Tariff Shocks and the Slope of the Phillips Curve

Tariff shocks are assumed to be a one-time level shift of the price level – inflation accelerates but only temporary before falling back to its long-run trend. But is this always true? In this note, we discuss under what assumptions tariffs can affect the slope of the price level, that is a persistent (or permanent) inflation shock. For each theoretical case, we provide an intuitive example.

TL;DR

What follows is pretty “wonkish” and on the long side. In great summary: the tariffs shock is a shock to the price level – inflation is expected to accelerate and then to gradually converge back to its trend. However, there are also theoretical reasons to think that the shock will also affect the trend. In other words, it cannot be excluded that we will remain above target going forward.

Introduction

 

In modern macroeconomics, inflation dynamics are often captured by an expectations‑augmented Phillips curve. A Phillips curve (in its “primitive form”) is a relation between current inflation and the real marginal cost of firms, augmented with future expected inflation and a variable capturing supply or cost‑push shocks. (If the reader is more familiar with the Phillips curve expressed using the output gap, we remind her that the output gap is a good proxy for the marginal cost only under specific assumptions). In any case, a tariff on imports acts as a  supply shock: it raises firms’ marginal costs and therefore raises inflation at any given level of marginal cost. The usual assumption is that such a shock shifts the Phillips curve up (an intercept-effect) without altering its slope. The aim of this note is to examine theoretical reasons why a tariff shock may also change the slope of the Phillips curve, i.e., the responsiveness of inflation to the marginal cost.

When Tariffs Affect the Slope of the Phillips Curve

Tariffs on Intermediate Inputs and Global Value Chains

The literature draws a distinction between tariffs on final consumer goods and tariffs on intermediate inputs. For instance, using global input–output data, Cuba‑Borda and colleagues (2025) estimate trade‑cost shocks and find that higher trade costs for final goods produce large but short‑lived inflation spikes, whereas increased costs for intermediate inputs trigger more persistent inflation. In their multi‑country model, this persistence arises because tariffs on imported inputs propagate through global value chains and raise marginal costs across multiple sectors. When cost pressures persist, inflation becomes more sensitive to marginal cost: a given increase in marginal cost leads to a larger rise in inflation over time. Under such shock, the slope coefficient effectively increases, making the Phillips curve steeper. Indeed, if tariffs persist or raise input costs repeatedly, they also increase the marginal cost over time, not just once. Persistent increases in the marginal cost effectively raise the slope because inflation becomes more responsive to cumulative cost pressures.

Intuitive example. Imagine a smartphone producer that imports semiconductors. A tariff on chips raises the cost of every device. The producer passes some of this cost to wholesalers, who pass it to retailers, and ultimately consumers. If chips are used widely across electronics, the tariff raises costs in many industries. Because these cost pressures persist until production networks adjust (which takes time and does not happen all at once), a small increase in marginal cost now leads to sharper price increases. Inflation becomes more sensitive to marginal cost – a steeper Phillips curve – because each percentage point increase in marginal cost translates into a larger rise in inflation.

Reduced Import Competition (De‑globalisation)

Trade openness influences price‑setting behavior. Evidence from the United States shows that increased import penetration flattened the Phillips curve over time: between the late 1970s and the 2000s, the slope of the Phillips curve fell significantly, with around 40% of the flattening attributed to greater import exposure. A tariff that reduces import competition can reverse this process. When foreign competitors are less accessible, domestic firms gain pricing power.  As marginal costs rise, firms can raise prices more aggressively because consumers have fewer substitutes. Therefore, the responsiveness of inflation to marginal cost – the slope – increases. Indeed, in small open economy NK models (Galí and Monacelli (2005)), marginal cost includes the terms of trade, which measures the price of imports relative to exports. Tariffs worsen the terms of trade by making imports more expensive. This can amplify or dampen the effect of output on marginal cost, depending on how households and firms substitute across domestic and foreign goods.

Intuitive example. Consider a market for apparel where domestic producers compete with imported clothing. Without tariffs, a rise in marginal cost results in modest price increases because consumers can switch to cheaper imports. If a high tariff suddenly makes imported apparel expensive, domestic firms become the main suppliers. As their own costs (e.g. wage costs) rise, they can raise prices more aggressively because consumers have fewer alternatives. Inflation therefore becomes more sensitive to marginal cost and the slope of the Phillips curve steepens.

Production Networks and Sectoral Price Stickiness

Tariffs interact with the structure of production and nominal rigidities. Kalemli‑Özcan, Soylu and Yildirim (2025) embed tariffs in a New Keynesian open‑economy model with full input–output linkages. They show that when the production network is more granular—with many sectors and small input–output coefficients—the aggregate Phillips curve is flatter. Granularity diffuses sectoral shocks: a tariff on a peripheral sector barely moves aggregate marginal cost, so inflation becomes less sensitive to cost changes. Moreover, heterogeneity in price stickiness across sectors compresses the range of inflation outcomes and flattens the curve further. Conversely, tariffs on central sectors with inelastic supply can cause disproportionately large increases in marginal cost. In that case, inflation responds strongly to marginal cost, and the slope rises.

Intuitive example. Suppose a tariff is imposed on cork used in wine bottle stoppers. Cork is a peripheral input with close substitutes. The tariff raises the marginal cost of producing wine slightly, but because cork’s share in total cost is tiny and substitutes exist, the aggregate marginal cost hardly moves; inflation is less responsive to cost changes, and the Phillips curve flattens. By contrast, a tariff on semiconductors—a central input used in many goods with few substitutes—substantially raises marginal costs across numerous industries. Inflation then responds strongly to marginal cost, steepening the Phillips curve.

Non‑linearities

Nominal rigidities can create non‑linearities in the relationship between marginal cost and inflation. Forbes, Gagnon and Collins (2021) find that the Phillips curve is steep when inflation is high but becomes nearly flat when inflation is low and slack is large. While this non‑linearity is not specific to tariffs, it implies that a tariff shock occurring in a low‑inflation regime may have little effect on the slope because price and wage adjustments are limited. If a tariff shock pushes inflation above some threshold or triggers wage bargaining, the slope may increase.

Intuitive example. Picture a country where prices have been steady for years—nobody really expects them to move much. Companies don’t like to raise prices because shoppers might go elsewhere, and workers don’t bother asking for bigger paychecks since costs aren’t rising. Now imagine the government puts a tariff on imported textiles, making clothes a bit more expensive to produce. Stores don’t want to scare customers with higher prices, so they just take smaller profits. Because people don’t see prices going up, they don’t push for higher wages. The result? Inflation barely moves. But if prices were already rising in general, the story changes. In that case, stores would quickly pass their higher costs on to customers, and workers would demand more pay to keep up. Inflation would climb faster. So when inflation is low, the economy kind of “shrugs off” small cost shocks. But when inflation is already moving, those same shocks can set off a chain reaction that makes prices rise much faster.

Conclusion

Tariffs primarily function as cost‑push shocks that increase firms’ marginal costs and shift the Phillips curve upward. However, several structural channels, including input‑output propagation, import competition, network structure and heterogeneity, and non‑linearities imply that tariffs may also alter the slope of the Phillips curve – the sensitivity of inflation to marginal cost. Put it simply, we can expect inflation to accelerate now, peak in Q1 2026, and then converge back to its trend. However, we cannot exclude that the trend itself will be higher going forward and we will remain (well) above target.

Contact

Want something more tailored?

We provide tailored consulting on ad-hoc projects.