The Beauty Contest: How Cities are Shaped by What We Think Others Think (Part One)


Cities are complex adaptive systems. As human social creations, they are subject to all of the herd behavior and interrelated feedback loops that shape how humans behave in groups. This insight informs several central tenets of the Strong Towns approach: that local policy makers should make small, incremental investments instead of large, transformative projects; should design systems to adapt to feedback; and should aim for bottom-up action (chaotic but smart) over top-down (orderly but dumb). When we treat an aspect of the city, such as its housing market, like a simple machine—input in, output out—we tend to introduce fragility.

In particular, our cities can be shaped profoundly by the matter of what we think others think about a neighborhood or other place. A place can suffer a bank run on confidence if enough people decide it’s declining; on the flip side, the shared desire among enough people to turn a struggling neighborhood around can become a self-fulfilling prophecy. The sum total of individually rational behaviors can produce deeply dysfunctional collective outcomes.

Erica Barnett, a Seattle-based journalist, and Ben Klemens, a Washington, D.C.-based economist and author, are working on a book about how economic and game-theory principles can help explain sociocultural phenomena. In two guest posts, they will examine how a much-discussed urban phenomenon—gentrification—is best understood neither as top-down conspiracy nor as the inexorable consequence of supply-and-demand math, but rather as a social phenomenon profoundly affected by the question of what people think other people think about a neighborhood: a phenomenon called the “Beauty Contest.”

This is part one of a series about the “Beauty Contest” as a model for gentrification. Click here to read part 2.


This story takes place in Seattle, but it could be set in any large or midsize city where gentrification is changing the political, cultural, and physical landscape. For the purposes of this story, we’ll define gentrification as the process by which the residents of historically working-class, minority neighborhoods are displaced, often rapidly, by people with lighter skins and higher incomes. [Editor’s note: Read our own previous discussion of the different understandings of “gentrification” here.] How you choose to explain gentrification says a lot about your political worldview. Do gentrifiers act out of self-interest? Racism? A little of both? Or something else entirely?

In Part 1 of this two-part series, we use the example of one Seattle neighborhood to argue that it’s entirely possible to explain gentrification with something we can easily observe—signals, which communicate to prospective homebuyers what other people like them believe about whether it’s a good idea to invest in a certain neighborhood. Then, in Part 2, we use accessible theoretical modeling to demonstrate two ways in which, even in the absence of conscious beliefs (such as racism, classism, or the belief that land values will rise over time) or preferences (a desire to live next to popular amenities), signals can still lead to gentrification.

But before we get to Seattle, we have to go back to 20th-century economist John Maynard Keynes, and an economic concept called the beauty contest. In 1936—a time when newspaper sweepstakes were popular—Keynes came up with the following thought experiment: What if newspaper readers were asked to choose the most beautiful woman from a roster of photos, and the sweepstakes winner was drawn from a list of everyone who chose the most popular woman? The trick to winning that kind of “beauty contest” isn’t to pick the woman you personally find most beautiful, but to guess which woman other players think is the most beautiful, knowing that they’re doing the same thing.

The crux of the beauty contest is that we don’t form our opinions about beauty, the value of a dollar, or a house, in a vacuum—we come up with those beliefs based on a long chain of assumptions about what we think other people think. Why do I think a piece of paper printed by the Treasury that says “$1” in the corner is worth a dollar? Because I think I can give it to somebody, Steve, because Steve values it. Why does Steve value it? Because he thinks Maria thinks it is worth something. And I believe Steve thinks Maria values it because she knows Paula will accept it. And so on.

If the piece of paper said something else (maybe the ID for a bitcoin wallet), the same process might not happen. Given multiple options, it isn’t always obvious who’s going to win the beauty contest, or which centrally located neighborhood is going to suddenly shoot up in value. What is certain is that as long as human beings make predictions about value based on these signals, they will result in predictable behavioral patterns—you may not know where gentrification is going to happen next, but there will be some sequence of signals that will precede it.

So let’s make this less abstract.

Consider buying a new condo, with an asking price of $500,000. Whether it is a good deal today at $500,000 is partly a question of whether it will one day sell for more or less than $500,000. Your purchase is a wager, based on the information available to you, that the value of your condo and similar properties in the area will continue to appreciate.

Your decision to buy also depends on what amenities are nearby now. Will your friends want to move in nearby? Will restaurants and a grocery store start filling up empty storefronts nearby, and will they be able to stay in business? Will you have access to reliable transit lines when you move in? In a decade? All these questions also factor into whether you choose to buy here or across town.

There is no way to poll all the people somehow involved in such questions, so we have to play a beauty contest game, guessing what our friends, local restaurateurs, and transit planning agencies are thinking, and what they will be thinking in a decade. Meanwhile, they’re all doing the same thing: just as I don't want to move to a neighborhood with no restaurants, no restaurant wants to open in a neighborhood where they aren’t likely to have customers.

In the 1950s, determining a neighborhood's status was easy: check the map. As part of an arrangement by the National Association of Realtors and the U.S. federal government to limit the ability of African American homebuyers to buy houses in “white” neighborhoods, maps were developed with clear red lines around black neighborhoods and explicit racial covenants or bank practices enforced the racial divides on those maps. Today, these formal covenants have been replaced in most US cities by zoning, which walls off certain areas—often those originally segregated by redlines—as designated single-family-only neighborhoods, which tend, by historical precedent and by design, to be mostly white.

Seattle’s Central District in 1968. (Source: Seattle Municipal Archives)

In Seattle, the legacy of historical segregation is still evident in many areas, including one of the city's historically black neighborhoods, an area called the Central District. Racial covenants excluded African Americans (and Jews) from most other parts of the city between the 1920s and the 1960s—stipulating, for example, that "No part of said premises shall ever be used or occupied by or sold, conveyed, leased, rented, or given to negroes or any person or persons of negro blood.” As a result, the population of the Central District was once more than 90 percent African American.

Recent years have seen the demographics in the Central District nearly flip. In the 1970s, the neighborhood was more than 70 percent black; by 2016, that number was 20 percent. Whites now make up about 60 percent of the area's population, and the trend is on track to continue.

Few neighborhood businesses more perfectly exemplify this demographic shift—or serve as more of a lightning rod for protests and accusations of racial insensitivity and white colonization—than a flashy recreational pot shop called Uncle Ike's. (Washington state voters legalized recreational pot in 2012.) Lit by cheeky neon signs—on the next block: “Hey, stoner, around the corner"—it sells recreational pot like fancy coffee: expensive and unabashedly bourgeois. A ten-pack of Magic Kitchen peanut butter brownie bites will set you back $28, and a bottle of Moss Cow Mule cannabis ginger ale goes for $20.

Situated between an old car wash (recently purchased by Uncle Ike himself, Ian Einsenberg, and rebranded as Uncle Ike’s Car Wash) and an African- American church, Uncle Ike's is a sign of the apocalypse for anti-gentrification activists, at the literal crossroads of old and new. Across one street: A doomed strip mall anchored by a liquor store, a barber shop called Earl’s Cuts and Styles, and Africatown, a local nonprofit working to build affordable housing and preserve the history of the neighborhood. Across the other: A sleek modern apartment box called The Central, where open-plan one-bedrooms go for $2,060 a month.

Uncle Ike’s wasn’t the first new business in the neighborhood to cater to a largely white crowd—a quaint, quirky bar next door called “The Neighbor Lady”, which a local news report said would “breathe new activity” into “the troubled corner,” preceded the pot shop by two years. But Ike’s opening in 2014 was quickly followed by an influx of new businesses catering to a similar crowd, which converged on the intersection as if summoned by a pot-leaf-shaped bat signal. Today, a strip that once housed a 25-year-old soul food restaurant called Thompson's Point of View and was notorious for gang-related violence is home to an airy, loft-style coffee shop, a salon/cafe owned by two women who have said they were priced out of a nearby neighborhood known in past decades as a haven for artists and hipsters, a bike shop that sells $3,000 electric bikes, and a doughnut shop specializing in gluten-free mochi doughnuts.

As long as you're going to frequent an area, why not live there? Plenty of newcomers to the Central District have been doing just that, and housing prices have responded. The median home sale price in the neighborhood immediately surrounding Uncle Ike’s has increased 34 percent, from $519,000 to $695,000, since the shop opened in 2014, outpacing price increases in the city as a whole by seven percentage points. In the four blocks immediately surrounding Ike's, the city has granted or is in the process of granting permits for more than 900 new apartments and townhouses, plus a rezone that will accommodate many more. If the area’s demographic history is any guide, those new units will be occupied primarily by white newcomers with higher incomes than the neighborhood's historical average.

Neighborhood advocates have done their best to fight back the tide. They’ve organized 4/20 protests outside the pot shop, diverted a Martin Luther King Day rally to the store’s parking lot to confront Eisenberg personally, and even written a mournful song, "Irony on 23rd," calling the pot shop a slap in the neighborhood’s face. (The "irony" in the song's title refers to the fact that a pot store for white hipsters set up shop in a neighborhood where black and brown youth were historically targeted and thrown in jail for low-level drug crimes.) Several hundred protesters managed, on one occasion, to force the shop to close temporarily; they demanded that Eisenberg turn over most of his property to their cause, for the legal defense of minorities with drug cases, build a community center, and provide financial aid to people displaced from the neighborhood. These efforts have been met with shrugs by the perceived gentrifiers, and the redevelopment of the neighborhood has continued unabated.

A planning department notice labeled with “Gentrification” as a protest gesture. (Source: Flickr via Creative Commons license)

Can all of these changes be attributed to Uncle Ike’s? Of course not. The pot shop is a single storefront on a block with more than a dozen buildings, in a neighborhood spanning dozens of blocks. It’s highly unlikely that a market-moving number of pot smokers decide where to live based primarily on proximity to their favorite supplier. The neighborhood’s proximity to  downtown is obviously an asset, and it’s just down the road from a historically gay neighborhood, Capitol Hill, where rent is higher and homes sell for an average of $677 a square foot. Eisenberg, for his part, thinks the Central District would have gentrified with or without him, given how close the neighborhood is to downtown Seattle; but the fact is, other neighborhoods that are equally close to downtown have not yet gentrified, suggesting that there is something else going on.

More than just a place to buy pre-rolled joints with names like "Bubba's Gift" and Body Buzz bath salts, Uncle Ike’s is a signal that radiates out to the bars, the coffee shops, the developers, and the yuppies. The new bars and restaurants and gluten-free doughnut shops are a response to that signal, forming a reverberating feedback loop of signals that leads to more and more businesses (and sleek new apartment buildings with downtown views) that tell white and wealthy people: You're welcome here. This  is a place for you. Gradually, the neighborhood starts to feel more like many other neighborhoods in Seattle: new, well-maintained, predominately white and middle-class.

So why does this sort of thing happen—not just at 23rd and Union in Seattle, but in similar neighborhoods around the country? Racism, classism, or conscious beliefs about future land values could be part of the reason. But even in the absence of those kind of conscious beliefs, the opening of a business like Uncle Ike’s could have still moved the market.

In Part 2, we’ll show how changes in a neighborhood can serve as signals using simulations to demonstrate how gentrification happens even in the absence of conscious preferences, and how changes in house prices can actually be more dramatic in the absence of overt bias or preferences.

(Photos by Erica Barnett unless otherwise specified.)



About the Authors

Ben Klemens (Twitter: @b__k) is a Washington, D.C.-based economist and author (Modeling With Data: Tools and Techniques for Scientific Computing; Math You Can't Use: Patents, Copyright, and Software) who has written about policy issues in publications such as The Wall Street Journal and The Washington Post. Erica C. Barnett (Twitter: @ericacbarnett) is an independent journalist in Seattle with 20 years of experience as an editor and news reporter for publications including Seattle magazine, Sightline, The Stranger, Next City, and PubliCola, a local news site that she co-founded in 2009. Barnett and Klemens are currently working on a book that uses the principles of economic theory to argue that many common sociocultural phenomena, from gentrification to stock market trends to how political parties evolve, can be understood through game-theory concepts like the beauty contest.