Maybe I could pick one as an example, although we're trying to take a similar measurement approach for all the programs in terms of better achieving the outcome. The example I'd give is the ASETS program, which is focused on indigenous training and achieving better outcomes in indigenous training in terms of employment.
The challenge we've had in that program is a measurement challenge we often have with programs. We will fund training, but when we're trying to identify the medium-term impact, say over three or four years, it isn't possible for us to go back to the training provider and say, “Well, what's happened to that individual three or four years later?” because they will have lost track of the individual, as would be true of the people we fund under the ASETS program.
We can see short-term impacts: for example, did they successfully complete the training; did they get employment? But we can't identify whether their salary improved over time or over what time horizon.
With the ASETS program, we're using the SIN to be able to access the CRA database. This is done on an anonymized basis. We will not know who the individuals are. We will be able to take all of the individuals who've accessed the program under the ASETS program and then identify what happened to their income levels two, three, four and five years later. We will not be able to report to Canadians on exactly what the results were for each individual, because that would breach the confidentiality of the SIN data, but we will be able to identify which types of training, for which types of individuals, led to which types of better outcomes. That will allow us to have a feedback loop where we can then work with indigenous partners to promote those programs that had better outcomes and move resources away from programs that had weaker outcomes.
It's this medium-term data that is one of our real challenges, just measuring the employment results and where people are in six months. They may be able to get a job, but it doesn't mean they're going to stay employed for a longer period, and we can't do that.
That's one of the examples of how we're doing it in terms of results measurements. I don't know if there are other examples folks want to give.