NBER Economic Fluctuations and Growth Program Meeting

Why It Matters

Recognizing level‑based pass‑through reshapes macro‑economic modeling and policy analysis, ensuring more accurate forecasts of price dynamics and distributional impacts of cost shocks.

Key Takeaways

• Pass‑through measured in dollars shows near‑complete cost transmission.
• Log‑pass‑through appears incomplete due to percentage scaling effects.
• Evidence spans coffee, sugar, rice, gasoline, and manufacturing.
• Shift‑invariant demand systems can reconcile level pass‑through with theory.
• Mis‑specifying pass‑through biases policy impact estimates and asymmetry analyses.

Summary

The NBER Economic Fluctuations and Growth program featured Kunal Sanani’s paper on “complete pass‑through in levels,” challenging the conventional view that upstream cost shocks are only partially transmitted to downstream prices. Sanani shows that when pass‑through is measured in absolute dollars rather than percentages, many commodity‑linked consumer goods exhibit a long‑run pass‑through coefficient close to one. Using BLS retail price series for coffee, sugar, flour, beef, orange‑juice concentrate and extending to gasoline and the entire manufacturing sector, the analysis finds that price adjustments converge to full cost recovery within six months, even though log‑based estimates suggest only 30‑40 % transmission. A simple illustration—a firm with $1 input cost and $1 other cost, markup two, raises input price by 20 cents and only raises its price by roughly the same amount—demonstrates “complete pass‑through in levels” while appearing incomplete in logs. The pattern holds across product quality tiers, as shown by rice price heterogeneity during the 2008 global rice crisis, where percentage pass‑through varied but level pass‑through remained uniform. The findings imply that standard macro‑trade models relying on CES or homothetic demand, which predict log‑pass‑through below one, may be misspecified. Shift‑invariant demand systems from industrial organization can accommodate level pass‑through, suggesting that policy simulations of tariffs, carbon taxes, or supply‑chain shocks need to incorporate this nuance to avoid biased estimates of inflationary effects and distributional outcomes.