That type of knee jerk conjecture is really weak. The data collected on shrinkage, as noted in the linked Reuter’s article, is noisy. You can’t differentiate lose due to theft or shipping mistakes or cliericsl error.
More importantly, and not mentioned directly in the boingboing article, was the cited number of rising organized theft was based upon an analyst from a security firm. The report was created in partnership with that firm. With the recent redaction, there is no mention of that firm.