Last year’s emissions rebound is entirely due to the base effect, and has no bearing on the successes or failures of climate policy. The global crash in emissions during 2020 was entirely due to a pandemic. Accordingly, the 2021 rebound is an inevitable artifact of rapid monetary and fiscal action that quickly rebooted the global economy. The system simply rebounded towards its last known position, in 2019: a robust and ongoing mix of clean energy deployment that’s heroically pushing uphill against a vast global dependency on fossil fuels. It has been a surprise to see so many headlines and ledes suggesting this rebound somehow makes it harder for the US to reach its 2030 goals. The rebound does no such thing.
There appears to be fairly broad confusion about the kinds of effects and changes that can be ushered in by financial crises. The Gregor Letter was quite adamant at the front end of the pandemic that the ongoing question, say, about the imminent peak in global oil demand would not be answered by a temporary collapse in economic activity. Quite the contrary. Whether the US meets its 2030 climate targets, or whether global road fuel has already hit a peak, are questions that will be answered by the trends already in place before the pandemic; the progress that managed to take place during the pandemic; and how much further policy and effort is made from this point forward. If anything, the pandemic made a mess of statistical trends in employment, consumption, and output. This muddies, and does not clarify, the path we’re following.
Interestingly, the pandemic period was surprisingly constructive if not downright amazing for further deployment of clean energy technology. Car markets in Europe and China bolted towards EV, coal retirements continued apace in the US, global solar and wind growth reached higher highs, and some of the early iterations of new cleantech—say, the decarbonization of steel through hydrogen—began to roll out. Culturally, Americans are now excited about the advent of electric pick-up trucks from Ford, GM, and Rivian. And, the corporate sector continues to clean up its own energy consumption, while also pursuing cleaner operations. The pandemic could have turned out far worse for the cause of climate. Instead, fossil fuel consumption took body blows. In the OECD for example, work-from-home and other constraints on commuting and office work has taken a big bite out of petrol consumption, and it’s now likely road fuel in the OECD will settle at a new, lower level.
The outlook for 2022 couldn’t be more encouraging on the decarbonization front. China looks set to put 6 million new EV on the road, hydrogen applications are flowering and will next appear in the short-haul flight segment, and one of BNEF’s lead forecasters, Jenny Chase, sees a good chance that the largest contribution to global power generation growth this year will come from solar.
The time has arrived to consider the impact of rising EV sales on global petrol demand. Calculations around this question are imperfect of course. EV are being deployed in a wide variety of models and sizes: everything from super-minis in Chinese cities to plug-in hybrids in the US, and then fleet and truck sales across many domains. This means the amount of oil that will be displaced by the EV rollout is quite variable. Let’s give one example. While China is clearly at the leading edge now of EV adoption, the Chinese drive fewer miles than their American counterparts (10,500 miles per year vs 14,000 per year in the US), and the average fuel efficiency of Chinese ICE cars are higher, at roughly 40 mpg vs 25 mpg in the US. We also need to ask: from what starting point should we calculate petroleum savings? Well, China had very good yearly EV sales from 2014 through 2019. But we are in a new decade now, European EV are taking off, and US EV sales are finally about to rise. So, restricting the start point to 2020, China put 1.367 million EV on the road in 2020, and 3.521 million new EV last year. This year, the country is expected to put a fresh 6.000 million EV on the road. Together, that amounts to 10.888 million EV.
Putting those numbers together—miles traveled per year, fuel efficiency, and number of new EV on road in 2020-2022—one finds that each EV displaces 262.5 gallons of petrol per year. A basic oil to gasoline ratio is 2 gallons of oil to produce each gallon of petrol, hence the oil displacement is 525 gallons of oil per year, or 12.5 barrels of oil per year, given that a barrel contains 42 gallons of liquid. Grossing this up means that 10.888 million EV on China’s roads will avoid 136.1 million barrels a day of demand annually, or 372.8 thousands barrels of oil daily. That may not seem like much for a country that consumes 14 million barrels each day—it amounts to just 2.66% of total consumption—but it’s enough to be felt. Why? If the US or China or Europe consumed just 1-2% less oil each year, the global oil market would notice. Going forward, it means that the trajectory of China’s oil demand growth is starting to change.
And it’s not just China. According to BNEF, global EV sales will reach 10 million units this year. Again, some of these will be tiny, others hybrid, but many will be pure 100% battery electric; not just in the passenger sedan and SUV segments but in fleets, utility vehicles, and trucks. Using a conservative average of 35 mpg per displaced ICE unit, at an average imputed distance of 12,000 miles driven annually per unit, finds that the average EV hitting the road this year globally will displace 685 gallons of oil (12,000 miles driven / 35 mpg = 342.85 gallons of petrol, or 685 gallons of oil). So 10 million new on-road EV this year will avoid demand of 0.4468 million barrels per day (10 million new EV units x 685 gallons of oil per unit / 42 gallons/barrel = 16.309 million barrels annually, or 0.4468 mbpd). Nearly a half million barrels per day? That is fairly major. Consider that public and private oil market forecasters are constantly adjusting their outlooks in units of 0.100 mbpd, in a 100 mbpd market that tends to fluctuate year to year in a range of half a million to 1.5 million barrels per day. For example, in the IEA’s latest 2022 Oil Market Report, Africa demand is expected to rise by 0.10 mbpd this year, and Europe’s demand is forecasted to rise by 0.569 mbpd. The avoided oil demand globally this year therefore will not match, but will approach the currently forecasted growth in Europe. That’s a big deal.
Data discussion: It’s possible that imputing 12,000 annual miles driven is too high, but equally that an average mpg of 35 is also too high. Bus, truck, utility, and government car fleets are also turning over to EV, and those are not just high mileage vehicles but are far less fuel efficient vehicles. It’s clear that China will put 60% of new EV on the road globally this year, which displaces ICE vehicles with far better mileage. However. Within China’s total vehicle market lies a 15-20% commercial vehicle segment that is also converting to EV. Finally the US, where the highest oil displacement opportunity lies, will finally start climbing the EV adoption curve in earnest. US EV sales are projected to advance from roughly 500 thousand units last year to 850 thousand units this year. The impact from the US therefore is about to land, as its EV market moves towards and likely above 2 million units in 2025.
Some conclusions. China and Europe are currently way ahead of the US in EV adoption. Their existing ICE fleets are far more fuel efficient than their US counterparts however, so only when EV sales tend to scale to higher levels is the impact on road fuel demand registered. Good news: both markets are absolutely soaring. The biggest hit to oil demand will arrive therefore when the US EV adoption curve, which is still languishing at low levels, begins to rise. That too is about to happen. Indeed, let’s posit a rough formula for the number of barrels per day of oil demand avoided for every EV sold in the US: 14,000 miles driven annually, at an average 25 mpg, equals 560 gallons of gasoline annually, or 1,120 gallons of oil annually, or 26.666 barrels of oil annually, or 0.07306 barrels of oil per day. The US already has about 2 million EV on the road, therefore avoiding 0.146 million barrels per day. Let’s say the US puts 5 million new EV on the road in the four calendar years of 2022 through 2025, which is probably conservative. That will avoid another 0.365 million barrels of oil per day. If this doesn’t seem like much, let’s recall that US oil consumption peaked 17 years ago, and oscillations in its demand of just 0.250 mbpd affect the global oil market.
The time has arrived for the energy agencies like EIA and IEA and other forecasters to start including. In fact, let’s tweet that out:

News Briefs: Georgia Power has accelerated its coal retirement plans, intending to replace the lost capacity with natural gas and renewables. While the utility’s natural gas plans are substantial, the company may conclude, as it moves through the capacity reduction/addition process, that increasing its renewables+storage portfolio is more economic. • The IEA has dialed back its estimate of 2019’s global oil demand to below 100 mbpd, at 99.55 mbpd. Its estimate of 2020 demand is now down to 90.9 mbpd, which means the pandemic crushed global economic activity for a nearly 9 mbpd decline. Last year saw just 5.5 mbpd of recovery, to reach 96.38 mbpd, and IEA forecasts demand will make it back to 99.71 mbpd this year. For those forecasting continued growth in global oil demand—growth that lifts off from the 2019 level and marches forward to sustained higher levels—there is nothing in these data to support that view. • US oil production is expected to recover its former highs in 2023, according to the EIA. The previous annual record was 12.3 mbpd of production. This year will see a recovery to 11.8 mbpd, and next year will see 12.4 mbpd. It’s worth stating: there are no government policies curbing US oil production that exist today that didn’t exist during the past decade. Obama, Trump, and now Biden administrations have at best recorded minor, symbolic shifts in posture towards US oil production output. Same with natural gas. The trend just couldn’t be more clear: for over a decade, the US has emerged steadily as a global giant of fossil fuel production growth, with the inevitable result that exports of petroleum products and now LNG have also steadily grown. • California is estimated to have put about 230,000 new EV on the road last year. (Data comes from the CNCDA, and other sources.) Additionally, this suggests EV are now comfortably above 10% share, in a market that sold about 1.85 million vehicles last year. Lest there is any doubt: global sales of ICE vehicles peaked five years ago. EV are now totally in control of growth in China, Europe, and sub-markets like California, and the UK. The national US market for ICE, while also having technically peaked, is the last piece of the puzzle to move, and is finally making a more pronounced turn right now. •
To fully heal, the global economy needs higher production throughput, and that means getting everyone who’s able back to work. In the US jobs report released Friday, some encouraging news arrived on this measure via an improving participation rate, which crawled back towards pre-pandemic levels. In the updated chart from the St Louis Federal Reserve, see how the rate has recovered to 62.2% compared to the high of 63.4% in early 2020. Note: the scale of the chart is non-zero, and thus magnifies tiny increments. If the y axis here is not clear to readers in its rendering, please see the link to the original.
The participation rate matters quite alot at this juncture in the recovery. Constraints on labor supply are an increasing factor in wage increases, supply chain liquidity, and overall inflation. The extent to which global central banks (the ECB and the Fed mainly) will be obligated to “fight” inflation is largely determined by the participation rate. Increase the rate, and supply of lumber, trade volumes, goods availability, and services readiness will all improve. And if production and output and throughput systems improve, then inflation will fall. So this steady improvement in the participation rates is key. Why? Because in the prior couple of jobs reports, the US Fed faced a nasty mix of conditions in which the supply of labor was bumping—or rather, appeared to be bumping—up against a limit. And if that condition persisted, then the US Fed would be forced to hike rates after concluding that maximum employment had been reached, despite the fact that maximum employment had not been reached. Again, getting able workers off the sidelines fixes this.
But there are some statistical quirks that showed up in the January report, largely due to methodology practices at the Bureau of Labor Statistics. The population estimates used to gauge the participation rate are revised for a new calendar year; but these are not linked or smoothed with the prior year. Moreover, January brings a series of headline revisions to the prior year’s monthly job numbers (the headline numbers you see in the media), and these showed that the Q4 slowdown was not in fact a slowdown at all. For example, you may recall that the December number, reported in early January, was heralded as a very weak number, with only 199,000 jobs created. Ah, but it was not weak after all. Revisions posted Friday by the BLS now indicate that December’s job gain was 510,000 jobs.
What to conclude? First, we must remember the adage that no single jobs report will prove that a trend has ended, or that one has begun. We can’t, for example, trust that the participation rate improved in January because statistical anomalies press downward on the informational utility of just one month’s data. For a fuller discussion of those details, see this blog post by an astute professional who works at the BLS. But taking the other side of this cautionary note, we must also pay respect to the revisions to 2021 monthly jobs data, which shows steady and stable jobs growth all throughout the year—with no dropoff. That single metric alone tells us, or at least provides a now plausible case, that workers are indeed continuing to come off the sidelines. Therefore, the participation rate can be expected to continue its recovery.
Here is another question that remains to be answered: if the global economy is set up to boom, then by how much will the US participation rate need to exceed pre-pandemic levels, to serve demand?
The most important economic data point right now is the US Consumer Price Index (CPI) which will be reported for January this coming Thursday. The US Federal Reserve has remained relatively calm in the face of rising inflation data but the annualized 7.00% print revealed last month was the kind of development that can really challenge one’s resolve. There is no question that Democrats, for example, are starting to freak out about the way inflation is depressing consumer confidence and the public’s sentiment towards the economy. Some sign therefore that inflation has peaked, is peaking, or about to peak is desperately sought right now by politicians, and markets. Where do we stand?
Well, if you look at the progression of CPI throughout last year, note that the final reading of the year (the 7.0% in December) does actually represent a slowing from the leaping pace of 5.4% to 6.2% in October, and the 6.2% to 6.8% jump from October to November. Something like this slowing will need to be seen again, and soon. Given that lumber prices, natural gas prices, and used car prices either slowed their advance or fell last month, there is some reason to hope for a better number this Thursday, 10 February.
On the other hand (couldn’t resist!) gasoline prices, rents, and continued price pressure in goods and food may counteract some of the improving trends in the inflation basket. Taking all this together, however, we can make the following conjecture: if CPI comes in at 7.0% or even 7.1%, that can plausibly be interpreted as a sign that inflation is peaking. But if CPI comes in above that level, then interest rate markets are going to start pricing in more rate increases this year. Note: markets have slightly dialed back their forecasts of policy rate hikes this year, owing to softening economic data not just here, but around the world. As always, what matters here are inflation rates, rather than absolute levels. There is nothing good about inflation running at the 7% level. But if it runs there a second month in a row after a series of steady and much larger month-to-month increases, then markets—being the forward looking mechanisms that they are—are going to price in falling inflation up ahead. Let’s hope that happens.
—Gregor Macdonald
The Gregor Letter is a companion to TerraJoule Publishing, whose current release is Oil Fall. If you've not had a chance to read the Oil Fall series, just hit the picture below.