Taking a Higher Toll

When E-ZPass electronic toll systems are installed, tolls tend to rise:

Technology Eases the Ride to Higher Tolls, by David Leonhardt, NY Times: ...I spent a good part of my childhood summers at the Jersey Shore, and the tollbooths on the parkway always seemed to be a cruel final obstacle between me and the beach. Every 15 minutes or so, our car would have to stop yet again to drop a measly quarter in a bucket.

The ride is very different today, thanks mostly to the electronic toll system known as E-ZPass. At four of the tolls along the Garden State, the system is so sophisticated that cars barely have to slow down. A little box attached to the car’s windshield sends a message to a computer reader looming over the road, and money is then deducted from an electronic account. ...

As a result of E-ZPass and its ilk, ... many [drivers] don’t notice the cost of a toll. Which raises an interesting question: If you don’t know how much you’re paying for something, will you notice when the price goes up? Or has E-ZPass, for all its benefits, also made it easier for toll collectors to take your money?

A young economist named Amy Finkelstein started thinking about these issues a few years ago... So she collected decades of toll records from around the country and found a clear pattern.

After an electronic system is put in place, tolls start rising sharply. Take two tollbooths that charge the same fee and are in a similar setting... A decade after one of them gets electronic tolls, it will be about 30 percent more expensive on average than a similar tollbooth without it. ...

“You may be less aware you’re paying the toll,” said Ms. Finkelstein, now an associate professor at M.I.T., “but you’re paying a higher toll than you used to.”

The implications of this go well beyond highways. We increasingly live in an E-ZPass economy, in which bills are paid online, corporate cafeterias are going cashless and people take along their debit card, instead of cash, when they leave the house. Last year, 55 percent of consumer spending was done electronically..., while checks accounted for less than 25 percent and cash only 20 percent...

The E-ZPass economy is indisputably more convenient. It saves time and frustration. But the old frustrations that came with cash also brought a hidden benefit: they forced you to notice that you were spending money. With electronic money, it’s much easier to be carefree.

Marketers understand this dynamic well, which is a big reason they promote refillable gift cards and other forms of money that don’t feel like money. Part of what’s so intriguing about Ms. Finkelstein’s work is that it suggests that government officials may be coming to understand the dynamic, too. ...

Ms. Finkelstein obviously can’t prove that electronic tolls cause prices to rise by making drivers less aware of them. ... But she makes a spirited case for her conclusion. She has considered a number of alternate explanations for the increases and says the evidence doesn’t support them. At the very least, electronic systems do seem to make it easier for toll collectors to increase prices. ... Ms. Finkelstein discovered that tolls don’t usually rise as soon as an electronic system arrives. The increases tend to come a number of years later, once electronic payment becomes old hat. ...

Let me try an alternative to the agents don't notice price increases when the payments are electronic story (i.e. that agents are irrational). One thing to note is that after the E-ZPass system is installed, waiting times fall, frustration falls, and the inconvenience of not having correct (or any) change also falls. Thus, the economic cost is lower even if the dollar cost of the toll stays the same, and this would cause the quantity of trips demanded to increase. Over time, with inflation, the real cost of the toll falls further, the number of trips increases further, and eventually the increased congestion could require a return of the toll to its original value (in terms of the full economic cost, and assuming congestion is the reason for the toll), or even a higher value if demand for road trips grows due to economic or population growth.

[Another possibility, though I don't think it quite fits here, is that under the old system changing the price of a toll is costly in terms of transaction and collection costs. That is, a $1.00 toll is much more convenient than a $1.05 toll (more currency and coins are involved in the payment, and making change for, say, two dollars also involves more coins and hence more time), but with an electronic system it doesn't matter. Thus, under the old system, instead of changing tolls in small increments between, say, one dollar and two dollars as would occur with E-ZPass, changes will be larger and more infrequent (perhaps in quarter, fifty cent, or even dollar increments rather than nickel increments). The result will be that E-ZPass tolls that are more flexible and rise sooner than under the old system though with this story, in the long-run, the average toll would be the same (net of differences in transactions and collection costs).]

Comments

[dws]

Do you ever go to carnivals? The rides are always paid for in tickets, typically $0.50 to $1 per ticket and then several tickets per ride. Your arguements in favor of ease and efficiency apply here as well, but I've always thought people wouldn't pay $2 for very many of the dinky little carnival rides if they paid in cash. I wish there was a database somewhere...

[djconnor]

Higher traffic toll roads are more likely to go EZ-pass than the lesser traveled. Doesn't this suggest that maintenance costs would rise faster?

Also, don't forget about our good friends Supply and Demand.

says...

Mark,

The article states that similar toll roads are used for comparison- but are they truely similar? Or to phrase it another way, why does one road convert to an electronic pass system and the other doesn't? Were other factors considered, such as funding sources/amounts (gas taxes, general funds, tolls only, etc.), operating costs, deferred maintenance/construction, population growth, or traffic pattern changes?

[ndd]

I happen to live in a state where there was a semi-organized campaign against EZ pass for exactly this reason -- and what was predicted then has come to pass.

Anyway, while economists might not yet have figured out how this fits into their paradigm, behavioral psychologists knew this as long ago as 1964. Here's a link:

http://links.jstor.org/sici?sici=0002-9556(198006)93%3A2%3C345%3ASEOIDO%3E2.0.CO%3B2-1

That delay in positive or negative reinforcement (here, paying the toll) interferes with learning is a well-known principle of operant conditioning. The organism simply does not learn so well to associate the behavior with the response.
Thus, drivers who have to pay the new, increased toll immediately, also learn immediately, and if they are unhappy, will probably let the pollticians know. Drivers who don't have to learn a new behavior at all (only the bill changes at the end of the month) will be slow to notice and may not complain at all. Raising tolls on EZ pass is therefore an easy call for a politician. Classic stimulus/response/reinforcment.

As an aside, the same problem exists with meting out punishment for crimes.

[btgraff]

"Over time, with inflation, the real cost of the toll falls further"

i am not sure why this would be so - in toronto we have a toll road called highway 407 with most people using something like an EZ pass, or license plates get sanned and they send you a bill. tolls have increased dramatically because this is a privatised road - they charge what the market will bear. tolls might only come down if tolls are regulated, or if the road is government owned and tolls are limited to recovering costs

[cm]

There is surely something to the not-perceived-as-money argument, but I doubt that's the factor driving the installation of cashless systems. I find it far more plausible that traffic has increased so much that cash-only collection would lead to unpalatable congestions. (Plus, the element that automating anything has been the big vogue since a few decades.)

As for rising tolls, that's just exploiting a captive user population. In the SF bay area, many well-paying jobs are on the peninsula, and at least a few years ago housing in outer areas has been far more affordable. Provided you continue to commute to the peninsula paying those bridge tolls (I know of people who are doing 1+ hour commutes - each way - daily). That may also be tied to the unpalatable congestion argument -- having to take a long drive to work plus standing in line for it perhaps irks people quite a bit more than with a more occasional and discretionary ride. Electronic toll collection provides relief on that front.

[cm]

Off-topic automation note, in the Netherlands the "preferred" method of selling public transit tickets is by vending machine. When you go to a counter with a person, you have to pay an additional service fee for the same ticket.

[Lord]

There is also a move towards peak traffic pricing with people traveling during peak hours paying more. This may lower the cost of off peak travel relative to fixed pricing.

[luci]

As someone who has worked as a toll-booth cashier, in addition to the effects already mentioned, EZPass also cuts down on the opportunity for theft from the workers.
Ahem.

[afterburner]

EZ Pass started during World War II, when FDR had Congress pass the Pay-As-You-Go tax plan. Now it's deducted from that little black box you call your paycheck! So easy you never feel you are giving all that money to the government. So congress can redistribute it for you. How benevelant we are!

[Brian H]

I'm with CM. People are willing to pay for the convenience, so the higher fare is still felt to be "fair". If asked whether they would like to go back to the bucket in return for paying just the lowered toll, I suspect you'd have to crank the EZ toll up quite a bit, maybe much more than 30%, before a majority said "Yes!".

A fascinating experiment would be to have two bridges near each other to the same destination, one EZ and the other not. If the tolls were the same, you can be sure that more than just speed of crossing would move most to the EZ bridge.

The "supply" response to this could be one of several things: Up the EZ toll till loads equalled; Drop the bucket bridge toll till loads equalled; Switch the bucket to EZ toll. Etc.

[Bernard Yomtov]

I'd like to get a clearer idea of the data Finkelstein used. I drive a toll road, the MA Turnpike, fairly often. One thing that surprises me is how many cars do not have FastLane (our version of EZPass). There are often long traffic backups at the cash booths while the FastLane lanes are open.

I'm not sure what, if anything, this means for her results, but unless drivers on the toll roads she studied behave differently than they do here there are still a lot of cash tollpayers even on automated roads, the "lack of awareness" idea may not be so solid.

[cm]

Brian H: Small point, I didn't claim, or at least didn't mean, that people are willing to pay more for electronic collection. My contention was merely that electronic was introduced to relieve unpalatable congestions.

[cm]

Bernard Yomtov: I presume you have to pay installation costs or rent for the transceiver, and setting up and maintaining the billing process, making participation only worthwhile for "power users" (and to begin with probably possible for "documented" individuals, need I say more). I don't know how much time you will actually save with electronic vs. cash, and many may be able to "beat" the peak of the commute. Hereabouts working cultures differ, but in many office environments you will not meet disapproval when you come in after 9 or even 10.

[Bernard Yomtov]

cm,

Yes. There is a bit of work involved, but it's minimal. I probably spent thirty minutes on it ten years ago and nothing since. I paid about $25.00 for the transponder, and dimly recall that I got a little of that back in credits the first few times I used it.

It's hard for me to believe that's not worth it. Many times I've saved upwards of five minutes, sometimes much more, at a toll booth. And I didn't burn gas idling. I'm not a daily commuter either.

[Patrick Crozier]

Expanding on a previous commenter's point about supply and demand, prices rise because demand increases but supply stays the same. Because no one is building any new roads.