The graph shows the new dwellings constructed per new person (per person of population growth). We do notice a recent decline in the number of dwellings being constructed nationwide compared to the population growth, which is reflected in the later graph showing increased occupancy rates. The direction of causation amongst these variables remains unclear, and in all likelihood, they are interdependent.
Regression with net new dwellings per person of population growth as an explanatory variable for change in the capital city price index gives a negative coefficient (-0.011) but really, has no explanatory power (r2 of 0.006).
That means that analysis of population growth and dwelling construction figures has no power in explaining housing price changes.