« Schwarzenegger Asked to Shut Down Prostitution-Related Website by British Authorities | Main | R-E-S-P-E-C-T? »

Monday, October 05, 2009

Why The “Best Places to Live” Are Often The Worst: The Law and Economics of Cities (Part 2)

This is the second post in a series discussing the “Best Places to Live” rankings issued by America’s financial magazines.  My contention is that they are not only flawed, but fundamentally wrong-headed, mistaking good things for bad.  Further, they reveal a distorted view of the value of cities that infects both general discourse and public policy on the subject.   (The basic logic of these posts comes from my new paper, The City as a Law and Economic Subject.)  

To understand why, you have to look at what goes into the numbers.  

One of the key variables the magazines use is the cost of living, most of which comes from the cost of property.   Given the diversity of preferences for what constitutes a good place to live and the difficulty of measuring a variety of relevant characteristics, this seems a good place to start.  After all, if we are trying to figure out the value of a city to any given resident, a property cost variable will give us a bunch of information about what the demand across the population for living in a city is (factors affecting the supply of housing will be the subject of my next post).  As my great colleague Tom Hazlett once noted when he was told about the attractions of a low cost of living:  “The cost of living is dying.  What you are talking about is a low demand for living.”   

Unfortunately, magazines interpret the cost of housing in a strange way – they take an increase in price as a negative sign about quality.  To understand how odd this is, imagine if you were in a liquor store trying to figure out what the best bottle of whisky was.  One way to do this would be to ask someone, but in this you would be subject to the peculiarities of that person’s tastes and the fact that the person you ask might not know any more than you do.   Another way to do it would be to trust everyone else in the world’s judgment and just use the price.  This is will give us preferences in the aggregate.   I suggest you try this at home: buy this $685 bottle of pure bliss and an off-brand $8 bottle of whiskey and see which you like better.  You certainly wouldn’t say: “On balance, higher priced whisky is worse.”   

This, of course, doesn’t mean that you always want to buy the $685 bottle or that there aren’t bargains.  For any product, you want to choose a price/quality level.  Further, there may be products that are less valued but that you like better.  The same is true for cities.   Living on a cliff overlooking the beach in Malibu may not be worth it you or possible.  You may not have enough money, you may prefer living in a place that has seasons, or you may hate actors.  However, as long as you believe people can move between cities (an assumption inherent in the idea of ranking cities in the first place), the price of property will include all sorts of information about how the population as whole understands the quality of a place, as people are trading other resources in order to consume housing in that region.  Even more specifically, prices reveal something about the quality of cities as cities. People are willing to pay more to live in certain places because of the quality of the atmosphere the place provides -- the quality of the labor markets, public policies, how nurturing of human capital place is, how cool the people are, etc.   Looking at the studies, this fits -- for instance, the places with the highest cost of living index on Kiplinger’s study were Honolulu and San Francisco, two places most people would describe as very nice. 

Now, it would be possible to conceive of a type of list that was price-independent, where other variables are used to generate a ranking of quality that expresses the author's view of good price/quality tradeoffs.  However, if prices are to be included, claiming that they are purely negative is perverse.  This is particularly true because the qualities that cities have -- the aforementioned labor markets, public policies etc. -- will affect the price of property, resulting in a list that cancels out that categories they do include, leaving only the categories they don't to shape the rankings.  So as a starting point, these magazines have taken something positive and set it as a negative.     

And then it gets worse.  Below, see why the end result of these lists is a list of towns that have bad neighbors. 

The usual next step in producing one of these lists is to include a list of variables related to the output of public policies – school quality, taxes etc.  This too is a bit odd.  It has long been known that people take these characteristics into account when buying a house.  As explained by famed economist Charles Tiebout, individuals “sort” to their preferred bundle of public policies.  If some cities or regions have purely better public policies than others, this should already be reflected in the price.  The “capitalization” of public policy into housing is a well-documented phenomenon.  As public policy variables are already incorporated into the value of a house, including them and housing values results in cancelling out their effect on the studies.  By inserting public policy variables and counting housing values as a negative, the magazines have therefore created an index of how much demand there is for places are when public policies are ignored.  

After this, the magazines throw a bunch of other characteristics of the city or region, either observable statistics like the average salary in the city, or ratings by experts about which city has the best dating scene or what have you.  These are trying to capture what economists call “agglomeration economies” or the gains from living near other people.  There are three types of gains that are usually included by economists in this category.   

  • The first is reduced transportation costs for intermediate goods, which explains why auto parts manufacturers usually locate near cities that produce cars.   (Research on this effect is part of what won the Nobel Prize for Paul Krugman.)  
  • The second is the depth of markets.  For instance, living in a city with a deep labor market provides gains for workers.  They can specialize in a deep market and also have the insurance against the firm-specific risk – they can get another job without moving.  Deep labor markets also provide gains for employers, who can easily replace workers and can get very specifically talented employees.  The same logic works for consumer goods or for non-market goods.  For instance, a deep dating market in a city provides gains for those singles with specialized tastes, reduces search costs and makes break-ups (the dating equivalent of “firm-specific risk”) less difficult because there are more people to date (the equivalent of insurance.) 
  •  A final category of gains is what economists call information spillovers, or all the things you learn from your neighbors.  In places like Sillicon Valley, these are extremely important – new ideas just spill from one firm to another because smart people are meeting with one another and talking and the ferment creates new ideas and new businesses.   Jane Jacobs famously noted that it is in cities that most new ideas get developed, and it is because of these types of interactions.  These spillovers also affect human capital.  Economists have long known that city residents are paid better than their rural equivalents and Ed Glaeser, a Harvard economist who is probably the nation’s leading urban economist, has shown that the reason for this is that city residents learn a great deal from being in cities and become more productive the longer they stay.  

All of these factors – reduced transport costs, deep markets and intellectual spillovers – affect the price of housing too.  This is why rents in downtown Manhattan are so high, as it is the center of the deepest high-end labor market in the country and one of the biggest centers for informal learning about business.    The other factors the magazines put in capture some, but not all of the qualities of agglomeration economies.   When people make their decisions about where to live, these things are factored in too and hence are part of the price, as everything people like affects the price of a house.  The average salary in the city is no different in this respect than public schools – people are willing to pay more to live in an area that has a good job market, just as they are willing to pay more to live in a town that has good schools.  Just like the public policy variables, putting these variables as a positive next to the price as a negative results in cancelling out.     

So what’s left?  The other part of the price comes from the agglomeration variables the magazines don’t include.  What’s left is the quality of the other people and businesses in the city.  The magazines usually don’t include factors like how nice the neighbors are and how much you are likely to learn from them.  Similarly, they ignore things like how nice and deep the local retail and consumption markets are – how cool the stores in an area and how good the restaurants.  These things affect housing prices but aren’t taken into account by the variables they have.  

However, again, the magazines are treating these positives as a negative because they are counting the price of housing as a negative.    Aside from questions related to weighting, the basic result is that Best Places to Live lists are really lists of the places that have a number of truly terribly qualities that the variables they do use don’t capture.  Because these variables usually relate to the quality of your neighbors, both businesses and people, these lists are actually lists of the places with the worst neighbors.  Congratulations residents of Huntsville, Alabama and Louisville, Colorado!  Kiplinger’s and Fortune Magazines have found that you are the worst neighbors in America.   Put that on your Tourism Board’s promotional materials!

Posted by David Schleicher on October 5, 2009 at 11:09 AM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Why The “Best Places to Live” Are Often The Worst: The Law and Economics of Cities (Part 2):


Wonderful Post to read , i got very informative Content over here.

Posted by: liquor store in dubai | Oct 15, 2018 8:39:57 AM

So, the next logical question is, "why are expensive places expensive?"

Suppose that prices were unavailable. How would you predict what prices would run? Also, since the question is really what "places" are best, rather than what "houses" are best, shouldn't one be looking at land price?

One of the facinating things about land prices is that they are not independent of development. Intensely developed places often rapidly gain value. For example: Las Vegas, Nevada and Vail, Colorado. Most stories of real estate fortunes involve development of previously low value property.

Posted by: ohwilleke | Oct 5, 2009 6:06:08 PM

On a formatting note, it would be helpful if there was some way to differentiate between the posts in this series at the beginning of the title. I nearly navigated away from PB because I figured I had already read this post - and by extension the ones after it - because I didn't initially see the (Part 2).

Posted by: Matthew Reid Krell | Oct 5, 2009 4:26:42 PM

David writes above: "The point is not to suggest that a pure ranking of prices
is the best way to value cities, only that what these lists consider to be
"bargains," contain some weird normative assumptions about what makes cities valuable."

I certainly agree with that. As I suggested in the earlier thread, I think these rankings are just based on a particular conception of what people want or should want, which of course readers are free to agree or disagree with. I tend to think of these lists of being "Places you probably have never thought of living that might actually be a good fit for you, depending on who you are." But that's pretty wordy, so they just say "best places to live."

Posted by: Orin Kerr | Oct 5, 2009 1:35:21 PM

Ken -- your second point is right on. It's the subject of the next post in this series....

Posted by: D.Schleicher | Oct 5, 2009 12:06:47 PM

But these aren't lists of "best bargains," but rather of best cities. Even so, they don't give us a guide at all to true value or future value. Under Elan's implicit, and Paul's explicit point, we would expect high ranking cities to show property appreciation over time. But they don't.

I'm not suggesting that a strong-form EMH works in property markets. God knows. Rather I'm suggesting that what these lists control for shows certain biases -- a point I'll make clear in a later post -- that reveal failures in the way we look at cities.

The point is not to suggest that a pure ranking of prices is the best way to value cities, only that what these lists consider to be "bargains," contain some weird normative assumptions about what makes cities valuable.

Posted by: D.Schleicher | Oct 5, 2009 12:05:14 PM

You're ignoring the fact that high housing prices don't mean that consumers in general find the place valuable to live in; they mean that the particular people who live there do. If that group of people is specialized enough (people with vacation homes; actors; politicians) the price they pay for housing has little relevance to the value most consumers would get from living in the area, yet the price is still what most consumers would have to pay. This indeed makes the price into a negative.

There are also other factors. For instance, consider a place which has high housing prices because of regulation that makes it more expensive to build houses (or because it's located in a place where building houses is difficult). And when you're talking about rents, an area could have a generally higher cost of living, which means it costs more for an apartment owner to maintain an apartment.

Posted by: Ken Arromdee | Oct 5, 2009 11:53:36 AM

This post misses the whole idea. The utility of a city (making the standard assumption of quasi-linear utility)is precisely the value o the benefit(s) given by the city minus the cost of that benefit. Therefore, price is indeed a negative. While it is quite likely (probable) that they miss some benefits of cities this is besides the point. If one assumes that price is a proxy for all of the benefits conferred by the city than in equilibrium all cities are equal and rankings are useless.
Essentially the idea is to see whether all of the benefits are priced in. While supporters of the EMH would claim that there are they are this is highly unlikely due to friction costs (for example search costs and 6% transaction costs).

Posted by: Elan | Oct 5, 2009 11:46:31 AM

In a way, this post is a defense of the ratings (!). Suppose that the magazines do a pretty good job of identifying a bunch of factors that make a city more liveable. And suppose that the market mechanism isn't working perfectly well because, e.g., of the cost of moving from city to city. Then, by including price as a negative and a bunch of good things that lead to higher prices as a positive, what the listings aren't capturing is the best places overall (price will do that for you) but the most under-valued places. At the limit, if they managed to get a measure of neighbor quality and the like, a wise home-buyer would buy a house in the best city on the list and wait for housing values to rise to quality-of-life factors. Following those recommendations is a way of doing arbitrage!

Posted by: Paul Gowder | Oct 5, 2009 11:41:06 AM

The comments to this entry are closed.