I can start.
I think there were a few one-off discrepancies in that regard, but I think there's a multi-stage process for reassessing those job numbers, and it's important to understand that within the program, the job numbers weren't actually an indicator used to choose which projects were or were not funded. They were indicators of the outcome afterwards.
There are two stages to calculating the number of jobs. There are initial assessments from the companies, and then, as we close out files, they're further evaluated. With respect to the core information used in evaluation of the projects—their costs and the detailed analysis in terms of whether they were eligible and whether or not they'd received funding from other things—I think we had a fair degree of accuracy in terms of those discussions.