The hidden traffic metric that makes cities worse

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
this video is sponsored by curiosity stream when you sign up at the link in the description you also get access to nebula a streaming video service that city beautiful is a part of here are two different residential developments in the same metro area they both have 50 housing units the first is an apartment building in a dense urban area with only one parking space per unit the other is a neighborhood of 50 single-family homes with large garages and driveways plenty of car storage basically which one of these developments with the same number of households will have the greatest impact on traffic in the region intuitively you might think that the neighborhood of single-family homes would have the greater impact those households probably own more cars and have to make a higher percentage of trips by car their trips are probably longer to households in the apartment building can make more trips by foot bike or transit they would have a lower overall impact on the region's car traffic at least that's what you would think but according to traffic impact studies done for each project the apartment building actually has a greater impact on traffic and because new developments have to pay fees to mitigate those impacts the apartment building no longer pencils out it doesn't get built in the new neighborhood in the suburbs does how is this result possible it's possible thanks to a traffic metric known as level of service in this video we'll discuss how level of service warps or understanding of traffic impacts and the ways in which transportation planners are moving away from it in favor of better methods so let's jump right into it what is level of service exactly and why is it so bad I mean level service sounds kind of boring and innocuous level service is pretty straightforward it's simply a measure of delay there's a version of level of service that measures delay and free-flowing highway traffic but the one relevant to cities is intersection level of service how is it measured it grates intersections on an A through F scale with a meaning an average delay at an intersection of less than ten seconds that means traffic is low and maybe the traffic lights are even times so most cars aren't even stopping at the intersection at all no delay whatsoever my grades see the average delay at an intersection with a traffic light is twenty to thirty five seconds that means most cars are stopping and queuing at the intersection but probably getting through the light in the first cycle by grade F cars are waiting longer than 80 seconds at an intersection with a traffic light or more than 50 seconds for intersections with stop signs traffic is so bad that you're probably not making it through the light on the first cycle and you have to wait a while before getting through you can see how this would be useful information to have I mean a city government probably wouldn't want to approve a project that would add a ton of traffic right next to an intersection that's already busy all the time but how does this level of service metric lead to less infill development and more greenfield development so when a new development project is proposed that is of a reasonable size cities typically require a traffic impact study what's required varies by jurisdiction but basically they ask the developer to contract a transportation planner and prepare a report that estimates how many trips the development would add to nearby streets they are then required to estimate the level of service impacts to nearby intersections cities need to know this because they charge developers fees to pay for needed infrastructure upgrades if a new apartment building would add enough traffic to turn an intersection from an L o SC to an L OS f for example the city may require the developer to pay for a widened intersection or new traffic signals or something else to mitigate the effects of this new traffic the problem is that most of the intersections in a dense city are already at a high a lowest grade and the addition of one or two new infill projects can trigger l OS grade changes that would require developers to pay fees for intersection improvements because city street networks are typically denser with more intersections a single infill project can cause l OS impacts and a lot of intersections at the same time compare this to a new neighborhood out on the urban fringe this type of development would likely contribute many more additional miles of travel on the region's road network but the impacts are typically more diffuse by the time that new projected traffic reaches the city grid there aren't any major impacts developers outlet suburbs won't have to pay fees for rebuilding intersections these fees can mean the difference between life and death for development project they create a barrier to building infill development the type of construction that most places encourage with other policies infill development can reduce overall vehicle trips and contribute to the kind of dense walkable neighborhoods that are in demand and if the developer the infill project decides to pay the fee and the city decides to take that money and why didn't an intersection or a city street it can ruin the livability of the area it can also make it worse for pedestrians and cyclists level of service is meant to help us reduce traffic impacts but it paradoxically increases overall traffic it prevents dense walkable development and it can make our urban cores more hostile to everyone except drivers and is actually not that useful of a metric despite how reasonable it seems people don't really care about intersection delay specifically they care about how long their trips are as a whole let's say that an infill project increases ello s at a nearby intersection by 30 seconds the new residents of that development may not care because their overall trip length is way lower than living out in the suburbs adding 30 seconds to a 10 minute commute is preferable to having no delay but commuting 20 minutes this isn't just some abstract theory but it plays out in real life over the last couple of decades downtown Denver has grown significantly and it has many more jobs and residents now than it did in the past as you might expect all of that growth increased rush hour delay the kind of delay ello s measures from 4.2 minutes to eleven point seven minutes on average but overall travel time decreased between 1982 and 2007 all of those people downtown live and work close to each other and have access to transit bicycle infrastructure their average base trip without rush-hour delays is significantly lower than before and their total average travel time is lower as a result so are there alternatives to using level of service yes the leading contender to replace it is called vehicle miles traveled or VMT the idea with VMT is to estimate the number of miles the vehicles from a given development will add to the region's roadways transportation planners can use simpler models than the ones they use for a level of service to estimate how many trips that development will generate and how long those trips are you can see how this shifts the advantage to infill development transportation planners can assume that more trips in the apartment building will be made by walking cycling or transit reducing overall VMT trips and cities are typically shorter again reducing the project's VMT VMT also aligns with many cities and states efforts to reduce greenhouse gas emissions for example the state of California is requiring projects that requires state environmental you to switch from el OS to VMT and this is a big reason California has ambitious emission reduction targets in ditching el OS will help the state encourage the kind of development that's better for cities in the environment so in the future we can hope that developments like this one can be built without the hidden regulatory barrier known as level of service once we get rid of it our cities can be denser safer more livable and more sustainable a while back I did a viewer survey and it told me that quite a few of you enjoyed the great youtuber Tom Scott guess what it turns out that tom has produced and released a new nebula original series called money it's what I would call a game theory game show now all of content a nebula is great including my own to nebula originals but money is just excellent my wife and I started watching and immediately got hooked we binge watch the entire series it features other youtubers as contestants to money is well worth your while money a subscription to nebula starts at just three dollars a month or twenty dollars a year in addition to Tom Scott's original you get exclusive content from all sorts of fantastic and thoughtful creators like real engineering minutephysics and tears ooh and like I said I have two nebula originals for you to check out one is on the history of the city of Rome and the others on the show Parks and Recreation as if that wasn't good enough you also get access to the streaming documentary video service curiosity stream let's say you're interested in learning more about cities they have videos about Rome future cities Masdar and more and all of these documentaries are really high-quality and included in that extremely low subscription price so please consider signing up when you do you're supporting a platform built by thoughtful creators that you probably already know and love to sign up go to curiosity stream comm slash city beautiful there's a link in the description thanks you
Info
Channel: City Beautiful
Views: 249,180
Rating: undefined out of 5
Keywords: city planning, town planning, urban planning, urban design
Id: UEE4bSWBHdQ
Channel Id: undefined
Length: 8min 49sec (529 seconds)
Published: Thu Apr 09 2020
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.