There are a number of cool little tricks and features in Google Maps that tend to fly under the radar. One such feature are neighborhoods, which Google Maps not only labels when you zoom in to the appropriate scale, but can outline as well.
Who cares, you say? Well here’s the thing: unlike zip codes, block groups, or census tracts, neighborhoods have no officially defined borders, at least not as far as the government is concerned. So how does Google decide where the neighborhood boundaries are, or even which parts of cities should be considered neighborhoods and which should not?
First, a quick comment about the nature of neighborhoods. In a geographic sense, neighborhoods may be described as an area of a city with distinct characteristics and a shared identity — that is, the people that live within a neighborhood consider themselves to be a part of a specific, geographically-defined community with a particular name. This name is sometimes derived from the historical legacy of the area, and most often given by the residents themselves. Neighborhoods may have very definite physical borders, such as a major street or canal, but quite often have fuzzy transitional zones that separate them from others in the city.
So, again, how does Google know what is and what is not a neighborhood, and, more impressively, where exactly their borders lie? Curiously, there is no definitive answer to this on Google’s own blog or elsewhere that we can find. There seems to be three possibilities here: either Google has done most or all of the work in house, they used a third-party such as Maponics (now part of Pitney Bowes), or they primarily relied on user-defined boundaries.
Between 2008 and 2017 Google’s Map Maker allowed users to actively contribute to Google’s map database, adding places, road segments, and other missing information. It’s unclear, however, how much of this information may have been used in defining neighborhood boundaries.
It’s quite odd that Google mentions nothing definitive about their methodology for determining neighborhoods, particularly since they they wrote about the challenge of defining neighborhoods when discussing the Bostonography project on their blog:
In almost every city, there is some disagreement as to where one neighborhood ends and another begins. Furthermore, as is often the case, reputable sources have differing neighborhood lines. In reality there are no physical lines on the ground clearly defining neighborhoods, however these hypothetical boundaries do have a real impact on local economies, politics, and identity.
At any rate, Google’s neighborhood defining system appears to have a bias toward compact, and well-recognized urban neighborhoods compared with sprawling suburban subdivisions. The latter Google often does not recognize as a unique neighborhood. However, the feature does appear to becoming more comprehensive over time, with a larger number and wider variety of neighborhood typologies recognized now than in the recent past.
In Chicago, for example, both the urban neighborhood of Oak Park and the suburban neighborhood of Carol Stream are recognized by Google:
Note, however, that Carol Stream is spread over a much larger area, encompassing a number of separate subdivisions. Naturally, this hints at the challenge discussed earlier: how to define a neighborhood. In the case of a compact urban neighborhood, there is often a distinct commercial center that acts as the area’s focal point and gathering place. In the suburbs, segregated land uses result in a hodge-podge of strip malls and separated residential subdivisions with no definable center. From this perspective it is easy to see why Google may not recognize your suburb as a neighborhood — without a definitive center to the average residential subdivision (or even clusters of subdivisions) Google is prone to clump large areas together under a single name.
So, you may not agree with Google, but given their unofficial status, neighborhoods — and their boundaries — are always going to be contentious. And perhaps with the way cities are constantly evolving, that’s just as it should be.