What does "induced demand" really amount to?Matthew Yglesias

www.slowboring.com

What does "induced demand" really amount to?

Matthew Yglesias

12 - 16 minutes

One of my resolutions for the new year is to encourage people in the pro-housing community to spend less time engaging with the bizarre arguments of Marxist geographers and other hard-left NIMBYs and more time addressing normie concerns about a world of housing abundance.

And the biggest, most important thing on that list is traffic.

I live in an unusual (for America) neighborhood where almost all of life’s daily errands — trips to the elementary school, the supermarket, the dry cleaner, the gym, haircuts and pedicures, etc. — can be done on foot. According to the Census, only 13% of our tract’s commuters drive to work.

This is way below not only the national average but the city average, so if D.C. got wise and allowed tall apartment buildings in our neighborhood, that would drastically reduce the per-person driving in the city. But 13% is not zero. And even though a family like mine drives much less than the typical American family, we do own and occasionally drive a car, so the overall volume of traffic would increase if our neighborhood was upzoned.

The basic anxiety is even more severe for in-demand communities that don’t have strong traditions of walking and transit use. A future version of Los Angeles with lots of dense infill construction would become much less car-oriented than the current version of the city, but it would still be pretty car-oriented. If you built tall mixed-use projects near all of LA’s metro stations, the residents of those buildings would drive less than the Angelino average, but they would still drive some. And this would create problems for existing residents.

I think in both cases, the downside of more traffic is actually relatively small compared to the benefits of a more economically dynamic city. People don’t like traffic jams, but they don’t like high taxes or high levels of homelessness either. But the traffic issue is worth tackling on its own terms both because it’s a problem in the cities we already have and because fear of traffic is a major impediment to building the cities we should have.

The New York Times recently did a big feature grounded in induced demand theory headlined “Widening Highways Doesn’t Fix Traffic. So Why Do We Keep Doing It?” which skirted around the kind of obvious answer that we do it because it lets more people drive to more places.

In other words, I think the idea of induced demand is just somewhat less paradoxical than people make it out to be.

Consider the traffic impact of a mass transit project rather than highway widening. In the aforementioned Los Angeles, a project is currently underway to extend the little stump of a purple line in the map below and extend it west to Beverly Hills, then down Wilshire Boulevard into Santa Monica and to the Pacific Ocean. This may never happen due to the usual litigation and delays, and it’s conceivable that once completed the project will be a low-ridership white elephant. But let’s stipulate that it gets built and that lots of people take advantage of this transit option.

Well, what happens next?

    Some of the people riding the new Wilshire Metro will be making trips they wouldn’t otherwise have made, taking advantage of new opportunities and making their life better.

    Some of the people riding the new Wilshire Metro will be making trips they otherwise would have made by car, reducing traffic.

    But now that there’s less traffic, some people throughout Los Angeles will make trips they otherwise would not have made due to fear of traffic jams, bringing the system back to congested equilibrium.

In other words, “Ambitious Mass Transit Projects Don’t Fix Traffic. So Why Do We Keep Doing Them?”

New York City has by far the biggest and most robust mass transit system in the country, but that hasn’t ended road congestion. And yet I think it’s obvious that a high-ridership extension of the Purple Line wouldn’t be a policy failure even if, via the induced demand mechanism, it didn’t end up impacting traffic congestion. The success would be that more people get to go more places. Some of those extra people would be metro riders and some would be drivers, but the point is that region-wide mobility would be improved. By the same token, the existence of serious traffic congestion problems in New York doesn’t show that the New York City Subway, NJ Transit, Metro-North, and the LIRR are all pointless. They’re vital to the existence of the region — without them, the whole metro area would be smaller and poorer.

In other words, I don’t think the induced demand critics of highway widening are wrong exactly. But they’re not really saying what they mean. This is what I think they mean:

    The pollution associated with driving cars is bad.

    Addressing that pollution via an appropriate gas tax seems politically challenging.

    Because the American political system is laden with veto points and NIMBY institutions, blocking highway projects is easier than raising taxes.

    Both new transit and new highways fail to solve traffic, but transit fails by leaving net driving flat while highways increase vehicle miles traveled.

    Therefore we should advocate for transit and not for highways. 

I don’t have a problem with that logic exactly. But when you live by the NIMBY, you die by the NIMBY. Just as the same NIMBY toolkit that blocks private development also blocks public housing, the NIMBY toolkit that blocks highway projects also makes it impossible to complete transit projects in a timely and cost-effective manner.

Beyond that, traffic congestion is a real problem and it deserves a solution.

The actual solution to traffic jams is well-known but underutilized — it’s called congestion pricing.

You charge people money to drive on crowded roads, and you charge them more money for driving at the most crowded times. Every time someone proposes this idea, it’s incredibly controversial and politicians seem to perceive it as very unpopular. And yet once cities implement congestion pricing plans, they basically never seem to reverse them. I think the biggest issue here is probably a natural bias toward the status quo.

But I also think that advocates often make things worse by turning congestion pricing into a revenue-generation scheme for mass transit. New York, for example, keeps teetering on the brink of implementing a central business district congestion charge, with the money going to the MTA. People tend to see the world as more zero-sum than it really is, so this plan evokes the notion of a fiscal transfer from drivers to subway riders. In the case of New York, that means tons of opposition from suburbanites. But more generally, it means opposition from most people, as the driving majority outnumbers the transit-riding minority.

I think a much more compelling plan would be to implement the congestion price as tax reform. In D.C., for example, there’s a general six percent sales tax, with higher taxes on soda, booze, restaurant meals, sports tickets, hotel rooms, and parking garages. I don’t know how much money a D.C. congestion charge could raise, but I would use the revenue to cut the sales tax. That way, the charge becomes a benefit not to people who don’t drive, but to people who drive a below-average amount — a much larger group — so that drivers who do pay the charge get the benefit of less traffic congestion.

Tax reform also seems to me to be the best way to answer the concern that congestion pricing is regressive:

    You’re swapping out one regressive tax for another.

    The poorest people don’t drive but do pay sales tax, so this shift benefits them.

    Among rush hour commuters, working-class people are more likely to have less flexible schedules and face more punitive consequences for being late, so they benefit the most from improving traffic. 

The upshot is that while this swap obviously can’t make everyone better off, it will make most people better off as they benefit either from lower taxes or faster driving.

This is often overlooked in the rush to turn congestion pricing into a piggy bank for transit agencies, but a congestion charge is a big win for mass transit even without dedicated revenue.

There are a bunch of different reasons for that, but a subtle one is that with congestion pricing, the bus will go faster. Private cars go faster, too, of course. But the key thing about the bus is that if it goes faster, you can increase the frequency on each route without spending any extra money. So a car trip and a bus trip both become faster. But the car gets more expensive while the bus becomes more frequent, so bus ridership naturally rises, both because of a push from people trying to avoid the congestion charge and also because of the pull of higher frequency and improved service.

And higher ridership, of course, means more revenue, which you can then invest strategically into further improving service.

A transit service that doesn’t exist obviously won’t get a ridership bump. But without the mental gymnastics around induced demand, it’s even clearer that agencies should prioritize high ridership when deciding which transit projects to build. And the reason to do that is in part because, in a world of road pricing, you also want to prioritize road projects that will induce a lot of demand. The reasonable concern about induced demand today is that if you fix the price of access to the road at $0, there’s no way to tell whether or not it was worth the effort to expand capacity. With pricing, you can look at the revenue impact of investing in more highways and decide what to do.

But “inducing” extra demand would be part of the point — the only thing worse than a new highway project that only causes people to drive more would be a new highway project that didn’t cause people to drive more and wasted a bunch of money. You’re still left with the problem that driving cars causes pollution, but we would then be in a position to address the pollution issue separately from the traffic jams issue.

A big source of trouble in American transportation policy is that way back in the day, the federal government instituted a policy of using gasoline tax revenue as the source of money for the Highway Trust Fund.

The idea was that this served as a kind of user fee for roads, and it made a certain amount of sense in an era when vehicle size was the overwhelming determinant of fuel economy. Basically, you would pay more into the HTF if you drove more and also if you drove a heavier vehicle — people who drive more presumably get more use out of the highway system, and people who drive heavier vehicles cause more wear-and-tear on the roads.

So far so good.

But of course, burning gasoline causes pollution externalities that have nothing to do with road use. And the way the American political system works is that it’s much easier for a Democratic Party administration to have the EPA raise fuel efficiency standards than it is to get Congress to pass an increase in the gas tax. EPA action requires passing a cost-benefit test, but that’s relatively easy to do because air pollution is bad. For any given amount of pollution reduction, it would be much, much better to achieve that reduction via higher fuel taxes than via EPA action, but the asymmetry in the number of relevant veto points means that fuel efficiency standards have risen a lot over the past 30 years and the gas tax has remained the same. So for the past 10 years or so, we’ve had a situation where gas tax revenue doesn’t cover the cost of the highway program. And it’s also a situation in which fuel economy is increasingly driven by engine technology rather than vehicle weight.

A Toyota Highlander Hybrid gets better gas mileage than a non-hybrid Corolla, even though it weighs 1,500 pounds more. The Highlander Hybrid is also much more expensive. And newer cars are more efficient than older ones. And of course electric cars are heavier than similarly-sized internal combustion engine cars and also more expensive, but use no gasoline whatsoever. So the gas tax is becoming both inadequate to fund the road system, increasingly regressive, and also increasingly not serving as a useful proxy for road use. We will eventually need to regulate gas-burning cars out of existence entirely and then have some kind of tax on vehicle weight that accounts for road damage and the pollution caused by tires wearing down. As someone who likes to advocate for political pragmatism, I have honestly no idea how that’s going to come about. But fundamentally, it is good to have escaped the dynamic whereby pollution is straightforwardly linked to road usage. The problem is that pollution is still bad, and charging for road use is still good.

The one thing that gives me hope that these seemingly hopeless issues can be addressed is that people really are going to be very upset if highway funding evaporates entirely. The last two major road funding bills — the FAST Act back in 2015 and the Infrastructure Investment and Jobs Act in 2021 — addressed the gas tax shortfall with gimmicks and deficit spending. That was fine at a time of low inflation and low interest rates, but it’s not going to work going forward, so Congress will have to come up with something. And as long as they’re at it, they may as well try to come up with something good.