How an Energy Shock Becomes a Monetary-Policy Problem
This chapter examines how a central bank diagnoses an energy shock before deciding whether to look through it. The immediate task is to identify what the event changes first—supply, demand, or both—and to judge how long those effects may last. That diagnosis connects an oil-price headline to output, employment, inflation, and the timing of a policy response.
Start with What the Shock Changes
A demand shock begins with spending. Households may consume more or less, firms may change investment, governments may alter expenditure, or exports may move. In the framework described by Federal Reserve Vice Chair Philip N. Jefferson, these changes initially affect demand without directly changing the economy’s underlying productive capacity.
A supply shock begins somewhere else. It affects the processes and markets that determine what the economy can produce: the available supply of labour and capital, the productivity of those inputs, or the structure of production and pricing. An interruption to energy supply is initially a supply-side event because it changes the cost or availability of an input used by households and firms. It can reduce real purchasing power and make some production less viable even when no one has chosen to spend less.
That distinction matters because monetary policy reaches the two shocks through different routes. When demand is running ahead of the economy’s capacity, higher interest rates can cool spending and relieve pressure on both prices and employment. When energy supply is disrupted, a rate increase cannot restore the missing fuel or reopen a damaged route. It can still influence the demand response and the expectations surrounding inflation, but it may do so at the cost of weaker activity.
The output gap provides a compact way to describe that relationship. It is the difference between actual output—usually measured by gross domestic product—and an estimate of potential output, the level the economy can sustain over the long run while maintaining price stability and maximum employment. A positive output gap means that demand is strong relative to the economy’s estimated capacity. Employment tends to be above its maximum sustainable level and inflation faces upward pressure. A negative gap means that production is below potential; employment tends to be weaker and inflation faces downward pressure.
The output gap is not a number that policymakers can simply read from a gauge. Potential output is hypothetical and must be estimated. The distinction between supply and demand can also break down when an event affects both. Still, the framework forces a useful question: is the economy mainly trying to cool excess spending, or is it absorbing a reduction in what it can produce?
A simple thought experiment shows why the distinction matters. Suppose households and firms initially try to maintain the same purchases, but the price or availability of energy changes. The economy now has less purchasing power and, in some sectors, a higher cost of producing the same goods. If spending does not immediately fall, demand may look resilient even as the economy’s productive capacity is under pressure. The result can be rising prices without the kind of excess demand that a rate increase is designed to suppress. If households then cut consumption because energy absorbs more of their income, the same supply event also creates a negative demand effect. The first classification remains useful, but it no longer describes the whole path.
The movement of potential output adds another complication. A fall in actual GDP is not the only way an energy shock can create a negative output gap. If the economy’s sustainable capacity falls because energy-intensive production is curtailed, actual output may remain stable while the distance between actual and potential output changes. Conversely, a decline in actual output can reflect weak demand rather than a lasting loss of productive capacity. Policymakers must therefore interpret the output gap alongside prices, production constraints, employment, and the source of the disturbance. The framework organizes judgment; it does not remove the need for judgment.
For the energy shock under discussion, the answer may change over time. A conflict can disrupt oil and gas supply, pushing energy costs up. That direct supply effect reduces real income because households and firms must devote more resources to energy. The loss of purchasing power can then weaken consumption and investment, creating a demand effect. Financial conditions can tighten as markets reassess growth, profits, and the possibility of further disruption. A supply shock can therefore create a negative demand response even as it raises the price level.
This combination is what makes the central-bank problem different from a simple overheating episode. Imagine inflation rising while output remains strong. Tightening would cool demand and could support both sides of the mandate. Now imagine inflation rising because energy costs have jumped while real incomes and production are weakening. Tightening may reduce the risk that the price increase spreads, but it can also intensify the contraction. Easing may cushion activity, but it can add demand to an economy whose supply has been impaired.
Shocks can also point in opposite directions before they have been fully identified. The energy event itself may lower demand, while a separate investment or consumption surge raises it. Trade-policy changes may alter prices in the near term and productive capacity over a longer horizon. This is why classifying the initial effect is a working diagnosis, not a declaration that the economy contains only one type of shock.
Timing Changes the Decision
The second part of the diagnosis is duration. A supply shock can be temporary, causing a sharp but short-lived fluctuation, or persistent, changing prices and activity for much longer. The difference is not merely a matter of historical description. It determines whether a policy response will arrive while the original problem is still present or after it has already reversed.
Monetary policy works with a lag. Suppose a conflict briefly interrupts supply, energy prices jump, and then alternative sources become available before a rate change has much effect. If expectations remain anchored and broader inflation does not spread, aggressively tightening policy could restrain demand after the energy problem has eased. The central bank would have responded to the first-round price movement but created an additional cost in employment and output.
That is the narrow case for looking through. It is a forecast that the initial price movement will reverse, or at least remain contained, before monetary policy’s main effects arrive. The forecast must also include the economy’s response while the shock is active. A lower oil price later does not automatically undo price increases that firms have already made, wage demands that have already changed, or spending decisions that have already weakened.
This makes the timing test more demanding than comparing today’s oil price with yesterday’s. Policymakers need to distinguish the level of a price from its rate of change and from the breadth of the response. A one-time jump in an energy component can raise the price level without producing a continuing sequence of increases. A volatile price path can keep changing expectations even when its average direction is unclear. And a supply disruption can continue to affect output after the price chart has turned if firms cannot secure inputs or if financial conditions remain tight. The relevant forecast is about the joint path of prices, expectations, and activity.
The timing of information matters too. Early readings may reveal the direct energy movement but say little about wages or firms’ broader pricing. Later readings may be more informative about persistence, but they arrive after households and businesses have already made decisions. Waiting is therefore not a neutral pause between two better-informed choices. It trades the value of additional information against the possibility that the economy will move into a less stable state before the information becomes clear.
Now reverse the timing. The disruption lasts, alternative supplies remain uncertain, inventories fall, or damage creates another price spike. Households begin to expect higher inflation and firms adjust prices more frequently. In that setting, waiting may buy information, but it may also allow the shock to become embedded in behaviour. The central bank would then face a more difficult task: not only responding to the original supply problem, but also restoring confidence that inflation will return to target.
Jefferson’s framework distinguishes cases in which inflation and the output gap move together from cases in which they move in opposite directions. When a positive output gap accompanies inflation above target, raising rates can cool excess demand, bring employment closer to a sustainable level, and reduce inflation pressure. When output is below potential and inflation is weak, lowering rates can support both employment and prices. These are relatively aligned cases.
An energy supply shock can instead produce a positive inflation impulse alongside a negative activity impulse. The policy objectives are then in tension. Tightening addresses inflationary pressure but may push employment below its maximum sustainable level. Easing reduces the damage to activity but may amplify demand and make the inflation problem more persistent. The appropriate choice depends on the relative costs of the two risks and on how likely each is to grow.
That likelihood is not visible in the energy price alone. Policymakers need to watch whether inflation spreads beyond energy, whether expectations remain anchored, and whether output and employment weaken materially. They also need to ask whether the shock’s first classification still fits. A disruption that began on the supply side can produce a demand slowdown; a separate demand force can offset or intensify it. New information changes the balance even if the original event has not changed.
The recent conflict illustrates why duration is hard to infer. The BIS reports that oil prices initially rose sharply and later moved toward pre-conflict levels, but with significant volatility. It also points to uncertainty over infrastructure damage, alternative supplies, and the Strait of Hormuz, which matters for other inputs as well as energy. A partial price reversal is evidence that some pressure has eased; it is not evidence that every supply risk has disappeared. The bulletin’s comparison with earlier wars shows that prices can retrench when conflict ends or substitutes appear, but comparison does not tell policymakers in advance which path the current episode will follow.
The practical consequence is that a central bank should treat “temporary” as a revisable assessment. The relevant question is not whether anyone can predict the exact end date of a conflict. It is whether the evidence currently supports a contained first-round movement strongly enough to accept the risk of waiting. If the answer is yes, the bank may look through the initial energy effect while monitoring the indicators that would invalidate that judgment. If the answer becomes less certain, the cost of delay rises.
This timing test also explains why two central banks can look at the same global event and reach different conclusions. One economy may have resilient output, low energy intensity, and well-anchored expectations. Another may depend heavily on imported fuel, face a falling currency, and enter the shock with inflation already high. The physical event is common, but the policy clock and the risks on either side of the mandate are not.
The diagnosis, then, has two linked parts. First ask what the shock changes and how supply and demand interact. Then ask whether its effects will fade before policy works or persist long enough to alter expectations and broader prices. The next problem is to trace that second possibility more closely: how does an energy-price increase move from a direct cost into core inflation, and which conditions amplify or dampen the passage?