$81 billion dollars. That’s how much traffic jams cost America in 2022. Millions of hours of motorists spent sitting, wishing they were anywhere but their car. But where is highway congestion the worst? Although they can’t seem to fix the problem, at least the government can tell us what stretches of US interstate suffer from the worst traffic. The Federal Highway Administration tracks delays on highways around the United States. They even rank the top 100 - find the full list here. I recommend taking a look at the complete list, but the top 20 is reproduced below:
How the ranking is calculated is interesting. First, note that this is technically the delay of freight traffic (trucks), not passenger cars. This isn’t optimal, as it means a highway with relatively few trucks buy many passenger cars will not appear on this list, while a highway with a relatively high number of trucks will be higher in the ranking than it should be. My guess freight is chosen over passenger vehicles due to data availability, but the delay for trucks should correlate pretty well with the delay for all vehicles, so I’m fairly happy with the statistic.
To calculate the amount of congestion, first the total time delay, measured in hours, is calculated. To do this the difference between actual travel times and free-flowing traffic time is measured throughout the day. So if from 2:15PM-2:30PM the actual travel time over a stretch of highway is 20 minutes, while the free-flowing traffic time is 5 minutes, then that adds 15 minutes, or 0.25 hours, to delay time. That delay time is the column labeled “Delay (Hours)” above.
Next, the delay time is multiplied by the number of trucks that are estimated to use that stretch of highway each day. This yields the “annual truck hours of delay” (not displayed on the table). Finally, the annual truck hours of delay is divided by the length of road to get “total truck hours of delay per mile” which is the column labeled “Delay/Mile” in the table above.
Whew.
Calculations aside, it’s interesting to see what stretches of road are the most congested. Unsurprisingly, the most congested stretch of highway in the US is in the New York area, and it isn’t even close. Trucks are delayed an incredible 263,116 hours per mile per year along the stretch of I-95 that crosses from the Bronx to Manhattan to New Jersey across the George Washington Bridge. Having driven this stretch of road myself several times, I can confirm that it is hell on earth.
It also makes sense that this is the most congested stretch of highway in the country. First, it goes through New York City, the most densely populated and most populated area in the US. Second, the Long Island Sound puts tremendous pressure on I-95. All traffic going from the Mid-Atlantic to New England, whether that be a truck going from Baltimore to Boston or from Richmond to Providence. Until a bridge gets built linking Long Island to Connecticut (one can dream, right?), I-95 will always be a chokepoint. Third, the Hudson River means that even local traffic that wouldn’t otherwise need to use a highway will need to get across the Hudson. The city could use a few additional bridges to better link New Jersey with Manhattan. Any of these three factors alone would make this stretch of I-95 congested, but combine all three and you have a highway that rarely has free-flowing traffic.
The rest of the top 20 isn’t too surprising, with the notable exception of one stretch of I-35 through Austin, Texas. What’s going on here? Austin is the 26th largest city in the US by metro area population, and it doesn’t have the waterway or topographical restrictions that sites near bodies of water or mountains have to deal with:
My best guess is that the lack of a bypass route similar to I-294 in Chicago makes I-35 such a bad choke point. Also, over 22,000 trucks per day use this stretch of highway, third most in the top 20. So maybe that high volume is responsible for the delay. If a reader has a better answer, please let me know in the comments.
Other than this stretch in Austin, the top 20 is what one would expect. Number 2 is the section of I-90/I-94 that runs along the western edge of downtown Chicago. This stretch of road has been under construction for years. Luckily, the project is nearly finished, so I’m sure once it’s done there won’t be any congestion… that’s how traffic engineering works, right?
Continuing down the list, the main drivers of congestion appear to be large populations or waterway limitations. Houston has plenty of space but also has 7 million people and little public transit. Cities like Seattle and San Francisco may not be the largest cities, but their proximity to the Pacific Ocean and other waterways acts as a giant funnel. One mystery is Lake Charles, LA at number 24. I assumed this was just a suburb of New Orleans, but it’s an independent city that isn’t that close to any large urban area. The congestion is there despite a bypass route. I think this might be a data error, as number 23 and 24 are tied.
Another surprise shows up at number 80. A short stretch of I-95 that goes through New Haven, Connecticut, is responsible for 26,805 hours of truck delays per mile per year. Along with Lake Charles, New Haven is one of the smallest cities to appear on the list. Unfortunately its location as the nexus between I-95 and I-91, as well as the lack of a bridge across the Long Island Sound, makes this a chokepoint.
Is the New Haven spot also a data error? Unfortunately, no. I only have to travel 7 miles of highway to travel between my home and work every day, and this 1.8-mile stretch of road is part of it. Which is A) incredibly bad luck and B) leads me to confirm that it is one of the most congested stretches of road I have ever seen. Bummer.
This was timely for me - I get a chuckle when people complain about traffic in Charleston, and its not even in the top 100! It could be the way the data is reported, but most of the trucks would be centered around the airport and seaports and my guess is they funnel together on I-26, which also leads out to Summerville.
As for Austin, I was surprised that was on there (I never thought it was that bad but didn't do a standard commute). Lady Bird lake downtown is a waterway restirction (you are wrong there) and I-35 is the main thoroughfare, but MoPac on the other side also exists. I think its a combination of limited public transit, rapid expansion that the roads havent caught up with, and most people being priced out of downtown and having to commute from large suburbs like Round Rock. Its also pretty confusing (there are two different levels, weird on and off ramps, some bad lane merges) that probably create a lot more congestion than a standard highway should.
Like you mentioned there are some questionable or anomalous data points in this list. If the FHWA is going to go through the trouble of coming up with the metrics you think they'd double check their work or provide some context. At least an asterisk denoted sections under construction or detours because of construction would help.
As an example construction started on i70 in Denver (#32) in 2018, which caused some traffic to move to 270 (#36). That construction wasn't completed until about a month ago. I'd be curious to check back on those trends from 2017-2024 to see what effect if any the road improvements made. As you said these project must improve traffic flow! (/s)
I'm sure the FHWA wouldn't be proud of this list, but I'm also curious if they track overbuilt corridors (whatever the opposite of a bottleneck is). I think it would be valuable to know when they thought a lane expansion was warranted, but turned out to be unnecessary or underutilized.