Somewhere in Texas…

Somewhere in Texas, a city is missing its economist…

My father sent me a link to this article, about how Dallas, along with many other cities across the nation, is turning off their traffic cameras that catch red light runners, because they are too effective at keeping people from breaking the law, resulting in significantly reduced revenue for the city’s budget.

On the Road

A couple years ago, I probably would have read the article with interest, gotten a good chuckle, and tossed it aside. But I see things differently now, either due to my studying to be a Chartered Financial Analyst or reading Freakonomics.

Either way, here’s my pro bono for Dallas…

This isn’t a pass/fail problem. This is an elasticity of demand optimization problem.

Now clearly you can make more money by charging people higher prices for things they either can’t do without (food) or feel like they can’t do without (cigarettes). But red-light running fits neither of these and is therefore an elastic activity.

I know the article didn’t say that they raised ticket fines, but charging people the same amount while increasing the probability of catching them seems like the same thing as: Average Price Paid = (Fine Amount) x (Probability of Fine)

So, before turning off the cameras…

    (1) … experiment with lowering the ticket fine

Just as businesses set prices in order to maximize revenue, so should Dallas. You don’t start selling a new product for $100 and then quickly discontinue it because it’s not selling in the volume you hoped for. Publicly announce a sale on red light fines and monitor resulting change in volume. Repeat until price point for revenue maximization is found.

    (2) … lower the odds of getting a ticket

It may turn out that traffic violations will continue to stay very low, even if the cost is $1 per infraction, if people know they will be caught. In that case you need to change tactics. Publicly announce that only 3 out of every 4 violations captured will actually be issued a ticket. And again, iterate to find the odds that yield maximize profitability. Many people just plain feel lucky even when the odds are against them. Works for Vegas.

    (3) … optimize both probability and cost

A no brainer, and just what it says. Streamlining both variables may result in even higher revenue than optimizing one alone.

    (4) … entice people to break the law!

Ultimately to maximize revenue, you need people to break the law as much as possible. So keep the optimized fine and probability points, but for each ticket issued throw in 10 free lottery tickets. You’ll probably find that you’ll then be able to raise both fine amount and probability of issuance.

    (5) … verify that remaining demand is not inelastic

Installing cameras definitely seemed to get rid of the elastic red-light runners, but what if the remaining come from an inelastic group? Those would be fathers rushing soon-to-be mothers to the hospital, out-of-towners unfamiliar with the cameras, or the rich and/or unintelligent. In any case, the point is they might be able to increase their revenue by 10x by raising the cost per infraction to $1000 from $100.

One final suggestion (which may already be in practice) is to allow lawbreakers to finance their fines. But then they’d have a new optimization problem on their hands. The interest rate…

4 thoughts on “Somewhere in Texas…”

  1. I assume you wrote that post with your tongue “firmly-in-cheek” but you do point out how nonsensical government can be.

    When did discouraging folks from breaking the law by ticketing them become a business? Why is “How profitable is that red-light camera?” even being considered here? Isn’t the safety of the citizens more important that the amount of revenue being generated? Now don’t get me wrong, if the whole system is losing money then, by all means, shut it down. But when they fixate on how profitable one intersection is over another, I think they need to have their heads examined. Then again, we are talking about government bureaucracy here…

  2. Thanks for writing Toby. Yes tongue definitely in cheek. 😉

    I was a bit amazed that while they do bring up other points (like, do the cameras end up causing more accidents?) the decision seemed overwhelmingly based on revenue and there was no attempt to hide this! At least there’s transparency…

    -Lumilog

  3. Average Price Paid = (Fine Amount) x (Probability of Fine)

    Please note: EXPECTED Price Paid = Sum[(Fine Amount) x (Probability of Fine)]

    Average Price Paid = Sum(Fine Amounts) / N

    With 45 days left until the exam, lets mind our semantics. Happy studying.

Leave a Reply

Your email address will not be published.