The per-unit costs will be lower. The value of redeveloped land only goes up when its bought out to build density. If you currently live a mile out of the urban core in some sprawl mcmansion your house is probably worth 5 million right now in a lot of cities. If the city proper dropped zoning requirements and let density happen the price of your house will vary between the "loss in desirability" from being closer to an expanding urban core and the potential value of redeveloping your house. Your lot could end up being worth an order of magnitude more if the demand is there to put a high rise on it.

Some will win huge as major contractors want to replace your quarter acre sprawl development or row house with 10+ story urban and are offering absurd amounts of money for all the lots. Some will lose if nobody has the capital to redevelop their area and density goes up around it. But if you get the ball rolling there is so much pent up demand for urban migration and sustainable living that we could probably go decades of continuous development in most cities without the money drying up. Its an untapped gold mine of economic growth.

This is only true if an increase in the housing supply does not lead to an increase in the population. At a national level, this is probably true enough to assume. But, at the level of individual cities, it is not nessaserrily true. I can easily imagine scenarios where an increase in housing causes an increase in the local population greater than the increase in housing capacity.

For increase supply to have no effect on price, the demand curve has to be highly inelastic, but this just isn't true. People very much do make decisions about where to live based on the cost.

At the margin, people may be more likely to move if you build more housing, but this would be because they can get a better deal on housing, not because there is some abstract idea of more housing.

Without a reduction in price/increase in housing quality, there is no mechanism that would cause more people to show up.

The main exception to this is if you have rent control capping prices, but this isn't really a big part of any market where this discussion is taking place.

I am thinking more in terms of lagging responses.

For instance, suppose we increased supply 10% today, and prices immidietly fall. Seeing the lower prices, bussiness begin making plans to open in the new city to take advantage of the lower COL workforce. In total, these plans would increase the demand for labor by 20% and take 5 years to be implemented.

At some point in those 5 years, people will have moved into the new housing driving the costs back up to their original levels.

In an ideal market, bussiness would stop moving in, as their original motivation to do so is gone. However, the market is not ideal. Plans had already been written, and the moving has already begun, so the demand for labor will increase even further, encouraging more people to move in, driving up the costs of housing.

I mean, rents won't immediately fall. We're incapable of producing that much housing quickly. We're limited by all sorts of construction infrastructure. It's going to take a decade to crawl out of this hole in the best case.

This also assumes that these hypothetical employers would be paying the same amount as the employers who were there before. Which is unlikely because the existing employers were already there paying increased wages, vs the new employers who waited for falling CoL/salaries to move in.

Also, raising salaries is often not enough to offset CoL expenses, since not all employees have the same CoL expenses, e.g. families, and those with higher CoL expenses will again move away / won't come to these jobs.

Also, in practice, the amount of jobs is not the binding factor in, e.g. the bay area. There are plenty of jobs out there, people are just a lot less willing to move there due to the insane CoL/shitty commute.

You could have some perfect storm where the induced demand perfectly offsets the new housing, but if you assume there is some elasticity at all points in the system, I think the default assumption should be that increased supply will drive down prices for everyone.