Tuesday, January 8, 2013

Conflict Zones: A Simple Study in How Increasing Vehicle/Pedestrian Conflict Makes for Safer Places

This felt necessary.  In an email discussion somebody about real and perceived safety in urban, suburban, and exurban areas, probably better classified as walkable and unwalkable places, somebody mentioned that NYC is experiencing record highs in pedestrian fatalities this year.  Up to 170.  Now this is mostly due to increased pedestrianization and the competition for space between new pedestrians and existing vehicles on roads in NYC that are systematically being made more safe.  One by one.

The conflict is inevitable, but ultimately the goal is to decrease the need and danger posed by the automobiles over time.  In that sense, there are both long-term goals, but short-term increases present even though the numbers are higher than ever.  If the amount of pedestrians is increased, the number of conflicts between cars and pedestrians will be increased as well until the cars are slowly decreased in the equation.  With this understanding, it's entirely possible that NYC's rate of pedestrian fatalities could be going down despite the total number going up.  This is fairly hard to figure since most fatality rates are calculated quite simply:  total deaths (by type) divided by total population (in 100,000s).

To fully calculate this we would have to have NYC's commuting rates by mode, by year, or at least decade.  I haven't found those yet, but I'm sure they're out there.

The larger point I want to make is about NYC's seemingly high pedestrian fatality rate in comparison to other places.  All things NYC seem high by scale.  It's a bigger place, with more people, therefore there will be more of just about everything.  Over that year time span where NYC lost 170 pedestrians in traffic accidents, Dallas is on pace for about 72.  170 is greater 72 (triangular dots meaning therefore) NYC is totes way more dangerous than Dallas!!!11!

These numbers begin to slowly fall apart when you 1) calculate the simple rate then, 2) really fall apart when you start looking at representative number of conflicts between pedestrians and vehicles.  Let's do so shall we...

Population (100,000)
App. Annual Ped Deaths


Perc. Ped Commuters

Ped. Adj. Fatality Rate

Perc. Ped+Transit Commuters

Adj. Total Ped. Fatality Rate

This is a table.  It includes population, approximate annual pedestrian fatalities, and most importantly commuting rates by mode to represent an abstraction of conflict points.  And this is different than the theoretical conflict points of traffic engineering which draws lines of theoretical movement and every time they intersect, voila! conflict point.  It doesn't say whether there are actually vehicles or pedestrians there (usually, it's been engineered in a way that no pedestrians would be there, therefore SAFER! Amirite?!).

When factoring simple rates of pedestrian fatalities by population, we see that Dallas has nearly 3x more pedestrian fatalities.  And we're all like, duh! People drive more.  But because people drive more, they also walk less.  So that means even less conflict points.  If there are less conflict points, it would theoretically be more safe.  Not so.

When factoring the population only for pedestrian commuters, the rate of pedestrian fatalities rises in Dallas to 17x that of New York.  Taken to an even further extreme to factor transit commuters, since there is nearly always pedestrian activity at both ends of a transit commute, this number balloons to 51x the pedestrian fatality rate in Dallas as opposed to New York.  Therefore, we can deduce that increased conflict between pedestrians and vehicles yields a safer overall environment even though there is still the possibility of danger evidenced by the 170 pedestrian deaths in NYC this year.

So what do we have to do?  We have to get more pedestrians.  And the only way to do that is to build street/block structure and infrastructural networks that are conducive to walking.  And in hollowed out sun belt cities that also means leveraging the real estate market to favor infill development rather than sprawl (and much of that has to do with the bones of the real estate market, the shape of the infrastructural network).  And the only way to do that is to replace a good portion of the dendritic highway/arterial network with a multiply interconnected, reticulated grid-like system.