I like this post and would like to see it curated, conditional on the idea actually being good. There are a few places where I’d want more details about the world before knowing if this was true.
Who owns this land? I’m guessing this is part of the Guadalupe Watershed, though I’m not sure how I’d confirm that.
I don’t know what that means, but it might be important.
How much does it cost to fill in land like this?
It looks like for pool removal there’s a cost of between $20-$130 per cubic foot yard (thanks johnswentworth). Making the bad simplifying assumption of 6ft of depth and 50 square miles that’s 8.3 billion ft^3 310 million cubic yards. Since the state of CA is very bad at cutting costs, let’s use the high end cost estimate which is about 1⁄8 of $1000 so that makes the cost estimate $1 trillion $300 billion.
With a trillion dollar price tag, this stops looking worthwhile pretty fast.
Spitballing about price estimates:
People have filled in things like this in the past, which suggests lower costs
Human effort may be much more expensive than it was previously
pool filling prices might include massive fixed costs and regulatory costs that wouldn’t scale with volume
The state could auction the land to a private company that might do a better job negotiating costs
If fixed costs are 90% of pool fillings and will be negligible by volume for this, and if we further use the lower bound of cost per filling, then we reduce cost by 60x to about $5 billion. Let’s call that an 80% confidence interval, where the low end is clearly worth it and the high end clearly not.
How much does it cost to build a bunch of housing there?
First Google result says $65k-86k per unit, though economies of scale might bring that down. Then the suggested 2 million units would cost ~$130-170 billion; potentially significantly more or less.
How much value does the housing create?
The cheapest rents I could see with a casual search was something around $900/bedroom/month in Fremont.
Rounding up to $11k/year, it would take 6-8 years to recoup construction costs, not counting maintenance.
At the low end of land filling costs, $16 billion, adds less than one year to the recoup timeline. At the high end around $1 trillion, it would take about 50 years to recoup the costs. $300 billion, that ~triples to ~20 years.
Reaching the end of this, I think I’m uncertain about how economical the idea is. This is mostly because of large error bars around my cost calculations.
An investment that pays off in value created 50 years down the line is probably worth it for society, but very unlikely to happen given the investment environment today.
My ending impression is I want this post curated, because I want city managers and real estate investors to run these numbers (ideally being nerd-sniped by my terrible naïve calculations) and make the decision for themselves.
It looks like for pool removal there’s a cost of between $20-$130 per cubic foot.
Minor correction: that source says filling in a pool is $20-80 per cubic yard, which would only be ~$1-3 per cubic foot. The higher numbers are for demolition, but that’s presumably dominated by the cost of the demolition rather than the fill—jackhammers are a pain in the ass.
It looks to me like you’re comparing the cost of constructing a unit to the price of renting a bedroom?
I don’t think you’re counting the cost of infrastructure construction?
Bulk fill is massively cheaper than the quantities you would use for pool removal.
Since this is a seismic area you can’t just use bulk fill: you need to do some amount of reinforcement/stabilization.
You’re talking about this as if the government would pay for all construction and then own all units and rent them out? I agree that governments are not typically interested in that kind of financing, especially in the US. If you sell the units, or the buildings, however, there is a thriving real estate market that does operate on these time scales.
I was trying to do a back of the envelope calculations of total cost of work and total value created (where I’m using cost of rent as a (bad) proxy for (capturable) value created).
I definitely wouldn’t assume that the government or any single agent would be doing the project, just that the overall amount of capturable value must be worth it for the investment costs, then different parties can pay portions of those costs in exchange for portions of or rights to that value, but I doubt adding in the different parties involved would make my estimates more accurate.
Do you have a source for cost of similar projects? My estimates are definitely very bad for many reasons.
I like this post and would like to see it curated, conditional on the idea actually being good. There are a few places where I’d want more details about the world before knowing if this was true.
Who owns this land? I’m guessing this is part of the Guadalupe Watershed, though I’m not sure how I’d confirm that.
What legal limits are there on use of the land? Wikipedia notes:
I don’t know what that means, but it might be important.
How much does it cost to fill in land like this?
It looks like for pool removal there’s a cost of between $20-$130 per cubic
footyard (thanks johnswentworth). Making the bad simplifying assumption of 6ft of depth and 50 square miles that’s8.3 billion ft^3310 million cubic yards. Since the state of CA is very bad at cutting costs, let’s use the high end cost estimate which is about 1⁄8 of $1000 so that makes the cost estimate$1 trillion$300 billion.With a trillion dollar price tag, this stops looking worthwhile pretty fast.
Spitballing about price estimates:
People have filled in things like this in the past, which suggests lower costs
Human effort may be much more expensive than it was previously
pool filling prices might include massive fixed costs and regulatory costs that wouldn’t scale with volume
The state could auction the land to a private company that might do a better job negotiating costs
If fixed costs are 90% of pool fillings and will be negligible by volume for this, and if we further use the lower bound of cost per filling, then we reduce cost by 60x to about $5 billion. Let’s call that an 80% confidence interval, where the low end is clearly worth it and the high end clearly not.
How much does it cost to build a bunch of housing there?
First Google result says $65k-86k per unit, though economies of scale might bring that down. Then the suggested 2 million units would cost ~$130-170 billion; potentially significantly more or less.
How much value does the housing create?
The cheapest rents I could see with a casual search was something around $900/bedroom/month in Fremont.
Rounding up to $11k/year, it would take 6-8 years to recoup construction costs, not counting maintenance.
At the low end of land filling costs, $16 billion, adds less than one year to the recoup timeline. At the high end around
$1 trillion, it would take about 50 years to recoup the costs.$300 billion, that ~triples to ~20 years.Reaching the end of this, I think I’m uncertain about how economical the idea is. This is mostly because of large error bars around my cost calculations.
An investment that pays off in value created 50 years down the line is probably worth it for society, but very unlikely to happen given the investment environment today.
My ending impression is I want this post curated, because I want city managers and real estate investors to run these numbers (ideally being nerd-sniped by my terrible naïve calculations) and make the decision for themselves.
Minor correction: that source says filling in a pool is $20-80 per cubic yard, which would only be ~$1-3 per cubic foot. The higher numbers are for demolition, but that’s presumably dominated by the cost of the demolition rather than the fill—jackhammers are a pain in the ass.
Thanks! Updated.
It looks to me like you’re comparing the cost of constructing a unit to the price of renting a bedroom?
I don’t think you’re counting the cost of infrastructure construction?
Bulk fill is massively cheaper than the quantities you would use for pool removal.
Since this is a seismic area you can’t just use bulk fill: you need to do some amount of reinforcement/stabilization.
You’re talking about this as if the government would pay for all construction and then own all units and rent them out? I agree that governments are not typically interested in that kind of financing, especially in the US. If you sell the units, or the buildings, however, there is a thriving real estate market that does operate on these time scales.
I was trying to do a back of the envelope calculations of total cost of work and total value created (where I’m using cost of rent as a (bad) proxy for (capturable) value created).
I definitely wouldn’t assume that the government or any single agent would be doing the project, just that the overall amount of capturable value must be worth it for the investment costs, then different parties can pay portions of those costs in exchange for portions of or rights to that value, but I doubt adding in the different parties involved would make my estimates more accurate.
Do you have a source for cost of similar projects? My estimates are definitely very bad for many reasons.