Gasoline prices and consumer sentiment

Gasoline prices (in case you’ve been hiding in a cave and didn’t know) have been on something of a roller coaster the last few years. And it looks as though we’re climbing back up another hill at the moment. How much are the recent increases in gas prices likely to weigh down American consumers?

U.S. average retail gasoline price, in dollars per gallon, plotted monthly using the data from the middle week of the month, January 2004 to June 2009. Data source: EIA.

Up until the fall of 2008, consumer sentiment had been closely following that roller coaster, with a sharp plunge in consumer sentiment accompanying the spiking gas prices associated with Hurricane Katrina in 2005, a second, broader drop in sentiment accompanying the second, broader bump in gasoline prices in 2006, and then a significant sustained decline in consumer sentiment as gasoline prices began their remarkable rise over 2007-08. The burden on consumers from that last run-up was in my opinion a key factor that precipitated the initial economic downturn over 2007:Q4 to 2008:Q3. The path of consumer sentiment is plotted as the solid line in the figure below. I’ve also plotted gas prices on an inverse scale (the dashed line in the picture below) in order to highlight visually the negative relation between sentiment and gas prices. The dashed line was calculated as 22/P for P the gasoline price in dollars per gallon, which you could interpret as how many miles you could drive for every dollar you spent on gasoline if you get 22 miles to the gallon.

Consumer sentiment versus miles per dollar. Solid line (left scale): Reuters/Michigan index of consumer sentiment, from FRED and
MarketWatch. Dashed line (right scale): miles per dollar spent on gasoline, calculated as 22 divided by the price from the previous figure.

That strong relation between gas prices and consumer sentiment continued as the falling gasoline prices (or rising miles per dollar) between June and September 2008 lifted consumer sentiment back up. However, the subsequent financial scares and credit problems in the fall introduced a very dramatic new dynamic, causing consumer sentiment to drop back down to the June 2008 lows by November despite plummeting gas prices.

So how should we assess the likely consequences of the fact that gas prices have now come back up significantly from their lows of December? The Edelstein-Kilian regressions employed in my paper from a recent conference at the Brookings Institution imply that a 20% increase in energy prices would historically be followed within 2 months by a 15-point drop in consumer sentiment and a 1.4% decline (relative to trend) in real consumption spending. From that perspective, the 46% (logarithmic) increase in (seasonally unadjusted) gasoline prices since December is quite worrisome.

On the other hand, since those December prices were 88% (logarithmically) below the July 2008 peak, consumers should have been giddy in December and still be significantly more sanguine now than they had been last summer, if the only thing on their mind was the price of gasoline.

Only problem is, consumers were anything but giddy in December. Credit and employment challenges have weighed far more heavily than gas prices over the last 9 months, and are presumably far more important than gas prices for determining what happens over the next few months as well.

But whatever may come next on the credit front or for unemployment, the impressive spring spike in energy prices can not be a welcome development for American consumers.

14 thoughts on “Gasoline prices and consumer sentiment

  1. DickF

    Cause and effect are difficult things to determine and become more confused when corelation is used as the “proof.” Who is to say that the CAUSE of high oil prices and the CAUSE of “credit and employment challenges” are not the same?
    Currently the price of oil is close to its historical relationship with gold. If the monetary authorities or congress do not screw things up there could actually be a slow recovery. It the congress took aggressive supply side action we would almost instantly be out of the recession. But, alas, congress and the monetary authorities will probably follow the same blueprint they have been following for almost 100 years now and that means the new “New Deal” will extend our stagnation for decades.

  2. Steve

    It seems to me there is a multiplier that exists concerning rising gasoline prices. Something of the order of of 2 to 1. That is for every dollar per gallon gas rises, 2 dollars is removed from consumer spending. The damage to the consumer market seems to linger as well. When the gas prices fall 1 dollar, it seems that only 1 of the 2 dollars that was lost is regained in consumer spending. With the rapid cycles we are seeing in gas prices, new spending habits are beginning to emerge, whether good or bad, that are taking hold on the consumer and are apparently not easily shaken. Maybe the old saying, fool me once, your fault, fool me twice my fault, is true.

  3. Marc V

    Forgive me if I’m a little economically challenged (both wallet and intellectually), but what “aggressive supply side action” were you thinking that Congress should take? I would love to see us brake the debt train, but Barry Obama has put his foot on the pedal and is yanking hard on the engine whistle chain.

  4. jg

    Nice (inverse) correlation graph on sentiment and gasoline prices, Professor. Makes sense.
    You have been asking for The Fed for inflation, and higher gasoline prices are a manifestation of such.
    Inflation of gasoline, heating oil, and food prices (grown with oil-based fertilizers) is going to be terrible in the face of falling employment and falling (real and nominal) wages.

  5. DickF

    Let me give a really quick lesson in supply side economics first. Economics is about one thing, creating optimal conditions for people to enter into free exchange. That’s it; both side benefit, it is win-win.
    The Laffer curve is the best tool to use to determine whether we are making the right policy decisions. If we make a change that increases total production and by extension total tax revenue then it is a good change. If it reduces total production and reduces total tax revenue then it is a bad change.
    Now to specific policy changes: first we should reduce personal tax rates. Optimal would be one tax rate between 15-20%. Such a change would immediately “stimulate” both investment and spending and would turn the economy around instantly. Now many will say that a 15-20% tax rate will not support our budget but the increased revenue from production would more than offset. This is one of the problems with the Fair Tax. The proponents use static analysis to determine the tax rate. Now understand a tax rebate has no real supply side effect. It must be a change in tax rates. People simply spend a rebate, but there is no real change in their lives because nothing has really changed. We need people to want to produce and consume again and feel they can do it without the government taking it away from them.
    The second move would be to lower business taxes especially taxes that hinder the use of capital, capital gains taxes and inheritance taxes. A capital gains tax is actually not a tax on capital gains but a tax on inflationary illusion. In capital gains there is usually no real appreciation only a different money valuation of the same capital. The tax is a tax on the incremental difference between the old valuation and the new valuation with no change in the capital, so the tax actually erodes the value of capital.
    Before I go to number three let’s have another lesson in supply side economics. Wedges to economic exchange act the same as tax increases when we look at the Laffer curve. What this means is that when government spending must be covered in the future by tax increases the spending moves the economy deeper into the prohibitive range of the Laffer curve. The Bush/Obama spending has pushed us deep into the prohibitive range of the Laffer curve.
    Three is to stop government intervention in the market and sell off all assets the government has confiscated. That means GM, but it also means Fannie and Freddie who own around 80% of the nations mortgages. When the government intervenes in the market, investors and producers stop investing and producing. The political whims of government push uncertainty to a point that investors will not act. There can be no planning because the potential for profit is uncertain and illogical dependent on political connections rather than the satisfaction of wants and needs of consumers.
    Then we must stop the government from attempting to provide goods and services. The government is horrible at providing goods and services, both wasteful and inefficient. Look at Medicare where costs are through the roof, payments to doctors are inadequate, and government paperwork generates enormous overhead. Look at the post office unable to compete with FedEx and UPS without government intervention and protection. The list could go on and on, but I hope you get the picture.
    We must understand that the nature of government is to look at consumers as a cost and so government reduces benefits to consumers to reduce costs. Business, on the other hand, recognizes consumers as their source of income so they increate benefits for consumers to keep them as customers to keep them from going somewhere else. When government wants to prevent consumers from going somewhere else they use coercion by passing a law rather than increasing services (increasing services would increase cost. See the rhetoric supporting Obama healthcare).
    Finally, end government regulations intended to protect businesses and investors from the discipline of the market. Government regulatons should be the enforcement of contracts and the protection of people all under the Rule of Law. I know that central planners are horrified that military spending is around 30% of our budget while social spending is only 60%, but in truth in 1960 (before Vietnam and after Korea) the percentages were reversed. That is the proper mix, more national security and less social security. The government should be protecting our nation. Providing wants and needs is best done by the producer in free exchange with other producers.
    Finally, anchor the currency. The best anchor would be gold. Perhaps the greatest friction in our economy is the unstable value of the medium of exchange. The debtor benefits at the expense of the creditor because of excessive credit creation, then the creditor is demonized when he attempts to raise interest rates to break even. But when the excessive credit creation finally runs its course and capital is destroyed, deflation hits and the debtor finds that he cannot service his huge debt because of declining assets values. This is where we are today. A floating currency is more destructive than most realize because it injects uncertainty into every transaction in the economy.
    We could get into more specifics but this is enough.

  6. aaron

    That fixed 22 MPG might distort things a little. MPG varies seasonally and with traffic conditions.
    Fleet MPG declined considerably from 2005 to 2008. It peaked around 21.3 MPG at the end 2005/beginning 2006 [this is total VMT (all road travel) over gasoline consumed]. It fell to 20.9. That’s almost a 2% decline in fuel economy over just a couple years (and our fleet rating probably improved during this time and we didn’t increase the amount of driving we were doing).
    It also improved during 08, after economy tanked and the roads cleared. Back up to about to 21 MPG, though a much smaller improvement given VMT fell below 2006 levels.
    I improved my MPG from ~26.7 to between 28 and 30 MPG by accelerating more quickly and due to decreased congestion (able to move around traffic better and travel at a higher, more fuel efficient speeds).

  7. Strat

    Before you ask “what will this gasoline price change do to consumer sentiment” it seems to me that you must first ask “is this gasoline price change due to a supply shift or demand shift?” If current gasoline price increases are due to a demand shift (caused by e.g., “green shoots”) then the question of their effect on consumer sentiment is really a question about negative feedback during adjustment to the new, generally higher macroequilibrium – i.e., “how do rising energy prices limit the increase in consumer sentiment that is driving them?”
    Your posting on the other hand asks about the effect of an (exogenous?) increase in gasoline prices on consumer sentiment. Where is this gasoline price increase coming from? Is this due to a supply shift?
    Although your graph supports your hypothesis about (exogenously) higher gasoline prices as a catalyst for the beginning of the recession, it also suggests that since September 2008 the short term macroeconomic outlook has driven both gasoline prices and consumer sentiment.

  8. tjgje

    I am curious–
    How much of the commodity price increase (gasoline inlcuded) can be traced back to the FED liquifying a variety of assets from a broader range of financial institutions? FI’s may not be willing to lend to me or you, but they may be willing to let their trading desks use that ‘extra’ cash, or they may be willing to lend it to mutual fund managers or hedge fund managers.
    Granted there are other reasons for the rebound in commodity prices, like China stockpiling commodities and refineries curtailing production. But, if the FED adds some ‘liquid’ fuel to those sparks then momentum and optimism take over and create overshooting.

  9. Marc V

    Thanks for the explanation. Supply-side sounds eerily like voodoo economics. I doubt you will ever see us get back to a gold standard in our current economic paradigm. We may see significantly more bartering in the near future if and when Obama and crew are done wrecking what little value our greenbacks have.
    It was frustrating around this time last year when I wanted to hear a candidate espouse the supply-side economics you gave, but none were available. Both McCain and Obama were campaigning on “I’m not Bush” as well as promising to provide new governement programs (but not on the backs of middle class taxpayers!). Eventually enough moderates moved towards Obama as McCain was viewed as another DC politican, in spite of his “mavericky-ness”. It was good to hear Palin inject some supply-side economics into the campaign, though the MSM was successful at scaring voters away from McCain/Palin.
    I don’t know who’s going to step up next year as a potential candidate promulgating supply-side economics, as most of the Republicans seem to be in the spend-spend-spend crowd.

  10. Bill Conerly

    Two years ago you taught us that gasoline prices have a pronounced seasonality ( The consumer sentiment numbers are not seasonally adjusted, but I’ve never noticed, or heard a discussion of, seasonality in this series. Perhaps you could give us a short tutorial on comparing a series with a pronounced seasonality with one that is not seasonal. Should the gas series be adjusted? Should both be adjusted? Should neither be adjusted?

  11. JDH

    Bill Conerly: Excellent point. The statistical analysis that I’m implicitly drawing on here relates consumer sentiment to the seasonally adjusted energy PCE deflator. I was originally intending to use the latter here as well, but chose not to since it was only available as of February, and I really wanted to talk about what’s happened since. I’m also a little leery of standard seasonal adjustment procedures when applied to something like the recent behavior of gasoline prices– very easy to confuse the big global moves in crude oil prices in the spring of 2008 (which in my mind had nothing to do with a recurrent seasonal factor) with the spring seasonal in gasoline prices.

    The high road is to use seasonally unadjusted data and build your concept of seasonality of either or both series directly into the model. That’s of course a lot of work, which is why we often instead just look at connections between seasonally adjusted series. But you’re very correct to point out that I’m not doing either of these here.

Comments are closed.