Here's something to keep in mind in the context of the past several years, as well as during the current runup in gas prices: They're more than likely higher than the press's reported "national averages."
On Friday, the Associated Press reported the following concerning gas prices: "The national average is $3.64 a gallon, up a cent and a half from Thursday, with the highest prices in California, the Northeast and the Midwest." It would appears that the press typically uses GasBuddy.com for its national average quote, which is currently just above $3.68. I really don't intend to knock the web site, whose primary mission is to help consumers find the cheapest gas prices in their neighborhood. But their quoted "national average" appears to really be the average of each of the 50 states plus DC giving each state equal weight, without any accounting for states' widely varying populations. And yes, the difference matters by enough that it's worth noting.
Here is the rundown of how I believe GasBuddy.com calculates its national average. I obtained the quotes at about 1:30 ET (highlighted states are those which exceeeded the national average):
The result of $3.679 is within a trifling 0.2 cents of what GasBuddy's quoted national average was at the time ($3.681) -- a difference which could be explained by the fact that it took about a half hour to obtain and record all 51 quotes.
Now let's look at the average national price based on each state's population, which is probably as good an indicator as any as to what a true national average would be:
The difference of about 6.7 cents isn't inconsequential, and it's obviously more reflective of reality than a simple 51-"state" (50 states plus DC) average. It's definitely something to keep in mind if (more likely when) GasBuddy's national average tops $3.94, because, assuming no significant changes in which states are more expensive and less expensive, it will mean that the population-weighted average has reached the psychologically meaningful $4 mark.
Assuming that GasBuddy.com hasn't changed its calculation methods and that the most and least expensive states haven't changed much, the above also means that the population-weighted average topped $4 for a brief time in the spring of 2011.
Separately, there may be a bit of reporting bias at GasBuddy.com if it calculates its averages strictly on user reports. It would seem that users would be more likely to tip others to a low prices than to higher ones.
Somebody should develop a computer-tablet-smartphone app for this -- and unless it can prove otherwise, the press should stop using GasBuddy.com as the last word on what average gas prices really are.
Cross-posted at BizzyBlog.com.