Reply to Overriding local zoning to allow multi-family units in suburban neighborhoods in VA

Forum Index » Metropolitan DC Local Politics

Reply to "Overriding local zoning to allow multi-family units in suburban neighborhoods in VA"

Subject:

Emoticons





More smilies

Text Color: Font: Close Marks

[quote=Anonymous][quote=Anonymous][quote=Anonymous]Economist here (not the PP). The Annenberg and Kung study has some nice features, but I wouldn't put much stock in its conclusions yet. I would characterize it as being more like a first step in setting some building blocks for future structural modeling.  There are still big gaps that need to be addressed.

The authors highlight the lack of migration in their model, and that's an important gap.  But, I think they have a bigger issue which is the lack of a model of amenities.  Their model concludes that is is more effective to improve amenities in lower-priced neighborhoods than to build more in high-priced neighborhoods, but housing is assumed to be of an identical form in LL locations, and amenities are implicitly assumed to be location-specific, but non-rival and therefore not a source of negative externalities.  In the real world, different types of neighborhoods have different types of amenities. Amenities in more urbanized areas tend to be effectively non-rival and non-excludable (sidewalks, public parks, coffee shops and restaurants).  But, these amenities are often only financially viable given sufficient density, and their utility to residents is itself often a function of density, because people like these things to be within walking distance.  Suburban areas have a very different set of amenities (larger private lots) that can impose large negative externalities on residents of other neighborhoods by using up much of the available land for private use.  Their model doesn't have anything to say about this, and their empirical methodology isn't able to take it into account at present.

In summary, amenities are heterogeneous, and spatial characteristics of different neighborhoods are heterogeneous in ways that are tightly linked to amenities. So, we can't really conclude much about the value of increasing density without constructing a model that also links the two.  My strong suspicion is that if you wrote down a monocentric city model that allowed for migration and for density to increase the quality of non-rival amenities, you'd fit the data at least as well as their model, but you'd also conclude that increases in density increase aggregate utility for both high-income and low-income residents. It would do this by improving amenities for high-income residents while also increasing affordability for lower-income residents who reside in more peripheral locations.[/quote]

Why do you find it necessary to write a lengthy treatise on here when your conclusion is nothing more than a suspicion? This is why many people think economists are blowhards.[/quote]

First of all, someone repeatedly cited this paper as supposedly strong evidence, even claiming the authors to be some of the world's top economists, without any acknowledgement of the serious limitations of the paper.  I'm just providing an honest assessment of the paper from someone who actually has the training to think critically about the model.  Sorry that I imposed on your pithy one-liner space.

Second of all, how exactly do you think that science works? People have hypotheses about how the world works, we write testable models that characterize those hypotheses, and then we test those models using the best data and empirical methods we have. If we already had all the answers, no one would need us.

What is different about Economics and other social sciences is that we're not aiming to build an exact and complete model of the world, because such a model would be as complicated as the world itself.  Instead, we build and use models that parsimoniously and accurately represent the key economic relationships we care about. That's an iterative process, it often involves a back and forth between multiple authors, and it's typical to start with a foundation that's too simple and then add features that seem empirically justified or that might alter the main conclusions of the model.

I have my own projects, so I won't be working on this particular topic. But, if I was asked to peer-review this paper, I'd provide the authors with the same assessment I gave you.  The paper is a nice foundation for future research, but it's still missing some key things that are likely to affect the conclusions of the model. That's not being a blowhard, that's just science.

PS: If you think that's a lengthy treatise you should really read more.[/quote]

Options

	Disable HTML in this message
	Disable BB Code in this message
	Disable smilies in this message

Review message

Search Recent Topics Hottest Topics