It looks to me as though $3 acted as resistance on several occassions in the not-too-distant past. Only twice has retail gasoline exceeded this level and stayed there for several months.

I don't disagree that at some point it loses its severity of impact. However, I've read myriad reports saying that for every 10 cent increase in gasoline over 11 billion dollars are drained from the American economy, or roughly $1B per penny increase. I'm not sure if this is contemporaneous data or the historical average. I question the statistic because two opposite forces are at work here. First, Americans are consuming less gasoline on a per capita basis. After the 2007-2008 energy shock fuel efficiency increased significantly whether this was in the form of higher MPG vehicle fleet rotation (cash for clunkers) or improved driving habits (combined trips, proper tire inflation, slower acceleration etc) is not really of concern. All that matters is that in America we're consuming less gasoline per capita. Second, acting against this increasing efficiency is the fact the economy is still in a very slow state of growth and bordering on the verge of recession according to the ECRI. While maybe $4/gal gasoline would be a minor impediment 5 years ago I question the consumers' ability to handle a sustained period of elevated gasoline prices in such a low growth environment. Keep in mind that the pay-roll tax cut the Obama Administration is hanging on to amounts to about an extra $40/week in the average Americans bi-weekly paycheck. This 'tax relief' is effectively zero when gas prices go from $2.75/gal to $4/gal meaning whatever stimulative effect the Administration hoped for is diminished. And lastly, income growth in the last 12 months was 2.8% nominal but when you factor in inflation real income growth was actually negative 0.06%. That doesn't leave much room for the consumer, whom is 70% of the economy, to spend on discretionary items. Now I'm not predicting a recession just yet but this is definitely a new development that is going in to the 'model forecast' and in my personal opinion it deserves close attention and should be followed regularly.