Over the past year, we have repeatedly given the quantitative answer that has stumped so many: where did all those overhyped US "gas savings" go, because they certainly did not go into the broader economy, or toward discretionary purchases, as countless economists had said they would. The answer: more gas. Gallup confirmed as much most last week when it reported that Americans' reported changes in spending have remained stable in most categories of goods and services over the past year - except for gasoline, with 35% reporting they spent more on gasoline in the August-September period. Paradoxically, Gallup found the inverse of what had become erroneous conventional wisdom: "not only were Americans not spending more, they are spending less than they did in the past year on discretionary purchases such as retirement investments, leisure activities, clothing, consumer electronics, dining out and travel." But while we knew the quantitative answer, namely that Americans bought more gas with their gas savings, we were missing the qualitative one. Courtesy of the NYT we now learn that not only did consumers not redirect their spending to other discretionary items, but engaged in an act that has stunned economists around the globe: they don’t just buy more gasoline; they bought more expensive gasoline! And this is how a product that was essentially a staple good, suddenly provided the satisfaction of a discretionary splurge, even though it is virtually the same just more expensive. The NYT explain this observation which is just the latest mockery of macroeconomist models, and once again shows why theory never applies to the real world. A new report by the JPMorgan Chase Institute, looking at the impact of lower gas prices on consumer spending, finds the same pattern as earlier studies. The average American would have saved about $41 a month last winter by buying the same gallons and grades. Instead, Americans took home roughly $22 a month. People, in other words, used almost half of the windfall to buy more and fancier gas. The refiners will be delighted: We know how that extra money was probably spent thanks to a separate 2013 study by the economists Justine Hastings of Brown University and Jesse M. Shapiro of the University of Chicago, who got their hands on detailed accounts of the purchases made by 61,494 households at an unidentified retail chain that also sold gas. Professors Hastings and Shapiro showed that households adjusted their gas consumption much more sharply in response to changes in gas prices than in response to equivalent changes in overall income. In the fall of 2008, for example, as gas prices fell amid a broad economic collapse, consumers responded as if the decline of gas prices were the more important event, significantly increasing purchases of premium gas. And this is where the head of every tenured economist living in their ivory academic tower, and tweaking economic models they themselves created and thus know the goalseeked answer apropri based on their own preset assumptions, explodes. This is not rational behavior. Americans spent about 4 percent of pretax income on gas in 2014. One might expect them to spend about the same share of any windfall at the pump — maybe a little more because gas got cheaper. Instead they spent almost half. Americans, in short, have not been behaving like the characters in economics textbooks. Inconceivable: after all academic central planners are in charge of the entire world - what would happen if suddenly it becomes common knowledge that the entire "New Normal" experiment has failed because the lifetime academic hacks inside the Marriner Eccles building don't realize their theoretical models have zero applicability in the real world? At least when it comes to the "premium gas" paradox, there is an explanation: Researchers have found that people treat money as earmarked for particular kinds of spending, a tendency behavioral economists call “mental accounting.” If someone is buying rounds at the neighborhood bar, people tend to treat the money they didn’t spend as “beer money,” and sooner or later they tend to spend it disproportionately on beer. As a result, they end up drinking more beer than they had originally intended. The JPMorgan study compares gas spending between December 2013 and February 2014, when prices averaged $3.31 a gallon, with gas spending by the same people in the same period one year later, when average prices were one dollar lower. The study found that the average American spent $136 per month on gas during the high-price period and $114 per month on gas during the low-price period. While the price of gas fell by roughly 30 percent, spending on gas declined by only 16 percent. The study, based on the spending patterns of about one million JPMorgan customers, does not track the kind of gas consumers purchased. It shows that people bought more gas as prices fell, and that the increase in consumption is not sufficient to explain the entirety of the increase in spending on gas. And perish the thought someone actually saved it, but no fear: the upcoming negative interest rates will surely fix that pesky glitch in the economists' model. Unless they don't, and economists end up scratching their heads at even more "irrational" behavior: Moreover, this behavior was prevalent: 61 percent of the households made at least one irrational gas purchase. People “treat changes in gasoline prices as equivalent to very large changes in income when deciding which grade of gasoline to purchase,” they wrote. At the end of the day, though, the joke is on the consumers themselves: as the FTC notes, for most modern cars "splurging" on premium gas is usually a waste of money. At least the refiners are laughing all the way to the bank, as economists the world over continue to scream that any minute now "irrational" consumers will finally make the spreadsheet's life easier, and engage in rational behavior.