Cumulative Model

Getting rid of traffic lights has been proposed for certain metro areas.

Getting rid of stop signs has also been proposed. I think this is a good idea. The vast majority of stop signs could replaces with yields, the rest eliminated, to the benefit of safety and traffic.

But the remaining traffic lights? I don't think we need them most of them. We don't need the red anyway. We should only have two traffic signals: Yield and Don't Yield.

With just green and yellow lights, we could help keep traffic moving and help it clear more easily when it does form.

Green means Don't Yield. Yellow means Yield. If you have a yellow, other traffic has the right of way. If there are cars in the interesection, you must yield to them. Changes to Yield would overlap other traffic. Like current yellow lights, if you are unable to stop, you may continue.

We are currently enduring a natural experiment on the effects of higher gas prices. While it has spurred movement toward more efficient technology, it has brought about some severe consequences that will need to be dealt with.

Last month it was reported that driving in the US was down 4.3% in March compared to last year. What everyone missed was that gasoline consumption wasn't. It was down less than 2%.

For the year, gasoline consumption is down little more than 1/2% [correction, DOE now shows close to 1% less fuel consumed for the year].

We aren't using less fuel, we're getting less done with the fuel we are using.

If the most efficient driving was being eliminated, it still couldn't explain the large difference in fuel efficiency. The driving being cut would need to be several times more efficient than normal to have such a negligible impact on fuel consumption. This is not plausible.

Among the reasons: Less efficient fuel mixtures may be being used; People are acting on bad advice. We've known for awhile now that accelerating faster is more fuel efficient (this is even before considering the beneficial effects on traffic), yet people believe the opposite; People may be driving more at high traffic times to generate needed income and be too tired and poor to drive at other times; And, during the economic slow down, communities may be neglecting good traffic management (e.g. not timing traffic lights properly).

We also need to consider whether higher prices will strengthen the movement toward more efficient technology or have little additional effect. (i.e. Has the move already happened and will further price pressure be of no value? After the crunch of the 70's, efficiency contitued to improve greatly despite falling prices.)

Additionally, we need to realize that in the mid-term, our current vehicle fleet and the infrastructure to produce more aren’t suddenly going to disappear. New tech won’t wash out these adverse effects.

[The gasoline consumption data can be found here: http://tonto.eia.doe.gov/dnav/pet/xls/pet_cons_wpsup_k_w.xls

It's in excel format. See U.S. Weekly Finished Motor Gasoline Product Supplied (Thousand Barrels per Day).

The Energy Information Administration defines Production Supplied as their calculation of consumption:

Products Supplied Approximately represents consumption of petroleum products because it measures the disappearance of these products from primary sources, i.e., refineries, natural gas processing plants, blending plants, pipelines, and bulk terminals. In general, product supplied of each product in any given period is computed as follows: field production, plus refinery production, plus imports, plus unaccounted for crude oil, (plus net receipts when calculated on a PAD District basis), minus stock change, minus crude oil losses, minus refinery inputs, minus exports.

More petrol data can be found here: http://www.eia.doe.gov/oil_gas/petroleum/info_glance/petroleum.html]

previous post

UPDATE: A mistake was made or the DOE revised their numbers. For the year, consumption is down close to 1%, not close to 1/2%.

UPDATE II: Large trucks do get less than 1/3 of the gas mileage of passanger cars (6.7MPG vs 22.9MPG 2005, DOT). However, they make up only 7.5% of total miles driven.(ANNUAL VEHICLE DISTANCE TRAVELED IN MILES AND RELATED DATA - 2005 BY HIGHWAY CATEGORY AND VEHICLE TYPE) If all driving cut was in passanger vehicles, it should still reduce fuel consumption by only slighly less.

Here are the weekly comparisons of 2008 over 2007, 4 week moving average 2008/2007, and the total consumption, year to end of week, 2008 over 2007:

Weekly | 4 Week Moving | Average Total
Week 1 1.12% 0.42% 1.12%
Week 2 0.62% 1.13% 0.87%
Week 3 -0.49% 0.42% 0.42%
Week 4 -1.65% -0.10% -0.10%
Week 5 -2.41% -0.99% -0.56%
Week 6 -0.45% -1.25% -0.54%
Week 7 -0.79% -1.33% -0.58%
Week 8 -0.83% -1.12% -0.61%
Week 9 -1.31% -0.85% -0.69%
Week 10 -0.29% -0.81% -0.65%
Week 11 -1.83% -1.07% -0.76%
Week 12 -1.45% -1.22% -0.82%
Week 13 -1.70% -1.32% -0.89%
Week 14 -1.96% -1.74% -0.97%
Week 15 0.98% -1.04% -0.83%
Week 16 0.57% -0.55% -0.75%
Week 17 -0.77% -0.31% -0.75%
Week 18 -0.35% 0.11% -0.73%
Week 19 -0.65% -0.30% -0.72%
Week 20 -0.78% -0.64% -0.73%
Week 21 -1.18% -0.74% -0.75%
Week 22 -3.85% -1.62% -0.89%
Week 23 -0.80% -1.66% -0.89%
Week 24 -3.54% -2.35% -1.00%

I've contended for a while now that high gas prices will lead to decreased efficiency.
It seems I've been right. I've been looking more closely at the data at the Energy Information Administration. I've found the appropriate data set (see Product Supplied, Finished Motor Gasoline).

The EIA defines product supplied as the appropriate measure for consumptions:

Products Supplied--Approximately represents consumption of petroleum products because it measures the disappearance of these products from primary sources, i.e., refineries, natural gas processing plants, blending plants, pipelines, and bulk terminals. In general, product supplied of each product in any given period is computed as follows: field production, plus refinery production, plus imports, plus unaccounted for crude oil, (plus net receipts when calculated on a PAD District basis), minus stock change, minus crude oil losses, minus refinery inputs, minus exports.

Gasoline consumption is down less than 1% (approximately 2/3 of one percent) compared to last year, while driving has been estimated to be down 4.3%.

This is a major decline in fuel efficiency. It cannot be explained by decreases in the most efficient types of driving alone. The driving cut would need to be several times more efficient than normal driving.

I’ve identified three reasons for the decline in efficiency I believe most likely(other than population growth combined with a lack of infrastructure growth):

3. Less efficient fuel mixtures (reformulated gasoline is making up a greater percentage of fuel).

2. Possible Giffen Behavior.

People are pressured to forgo luxury driving during off-peak hours, but must drive more during peak hours to produce a needed increase in income. And people are only willing to do so much driving in a day or week. People must drive more during congested times and are too tired to take the family out or take that country drive to visit grandma. Maybe mom and dad don't even want to be in a car any more.

1. Drivers acting on bad information (e.g. Accelerating slower rather than faster)

"It's not commonly understood by people who drive," Dr. Dougherty said. "They think that the way to get best fuel economy is to accelerate very gently, but that proves not to be the case. The best thing is to accelerate briskly and shift.

"Don't give it everything the car has, but push down when you're going to shift, using maybe two-thirds of the available power, and change through the gears relatively quickly."

Megan McArdle answers the unanswerable question.

Somehow, she decides it's zero.

Never mind that if you can't count it, you can't control it, and you certainly can't raise its price.

It'd be like trying to tax or set the price of cocaine, keeping it illegal and still prosecuting.