Reading Pass-Through Without Falling for a Constant
This series has followed one question from two complementary directions: why does an exchange-rate movement reach consumer prices differently across economies and states? The final judgment is that exchange-rate pass-through is not a constant attached to a currency. It is a conditional response shaped by the inflation environment, the behavior and policy regime of the exchange rate, market structure, fiscal and monetary credibility, labor-market conditions, the price component being measured, and the way the shock is identified. That judgment gives the reader a practical capability: evaluate a pass-through claim by restoring the conditions that make its number meaningful.
The response is a function of an economic state
A constant coefficient is attractive because it makes a complicated transmission chain easy to summarize. If a 1 percent exchange-rate movement always produced the same percentage change in consumer prices, a historical estimate could be carried directly into a forecast. The evidence does not support that simplification. The BIS study finds meaningful variation across country-periods, while the IMF's Singapore analysis shows that even within one institutional setting the response changes with labor-market tightness and with the CPI component being observed.
The more defensible mental model is a response surface. One dimension is the exchange-rate movement itself: its size, direction, and whether it is a market movement or an identified policy shock. Another is the price endpoint: import prices, headline CPI, core CPI, goods, or services. A third is time: the response may appear immediately, accumulate over several quarters, or weaken at longer horizons. Around those dimensions sit the conditions that affect pricing behavior—average inflation, distance from target, realized volatility, exchange-rate arrangement, product homogeneity, economy size, fiscal health, and labor-market tightness.
These conditions are not alternative names for one hidden variable. Inflation describes the general price environment; an inflation gap describes distance from a policy objective. Volatility describes how the exchange rate has behaved; a de facto regime describes how it is arranged or managed. Economy size and product homogeneity speak to pricing power and the ability to absorb costs in margins. Fiscal health speaks to the credibility of the wider stabilization framework. Labor tightness changes domestic demand and the cost structure of services. Keeping the dimensions separate is what makes the synthesis useful: a country can be low-inflation but fiscally fragile, have a flexible regime but repeated depreciation, or have low exchange-rate volatility and a tight labor market.
The BIS global results locate the center of this surface without pretending it is flat. The full-sample one-year CPI pass-through is 13.4 percent, but the group estimates are 7.6 percent for advanced economies and 17.1 percent for emerging market and developing economies over their full samples. In the 2010s, the advanced-economy estimate falls to 4.4 percent while the EMDE estimate is 20.0 percent. Those figures establish that time and broad country group matter in the reported data. They do not establish that income classification itself is the cause, or that any individual economy should inherit its group's mean.
The non-linear driver analysis explains why a group average is only an orientation point. At average inflation below roughly 5 percent, the BIS fitted pass-through remains modest; at 20 percent, it is about twice the level near 2 percent. Volatility follows a U-shape, with a fitted minimum near an 8 percent NEER standard deviation over the 40-year sample. That means a lower-volatility economy is not automatically a low-pass-through economy, and a higher-volatility economy is not automatically a cautious one. Expectations about persistence and reversal determine how the observed movement is translated into a price decision.
The same logic extends to market structure. Larger economies show lower fitted pass-through, consistent with more pricing-to-market and greater scope for firms to adjust margins rather than local prices. More homogeneous imported goods show higher fitted pass-through; the paper's proxy moves from roughly 15 percent when food and fuel account for no more than about one-quarter of imports toward about 30 percent at higher shares. These relationships make the response surface concrete. A currency movement meets firms with different room to change markups, substitute inputs, differentiate products, or wait for information.
That is why a linear average can fail in two ways. It can conceal a threshold, such as the sharp rise associated with a larger inflation gap or higher average inflation.