Every Oil Shock Since 1973: What History Tells Us About 2026

The Hamilton Framework: Net Oil Price Increase (NOPI)

Economist James Hamilton at UC San Diego developed the Net Oil Price Increase (NOPI) framework through his 2003 paper analyzing the relationship between oil prices and recessions. The NOPI measures the increase in the price of oil above its highest level in the prior three years. This approach captures the economically meaningful shock — new prices relative to the recent historical baseline that businesses had adapted to — rather than absolute prices, which can be misleading across decades with different real interest rates and commodity super-cycles.

The formula is straightforward: NOPI(t) = max(0, P(t) - max(P(t-36 months))) ÷ max(P(t-36 months)). When oil rises from a baseline of $50 to $75, NOPI is 0.50. The framework captures how much "newness" the price increase represents for the economy.

Hamilton's empirical finding: a NOPI of 0.10 (10% increase above the prior 3-year high) reduces GDP growth by approximately 1.5% on an annualized basis. The relationship is nearly linear up to NOPI of approximately 0.35, after which the relationship flattens slightly as demand destruction accelerates. The -0.15 elasticity coefficient — meaning a 100% increase in oil prices (NOPI = 1.0) reduces GDP by 15% — has proven robust across multiple recessions and time periods.

The current NOPI reading is extreme. Brent crude at $105 per barrel represents a roughly 55% increase above its 3-year trailing high of approximately $68 (June 2023). The NOPI is therefore 0.55. Using Hamilton's -0.15 coefficient, this implies an 8.25% GDP drag — though this is an annualized calculation that assumes sustained elevated prices. If oil stabilizes at $105 for 6 months, the annual GDP impact would be approximately 4.1%.

1973-1974: The Arab Oil Embargo

The Yom Kippur War, initiated on October 6, 1973, triggered the Organization of Arab Petroleum Exporting Countries (OAPEC) to impose an oil embargo on nations supporting Israel — specifically the United States and the Netherlands. The embargo cut approximately 5 million barrels per day of global supply, roughly 10% of world production at the time.

The immediate effect was dramatic. Crude prices surged from $2.59 per barrel in September 1973 to $11.65 by year-end — a 350% increase. In NOPI terms, oil had been running at approximately $2.50 per barrel; the spike to $11.65 represented a NOPI of 3.66. This was the largest NOPI shock in modern history.

The recession that followed was severe. US GDP contracted 3.2% in 1974, the worst year since the Great Depression. Unemployment rose from 4.8% to 9.0%. Inflation hit 11% by year-end. The stagflation created a policy dilemma identical to today's: Fed Chair Arthur Burns faced rising inflation and rising unemployment simultaneously.

The embargo lasted until March 1974 — six months. But the damage was permanent: inflation expectations became unanchored, and recovery was slow. NOPI modeled the shock correctly: a 366% NOPI shock should produce severe recession, and it did.

1979-1980: Iranian Revolution Production Loss

The Iranian Revolution, culminating in the overthrow of Shah Mohammad Reza Pahlavi in February 1979, removed Iran from global oil markets. Iranian production, which had been running at approximately 5.7 million barrels per day, collapsed to below 1 million bbl/day within weeks.

This was a pure supply shock of roughly 4.7 million bbl/day lost capacity — larger in absolute terms than the 1973 embargo but occurring in a market 30% larger. Brent crude surged from $24 per barrel in December 1978 to $42 by June 1980 — a 75% increase. The prior 3-year high was approximately $14 per barrel, making the NOPI 2.0.

The Fed, now under Paul Volcker, responded with aggressive rate increases (funds rate hit 20% by mid-1981) to break inflation expectations. The recession that followed was deep: GDP contracted 2.7% in 1980, unemployment hit 9.7%, but the inflation kill was fast and permanent. By 1983, inflation was below 3%.

Hamilton's framework predicted this correctly: NOPI of 2.0 implies 30% GDP reduction on an annualized basis, a prediction consistent with the observed recession severity. The policy response mattered for duration but not for the fact of recession — that was determined by oil shock magnitude.

1990-1991: The Gulf War Shock

Iraq's invasion of Kuwait on August 2, 1990, removed approximately 4 million bbl/day of Kuwait's production and threatened Saudi Arabia's 8 million bbl/day capacity. The market feared supply loss of up to 10% of global production. Brent crude spiked from $15 per barrel pre-invasion to $40 within weeks — a 167% jump.

However, the geopolitics shifted quickly. Saudi Arabia remained in the US security umbrella. A coalition military campaign pushed Iraq out of Kuwait within six months. Production was restored by mid-1991. The NOPI, while serious (approximately 1.67 above the $24 three-year baseline), was temporary.

The recession of 1990-1991 was mild — GDP contracted 1.6%, unemployment hit 7.8%, and recovery began by Q2 1991. Fed Chair Alan Greenspan responded with aggressive rate cuts (from 8.25% to 2.75% between July 1990 and July 1992), which cushioned the employment impact. The short-lived NOPI shock combined with rapid policy easing produced a mild recession followed by a fast recovery.

This 1990 case shows that duration matters. A temporary oil shock with appropriate policy accommodation is recessionary but not catastrophic.

2008: The Oil Shock Within the Financial Crisis

The 2008 financial crisis was fundamentally a credit shock, not an oil shock. However, oil prices spiked in the run-up to the crisis. Brent crude rose from $33 in January 2007 to $147 in July 2008 — a 345% increase. The NOPI, measured from the $31 baseline of 2004, was approximately 3.74 at the peak.

But the timeline reveals the relative importance of the oil shock. Oil peaked in July 2008 — after Bear Stearns' collapse in March, after the first major unemployment rises in spring 2008, but before the Lehman Brothers collapse in September. In other words, the recession was well underway before oil peaked. The oil shock amplified the financial crisis recession but did not cause it.

By year-end 2008, oil had collapsed to $30, and the NOPI returned to near-zero. The GDP contraction of 2.5% in 2008 cannot be attributed to oil shock mechanics; it flowed from credit destruction and the loss of $16 trillion in household wealth.

This case shows the interaction between multiple shocks: an oil shock layered on top of a financial shock amplifies damage. The current environment has some of this quality — oil shock (energy), supply chain shock (aluminum, helium), and labor market deterioration combining for elevated risk.

2022: Russia-Ukraine War, High Oil, and No Recession

Russia's invasion of Ukraine on February 24, 2022, created an oil shock through supply disruption and energy market uncertainty. Russia and Ukraine together export approximately 5 million bbl/day of oil and petroleum products. Brent crude spiked from $92 to $137 within weeks — a 49% jump. The NOPI, measured from the $77 three-year baseline (June 2019), reached approximately 0.78.

The US did not enter recession in 2022 despite this NOPI shock — the most significant shock without a recession outcome since the 1970s. Why? The answer points to the constraints on the current shock.

First, the US is now a net energy exporter (though still importing crude for refining). Energy production added to the economy rather than reducing it. This was not true in 1973, 1979, or 1990, when the US was a major net importer.

Second, the oil shock proved temporary. By late 2022, prices had fallen back to $85, and NOPI had returned to manageable levels. The supply disruption was managed through spare capacity releases from the Strategic Petroleum Reserve and temporary demand destruction.

Third, the shock did not coincide with other major adverse signals. The yield curve had not un-inverted. The Sahm Rule had not triggered. Credit spreads were not widening. The oil shock was isolated.

The 2026 Shock: Which Historical Case Applies?

Current NOPI is 0.55 — roughly between the 1990 Gulf War magnitude (1.67) and the 2022 Russia-Ukraine shock (0.78), placing it squarely in "recession-risk" territory. But the critical question is duration: will this shock be temporary like 1990/2022 or persistent like 1973/1979?

The geopolitical situation differs from prior cases. The Israel-Iran conflict is not a temporary war with a clear resolution timeline. It could escalate, de-escalate, or settle into a persistent low-level conflict state. Iranian steel is offline for years. Gulf aluminum smelters are damaged with 3-5 year repair timelines. These are multi-year supply losses, not temporary disruptions.

Critically, the 2026 shock coincides with other adverse signals: yield curve un-inversion, Sahm Rule rising, credit spreads widening, consumer sentiment falling. This conjunction is most similar to 1973-74 and 1979-80, not 1990 or 2022. In those earlier cases, oil shock combined with other macro deterioration produced severe recessions.

The difference in the current case: if the Fed cuts rates (as current guidance suggests), the policy response will be more Greenspan-like (1990) than Volcker-like (1979). This matters for amplitude and duration but not direction. A recession with Greenspan-style accommodation would be milder than Volcker's path, but still a recession.

Hamilton's framework suggests 0.55 NOPI at 4+ months duration implies 8% GDP annualized drag, translating to roughly 2% negative GDP growth if sustained for one quarter. This alone would not trigger recession NBER definition (two consecutive quarters of negative growth), but combined with the labor market deterioration tracked by the Sahm Rule, the probability rises substantially.