The public and Catholic school boards in Hamilton have closed and merged several primary schools between 2006 and 2016.
This is an effect of government policy that provides provincial funding for new schools in urban periphery growth areas, contingent on the closure of under-capacity schools in urban core areas.
Core schools are under-capacity simply due to the long-run (multi-decades) decline in fertility in the core.
If we want to later examine the effect of school closures on house prices, there's an interesting question to address first: should a school closure lower local house prices, or raise it? Hedonics says the former, but it's not so simple.
It's been suggested to us that, in core neighbourhoods undergoing gentrification, house prices will be going up at a faster rate than the city average (the houses were cheap, now their prices are catching up). If the gentrifying population moving in is young singles and couples with no children, then that neighbourhood's children-per-unit-area will be decreasing, thus their school is more likely to close; but because of gentrification we'll actually see prices rise faster than average! Thus, we'd find that school closures boost house prices, which is the opposite of what hedonic theory would suggest.
Now, there's a complication: what exactly is gentrification? Yes, it's a change in average income of a neighbourhood, as the working class is replaced by wealthier residents. But:
There are other neighbourhoods in Hamilton that are also working class, but mostly senior residents with no school-age children left. These areas are also gentrifying right now - but with an influx of professional families with children, moving into town from the GTA.
In those neighbourhoods, we'll be seeing a rise in children-per-unit-area, meaning less likelihood of school closure, with house prices increasing faster than average. In that case, we should see a result consonant with hedonic theory. But obfuscated.
So there's the problem: a simple hedonic analysis will have to take into account the different populations of neighbourhoods. Nothing is as simple as economic theory says.
The code in this repository will hopefully eventually investigate this, maybe perhaps. If I can just get started.