This is an old post, based on a study I did in a research methods course during my undergrad. Don’t let “undergrad” scare you, I went a bit overboard with this project, and it’s worth reading and considering.

Some stuff has changed since doing this study, and I’m working on editing it accordingly.

It’s posted here without a link to the original post because it’s posted on my personal blog and some of that stuff is…well…personal.
Anyway, here:
I hate driving. It’s no secret. I wouldn’t say I hate it with every fiber of my being. But I hate it with most of them. And those I don’t hate it with, they’ll come around.
Because of my hatred for driving, I tailor my driving habits for efficiency. I just want to get it over with. So when coming up to a red light, I take my foot off the gas, and try to time it so that the light is green before I actually stop. Better on gas, fewer grey hairs. Similarly, when actually stopped, I keep an eye on the OTHER light. Once it’s red, my foot is off the brake. That way the car is just about moving, and once my light turns green, I’m already going.
This has affected my observations of other drivers. I watch them doing stuff. Needless stuff. And it infuriates me. They’re wasting so much time. They’re so distracted and potentially endangered. And inevitably, they’re pumping out emissions and wasting money.
People tell me I’m strange.
I had been searching for years for a study that looks at this stuff. At the costs of driver distraction, either in gallons or dollars. I searched for traffic in Providence, I searched for articles on driver distraction, on fuel burned while idling.
It just didn’t exist. So somewhere in my brain I was cooking up how I’d want to do it were I ever to get the opportunity. And I never did find anything like it.
So I did the study myself, thank you.
I put together an internet survey to get some stuff together; asked some questions about drivers. I didn’t care who was driving. You could be my girlfriend, my brother or Leonard Nimoy for all I care–if you’re driving in my city you’re fair game.
I wanted to know a few things. What kinds of cars are being driven and their fuel economy. I wanted to know average mileage daily. I wanted to know what distracted drivers. And I wanted to know what people perceived was distracting to others. Also, for my own well-being, I wanted to know if people felt that the distractions of others hindered their commutes.
From there I made some data collection sheets.
These have fields for the time of day, intersection, and number and kinds of lanes going in each direction. They also have fields for the common types of cars in Providence, with a field for “other,” in case some dweeb pulled up on a Vespa. They also have fields for several distracting behaviors, and another “other.” Another is “none.” Lastly, a field for interactions between drivers as well as time elapsed for each car.
So I sit down at the intersections with my sunglasses on so nobody knows I’m staring at them. Three cars pull up. I mark 1, 2 and 3 in the boxes that correspond with the vehicles observed, and then quickly glance to get an idea of what’s going on in the cars. Because once the light changes, I only have seconds before it’s too late to know what was going on. Don’t wanna have to think later. Do it first.
The light turns, and I hit my stopwatch. If that first guy is on his phone, but starts driving right away, he isn’t distracted–he’s multitasking. If he’s on his phone and sits there for 11 seconds, he is. Also, if he isn’t distracted but is waving on the guy who wants to turn left, that’s not distraction either. It will add to the average time spent clearing the intersection while not distracted, though. And that stuff is important later when I crunch the data.
The whole ordeal takes about 12 seconds total. But in that 12 seconds, I’m recording the type and position of three vehicles, what they’re up to, how they’re interacting and how long it takes each of them to clear the intersection.
So I did this, 125 times. I observed 277 vehicles at five intersections. So far. I’m gonna keep this going.
I took my data sheets and started punching stuff into Excel.
First, I made a few columns. One each for the first three undistracted drivers to come up to lights. One each for the first three distracted drivers to come up to lights. Each column had every time observed, for every car I saw. 277 so far. But I’m not done with the study yet.
I put them in ascending order and found the mean of each row. Then I compared each position (1-3) with its counterpart, distracted vs undistracted, using a T-test. Basically it’s a test to compare actual differences in data sets to see if they’re significant. Fancy bar graphs that show one huuuuuge column, and one teensy one, might not actually represent statistical differences. You can manipulate data to apparently show what you want it to without technically lying. I wanted to avoid that, so I did the test.
In every case, the tests indicated that the difference was too great to be attributed to chance alone. Meaning there is a strong correlation with distraction and delay. I “figured,” but “figuring” isn’t science. So I try to avoid assumptions.
So ok. Cool. What did I do next? I took the undistracted time for driver one, and subtracted it from the distracted time. That’s the extra time spent when distracted, on average, for the first car to a light. I repeated the same for positions two and three. Then I took those three times that were left–the extra time added by distraction for the first three drivers–and found the mean.
That mean? 2.81 seconds. Now we’re getting somewhere.
But it seems like nooothing. Right? Well yeah. I mean, who cares if they’re held up 2.81 seconds? Well, for one thing, that’s an average of multiple positions at a light. It’s lower for the first, and higher for the third. And your distraction compounds as it trickles through a row of cars. But I didn’t look into that part about the “trickling,” it’s just clear that A leads to B, so I said it.
But yeah. So 2.81 seconds. What do you do with that number?
Well, remember I took the surveys and paid attention to the cars? They helped me establish a fuel economy. From survey/interview responses, I came up with an average fuel economy–for Providence–of 26.5 mpg. Apparently nationwide it’s 24.1, so my study might be on the conservative side of things. I also compared the amount of drivers distracted versus undistracted, and calculated that just over 24% of drivers were distracted.
Ok! Neat! So I know the average fuel economy in Providence, the relative frequency of distraction and the average delay time for distracted drivers.
Now what?
Ok. So They’re not driving. They’re idling. Sooooo, what does that translate to as far as fuel wasted? Well, according to this study, at more or less that fuel economy, you burn between 1/4 and 1/2 of a gallon per hour while idling. But this is 2.81 seconds. Who cares?
Well. What about the quantity of drivers in Providence? Let’s look at the city, and then the county, just to get an idea of scope.
The city of Providence has 67,169 workers over the age of 17 according to the 2,000 US Census. Of those, 50,992 commute daily to work using their own car. Seems like a lot, but just Rhode Island Hospital employs like 6,000 people, and that’s one building.

Of those drivers in Providence, 24%–or 12,228–are distracted over the course of a day. Wow.
So now you have 12,228 drivers taking an extra 2.81 seconds on average to clear intersections, at roughly 26mpg.
12,228 x 2.81 is 34,360.68 seconds–or about 10 hours–per day. Ten cumulative hours per day are spent distracted. And at 26mpg, they’re burning 1/4-1/2 tank of gas an hour. That’s 2.5-5 gallons per day, or 905-1,825 gallons yearly. Just in Providence. Just work commute. And only taking into consideration stopping at one light.
Also, that’s only the city of Providence. The county has 246,593 people driving to work. That’s 59,182 distracted drivers and about 23 hours of distraction and up to 8,395 gallons yearly. That’s a tanker truck. And still just work commute. Still one light. And we haven’t left the county yet.
Is this sinking in? Go from a city to a county and you’re quintupling the waste. On my round trip to school today, I hit 15 lights. That’s probably on the high end–going to work I only hit 1 light the entire time. But seriously–I’m gathering that data now.
What if the average amount of lights is just five? We can all agree that our average drive probably involves five red lights. Round trip? You know it does. If that’s the case, you’re at 41,975 gallons before you leave the county, and before you start looking into OTHER reasons for driving.
I’m working on getting statistical data from the Providence Place Mall parking garage. I want to know how many people get one of their parking passes yearly. I’m betting it’s close to or over 200,000, easily. Why not? How many employees do they have? Their garage’s capacity is something like 4,000 cars. I’d be they fill it a few times a day. Maybe not all at once, but it doesn’t seem far fetched to me.
Seriously, in the city you have 50,992 people going to work. Supply and demand. How many customers do you have if you need that many workers to sustain it? I don’t know the ratio, but I’ve worked retail. There are a lot more customers than employees. A. Lot.
I don’t know what to do about this. Five lights and a 4:1 ratio of customers to employees and that 42k gallons quadruples. What about when you expand this area to all of Rhode Island? Of New England? The East Coast? You can see how quickly this adds up.
If 42,000 gallons cumulatively wasted in 2.81 seconds in Providence County is an extremely conservative estimate (remember–this only takes one reason for driving into consideration), then this is a huge waste factor. The lowest gas price in Rhode Island at the time of completing the first draft of the paper attached to this study was $3.35. Lower than national average. So here’s another conservative estimate for you:
$140,616.25 is wasted at the pump, yearly, by distracted drivers in Providence County, Rhode Island.
Conservative is the operative word there.
This is affecting gas prices. This is affecting the environment. This is affecting the advertising we see on TV and billboards. This is affecting the actions of legislators–both those in favor of and those opposed to related measures. This is affecting your taxes.
8,400 gallons?
$140,000?
$140,000 is nothing.
Those 2.81 seconds are the tip of a very, very large iceberg.