Apparel Sourcing Intelligence - Worldwide

What’s the use of junk statistics?

Do you believe hundreds of people died in in Bangladeshi garment factory accidents this year?

Well the Bangladesh Occupational Safety, Health and Environment Foundation (a trade union-funded pressure group) says it does.

Well, not quite. Its definition of an accident at work is anything that happens travelling to or from work, or as a result of a riot anywhere near your factory. Or, to be more precise, anything reported in the papers as happening. And it won’t answer my emails asking quite how many garment workers actually died.

But who cares? Now there’s another junk number activists can throw around: not only does the garment industry pay rotten wages – which is why the queues outside Bangladeshi factories trying to get jobs are so long. Now you can claim it kills hundreds of people a year too.

Now one death in a factory is one too many. And I’m all in favour of any attempt to name and shame factory owners who lock fire exits, allow poor safety standards around cutting equipment, or do anything else to endanger their workers. But silly numbers, plucked out of the air, about “hundreds of fatalities” do nothing to help serious attempts at improving worker safety.

Here’s who they do help. They help the enemies of workers in poor countries: the near-racist newspaper columnists who seize on every example of poor working conditions to persuade their gullible readers to buy clothes made in the West – and put another few hundred Bangladeshies ouyt of work at the same time. Junk statistics help hypocritical “NGO workers” who draw nice salaries to campaign against non-existent abuses. And they help businesses competing with Bangladeshi workers’ employers.

But they’re a barrier to real change. Next time – and let’s hope it’sa long way off – a factory collapses because corrupt officials have allowed it to be built where it shouodn’t, or a fire breaks out, an overload of junk statistics will be yet another reason the right steps won’t be taken.

On a subject as sensitive as this, junk statistics really can kill