent priority in Bangladesh. The Accord, the Alliance and the Bangladesh government have all agreed on the specific numerical strength assumptions to be used in these inspections.  
When the Accord engineers conducted their initial assessment at Sinha Knitting, using the agreed assumptions for concrete strength, calculations showed that the building’s columns could be severely over-stressed – to the point where evacuation of the buildings would have been required, pending further analysis. In the end, evacuation proved unnecessary, because the building owner, facing the prospect of evacuation, belatedly shared data from actual concrete strength tests, performed at the time of construction. These test results showed the concrete strength to be much higher, at this particular factory, than the consensus assumptions – and the Accord engineers, on this basis, concluded that the building is safe. For reasons unknown, the owner had never previously revealed this data to any inspectors.  
While the issue at Sinha Knitting is now resolved, the inspection process revealed very disturbing information about the Alliance inspection previously conducted by VF Corporation. The Accord’s Chief Safety Inspector reviewed those inspection results and found that VF’s inspectors failed to use the agreed numbers for concrete strength, but instead used assumptions that were much more lenient. (At the time, the VF inspectors had no knowledge of the concrete test data possessed by the factory owner, so those test results were not the basis of the assumptions used by VF.) VF concluded, based on its less stringent assumptions, that the building was safe – giving it a grade of “A,” the highest grade.

Even more disturbing, a third inspection had been carried out by engineers from BUET (a local university), working on behalf of another brand that is not part of either initiative. The BUET inspectors used the same exact numbers for concrete strength as VF’s inspectors. Yet they concluded that the structure had very significant structural problems and was only marginally safe for occupation – a fundamentally different conclusion from the one reached by VF. Thus, VF’s inspectors not only used the wrong assumptions, they applied them improperly. The effect of both mistakes was to greatly underestimate the potential risk to workers. 
The Accord’s Chief Safety Inspector wrote to a senior Bangladeshi government official on the matter, saying: “The fact that values other than the agreed values are being used is very concerning. This could lead to disastrous results.” He further noted that “even though the same numbers are used by BUET and Alliance, BUET found the internal columns to be inadequate, and Alliance gave the structure an ‘A’.” This case provides a good indication why the Chief Safety Inspector has chosen not to give consideration to the Alliance’s brand-run inspections.  
It is fortunate that at Sinha Knitting the actual strength of the building’s concrete proved to be higher than the consensus assumptions; however, in many buildings, this won’t be the case. That is why the assumptions exist and why it is imperative that they be used consistently. The workers at Sinha Knitting were lucky, but there are dozens of factories in Bangladesh, not covered by the Accord, where workers are currently relying on inspections done by VF. And there are many more where inspections were controlled by Walmart, Gap or another brand. We hope that inspections at these factories, unlike VF’s inspection at Sinha Knitting, have utilized the proper assumptions for concrete strength, but it is unclear whether this is the case. That is one of many reasons why we continue to urge all brands and retailers sourcing from Bangladesh to sign the Accord and submit to genuinely independent inspections of all supplier factories. 

Scott Nova 
Worker Rights Consortium 
5 Thomas Circle NW, 5th Floor 
Washington DC 20005 
ph  202 387 4884 
fax 202 387 3292