02 April 2020

Where people stayed home

Map from the New York Times.  Data from the week of March 23.  States and cities with known stay-at-home orders are crosshatched and outlined in black.

Also interesting is that data like this can be derived from cellphone data of 15 million people. 
The location data, from Cuebiq, a data intelligence firm, measures the range that people travel each day. It cannot predict where outbreaks will spread, and it does not track how many interactions people had while they were traveling. Not all travel is problematic: A person driving for a few miles to pick up groceries would not be violating stay-at-home orders. And people in cities can infect others without traveling far.


  1. I'm so proud of Ohio!

  2. I'd be interested to see this map overlaid with other maps -- such as the US election map, as suggested above. Or education levels. Or other possibly correlating factors.

  3. The article notes that not all travel is equally problematic, but is misleading by not including the maps that show *where* it is problematic, as pointed out in this twitter thread and several replies, both maps and people discussing their personal experiences in different areas of the country: https://twitter.com/sarahkendzior/status/1245710184917487617.

    In the Western Interior, distances to grocery stores, etc. are obviously long, and most of that travel is going to be taking place in cars and not interacting with other people, as shown in this map: https://twitter.com/AndyRowell/status/1245746906359201792. I disagree with the author of that tweet that the first map in the thread is misleading, it is merely highlighting something different.

    What the first map in the thread (here is the link again, for the sake of unambiguity: https://twitter.com/sarahkendzior/status/1245710184917487617.) highlights is that the areas in the South where people are still moving are in food deserts (both a long distance to get food, although not as long as the threshold used for the other map, *and* no car). In these areas, it is very difficult for people to get food while remaining isolated and/or travelling safety. It is not a coincidence that this area overlaps a good deal with the hungriest (https://twitter.com/canoncity7/status/1245743866948521990) and poorest (https://twitter.com/canoncity7/status/1245744249255219201) states in the US. This is also (not coincidentally) the heart of the "black belt" (https://twitter.com/CharlesMBlow/status/1245680178455609344).

    Not including this additional context is outright irresponsible. If the intention was merely to show disease risk spread in different, the NYT should have corrected for or included another map showing population density. However, much of the rest of the article talks about people not following stay-at-home advice or orders, implying that these are areas where people are not taking the threat of COVID-19 seriously and callously or carelessly putting people at risk. This is largely what people seem to be taking away from the map as well, portraying those areas as ill-informed or stupid, even to the extent of implying that they deserve death (https://twitter.com/AaronMarx/status/1245711516550205440). Additional context shows that those are poor and vulnerable communities - who are outright unable to follow public health best practices on top of decades of systemic medical neglect - getting hit hardest.

    If you want to link this to politics with not-so-subtle implications that this is happening because those areas are poorly educated "red states", remember that Fox News-watching, Trump-voting far right Rupublicans are predominantly *not* the rural poor (and that "rural poor" does not mean "white", especially in the South) and that they control red states by disenfranchisement, voter suppression, and gerrymandering of the actual poor and vulnerable, not true majority rule.

    Finally, these maps - especially the one showing when average travel first dropped below two miles - fail to take into account where coronavirus outbreaks actually *are*. Obviously people in Seattle went into physical distancing and quarantine sooner and are following it more strictly, they were hit by the pandemic sooner. Likewise, if the goal is to model risk for spreading the disease, the map should include centers where there is already a high number of people infected that it could spread from. I imagine there is also long-term aggregate data of how people tend to move in and out of those centers that could be used to model where the disease might go next.

    This is poor journalism and worse GIS by failing to provide necessary context for understanding the patterns shown on these maps and interpreting what they mean in practical terms - and you can see that a disturbing number of people are coming away from reading the article blaming the people who are most vulnerable.

    1. Thank you for this! I live in Mississippi. For years I lived where the nearest real groceries were 12 miles in one direction and 16 miles in the other! Where I live now is a small town of 6,000 people but it is still a 4 MILE round trip to the grocery (I live on one side of town and the grocery is on the FAR side opposite me). Their arbitrary 2 mile limit is clearly arrived at by urban dwellers who have NO CLUE what it is actually like to live in a rural area!

    2. Damn that's good criticism.

      I want a map that well-represents the population's capacity for data-literacy. THAT would be provocative...

    3. To be honest, I don't know if it makes sense to map that geographically (although data literacy is certainly ill-taught in many places) because the more people from different areas with different life experiences you have together, the better everyone's effective data literacy gets. The "sniff test" is honestly one of the most important things in data literacy, because it lets you catch things that seem suspect and that you should take a closer look at. Different people will catch different things that smell fishy based on their different life experiences, so the more people with different backgrounds you have together, the more dodgy uses of data you can catch. Then, once someone has pointed out that something needs a closer look, everyone can chime in with their relevant expertise to pick it apart. That's what I was doing in the post above: synthesizing a lot of critiques of the map by different people in the thread I linked, together with my own background knowledge of cartography and GIS. (For those unfamiliar with the acronym, GIS stands for Geographic Information Systems: the storage, display, and analysis of spatial data.)

      There are a lot of problems with social media, especially in it's current form (I would love to see what social media could become if it was designed as a genuine public forum rather than as data-gathering and money-making platforms for a handful of corporations), but I think this is one of social media's biggest advantages and a genuine benefit it brings to society.


Related Posts Plugin for WordPress, Blogger...