Friday, April 27, 2012

11 April 2011 – Financial disincentives vs. non-discretionary spending


President Obama raised hackles last week when he offered a glib response to a questioner at a town hall meeting in Pennsylvania who was concerned about the rising price of gasoline.  The official White House transcript of the event (note A) reads as follows:

Now, I notice some folks clapped, but I know some of these big guys, they’re all still driving their big SUVs. You know, they got their big monster trucks and everything. You’re one of them? Well, now, here’s my point. If you’re complaining about the price of gas and you’re only getting eight miles a gallon -- (laughter) -- you may have a big family, but it’s probably not that big. How many you have? Ten kids, you say? Ten kids? (Laughter.) Well, you definitely need a hybrid van then. (Laughter.)

“Laughter”, eh?  For the record there is no hybrid vehicle on the market today that will carry a driver and ten children.  And while the President loves to criticize SUVs, he might want to look out the back window of his three-tonne armoured limo - which gets considerably less than 8 miles to the gallon, I might add - at the dozen or so Chevy Suburbans that follow him everywhere, and ask himself how many of those Secret Service folks, D-Boys and assorted strap-hangers you could cram into a Chevy Volt.

But I digress.  The point is that this town hall exchange makes it clear that the Obama Administration, like many other political entities (on both sides of the spectrum) appears to believe that consumer behaviour that they define as undesirable - in this case, consuming fossil fuels for transportation purposes - can be altered through financial disincentives.  This belief betrays something about their attitude re: the free market, because the same thesis has traditionally been applied to so-called “sin taxes” on products like tobacco and alcohol.  The comparison is strangely apt, because sin taxes demonstrably affect consumption only at the margins.  While casual consumers may reduce consumption of “sinful” products in response to price increases, individuals who are physically addicted to cigarettes and alcohol are not disincentivized by growing costs.  Rather, they support their increasingly expensive addiction by reducing expenditures in other areas, often with catastrophic personal and societal results.

This raises an interesting comparison.  In contemporary Western society, how big a role does consumer choice really play in fossil fuel consumption?  In other words, is consumption of fossil fuels a casual purchase, or are we more akin to addicts?  I use the latter term deliberately, as Western society is routinely and roundly condemned for being “addicted to fossil fuel.”  Obama, in remarks addressing last summer’s oil spill in the Gulf, cited “America’s century-long addiction to fossil fuels” as an argument for favouring “green energy” over exploiting offshore and other domestic oil resources.(note B)  If this is true - if we are in fact “addicted” to fossil fuels - then choice is less of a factor in energy consumption than people think, and it is illogical to expect financial market disincentives to alter consumer behaviour.  That, after all, is what “addiction” means.

This is fairly easy to check out, actually.  All we need to do is take a look at US fuel consumption patterns over the past several decades. 

Figure 1 - US Fuel consumption patterns 1991-2011
(source: Energy Information Administration, Note C)

For ease of comprehension, I’ve only listed the two most significant types of fuel consumed in the US - distilled motor gasoline (including diesel) and distilled fuel oil - over the past twenty years.  Three features pop out of this graph.  First, fuel oil consumption peaks in the winter months and bottoms out in the summer months; while gasoline consumption peaks during the summer months and bottoms out in the winter months.  No surprises there; winters are cold, and summers are when goods are shipped and people travel.  Second, fuel oil consumption in Feb-Mar of 1994 was atypical, which is also not surprising, as 1994 was an abnormally cold winter in large areas of the US.(Note D)  Third, the effect of the current recession on fuel consumption is clearly evident; gasoline consumption has declined somewhat, and fuel oil consumption has declined enormously, since the recession began in the fall of 2008.  However, it is worth noting that the proportional decline in consumption is much less for gasoline and diesel - transportation fuels, in other words - than it is for fuel oil.  One reason is because fuel oil has many industrial uses beyond heating homes, and the decline in fuel oil consumption reflects the enormous increase in unemployment and the decline in industrial output associated with the recession.  The divergence in the two curves post-2008 demonstrates that transportation fuel consumption is less likely to be affected by a severe economic downturn than fuel oil consumption.  To put it another way, the recession hasn’t caused anywhere near as large a drop in transportation as it has in industry.  Even if they aren’t working, after all, people still have to move, and they still have to eat.

Okay, so how robust is fuel consumption in the face of price changes?  If consumption of transportation fuels is a matter of choice, then, ceteris paribus, as practitioners of the ‘dreary science’ say, an increase in price should result in a concomitant decrease in consumption, right?

Figure 2 - US Fuel (diesel and gasoline) prices vs. total fuel supplied 1991-2011
(source: Energy Information Administration, Note C)


The blue and yellow lines in figure 2 represent, respectively, the price per gallon of gasoline (all grades) and diesel fuel in constant 1982-84 dollars (I had to use 82-84 dollars as that was the only consumer price index data available from the Bureau of Labor Statistics for the whole of the period in question - Note E).  The pink line represents the total of diesel and gasoline supplied for consumption, while the black line is a four-period moving-average smoothed trendline for total fuel supplied (I used four-period smoothing because the data was in weekly increments.  The smoothing establishes monthly patterns, offset to the right by one month.  As you can see, the black line more or less replicates the pattern shown by the yellow [transportation fuels] line from Figure 1).

Okay, what does figure 2 show?  Well, as the pink and black curves demonstrate, transportation fuel consumption cycles annually, high in the summer period, and low in the winter period.  Apart from a gradual increase over the past 17 years, the only significant change was the decline in consumption due to the current recession.  Post-2008, however, we can see a slight recovery during peak consumption periods (i.e., summers) but not during low consumption periods (i.e., winters).  The blue and yellow lines, by contrast, show fuel prices in constant dollars.  Gasoline and diesel prices map so closely on an annual basis that there’s little point in treating them separately. 

We can extract a number of interesting lessons from fuel price behaviour over the past 20 years.  First, fuel prices adjusted for inflation were pretty much the same in 2004 as in 1994.  There was a noticeable bump from 2000-2002, which correlates with the economic disruptions due to the Dot-Com scandal and 9/11; a major increase in 2008; a major drop-off shortly thereafter; and a rapid increase from the recessional trough to the present.  The massive increase took place from summer 2007 to summer 2008, at the same time as grain shortages resulting from cropland diversion to ethanol production led to massive increases in worldwide food prices; and the massive decline correlates precisely with the collapse of the sub-prime mortgage industry in the US that touched off the current recession.  The more recent increase in prices has been linked to increasing demands from industrializing nations like China and India; the 2010 oil spill in the Gulf of Mexico; supply concerns resulting from ongoing regulatory constraints on resource exploitation in the US; and, more recently, the explosion of popular discontent in the Middle East.

What’s interesting in this graph is the obvious lack of a significant correlation between the price of transportation fuels and fuel consumption patterns.  Look at the first half of the chart.  The 2000-2002 price bulge produced no detectable change in the consumption curve.  If fuel was truly a pure market commodity, a “nice-to-have” good that was subject entirely to consumer choice, then logically an increase in price should have led to a decrease in consumption, or vice-versa (or, if you like, the other way around - an increase in consumption should produce an increase in demand and therefore in price, and vice-versa).  No such thing happened.  One of two interpretations is possible: either the cost of fuel for the average American consumer represents proportionally such a miniscule percentage of total household expenditures that price changes are relatively meaningless (i.e., fuel is cheaper than dirt); or the purposes for which fuel is purchased are so important that price is relatively meaningless, and the consumer will continue to consume it in accordance with existing patterns, regardless of price changes (i.e., fuel is a necessity).

Well, these are also easy interpretations to check out.  According to a recent Reuters report, increases in motor fuel prices are going to raise the average US household’s expenditures on motor fuel by $700 (28%), to $3235, for 2011.(Note F)  Given that the median income in the US in 2009 was $50,000 (US Census Bureau), motor fuel costs represent more than 10% of the average family’s after-tax income.  This is a significant proportion of the average family’s expenditures and eliminates the “fuel is cheaper than dirt” argument.

Does this mean that fuel is a necessity and that consumption, therefore, is less dependent on price than we think?  Take a look at the right half of the graph in Figure 2.  Looking closely at the black moving-average trendline, we can see a very slight deceleration in the aggregate increase in fuel consumption from 2004-2008.  During this period, the real cost of fuel doubled.  The fact that a doubling in price correlates with only a barely-detectable decline in consumption further reinforces the impression that fuel consumption is poorly linked to fuel price.  This in turn suggests that consumer decision-making about fuel purchases is driven largely by factors other than price.

As a gross error check on this reasoning, think about your own driving patterns.  How many of the kilometres that you drive in a week are truly “discretionary” - i.e., how much of the driving that you do is done for purposes that could easily be eliminated without inflicting a significant impact on your personal life, to the point of forcing a “step-change” in your behaviour?  How much of your driving is because you want to, not because you have to?

If the bulk of transportation fuel purchases are indeed effectively non-discretionary, then there is only a very limited extent to which consumer behaviour patterns vis-à-vis transportation fuel consumption are subject to alteration through the imposition of artificial price increases regardless of how they are accomplished; direct taxes (like carbon taxes levied at the pump) or indirect taxes (like price increases resulting from trading in carbon credits purchased by petroleum distillers, with costs passed on to consumers), even if enormous, will not have the desired effect.  This is the difference between a luxury and a necessity.  If the price of caviar goes up, chances are you’ll eat less caviar; and if it goes up enough, you might well phase it out of your diet.  Nobody needs caviar, after all.  If the price of wheat goes up, though, chances are you’ll still buy bread; you’ll just have to adapt by reducing consumption in areas that are more discretionary.  You’ll see fewer movies, put off buying that X-Box, or not purchase the sofa you’ve been saving for.  You don’t really have a choice about buying bread, you see, because the alternative is starvation.

And as the data demonstrate, you largely don’t have a choice about driving your car.  The little discretionary things - like that trip to Disney World you wanted to take the kids on - will disappear.  But that’s less than 10% of the gas you’ll buy in a year.  And a 10% reduction at the margins of consumer behaviour will have a negligible impact compared to much larger effects - like the massive reduction in fossil fuel consumption imposed by a crushing economic crisis; or the massive increase in fossil fuel consumption occasioned by increasing demand for fuel in the growing economies in Asia.

When 90% of your purchases in a given sector are non-discretionary, you’re an addict.  So the bottom line, I guess, is that for all practical purposes, we ARE addicted to fossil fuels - at least in the sense that price changes do not appear to significantly impact consumption patterns (looks like I agree with Obama on something.  Somebody write the date down).  Which is why trying to modify behaviour through price manipulation won’t work.  Forcing consumers to pay higher prices for largely non-discretionary commodities like transportation fuel will not lead to a significant reduction in fuel consumption; it will simply lead to a lower standard of living by forcing them to make sacrifices in discretionary areas.  Spending an extra $700 on gas simply means that the family won’t be able to afford the nice-to-haves, like sending little Johnny to hockey practice.  Which of course means that Mom won’t have to drive him there, which means that she’ll save the expense of the fuel that would have been consumed, which in turn means a concomitant reduction in those greenhouse gas emissions Obama was talking about when he advised that Pennsylvania man to buy a hybrid to transport his 10 children in.

Of course, any family that is having difficulty affording a $700 increase in annual fuel costs probably isn’t in a position to spend $40k on a Prius.  But, hey…details.