Oil Prices and Growth: Is This Time Different?
Global growth expectations have dropped with the spike in oil prices, which have historically often been associated with surges in inflation, reductions in real income, and eventually recessions (Exhibit 1).
However, the relationship between oil prices and growth has changed for two major reasons. On the one hand, the 50% reduction in the energy goods and services share in consumer spending since the early 80s in the US and Canada has reduced the negative effect of oil price increases on GDP growth via consumption (Exhibit 2, left panel). On the other hand, the sharp drop in energy capex as a share of GDP compared to the early 80s and the 2010s in the US and Canada has reduced the positive effect of oil price increases on GDP growth via business investment (Exhibit 2, right panel).
In this Analyst, we revisit the impact of oil price changes on GDP growth in the US, the Euro area, and Canada accounting for various structural changes. We estimate how the impact of a $10/bbl change in the real price of oil (expressed in January 2022 PCE dollars) on GDP growth over the next year has changed over time.1 We estimate the
total impact of oil price changes on GDP as the sum of the impacts on 1) consumption in all three economies via changes in inflation, 2) energy business investment in the oil producing US and Canadian economies, and 3) spillovers in Canada to consumption (via wage growth and thus income) and to non-energy business investment.
We estimate the consumption effect of higher oil prices via price inflation in three steps.
We first estimate the impact of higher oil prices on headline inflation using regressions of energy inflation and our rules of thumb for spillovers to non-energy inflation categories. Second, we translate these changes in energy and non-energy inflation into changes in real spending power using data on spending shares, which change significantly over time (Exhibit 2, left panel). Third, we assume that the cumulative decline in real consumption equals the loss in real income. This multiplier of 1 is broadly consistent with household-level studies of the propensity to consume out of oil driven real income shocks and CBO estimates of the multiplier out of tax cuts for low and middle income households. We also use our fiscal impulse model for the timing of the spending effects.
Exhibit 3 shows our results.
Looking across countries, we estimate that a $10/bbl rise in the real oil price subtracts 0.15-0.20pp from growth via higher inflation and lower real household income in the US, and 0.15pp in the Euro area and Canada.
Why are these GDP effects via consumption so similar across countries despite a much higher share of energy goods and services in consumer spending in the Euro area (11%)than in Canada (6%) and especially the US (4%)? The main reason is that European consumer energy inflation responds less to oil prices than in the US and Canada (Exhibit
4, left panel). Much higher taxes per liter of gasoline in Europe (right panel) and a larger reliance on natural gas and coal help explain the smaller oil sensitivity of energy inflation in Europe.4 Additionally, household consumption accounts for a smaller share of GDP in Europe than in the US and Canada. Why are the US GDP effects via consumption only slightly more negative than in the Euro area if US oil consumption per capita is just above 2 times higher than in Europe? The simple reason is that GDP per capita is also nearly 2 times higher in the US (at least on a current $ market basis) than in Europe.
Looking over time, we estimate that the fall in the energy consumption share has halved the negative growth effect via consumption per real $10/bbl since the early 80s in North America (Exhibit 3). Although much of this reduction happened in the 80s and early 90s (before European data are available), this trend has continued since then in Canada and the US.
In contrast, the negative growth effect via consumption has been broadly stable in Europe, where the consumer spending share of energy items has actually trended up (Exhibit 2, left panel). In turn, the trans-Atlantic divergence in energy consumer spending shares largely reflects increases in European spending on natural gas (due to faster price growth), and electricity (due to rising taxes), and slower non-energy nominal consumption growth.
Energy Capex Effects
The main offset to the growth drag from reduced consumption is the boost to the energy capital expenditures from higher oil prices. We estimate this impact of oil prices on growth via energy capex in both the US and Canada, which are major oil producers, using a simple model. We estimate the positive oil price growth effect via energy capex as the product of 1) the energy capex GDP share—which varies significantly over time (Exhibit 2, right panel)—and 2) the effect of real oil price changes on energy capex growth. Using subsamples, we find that the sensitivity of US energy capex to real oil price changes has declined substantially recently relative to the two shale capex booms of the 1980s and the 2010s. In particular, growth-oriented oil companies kept boosting their capacity on surging oil prices in 2010-2014.5 Since then, oil producers’reinvestment rates have dropped from 140% in 2015 to 40% in 2021 (Exhibit 5, left panel). This shift reflects investor focus on higher returns and ESG. A March Dallas Fed survey suggests that even the recent changes in oil prices and geopolitics are unlikely to fundamentally change the value-orientation of US public oil producers (right panel).
Taken together, we estimate that the positive growth effect via energy capex from a $10/bbl rise in the real oil price has declined by two-thirds since the 2014 peak in the US and Canada to 0.05pp and 0.1pp, respectively (Exhibit 6).
Canada Spillover Effects
In Canada, we estimate two additional positive growth effects from higher oil prices. We find that Canada’s large oil sector generates positive spillovers to household income and thus consumption, and to investment in downstream and upstream sectors.
On the consumption side, we estimate that a $10/bbl increase in the real oil price leads to a cumulative 0.25% boost to wage growth (Exhibit 7). Together with our assumption of a multiplier of 1 and a roughly 50% labor share, this implies an additional boost to growth via the consumption wage channel of 0.1-15pp per $10/bbl increase in the real oil price.
On the investment side, two different approaches both find a boost to GDP growth via non-energy capex of around 0.1pp per $10/bbl, that is relatively stable over time. The first approach regresses annual non-energy capex growth on supply-driven changes in real oil prices.6 The second approach defines the non-energy capex effect as the difference between the estimated contributions from the effects on total capex (estimated from regressions on quarterly data) and on energy capex.
New Rules of Thumb
Exhibit 8 shows our oil growth rules of thumb, which add up the various consumption and investment effects, shown separately in Exhibit 9. Specifically, we estimate that a $10/bbl rise in the real oil price lowers growth by 0.1-0.15pp over the next year in the US, by 0.15pp in the Euro area, but boosts growth by 0.15-0.2pp in Canada (averaging the 2021 estimates).
The negative effect of oil prices on US growth is now significantly smaller than during the recessions of 1990-1991, 2001, and 2007-2009. However, the negative effect on US growth is now significantly larger than in the 2010s and 1980s, when energy capex accounted for a larger share of GDP and was more responsive to oil price changes. Similarly, we find that the net positive effect of oil price increases on Canada growth rose from the early 90s until it started falling in 2015 along with energy capex. Finally, we estimate that the Euro area growth effect has been relatively stable, fluctuating between 0.15 and 0.2pp per $10/bbl.7 The intuition for the relatively similar latest US and Euro area effects is that the US energy capex offset has declined in importance, and that taxes tend to stabilize energy inflation in Europe.
We conclude by estimating the effect of recent oil price changes on sequential GDP growth. Our oil growth impulses incorporate our commodity strategists’ forecast that Brent picks up to $125/bbl this quarter and stabilizes at $115/bbl next year. Our oil-only impulses do not incorporate the additional drag from disproportionately large increases in natural gas prices and potential gas disruptions in Europe.
We estimate a negative oil price effect on 2022 Q4/Q4 GDP growth of ½ pp in the US and ½-¾ pp in the Euro area. The estimated oil drag on annualized growth peaks in both the US and the Euro area this quarter. In contrast, we estimate a sizable and front-loaded ½ pp boost to 2022 Q4/Q4 growth in Canada. Generally, our oil impulses also peak in Q2, which further increases the importance of upcoming data releases on consumption and investment.
The risks around our estimates are two-sided.
The negative growth effects from higher oil prices could be larger if the inflation boost is disproportionately large, higher input costs depress non-energy investment, the net impact on monetary policy is hawkish, or significant supply-side damage occurs. In contrast, the negative growth effects via consumption from higher oil prices could be smaller if healthy balance sheets or government support of energy bills limit the consumption drag. Finally, the positive growth effects could be larger if geopolitical changes boost the price sensitivity of energy capex in fossil fuels, LNG, or renewable energy.
With these risks in mind, we conclude with three implications for the growth outlook.
First, the sizable oil drags in Europe and the US support our below consensus 2022 growth forecasts in the US (1.9% on a Q4/Q4 basis and 3.1% on an annual basis) and the Euro area (2.5% on an annual basis and 1.5% on a Q4/Q4 basis). Second, these direct oil effects are likely insufficiently large to lead to a recession on their own. However, our Europe economics team estimates an additional 1.2pp drag from the disproportionate surge in natural gas prices and related disruptions in the baseline scenario (Russian gas exports to Europe through Ukraine cease). This additional natural gas drag rises to around 2½ pp under a full stop of all Russian pipeline imports to Europe, which would likely trigger a recession. Third, the substantial oil boost in Canada supports our above consensus 2022 growth forecast of 3.9% on a Q4/Q4 basis (and 4.2% on an annual basis).