Perhaps I could try to provide some context around this.
Statistics Canada doesn't have statistics for poverty. It has statistics for low income. We produce three different measures of low income, and they each have different perspectives on what it means to be in low income.
The ones that were presented to you today by Professor Veall and by Sébastien were relative low-income rates. The idea there is that low-income thresholds are indexed to the median standard of living of the population.
The relevant question, then, is whether or not the median standard of living of the population rises or falls relative to prices. Arithmetically, if the median income of the population were to increase faster than prices and OAS were only indexed to prices, then OAS would represent a smaller portion of the median. If, however, the median were to rise more slowly, then the OAS could increase relative to the median.