We’re all agreed that there’s a housing crisis in Canada. And there’s a fairly broad consensus on at least some of the contributing factors, including out-of-control immigration and weak construction numbers. But I, for one, don’t think that anyone has yet presented a solid, comprehensive solution - including the Liberal campaign’s truly terrifying proposal to pour billions of dollars into building and managing public housing (on top of the billions of dollars they’ve already fruitlessly sunk into programs like their National Housing Strategy Act).
Part of the problem is that there might not be a solid, comprehensive solution - even taking into account my own survey of smart approaches from a few months back.
But is there any way for governments to make themselves even minimally useful? Given that there isn’t going to be any single magic bullet here, I’m going to suggest that there might just be something fast and cheap where they could make a difference. And the inspiration for this idea came from a recent UK failure.
The people behind the Tract project recently released a valuable postmortem documenting the collapse of their startup. The idea behind Tract involved harnessing software tools and AI integrations to allow developers to:
Identify and secure land with long-term development potential before it received planning permission
Identify small, overlooked sites, secure planning permission and, optionally, off-load them to next-stage developers
Integrate project elements with map layers with planning data
Assist planning consultants and developers in drafting planning and application documents
Provide an integrated database of planning information including site constraints, local policies, past applications
In addition, Tract built a free application for landowners to encourage them to reveal their openness to development, thereby generating leads for developers or Tract itself as a tech-enabled land promoter.
The Tract postmortem identified problems that mostly came down to unexpectedly weak consumer demand, existing market segment competition, and overambitious spending. But the technology itself seems to have worked just fine.
The goals driving Tract are actually compelling, and particularly in the Canadian context. For instance, a Tract-like application could analyze aggregated data from MLS, municipal zoning, and land registries to identify underutilized or overlooked sites faster than manual processes. This could unlock land for small and medium-sized developers, who are often priced out of prime markets. Identifying 1,000 viable sites in just the GTA, each yielding 10-20 units, could add 10,000-20,000 homes annually.
Automating document drafting could cut weeks or months off pre-application and full permitting timelines. For instance, generating pre-app statements in hours rather than days could reduce the median approval time (currently 12-18 months in urban centers) by 10-20%. Shaving 3 months off 5,000 applications could bring 50,000 units (assuming 10 units per project) to market sooner, easing supply pressure.
Free or low-cost appraisals and automated document creation could lower entry barriers for smaller players, reducing reliance on expensive consultants (often $500-$1,000 per appraisal in Canada). This cost-saving could make marginal projects viable.
Is this post about subtly fishing for investors for my new stealth startup? Perish the thought! In the immortal words of Samuel Goldwyn, “you can include me out” of that one. But why couldn’t such an application exist as a public utility, complete with government-enabled integrations into official data sources like CMHC backends? And why couldn’t it use a cheap(ish) off-the-shelf AI model like Grok or Claude rather than having to customize something new?
Just assemble a handful of AI-friendly engineers and a project manager who’s unspoiled by public service, set them up with key contacts inside various levels of government, and give them a few months’ access to some empty office space. There’s a non-zero chance that a useful, publicly-available, housing project incubator app will come out the other end. And the costs would, by government standards, be trivial.
Why shouldn’t government resources be put to work in the service of the private sector? It’s not like it’s never happened before: what do you think Industry Canada is for?
As someone not from the GTA or the lower mainland / Vancouver area I find, rightly or wrongly, that much of this home shortage conversation is driven solely by the problems in these two geographic areas. It appears to be more of a big city problem which therefore needs big city solutions. Those typically revolve around densification as the solution and ramming as many people as possible into the smallest possible area. What frequently gets lost is even when these social engineering projects proceed unless they are funded to a large extent by tax payer dollars and subsidies the end result still does not result in any great improvement.
Well, as long as the developers hired for that housing incubator app aren’t anyone who had anything whatsoever to do with ArriveCan. Because that could end up being waaaay more expensive than you suggest.