NextFin News - A Federal Reserve study released in October concludes that the El Niño weather pattern now strengthening across the Pacific poses material upside risk to US inflation while simultaneously lifting employment, a rare combination that could complicate the central bank's policy path into 2027. The study, by Brigitte Roth Tran, a senior economist at the Federal Reserve Bank of San Francisco, estimates that a very strong El Niño adds roughly 1 percent to the headline PCE price level within about 18 months of onset, while producing widespread, multi-percentage-point gains in nonfarm payrolls.
The timing is what makes the finding market-relevant. The current episode, declared by the World Meteorological Organization on June 2, 2026, is already classified as very strong and is forecast to remain so through the first quarter of 2027. Core inflation, as measured by the Federal Reserve's preferred gauge, stood at 3.1 percent year over year in September 2026, well above the 2 percent target, and the Federal Open Market Committee held its benchmark rate at 3.75 percent to 4.00 percent on October 28 by a 10-2 vote, with Lisa D. Cook and Beth M. Hammack dissenting in favor of a further quarter-point increase. In that setting, a weather-driven supply shock that pushes prices up and employment up at the same time is not a clean input for a central bank trying to fine-tune demand.
The study's central tension is this: the inflation effect is large but highly conditional, while the employment effect is smaller in the near term but far more reliable. That asymmetry is the key to reading what the research actually implies for 2027.
The Mechanism: Why a Pacific Weather Pattern Hits US Prices and Payrolls
El Niños are not ordinary weather shocks. They are a global phenomenon that unfolds over months, can be forecast months in advance, and have effects that vary sharply by season and region. Those features make the standard tools of the weather-economics literature ill-suited to measuring them, which is why Roth Tran builds a new method: a modified stacked local-projections framework that estimates impulse responses separately for each calendar month, using the Relative Oceanic Niño Index (RONI) as a continuous treatment measure rather than a simple on-off switch. The sample covers 25 El Niño episodes between 1950 and 2025, with PCE price data from the Bureau of Economic Analysis and employment data from the Bureau of Labor Statistics.
The transmission runs through two distinct channels, and they operate on different clocks.
The inflation channel is a mix of direct commodity effects and slower-moving services pass-through. In the first summer of the ENSO year, headline PCE rises temporarily by about 30 basis points relative to April, driven by energy, food, and core services. Nearly half of that initial bump comes from gasoline prices, which track world oil higher in May and June. But the more consequential move arrives later: PCE begins rising again the following March and reaches roughly 1 percent above baseline by the end of the following November. That second-year increase is driven primarily by core services, with food contributing an additional 15 to 20 basis points to the headline figure. Within services, financial services and insurance add about 20 basis points by December of the ENSO year, and health care adds a similar amount to the longer-run rise.
Why would a weather pattern lift financial and health services prices? The paper does not claim a single smoking gun, but the mechanism is consistent with a broader disruption channel: flooding in the southern United States, supply-chain delays, insurance claims processing, and disaster-recovery activity all raise the cost of delivering services that are normally insulated from commodity cycles. The food effect is more intuitive, though the paper notes it is modest in headline terms because food is a small share of PCE; measured directly, food prices rise about 1 percent by June and roughly 2.5 percent over the full 18-month window.
The employment channel works differently, and in the opposite direction from what a naive supply-shock story would predict. In the early stages of an El Niño, employment responses are modest and offsetting, with construction rising while education and health services, financial services, and mining and logging decline. But starting in April at the end of the ENSO year, employment increases take hold and are widespread across sectors and states, persisting at least through the following November. Several sectors see gains on the order of several percentage points by the end of the 18-month period. Construction gets a temporary winter boost, consistent with milder weather in the northeast and northwest allowing building activity to continue; the 1997-98 episode saw single-family starts rise 6 percent from January to February, with construction income and employment up 25 percent over the December-February stretch.
The mining and logging sector is the notable outlier, posting a persistent decline that the paper links to lower oil prices and fewer winter heating degree days. That single sector's weakness is the clearest evidence that the employment gain is a reallocation story as much as a net-addition story.
The Critical Caveat: The Inflation Is Not Really From El Niño
Here is the finding that most headlines will miss. The large second-year increase in PCE prices is not, strictly speaking, caused by the El Niño itself. When Roth Tran adds one-year leads of the following year's El Niño or La Niña realization to the model, the second-year price increase "goes away entirely." In other words, the longer-run inflation is driven by the La Niña that typically follows an El Niño.
This matters because it changes the nature of the risk. The paper's baseline specification deliberately keeps the follow-on La Niña in the estimated effect, treating it as part of the total impulse response on the grounds that the La Niña is plausibly caused by the El Niño. The author writes:
Thus, this second year inflationary effect is highly uncertain and actually owes to the La Niña that follows.
But because La Niñas follow El Niños so frequently, she treats them as part of the expected price effect "unless or until there are meteorological forecasts indicating that a La Niña is unlikely to follow."
The practical upshot: the inflation number that matters for 2027 depends less on how strong the current El Niño is than on what happens in late 2027 and 2028. If a La Niña materializes, as is historically common, the full ~1 percent price-level effect is in play. If the ENSO cycle instead returns to neutral, a meaningful share of the second-year price pressure evaporates. That is why the paper's bottom line on prices is framed as "upside risk" and "heightened uncertainty" rather than a central forecast.
The employment effect carries no such asterisk. Unlike the price response, the payroll gains are not sensitive to whether a La Niña follows, and they are not sensitive to alternative empirical specifications. That asymmetry is the paper's most policy-relevant result: the labor-market boost is the more durable of the two effects.
Second-Order Read: Why This Makes the Fed's Job Harder, Not Easier
The first-order read of the study is straightforward: prices up, jobs up. The second-order question is what that combination does to the monetary-policy reaction function, and the answer is that it degrades the signal content of the data the Fed relies on.
A central bank fighting inflation wants to know whether a given inflation print reflects excess demand, which higher rates can cool, or a supply shock, which higher rates cannot fix without collateral damage. An El Niño that raises both inflation and employment simultaneously blurs exactly that distinction. Strong payroll gains would normally argue that the economy can tolerate restrictive policy; rising prices would argue for keeping policy tight. But if both are partly weather-driven, the Fed is being asked to calibrate demand management against a supply shock whose magnitude is only revealed with a lag of many months.
The timing of the lag is the real problem. The initial 30-basis-point summer bump is visible quickly. The more consequential second-year rise begins in March of the following year, roughly 10 months after the ENSO year opens, and peaks toward November. By the time the inflation is fully in the data, the weather event that caused it is long past, and the La Niña that may be driving it is only just becoming forecastable. A policymaker reacting to the late-2027 inflation print would be setting policy for conditions whose origin lies in a 2026 ocean-temperature anomaly.
There is also an expectation-formation channel worth flagging. The paper documents that economic agents appear to use El Niño forecasts in their decision-making: firms changed order placements by summer 1997 in response to forecasts, construction employment responds to precipitation forecasts up to six months ahead, and fisheries respond to three-month-ahead ENSO forecasts. If firms and households anticipate the inflationary and employment effects, some portion of the impact gets priced in early, which would show up as a front-loaded response in the data. That cuts both ways: it could make the early inflation bump more persistent than a one-off weather story would suggest, or it could mean the later effects are already embedded in wages and contracts by the time they appear in the aggregates.
The market implication is subtler than "El Niño means higher rates." The more precise read is that El Niño raises the variance of the inflation outlook at exactly the moment when the Fed is trying to declare victory and normalize policy. With the funds rate at 3.75 percent to 4.00 percent, core PCE at 3.1 percent, and real GDP accelerating to a 3.4 percent annualized pace in the third quarter of 2026, the baseline policy path already assumes a gradual return to target. The study adds a tail risk that is skewed to the upside for prices and the upside for jobs, which is precisely the skew that keeps a hiking bias alive even when the central forecast looks benign.
The Counter-Thesis: It Is Noise, Not a Regime Shift
The strongest case against taking this study as a policy signal is that it identifies a temporary, mean-reverting shock and risks over-reading it. Three arguments carry real weight.
First, the sample is thin. Twenty-five El Niño episodes since 1950, estimated with annual treatment measures, leaves limited degrees of freedom. The paper itself reports that statistical significance is sensitive to alternative specifications, and its own sensitivity table shows that some models fail to clear the 68 percent confidence interval on later-horizon price effects. A relationship that is strong in point estimates but fragile in significance is a warning sign, not a settled fact.
Second, the attribution problem cuts against the headline number. If the second-year inflation is actually a La Niña effect, then attributing it to El Niño for policy purposes is misleading. A central bank cannot respond to a 2026 El Niño in 2027 on the assumption that a 2027 La Niña will follow; the two are correlated but not mechanically linked. The paper's own framing, "highly uncertain," is the correct one.
Third, and most fundamentally, the economy has changed since the 1950s. Forecast quality and lead times have improved dramatically, building codes and resilience have reduced disaster damage, and yet global supply chains have made the economy more sensitive to overseas shocks. The paper lists these caveats explicitly and declines to claim that the historical impulse responses will repeat mechanically. Roth Tran writes:
Particularly strong El Niños like the currently still developing 2026-2027 episode could cause large-scale or widespread disasters that are disruptive and damaging in a manner not reflected in this analysis.
This acknowledges that the model may understate tail outcomes even as it may overstate the central tendency.
These counter-arguments do not defeat the study's core contribution, but they do bound it. The right interpretation is not "El Niño will add 1 percent to inflation in 2027." It is: El Niño widens the distribution of possible inflation and employment outcomes, with the skew to the upside, and the width of that distribution depends on a La Niña that has not yet occurred.
The falsifying signal is concrete. If core PCE prints at or below 0.2 percent month over month for two consecutive months during the second half of 2027, while nonfarm payrolls fail to show the widespread sectoral gains the model predicts, the upside-risk thesis is wrong and the effect is smaller than the historical impulse responses imply. Conversely, if core services inflation accelerates above 0.35 percent month over month starting in March 2027 alongside broad-based payroll gains, the study's mechanism is being confirmed in real time.
What to Watch: A Time-Horizon View
Short term (through mid-2027): Expect the initial, already-underway effects. The summer 2026 gasoline-driven bump of roughly 30 basis points is largely in the rearview mirror; the question is whether it leaves a residual in services inflation. Watch gasoline prices, insurance prices, and the monthly PCE services print. This is the window where anticipation effects dominate.
Medium term (late 2027 through 2028): This is where the second-year price rise would appear if a La Niña follows. The trigger to watch is NOAA's ENSO outlook for late 2027: a La Niña forecast materially raises the probability of the full ~1 percent price-level effect; a return to neutral cuts it. Employment gains should be visible across sectors and states by May 2027 and persist through November 2027, with construction leading early and services following.
Long term (structural): This is a cyclical, mean-reverting shock, not a regime shift. El Niños end; La Niñas end; the price level may be permanently higher by a fraction of a percent, but the inflation rate returns to baseline once the episode passes. The structural question is separate: whether a warming climate makes very strong ENSO episodes more frequent, which would raise the average frequency of these supply shocks without changing their individual dynamics. The study does not answer that question, and monetary policy cannot either.
Base case: a strong-to-very strong El Niño through the first quarter of 2027, followed by a La Niña, delivers the full impulse response, adding roughly 1 percent to the PCE price level by late 2027 and lifting payrolls by several percentage points across most sectors. Upside case for inflation: a historically strong La Niña follow-on combines with elevated oil prices to push the price effect above the historical mean. Downside case: ENSO returns to neutral without a La Niña, and the second-year price rise largely fails to materialize, leaving only the modest initial bump.
For investors and policymakers, the asymmetry is the point. The employment effect is the reliable leg; the inflation effect is the conditional one. A labor market strengthened by a weather pattern is still a tighter labor market, with all that implies for wages and services inflation. But an inflation print driven by a weather pattern is a signal to look through, not a signal to tighten, unless it begins to feed into expectations.
The study's real contribution is to put a number on a risk that markets had been treating as unquantifiable. Whether that number becomes a forecast or a footnote depends on a La Niña that has not yet formed.
Data as of October 30, 2026. The Federal Reserve Bank of San Francisco working paper cited is a research product of one economist and does not represent the views of the Federal Reserve System.
Explore more exclusive insights at nextfin.ai.
