This article from Strong Towns has a good explanation of why low-density development is not the answer to the housing supply issue.

this style of development works extremely well for a specific type of private developer… developers like Ross Perot Jr. are masters of the assembly-line approach: secure cheap land on the fringe, install infrastructure, and build tract housing as quickly as possible. At this scale, the profits are enormous, and the risks are low. The federal government provides generous support through mortgage guarantees, tax preferences, and highway spending, and buyers keep lining up for new homes.

But while the private sector gets the cash, local governments get the bill. Sprawling developments create long-term infrastructure liabilities—roads, water lines, sewer systems, schools, fire protection—that far exceed the revenue they generate. Local governments, which are really just collections of us acting together, are left trying to maintain and operate systems that are fundamentally unaffordable.

As Mayor Eugene Escobar of Princeton, Texas, put it, his town boomed with affordable homes, but now it’s struggling with traffic, overburdened infrastructure, and a lack of basic amenities. The city’s leadership is trying to build a real downtown, attract jobs, and create public spaces—but they’re doing it after the fact. That’s not planning. That’s triage.

Some suburbs seem to persist for long periods of time. But they are ones located within commuting distance of urban centers with high-paying professional jobs, and the zoning serves to keep the median income in those successful suburbs very high, and therefore able to support the very high infrastructure costs per resident or per square mile. There aren’t enough of these highly affluent people for all suburbs to work like this, so for every successful one there are many turning into slums.

What seems to be suggested instead is a gradual process of intensification from the middle out, so that populations, incomes, and tax revenues can keep rising over time as value is continuously created. This makes sense to me. I think there may be a more linear model though that could work for U.S. suburbs, where the intensification happens along a transportation corridor with progressively less dense development as you move back from highway. This way, you get a long linear downtown with access to transportation and other infrastructure at a low unit cost. People could live in relatively low-density neighborhoods if they want to and still not be too far from work, school, and inter-city public transportation. And these commercial corridors already exist in the form of arterial highways, water and power lines, big box stores and car dealerships separated by oceans of parking.