An interesting graph came up on the TransportBlog twitter feed a couple of weeks ago, courtesy of the Wall Street Journal:
Pretty self-explanatory, but what you’re seeing is that gasoline (or petrol) consumption in the US grew until the mid-2000s, and has since fallen. The projections from the government’s Energy Information Administration, which a decade ago were assuming that people would keep consuming more and more petrol, started to change tack in 2010 (assuming pretty flat demand) and even more in 2014 (assuming that demand would keep falling). This is the long, slow decline of petrol demand in the US, although some might think that decline actually looks pretty fast.
There’s another graph that goes along with the one above – they both come from this report, page 246ish. The other graph looks at consumption of all oil products (“petroleum” as they call it in the US).
Here, the story is similar but with a slower decline expected (fuels like diesel and other uses of oil are likely to be less affected by the demand downturn).
The US report gives the following explanation (forgive the wall of text, apparently paragraph breaks were high on the government’s cost-cutting list when the GFC hit):
The decline in petroleum consumption, starting in 2006, was unexpected. In the case of energy, industry-standard benchmark projections are produced annually by the Energy Information Administration (EIA) in its Annual Energy Outlook. Revisions to those projections include the effects of unforeseen developments in the energy sector. Figure 6-2a shows U.S. petroleum consumption since 1950 and projected consumption from the 2006, 2010, and 2014 editions of the Annual Energy Outlook. Only nine years ago, EIA projected an increase in petroleum consumption during the subsequent 25 years. But events dramatically affected subsequent projections: by 2010, EIA had reduced both the level and rate of growth of its projection; its 2014 outlook now projects petroleum consumption to decline through 2030 after a slight increase over the next five years. The reversal in projected petroleum consumption is led by the reversal in actual and projected gasoline consumption (Figure 6-2b): the 2014 EIA projection of consumption in 2030 is 44 percent below the projection made in 2006. Actual gasoline consumption declined between 2006 and 2010 mainly due to the recession and rising fuel prices, but much of the revision to the 2030 levels reflects the largely unexpected fuel economy improvements stemming from the Energy Independence and Security Act of 2007 and the Administration’s subsequent tightening of those standards. The 2014 projections further reflect the 2012 light-duty vehicle fuel economy and greenhouse gas emissions rate standards, which apply to model years 2017 through 2025 [JP note: the Obama administration tightened the Corporate Average Fuel Economy standards in the US for the first time since the 80s, incentivising US manufacturers to improve their woeful fuel economy]. The Administration’s fuel economy and greenhouse gas standards for medium and heavy-duty trucks also contribute to the reduction in projected petroleum consumption between the 2010 and 2014 Outlooks.
Let’s look at the same stats for New Zealand. It’s a little tricky, since the MBIE hasn’t produced forecasts of oil and transport demand since 2011, but long story short they were still projecting pretty steady increases for transport fuels (mainly petrol) and oil demand at that time.
Historical statistics, though, show that NZ’s petrol consumption peaked in 2007 and have begun a slow decline since then (although 2014 figures, not yet out, are likely to show a slight uptick from lower petrol prices towards the end of the year).
It seems pretty likely that New Zealand is going to see a long, slow decline in petrol demand as well.
Even petrol companies are now talking about how fuel use and driving are falling. The Herald reports:
Z Energy says petrol consumption is falling in relation to increasing availability of broadband.
While more fuel-efficient vehicles and rising petrol prices have also contributed to consumption falling from its 2007 peak, Z’s chief executive, Mike Bennetts, said demand was more sensitive to broadband connectivity than these traditional factors.
“People are doing less discretionary motoring and that may be about the price but what we have found is quite a strong link between broadband connections and fuel consumption,” he said.
“People are doing online shopping and Skyping granny rather than making the fortnightly visit.”
A 1 per cent improvement in broadband connectivity is estimated to cause a drop of 200 million litres a year in national fuel demand, more than the impact of GDP growth, population, fleet turnover, vehicle efficiency and the petrol price.
This is very much in line with some of the things we’ve been saying for a while. The world is changing and the Internet in particular is having an impact on how much and often people travel. This is especially the case amongst young people. This also matches what we’ve seen in other transport metrics with the number Vehicle Kilometres Travelled per capita (VKT) falling back to below its 2001 level.
The Ministry of Business, Innovation and Employment (MBIE) publish regular information about fuel consumption and this is shown in the graph below by fuel type.
Z are also predicting that petrol consumption will continue to fall
Fuel consumption peaked at 4.2 billion barrels a year in 2007 but is estimated by Z and the Ministry of Business, Innovation and Employment to fall to 3.6 billion barrels within the next 10 years.
Z has between 25 per cent and 30 per cent of fuel sales, and data compiled for an investor day shows light vehicle travel per person has fallen 6 per cent since 2005.
However, diesel use is increasing and while historically pegged to GDP growth is now surpassing that.
Undoubtedly there has been some shift from petrol to diesel which has contributed to the faster than GDP growth.
Of course one thing that isn’t mentioned in the article but will likely be having a small but growing impact will be that of the growth in PT in Auckland which has tended to outstrip population growth in recent years. About the only thing certain in transport these days is that we can no longer just assume things will just be a continuation of past trends.
OK, I admit I haven’t come up with the catchiest title for this post, but it’s a good TL;DR. As I’ve written previously, world oil prices were pretty dang low for 20-odd years: from the mid-80s to the mid-2000s. The price of a litre of petrol in New Zealand reached its lowest point in 1999, and started to rise from then on – with major rises in the last decade.
Petrol Consumption Trends
Our consumption of petrol grew and grew, until 2004. Since then, it’s been going in the downward direction, and more so since the 2008 recession began. Other bloggers have written about related issues such as the decline in drivers’ licenses being issued, less distance being travelled, and lower vehicle ownership.
The trends weren’t confined to New Zealand, either. Kent has written about declining travel in the US, and in fact there have been declines, or at least less growth, throughout the OECD. And it all started happening around 2003-2005.
Various explanations have been floated for this. Could it have been the millenials, too busy spending time on Facebook and smartphones to drive? Was it the recession? Perhaps online shopping?
Personally, I don’t think it’s a coincidence that vehicle travel and fuel consumption started to decline all around the OECD, at around the same time, just as oil prices started to go up dramatically. I consider these higher prices to be the main factor in reduced driving – certainly up until recession hit in 2008. Higher prices were the main kicker which may have started other shifts, e.g. young people becoming less likely to get a license.
Furthermore, fuel prices are likely to stay high, and in fact to increase over time. You would hope that governments would take this into account when determining its transport policy – although ours doesn’t seem very interested in doing so.
Some countries have reported more of a decline in driving amongst young people, and I don’t think this is a coincidence either. Young people would have been particularly hard hit by high prices – their lower incomes mean they had to cut back more, and their travel was often discretionary. Let’s say you are young enough to live at your parents, and you get $20 a week to do whatever you like with. Well, that $20 buys half as much petrol as it would have ten years ago. You can blow your entire pocket money just in getting to a few friends’ houses. It’s no wonder that fewer people are getting licenses. Another factor is that, post-2008, many countries are struggling with high youth unemployment.
In 2007, the excellently named Booz Allen & Hamilton did some research on just how much fuel prices affect travel. They estimated the short run elasticity of petrol consumption at -0.15, or -0.2 in the medium term. Elasticity is a commonly used concept in economics, and you can check out the Wikipedia page for more info – but in brief, an elasticity of -0.2 means that a 10% increase in the price of petrol is associated with a 2% decrease in petrol consumption. If something goes up in price, we buy less of it, right? Petrol is one of those things that we’re not very price sensitive to – the demand is “inelastic” – but it’s naive to think that higher prices don’t have an effect.
And, over the last decade, the increases in petrol prices have been massive. In real terms, they’ve gone from $1.27 in June 2003 to above $2.04 in June 2013. That’s a rise of 60%. In nominal terms, the increase is even larger. And if you look back to June 1999, when prices bottomed out, there’s been a real increase of 80% since then.
Holding other things constant – i.e. forgetting about economic growth, population growth and so on – the numbers from Booz Allen & Hamilton suggest that per-household petrol consumption would have fallen by 12% since 2004. That’s not too different from what we’ve actually seen.
Notwithstanding that elasticity estimates can be a lot less accurate when you get such large percentage changes, we’re certainly talking the right order of magnitude. I think a regression analysis would be likely to show that the main factor in decreased petrol use and driving is, indeed, higher prices.
Taking It Further
I want to point out that a lot of the non-academic debate on changing travel trends has focussed on things other than prices. In fact, even some of the ‘academic’ stuff seems to have ignored prices – see the paper here, although that only presents arguments rather than any kind of quantitative analysis. As an economist, I’m quite stunned by this. Transport fuels are a major part of household budgets (more than $2,300 a year on average), and prices have risen massively in the last ten years, and people don’t think that’s had an impact on the way we get around? I know I drive differently than I would if petrol was still $1.20 a litre. My daily patterns mightn’t change much, but I’d probably go on more road trips, more spontaneous drives in the weekend and so on. I’ll be looking at this more in the future.
Most people who drive will likely have noticed that the price of petrol in recent years has not exactly been as friendly on the wallet as it was a decade or so ago. In fact other than a three week stint in July last year where it only just dipped under, the average weekly price of regular 91 hasn’t been below $2 per litre since February 2011. The price per litre is shown in the graph below.
What’s interesting is to see the impacts that this is having on peoples spending.
Stats NZ release monthly figures on the amount of money spent on electronic cards across the country. This doesn’t include purchases using cash, vouchers etc. but does give a break down by various industries and so provides a good indication as to where people’s money is being spent. The data is broken down by
- Core Retail
- Non-retail industries (excl services)
The fuel industry sits under retail and as it’s broken out it means we can get an idea of just what impact the cost of increasing fuel prices are having on peoples spending – and the result isn’t good.
First of all here is the percentage of retail spending that went on the fuel industry based on the 12 month rolling totals and it’s gone from 10.5% a decade ago to just over 16% now. In other words we are now spending a much greater proportion of our money on fuel than we have in the past.
The next graph shows the change in the fuel industry spending compared to the rest of the retail category indexed back to September 2003. Again it shows that spending on the fuel industry is rising faster than our spending elsewhere.
However I understand the fuel industry figures are likely to included spending in stores associated with petrol stations i.e. people buying inconvenience items. The graph below how the percentage on the fuel industry compared to the average price of petrol and as you can see the two generally tend to follow each other. This graph is based off the monthly results so the big downward spikes will be the impact of Christmas increasing retail spending in other areas.
So why is this important? Well the money that is spent on transport – which is probably mainly used for commuting – is money that can’t be spent elsewhere in the economy. All up New Zealanders spend about $52.5 billion on cards in the 12 months to the end of October (which is double what we spent a decade ago). Imagine if over the last decade we had better developed options to allow people to get around without needing to always drive that had allowed our spending on fuel to stay at 10.5%. The difference between that and the percentage we spend now would be about $3 billion a year (on cards alone) which is almost 40% of what we spent on hospitality over the same time period. Imagine what that would do to our liveability and happiness if people could afford to spend more on activities or what would it do to lower income families who are struggling.
* subject to various caveats below!
I’ve just been looking at some inflation data for the USA, and I was a little surprised to see that, in real (inflation-adjusted) terms, gasoline prices reached an all-time high last year. Higher than the oil shock days of the’70s – early ’80s, in fact.
I checked this against some other sources, and it seems to stack up – the Energy Information Administration and some other chap reached the same conclusion (although the other guy reckons they were higher in 1918 – my data doesn’t go back that far). Both of those sources suggest that 2012 prices were only a little higher than they were in 1980/1981, and that prices in 2008-2011 were lower than 1980/1981 – so their data might be a bit better than mine (which is CPI data for urban areas – maybe they have slightly higher taxes on gasoline than the national average). Interesting nonetheless!
Of course, “real” incomes in the US have risen since the early ’80s, so gasoline is arguably a bit more affordable today than it was then. But it’s still easy to see why vehicle travel in the US has dropped off in the last few years, and while they’re finally investing in public transport.
The situation’s a little different in NZ; “real” petrol prices aren’t quite as high as they were back in the day, thanks to our strong dollar. That could all change, though, if the dollar drops – or if oil demand keeps rising.
Over the recent week, the price of petrol has moved up close to its all-time record high, in nominal terms, and is sitting at around $2.20 per litre. Yet at the same time the New Zealand dollar is also close to an all-time high vs. the US dollar, sitting at around 85c. This raises an interesting question of what petrol prices would be like if our exchange rate wasn’t so good for imports. The good thing is we can make an educated guess thanks to the work the Ministry of Business, Innovation and Employment do in their weekly oil price monitoring. At a high level the price of petrol is made up of three components, the cost to import the fuel, our taxes and levies and the importer margin (which covers everything from distribution and retail costs to profits).
It is the first of these components – the cost to import the fuel – that is most directly affected by the exchange rate, and handily in its weekly data the MBIE include the average weekly exchange rate. From that we can work out what the cost per litre would be in US dollars and then reconvert that to NZ dollars at a different exchange rate. I worked it out for each week over the last three years for two scenarios – if our exchange rate vs. the US was 5c and 10c less than it is now. I have then applied the extra GST that would be paid to that figure to get a result. Averaged out over 3 years, this shows that if all other components stayed the same, with the only thing changing being the exchange rate, an exchange rate drop of 5c would see petrol increase by ~7.5 cents per litre. A drop of 10c would see petrol increase by ~16c per litre.
There are a few key things that could cause the dollar to drop. The first is if the NZ economy weakens, the second if the US economy improved. Not only would this see our exchange rate drop but it would also see the demand for fuel in the US increase. The third thing that could happen is a major economic or political shock overseas, for example a political crisis in the Middle East. As we know these things can happen quite fast and when they do, investors run away from higher risk currencies like the NZ dollar to the perceived safer economies such as the US, Japan and Europe.
In the case of a middle east crisis, that would have a double whammy of increasing the cost we have to pay not only due to the exchange rate but also in the raw cost as investors become worried about supply issues (there could actually be a triple whammy if we also had improvements in offshore economies, particularly the US as that would also drive up demand). If such things were to happen it would be hard to say just what price our petrol could rise to, but it is interesting to note that we have seen it before.
Back in the middle of 2008, just before the GFC hit, higher raw costs combined with a lower exchange rate (9-10c lower than now) saw the importer cost peak to 20-30c per litre higher than it is now. At the time the price at the pump reached $2.12 and the reason it was less than we have now is due to the other two main components – the taxes and levies, and the importer margin.
Looking at the taxes first, a couple of key things have happened since that time 4½ years ago. Increases in the Fuel Excise Duty (FED), along with the introduction of the Emissions Trading Scheme have seen fixed taxes increase by almost 9.5c per litre. Another key aspect is that over this time period GST has increased from 12.5% to 15% and of course GST is applied on top of both the importer costs and the other taxes.
There has also been quite a change in the level of importer margin. Remember that this isn’t just the profit margin but is used to cover the rest of the oil companys’ local operations, such as distribution and retail costs. Back in 2008 it was as low as 9c per litre however the long term average prior to that (and for a few years after) saw the margin sitting at around 14c per litre. In fact it didn’t reach over 20c per litre until the middle of 2010. In recent years this has changed – the average over the last few years has been ~21c per litre with the average over the last year up at ~24c per litre. I don’t know if the increase is due to a rise in the bottom line for these companies or if it is due to increased costs or perhaps compensation for the proliferation of discount fuel vouchers that are now in the market.
Thinking about the future there is one other factor that needs to be taken into account. Just before Christmas the government announced that starting 1 July this year – petrol would increase by 3c per litre every year for three years. It is to help pay for many of the uneconomic RoNS projects and is also partly to combat falling tax revenues due to people driving less. What all this means is that in a few years time we could quite easily see the price of petrol increase by as much as 30c per litre. That would put the price of petrol up around $2.50 per litre. This isn’t meant to be a kind of alarmist post and is really just to highlight that petrol could quickly rise. If this were to happen, it would be interesting to see what impact that would have on people’s preferences to drive. It’s also a great reason why we should be investing in modes of transport that give people more choice in how they get around.
One last point before anyone starts commenting that we are over taxed for petrol. The MBIE also produce this graph quarterly which shows that the fuel taxes we pay are still much lower than most other developed countries.
I have added two more graphs to our transport stats collection showing the change in petrol prices over time. The data for both comes from the Ministry of Business, Innovation and Employment from their weekly, quarterly numbers. The quarterly numbers include the inflation adjusted figures which is very handy for comparing them over a long period of time and shows that in real terms, we still haven’t reached the peaks of the 80’s yet.
Edit: updated the first graph to add in a break down of the fuel prices.