If I'm understanding Alexander's argument, he believes that a denser population itself makes a city even more desirable to live in. To me, that sounds like a positive externality, which strengthens the argument for policies which encourage more housing in dense areas. Is that a sensible way of thinking about agglomeration, or is that already "baked in" to the upwards sloping demand curve you drew?
In my example, it’s a positive externality too. My moving there increases the value for you.
The problem with just drawing it as an externality is that, policy changes aside, in that world, building more housing decreases prices. That’s not Scott’s argument.
I think this argument is unnecessarily complicated. A denser Oakland with the better amenities Scott assumes--greater consumption variety, more walkable layout, agglomeration effects on productivity--is not the same good as it was prior to the density. Improving the quality of the good mid-discussion is not ceteris paribus, which is a crucial assumption for the partial equilibrium analysis most are doing when they discuss more supply leads to lower prices. Of course there could be general equilibrium affects that *appear* to raise the price of the same good, but that is caused by secondary effects that feedback into the initial analysis, some of which can even change the good in question (as with neighborhoods). A denser Oakland is a different Oakland (if amenities increase as Scott assumes), just like an iphone 14 is not the original iphone. No upward sloping demand curves are needed to explain any of this.
If you're going to go that route, than you can't even construct the demand curve for apples in the way we explain it to students, since after you've purchased/eaten one, the second one is a different good.
Also, this is a market level analysis. By construction, it is what the analyst is calling "the same good." No two people define goods the exact same. I choose to define the housing by objective measures, since that's what I have.
I agree it is a bit convoluted. But I wanted to find a way to explicitly separate out the amenity increases from (as the quote I pull says) building housing. You have to say "both curves are shifting, even though Scott says the cause is the building" or you have to say "these are two different goods, so actually the price didn't change at all" even though Scott is arguing the PRICE changed.
Maybe this is evidence of Scott's confusion but I'm presenting a way to square Scott with supply and demand.
I think we agree that Scott's analysis is confused. I disagree with your apple example, but we can get into that another time.
Back to Scott's housing example, I'd actually hypothesize that he gets the price effects reversed. A marginal increase in housing in a market like Oakland (MSA or city,? not clear so I'll assume city), which has say about 180K units (167K households according to Census, assuming each household has its own unit and there are some vacant https://www.census.gov/quickfacts/fact/table/oaklandcitycalifornia/AFN120217) would likely not change any amenities. 10K new units is about a 6% increase in the housing stock. Spread over the entire city, I doubt much changes e.g., infrastructure, parks, services, retail, etc.
On the other hand, doubling the housing stock in a short period of time, holding city limits constant, would substantially change the kind of city Oakland is. Much denser, more walkable, more congestion, more stores, variety, agglomeration/productivity changes etc. Such an increase is when Scott's amenity argument would matter if in fact density improves amenities and desirability on net. Not when the supply increases by 10K.
I don't think upward-sloping demand is the argument he is making. Rather, the way I interpret him is that increased density will cause a shift of the D curve to a D' curve, and that the intersection of D' and S' will be at a higher price than that between D and S. This would imply that the price would initially drop (to the intersection of S' and D) after an increase in supply, but once increased density had shifted the demand curve, the original price would be surpassed.
Maybe. But the whole point of having the separate curves is to be able to talk about them shifting separately. Then we can shift both if we want.
In my framing, I’ve already built in the demand response into the slope of the curve, instead of shifting that too. So the cause is the supply shock. In your example, both MUST shift so we aren’t able isolate a cause. We are left with the ambiguity of the first quote I use.
The difference is in the timing. The empirical research that's been referenced in the discussion suggests that in the short term, prices drop when supply goes up, as one would expect, so it seems natural to posit a downward-sloping demand curve. The increased density then enables higher productivity and increases the number of people who consider the locality their home, which shifts the demand curve, but this is not instantaneous, and there is no guarantee that these second-order effects on demand will outweigh the first-order effect on supply.
Maybe, but even in the long term, assuming density is the driving factor behind price makes some history very confusing! Short of a natural disaster (or some other exogenous force that reduces supply), why would a city ever shrink? Pick any city in the rust belt. Or try to explain why Manhattan population peaked in 1920, but prices are up.
To wade in on the impact of which effect dominates - this is from our discussion of housing:
"Anagol, Ferreira and Rexer (2023) looked at recent zoning reform in Sao Paulo, Brazil. Sao Paulo is the 4th largest metropolitan area in the world with a population of 21mln. A reform in 2016 allowed for a larger building density on each block – on average, this allowed for a 36% increase in building construction on a given lot. The authors found that this reform increased housing supply by about 1.9% and reduced home prices by 0.5% in Sao Paulo."
So densfication had limited impacts on prices. I think what's most important to look at regarding housing affordability solutions is actually how does welfare change. Once we start looking at welfare changes, it is understandable why upzoning will face significant political push back.
"We develop a framework to estimate the economic value of a recent zoning reform in the city of São Paulo, which altered maximum permitted construction at the city-block level. Using a spatial regression discontinuity design, we show that developers respond to the reform with short-run increases in filings for multi-family construction permits in blocks with higher allowable densities. In the medium-run we observe increases in housing availability and reductions in house prices in neighborhoods that were allowed more densification. Welfare is then estimated with an equilibrium model of housing demand and supply that allows for endogeneous housing regulation. We finding that, in the long-run, the reform produces a 1.9% increase in housing stock and a 0.5% reduction in prices, with substantial heterogeneity across neighborhoods. Consumer welfare gains due to price reductions are small, but increase 4-fold once accounting for changes in built environment, with more gains accruing to college educated and higher income families. However, nominal house price losses faced by landlords and existing homeowners overshadow all consumer welfare gains."
At the end of it all, the measurement of welfare is important. Naturally, welfare measures are subjective (i.e. whose welfare, how do weigh everyone) but it is important in terms of which solutions would work.
One solution that we wrote about that might be Pareto improving is developing transit lines.
One interesting thing about Scott's argument (which I didn't notice initially) is that the plots of housing cost vs density actually normalize housing cost by income. This is curious because income increases with higher density because of agglomeration effects. If the argument is that agglomeration effects explain why higher density leads to higher housing costs, then I think you have to be claiming that agglomeration effects are stronger for housing cost than for income.
Which, I guess, is possible.
But a much simpler explanation seems to be that there's an interaction between housing regulation and density/growth. If your city isn't dense and it's not growing, then it doesn't really matter if all you can build is single-family homes, and housing would still be cheap. But if you're dense and growing *and* you can't build, then your price would be going up a lot.
The Los Angeles urbanized area has been America's densest for years. But this is an average density for a large area. The point is that all the densities being discussed are averages. And they are misleading.
If I'm understanding Alexander's argument, he believes that a denser population itself makes a city even more desirable to live in. To me, that sounds like a positive externality, which strengthens the argument for policies which encourage more housing in dense areas. Is that a sensible way of thinking about agglomeration, or is that already "baked in" to the upwards sloping demand curve you drew?
In my example, it’s a positive externality too. My moving there increases the value for you.
The problem with just drawing it as an externality is that, policy changes aside, in that world, building more housing decreases prices. That’s not Scott’s argument.
I think this argument is unnecessarily complicated. A denser Oakland with the better amenities Scott assumes--greater consumption variety, more walkable layout, agglomeration effects on productivity--is not the same good as it was prior to the density. Improving the quality of the good mid-discussion is not ceteris paribus, which is a crucial assumption for the partial equilibrium analysis most are doing when they discuss more supply leads to lower prices. Of course there could be general equilibrium affects that *appear* to raise the price of the same good, but that is caused by secondary effects that feedback into the initial analysis, some of which can even change the good in question (as with neighborhoods). A denser Oakland is a different Oakland (if amenities increase as Scott assumes), just like an iphone 14 is not the original iphone. No upward sloping demand curves are needed to explain any of this.
If you're going to go that route, than you can't even construct the demand curve for apples in the way we explain it to students, since after you've purchased/eaten one, the second one is a different good.
Also, this is a market level analysis. By construction, it is what the analyst is calling "the same good." No two people define goods the exact same. I choose to define the housing by objective measures, since that's what I have.
I agree it is a bit convoluted. But I wanted to find a way to explicitly separate out the amenity increases from (as the quote I pull says) building housing. You have to say "both curves are shifting, even though Scott says the cause is the building" or you have to say "these are two different goods, so actually the price didn't change at all" even though Scott is arguing the PRICE changed.
Maybe this is evidence of Scott's confusion but I'm presenting a way to square Scott with supply and demand.
I think we agree that Scott's analysis is confused. I disagree with your apple example, but we can get into that another time.
Back to Scott's housing example, I'd actually hypothesize that he gets the price effects reversed. A marginal increase in housing in a market like Oakland (MSA or city,? not clear so I'll assume city), which has say about 180K units (167K households according to Census, assuming each household has its own unit and there are some vacant https://www.census.gov/quickfacts/fact/table/oaklandcitycalifornia/AFN120217) would likely not change any amenities. 10K new units is about a 6% increase in the housing stock. Spread over the entire city, I doubt much changes e.g., infrastructure, parks, services, retail, etc.
On the other hand, doubling the housing stock in a short period of time, holding city limits constant, would substantially change the kind of city Oakland is. Much denser, more walkable, more congestion, more stores, variety, agglomeration/productivity changes etc. Such an increase is when Scott's amenity argument would matter if in fact density improves amenities and desirability on net. Not when the supply increases by 10K.
I don't think upward-sloping demand is the argument he is making. Rather, the way I interpret him is that increased density will cause a shift of the D curve to a D' curve, and that the intersection of D' and S' will be at a higher price than that between D and S. This would imply that the price would initially drop (to the intersection of S' and D) after an increase in supply, but once increased density had shifted the demand curve, the original price would be surpassed.
Maybe. But the whole point of having the separate curves is to be able to talk about them shifting separately. Then we can shift both if we want.
In my framing, I’ve already built in the demand response into the slope of the curve, instead of shifting that too. So the cause is the supply shock. In your example, both MUST shift so we aren’t able isolate a cause. We are left with the ambiguity of the first quote I use.
The difference is in the timing. The empirical research that's been referenced in the discussion suggests that in the short term, prices drop when supply goes up, as one would expect, so it seems natural to posit a downward-sloping demand curve. The increased density then enables higher productivity and increases the number of people who consider the locality their home, which shifts the demand curve, but this is not instantaneous, and there is no guarantee that these second-order effects on demand will outweigh the first-order effect on supply.
That's fair. But then that just means I'm talking about the long run and so are people like Scott.
Maybe, but even in the long term, assuming density is the driving factor behind price makes some history very confusing! Short of a natural disaster (or some other exogenous force that reduces supply), why would a city ever shrink? Pick any city in the rust belt. Or try to explain why Manhattan population peaked in 1920, but prices are up.
To wade in on the impact of which effect dominates - this is from our discussion of housing:
"Anagol, Ferreira and Rexer (2023) looked at recent zoning reform in Sao Paulo, Brazil. Sao Paulo is the 4th largest metropolitan area in the world with a population of 21mln. A reform in 2016 allowed for a larger building density on each block – on average, this allowed for a 36% increase in building construction on a given lot. The authors found that this reform increased housing supply by about 1.9% and reduced home prices by 0.5% in Sao Paulo."
So densfication had limited impacts on prices. I think what's most important to look at regarding housing affordability solutions is actually how does welfare change. Once we start looking at welfare changes, it is understandable why upzoning will face significant political push back.
What estimates of changes to welfare are you referring to?
For example from the Anagol et al. paper -
"We develop a framework to estimate the economic value of a recent zoning reform in the city of São Paulo, which altered maximum permitted construction at the city-block level. Using a spatial regression discontinuity design, we show that developers respond to the reform with short-run increases in filings for multi-family construction permits in blocks with higher allowable densities. In the medium-run we observe increases in housing availability and reductions in house prices in neighborhoods that were allowed more densification. Welfare is then estimated with an equilibrium model of housing demand and supply that allows for endogeneous housing regulation. We finding that, in the long-run, the reform produces a 1.9% increase in housing stock and a 0.5% reduction in prices, with substantial heterogeneity across neighborhoods. Consumer welfare gains due to price reductions are small, but increase 4-fold once accounting for changes in built environment, with more gains accruing to college educated and higher income families. However, nominal house price losses faced by landlords and existing homeowners overshadow all consumer welfare gains."
At the end of it all, the measurement of welfare is important. Naturally, welfare measures are subjective (i.e. whose welfare, how do weigh everyone) but it is important in terms of which solutions would work.
One solution that we wrote about that might be Pareto improving is developing transit lines.
One interesting thing about Scott's argument (which I didn't notice initially) is that the plots of housing cost vs density actually normalize housing cost by income. This is curious because income increases with higher density because of agglomeration effects. If the argument is that agglomeration effects explain why higher density leads to higher housing costs, then I think you have to be claiming that agglomeration effects are stronger for housing cost than for income.
Which, I guess, is possible.
But a much simpler explanation seems to be that there's an interaction between housing regulation and density/growth. If your city isn't dense and it's not growing, then it doesn't really matter if all you can build is single-family homes, and housing would still be cheap. But if you're dense and growing *and* you can't build, then your price would be going up a lot.
The Los Angeles urbanized area has been America's densest for years. But this is an average density for a large area. The point is that all the densities being discussed are averages. And they are misleading.