There's another thing that makes a revenue neutral shift to LVT -- cutting other taxes on labor and capital -- even more tantalizing. That is, there's some evidence to suggest it's at least partly the case that All Taxes Come Out of Rents (ATCOR for short). For example:
Meaning that wage tax, BIRT, and tax on buildings reduces the land rents, ie, the amount of money we could collect via LVT. If you cut those taxes on labor and capital, the land rents we collect should increase. Intuitively, cutting those taxes will make Philly a more attractive place to live and work, and hence make land more valuable here.
>Suburban commuters who use city services but pay minimal city taxes
Might be worth clarifying that since they don't vote here, pols aren't worried about their votes; they're worried that they'd no longer want to commute and work here. I think pols are wrong about that, of course.
100%! Inherently landlords will raise rents to the upper limit of folks' income (which is ad infinitum in a housing crisis, because people have to live somewhere). Not a lot of room in someone's finances to deprioritize a place to live (vs nowhere to live).
Re:suburbanites, totally with you that politicians are catering to them for their money, not their votes—but that's enough for them to make decisions that make it both a) worse for people living here, and b) worse for everyone commuting in, actually.
We're of the same mind on this, of course—with any luck, some of the more with-it CMs will take it and run with it.
@JamieGauthier
@IsaiahThomas
@NicolasORourke
Alas, these not-actual-tags are but aspirational. One day, Russell. One day!
Totally! Given that the tax rate is uniform between land and property in present state, the 20% rate used by the city is, in a word, arbitrary—primarily because, regardless of the breakdown (60/40, 80/20, 100/0) between building and land on a $200,000 property, the tax shakes out to the same number.
Obviously, there are scenarios within the current setup where the building/land split is relevant—eg: when comparing the same building in two different parts of the city—but for the purposes of calculating LVT, what's most important is the magnitude of difference between the lowest- and highest-valued areas in the city.
To put it differently, it'd be of little consequence what that split is when hashing out the details of a prospective LVT rate—if we take the 40% number per the example, great! If we take 20%, we double the rate like you said. The end bill to the property owner would be the same for the purposes of the hypothetical. Once we're no longer considering the building for calculating taxes, we don't need to know the relative value building-to-land.
As noted above, what's of more significance is the multiplier between higher- and lower-valued plots in the city. That is, the per-square-foot value between Center City and less-productive sections of the city. That is what determines how much pressure an owner in each part of the city feels to optimize efficiency by building more on the plot.
We would of course need to firm up those numbers in a transition to an LVT-only world, but for illustrative purposes, we can work in this hypothetical with the understanding that the rate could be set to achieve a goal revenue for the city. Ultimately, it's a calculation to balance revenue with ensuring the difference between higher- and lower-valued parts of the city meets folks' needs.
Love it!
There's another thing that makes a revenue neutral shift to LVT -- cutting other taxes on labor and capital -- even more tantalizing. That is, there's some evidence to suggest it's at least partly the case that All Taxes Come Out of Rents (ATCOR for short). For example:
https://www.reddit.com/r/BasicIncome/comments/1zcmyb/all_taxes_come_out_of_rent_and_rent_absorbs_all/
https://schalkenbach.org/rsf-responds-to-the-new-york-times-article-titled-the-georgists-are-out-there-and-they-want-to-tax-your-land/ (ctrl f for atcor)
Meaning that wage tax, BIRT, and tax on buildings reduces the land rents, ie, the amount of money we could collect via LVT. If you cut those taxes on labor and capital, the land rents we collect should increase. Intuitively, cutting those taxes will make Philly a more attractive place to live and work, and hence make land more valuable here.
>Suburban commuters who use city services but pay minimal city taxes
Might be worth clarifying that since they don't vote here, pols aren't worried about their votes; they're worried that they'd no longer want to commute and work here. I think pols are wrong about that, of course.
100%! Inherently landlords will raise rents to the upper limit of folks' income (which is ad infinitum in a housing crisis, because people have to live somewhere). Not a lot of room in someone's finances to deprioritize a place to live (vs nowhere to live).
Re:suburbanites, totally with you that politicians are catering to them for their money, not their votes—but that's enough for them to make decisions that make it both a) worse for people living here, and b) worse for everyone commuting in, actually.
We're of the same mind on this, of course—with any luck, some of the more with-it CMs will take it and run with it.
@JamieGauthier
@IsaiahThomas
@NicolasORourke
Alas, these not-actual-tags are but aspirational. One day, Russell. One day!
Also
>Yes, yes—I'm still on my LVT bender.
Whomst amongst us. :-)
Land probably doesn’t represent 40% of property value unless it’s much higher than 40% in Center City.
OPA estimates 20% citywide.
Take typical $575,000 new townhouse, with no garage. Land is roughly $130,000 or 22%. So you’d need to double your tax rate.
Totally! Given that the tax rate is uniform between land and property in present state, the 20% rate used by the city is, in a word, arbitrary—primarily because, regardless of the breakdown (60/40, 80/20, 100/0) between building and land on a $200,000 property, the tax shakes out to the same number.
Obviously, there are scenarios within the current setup where the building/land split is relevant—eg: when comparing the same building in two different parts of the city—but for the purposes of calculating LVT, what's most important is the magnitude of difference between the lowest- and highest-valued areas in the city.
To put it differently, it'd be of little consequence what that split is when hashing out the details of a prospective LVT rate—if we take the 40% number per the example, great! If we take 20%, we double the rate like you said. The end bill to the property owner would be the same for the purposes of the hypothetical. Once we're no longer considering the building for calculating taxes, we don't need to know the relative value building-to-land.
As noted above, what's of more significance is the multiplier between higher- and lower-valued plots in the city. That is, the per-square-foot value between Center City and less-productive sections of the city. That is what determines how much pressure an owner in each part of the city feels to optimize efficiency by building more on the plot.
We would of course need to firm up those numbers in a transition to an LVT-only world, but for illustrative purposes, we can work in this hypothetical with the understanding that the rate could be set to achieve a goal revenue for the city. Ultimately, it's a calculation to balance revenue with ensuring the difference between higher- and lower-valued parts of the city meets folks' needs.