Rising oil prices over the last two years have put the issue of demand destruction back on the agenda, as producers, traders and analysts try to estimate how consumers will respond.
Demand destruction, a permanent downward shift on the demand curve in the direction of lower demand of a commodity, always becomes a topic of discussion during this stage of the price cycle, and the current discussion resembles previous episodes of high and rising prices in 2005-2008 and 2011-2014.
Brent prices have surged by $47 per barrel (170 per cent) from their low point in early 2016 and are now trading close to $75 per barrel.
Over the same period, weighted-average U.S. gasoline pump prices have risen by almost $1.13 per gallon (61 per cent) and now stand just a few cents below $3 per gallon.
Crude and petrol prices are still well below the levels of $115 per barrel and $3.80 per gallon where they stood just before oil prices started slumping at the end of June 2014.
But crude and fuels are no longer particularly cheap and most traders and oil exporting nations expect prices to increase further over the next year.
In real terms, oil prices are close to the average level for the whole of the last cycle from late 1998 through early 2016.
As the price-cycle matures and prices move towards their next peak, the focus on consumer responses is set to intensify.
In an early sign of political sensitivity in consuming countries, US President Donald Trump commented on rising oil prices via a message on his Twitter account on April 20.
“Oil prices are artificially Very High! No good and will not be accepted”, the president wrote with his customary directness.
In contrast, Opec officials have indicated they see no adverse impact on oil consumption as a result of price increases so far.
“I have not seen any impact on demand with current prices. We have seen prices significantly higher in the past – twice as much as where we are today,” Saudi Arabia’s Oil Minister Khalid Al Falih said in Jeddah.
“Reduced energy intensity and higher productivity globally of energy input levels leads me to think that there is capacity to absorb higher prices,” he said on April 20.
This part of the cycle is normally characterised by a game of "guess the threshold at which rising prices start to destroy oil demand".
In recent weeks, some analysts have suggested demand destruction will begin if and when prices rise above $80 per barrel while others put the threshold as high as $100.
Others express the same idea by suggesting $3 per gallon or even $4 is the psychologically important limit for US motorists.
But identifying a specific price threshold is probably the wrong way to think about the issue of prices and consumption. In reality, there is a continuum of consumer responses to price - ranging from demand stimulation to demand destruction.
The lower prices fall and the longer they are expected to stay there, the more consumption tends to be stimulated. The higher prices rise and the longer they are predicted to stay up, the more consumption tends to be destroyed.
The response of consumption to prices is continuous but highly non-linear. The response also takes time to materialise, as consumers slowly adjust their behaviour and buy new equipment, and it takes even longer to appear in the official consumption statistics due to reporting delays.
Adding to the complexity, oil consumption is also responsive to other factors, including economic growth and incomes; car ownership and vehicle fleet growth; average miles travelled and average miles per gallon.
Some of these factors are themselves more or less related to oil prices, at different timescales, which makes the analysis even more complicated.
For example, oil prices have an impact on choices about fuel economy when new vehicles are purchased.
As a result it is notoriously difficult to estimate the price-elasticity of oil demand and economists have generated widely varying estimates.
But the bottom line is that oil consumption does respond to price changes and the response is not geared to any particular threshold.
The relationship between prices and oil consumption is evident in the global statistics, at least for the high-income countries in the OECD, although it is not so clear for low and middle-income countries outside the OECD.
Oil consumption in non-OECD countries has increased every year since 1970, with the single exception of 1993.
In these countries, rising consumption has been driven by fast economic growth, rising household incomes and increasing vehicle ownership, which has dominated and masked any price effects.
By contrast, in the OECD, growth in incomes and vehicle ownership has been more moderate and the impact of prices on consumption is readily apparent.
OECD oil consumption fell in 1973-74, 1980-83, 2006-2009, 2011-2012 and 2014, all periods associated with high real oil prices.
Conversely, OECD consumption rose very rapidly between 1970 and 1973 and again between 1986 and 1999, when real prices were relatively low.
There are some nuances, including the elimination of oil from heating and power generation during the 1970s and 1980s, and the complicated interaction between the oil shocks and recessions.
But the basic relationship between prices and consumption for the OECD is clear. Oil prices have not usually risen high enough to reduce total global demand because non-OECD consumption has continued growing. Yet high prices tend to temper demand growth through their impact on OECD consumption.
The same basic relationship can be traced between US petrol prices, traffic volumes and petrol consumption, punctuated by the occasional recession.
The decline in petrol prices contributed to a notable acceleration in US petrol consumption growth in 2015-2016 compared with the preceding years.
But petrol consumption was flat in 2017 and is expected to grow by just 30,000 barrels per day in 2018, according to the US Energy Information Administration.
The escalation of oil prices since the start of 2016 has probably started to restrain consumption growth (compared with a baseline in which prices had remained at $30 per barrel).
So far, the demand restraint from increasing prices has been offset by synchronised global growth, especially in the middle-income countries that account for a rising share of oil use. If prices continue to increase, however, there will come a point at which consumption growth starts to slow in a much more pronounced fashion.
Unfortunately, experience suggests the extent of the demand deceleration will only become apparent after it is already well underway.
And the slowdown in consumption growth will continue even once prices stop rising, given the long lags in the system. Between 2011 and 2014, when oil prices averaged over $100 per barrel, declining consumption in the OECD and slowing consumption growth in the non-OECD created the conditions for the last oil slump.
If oil prices continue to increase, as most hedge fund managers and oil-exporting nations expect, the same scenario could play out again between 2019 and 2021.