If the chair allows it, just briefly, I don't want to leave the impression there were no front-end controls. We had historical data on the size of businesses' payrolls, and we found that sufficient to flag businesses for risk. If there was a significant discrepancy between the historical payroll paid, not at an employee level but in the aggregate, and what was being claimed on the wage subsidy, that resulted in these manual reviews that represented 40% of the dollars. Those manual reviews were done by the level of auditor who usually deals with it. In some cases they went to the most senior-level auditors in the agency.
I don't want the impression to be that because we didn't have the SINs, we didn't match the historical payroll, because we did.