(Simon Dawson/Bloomberg)

This week the advocacy group Smart Growth America released a report tallying America's 47,025 pedestrian deaths over the decade from 2003 to 2012. The total number of such fatalities has been on the rise since 2009, reversing the recent pattern, and pedestrians are now involved in a rising share of all traffic deaths, as this graph from the report illustrates:

Smart Growth America

That second trend, shown in the green line above, partly reflects the fact that while walking can be dangerous, riding in a car has become historically less so. This graph from the National Highway Traffic Safety Administration, charts the decline of motor vehicle fatalities since the 1950s, the overwhelming majority of which have involved people riding inside cars:


So, why are we making clear safety advances for people riding in vehicles but not for pedestrians traveling along some of the same streets? In large part, we've been able to address the first problem with technology and laws -- with airbags and mandatory seat belts and steeper penalties for drunken driving. Pedestrian fatalities, however, aren't a problem of technology, or even of the law (although many advocates would like to see tougher charges for drivers who cause those deaths). Rather, pedestrian fatalities are a problem of design.

And that makes them much harder to address. That means instead of retrofitting our vehicles with better technology, we need to retrofit streets and neighborhoods in some fundamental ways. We need to put sidewalks where they don't exist. We need to engineer roads that aren't so easily traveled at 50 miles an hour. We need to design intersections that prioritize both people and cars. We need to build places as if we actually think people might want to walk there -- something we've seldom done along major arterial roads and commercial strips.

This problem of design is apparent in Smart Growth America's rankings of the metro areas with the highest pedestrian fatality rates. Below we've charted the 25 most dangerous metros by a "Pedestrian Danger Index" used in the report. It's a measure of average annual pedestrian fatalities per 100,000 people in the population (between 2008-2012), divided by each metro area's share of commuters who walk to work. The Census Bureau's pedestrian commuter data is an admittedly imperfect proxy for the prevalence of walking in any community (you may recall it from this report). But this is the only reliable, comparative data the government collects. And it's used here to partially correct for the possibility that pedestrian deaths may appear higher in cities that simply have more pedestrians:


The majority of the metros on this list are in the South or Southwest. What they have in common, though, isn't necessarily climate; it's car-dependence. Nearly all of these cities have grown up in the age of the automobile, at a post-World War II time when we focused less on designing places for pedestrians because society had broadly acquired the luxury of driving instead.

Contrast these cities with Boston (ranked 50th out of the 50 largest metros), New York (48th), Chicago (44th) or even Washington (35th), more compact, older cities that were first plotted and built in the 19th century, or even earlier. Those cities -- although they've long since sprawled at their edges, as well -- are more likely to have smaller blocks, tight street grids and narrower roads originally meant for people on foot or slower traffic (by, well, horse). Urban planners tend to talk about these places as being more "human-scaled" precisely because they were built with pedestrians rather than cars in mind.

It's no wonder in the 21st century that these places remain friendlier to pedestrians (even as, in a city like New York where fewer people drive and more of them walk, pedestrians make up a higher share of all traffic deaths). The challenge now is how to layer pedestrian-friendly infrastructure on top of those places that weren't originally built for them, whether that's in the suburbs of Washington or the heart of Orlando.