Dear Utah County,
Imagine for a moment that you live in Utah County in 2050. Let’s say you just got a new job in SLC and you’re chatting with a friend about what your new commute would be like. You’ve identified two potential commute options:
Option 1: You take a typical car to work. Maybe it’s electric, maybe it’s gasoline, but you’ll drive it approximately like you do now. You’ll enjoy breathtaking views of I-15s offramps, freeway medians, rotting roadkill, and of course the other cars in front of you because you’ll have to pay full attention to the road at all times on your 1-hour+ journey to our state’s capital city. Yes, it is the future, and your drive to SLC is more than an hour because population has grown. That’s on the days with no car crashes, of course.
Option 2: You take a futuristic mode of travel. This future travel method only takes 30 minutes to reach your new SLC job, and the advanced technology is so good that it’s self-driving. You could read a book, you could catch up on work, you could enjoy your breakfast. Of course you’ll miss the stunning views of freeway offramps, instead you’ll have to endure views of open fields, rivers and streams, wildlife that’s alive, and our Utah mountains. Part of the route will go next to the freeway though, so you’ll at least get a fleeting view of those gorgeous off-ramps.
Which option would you pick? Before you read on, sit on that thought for a moment and ask yourself why you picked the option? Was it close?
Now seems like a good time to clue you in on the name of your Option 2 technology, it starts with an “F” and ends with a “-rontrunner”.
Rather depressingly, this isn’t a fictional future I created. No, it’s a fictional future UDOT recently created, and which UDOT is using to justify a freeway expansion project which will cost an estimated $3.7 Billion. They’re also planning to bulldoze many homes and businesses1 to make way for the freeway, far more than would be practical to outline here. All funded by tax dollars, of course.
According to a painfully long report which UDOT spent years producing called an EIS, found here, thousands upon thousands of your fellow Utahns would be stuck, unable to pick between these two equal options. Equal in the eyes of UDOTs traffic model at least (we’ll address why here in a bit). Traffic models whose details reveal an utterly ridiculous assumption that (in my opinion) effectively invalidates their entire prediction. Don’t worry though, it’s only the core prediction used to justify the entire $3.7 Billion project and all it’s downsides. Did I mention this proposal has been wildly unpopular and broadly opposed by Utah citizens as well? Why listen to how your citizens want tax money spent when you can listen to yourself instead!
At its core, the model assumptions misunderstand what causes traffic.
Increased traffic happens when more people choose to drive at a particular time on a particular route.
Intuitively it makes sense that, say, a 50% increase in population would equate to a 50% increase in traffic. That’s the more people part of that earlier phrase, 50% more people. That intuition is wrong, however, because it left out the other part of the equation: human choice. If all else were held equal then yes, traffic would increase by 50%, but this is the real world, all else is not held equal. More people can equal all sorts of different outcomes, depending on the choices of the population.
A change in travel choice can come in many forms. If your regular route to the grocery store gets more traffic, you might take a different route. If traffic gets too busy at your usual grocery shopping time, you might switch times.
Then there are the bigger choices. Not many people living in Richfield commute to SLC for work. Why? Well, a 2 hr commute isn’t worth it to most people, most people would either move closer to SLC or get a job closer to Richfield. The same is true as freeway traffic worsens, people look for other options.
And the last, most vital change in choice is the change in method. If your kids school drop-off traffic gets too bad at the local elementary school, you might walk them to school instead. In that case it wouldn’t just be a change in route or timing via car, it would be a change in the transportation method. By the same token, if commute time via car goes from 20 minutes to 60+ minutes while the commute via train stays at a steady 30 minutes, train ridership will increase dramatically. This is because most people would choose a 20 minute car ride over a 30 minute train ride, and most people would choose a 30 minute train ride over a 60+ minute car ride.
This dances around the last key point: Better alternatives will actually reduce car traffic times. Yes, you read that right. As stated earlier, increased traffic happens when more people choose to drive a particular route at a particular time. Well, if a better alternative to driving is available, people will choose the better alternative, and thus their cars will not be on the road to create traffic in the first place. After all, traffic is made of people in their cars. If more people are on transit, on their bikes, or walking (instead of in their cars), there’s less traffic for the remaining people who do choose to drive. This drives us (see what I did there) to the answer of why UDOTs traffic model is so far off: it’s built to predict traffic levels in a world where transit, biking, walking, and all other non-car transportation options are broadly terrible.
UDOT states that their model (actually the model they are borrowing from the Mountainland Association of Governments and Wasatch Front Regional Council) “…uses existing travel data and then predicts future travel demand based on projections…”. They continue to state “The model is calibrated to actual, observed traffic conditions…”. Do you see the problem? The model is calibrated to historical data. Historically, would you say you’ve had decent transit, biking, or walking alternatives to driving on your daily commute? On your trips to the grocery store? For most people in the recent past, driving has been the best option for almost all daily transportation, and the model was trained on that. The model has been designed using data from a world where driving is the most common choice to get from Farmington to SLC, so it’s incapable of predicting a world where transit would be the most common choice.
Think of it this way: say UDOT of 1924 had made a model to predict transportation needs in 1950 (as opposed to 2024 and 2050). Being prudent model creators, they use the most recent data from the 1910s (2010s equivalent) on through their present day, 1924. Their model, only knowing data from the past, would predict a whole lot of horses and buggies riding around in 1950. They’d also predict a great deal of train rides, as that was a pretty commonplace form of intra- and inter-city transportation in Utah in the 1910 and 1920s.
This would seem an intuitive, smart approach, and the 1924 reports would truthfully claim the model “uses existing travel data and then predicts future travel demand based on projections”, that it would be “calibrated to historical data” but their historical data would be completely incapable of predicting the future choices of future Utahns.
In reality, people wouldn’t often choose horses and buggies for transportation in 1950. Governments would instead choose to make driving an easier and more convenient alternative, ripping up train tracks to make way for larger, paved, roads. This would simultaneously make driving a better alternative and make train rides a worse alternative, so of course people would change their behavior and drive more. Such a model would give preposterous predictions about horse and buggy traffic levels, predictions which would be absurd on their face, just like our current model’s prediction of 1hr+ commute times via car when the train only takes 30 minutes.
We only think of driving as the default because for so long we’ve chosen to build our infrastructure and our communities with cars as the dominant option. For decades we have built our roads and communities so that more and more and more people would find it more convenient to drive when they travel, and less convenient to take other transportation methods. We made a world where driving would be the easiest alternative for most people. We made a world where more people would drive on any particular route at any particular time. The result has been worsening traffic and billions upon billions spent in a sisyphean attempt to keep up with it by making driving even more convenient and other transportation even less convenient.
If we want to improve traffic, on the freeway or anywhere else, we need to invest those billions into making driving only one of several good options for transportation to any particular destination at any particular time. I hope you’ll join me in advocating for smarter infrastructure spending and more freedom to choose between multiple quality transportation options. If we work smart and work together, we can make our towns, our county, and our state into a better place.
Sincerely,
Your Neighbor in Vineyard