Supply Chain Disruptions and Pandemic-Era Inflation

Featured in print Digest
This figure is a stacked vertically bar graph titled, Decomposition of US Inflation, 2017 to 2023. The y-axis is labeled, Standard deviation change in price, normalized to a mean of 0. It ranges from negative 5 to 5. The x-axis is time and ranges from 2018 to 2023.  The graph shows the decomposition of US inflation from 2018 to 2023, with three categories of bars: Supply chain, Productive capacity, and Aggregate demand. The graph also includes a line representing US quarterly goods inflation. The supply chain bars hover between -1 and 0 standard deviations from 2018 to 2020. In 2021, the values spike up to be between 1 and 3 standard deviations. In late 2022 and early 2023, the values return to hover between -1 and 0 standard deviations. The Productive capacity bars range between -2 and 0 standard deviations from 2018 until 2020. From 2020 until 2022, they range between 0 and 1.5 standard deviations. In 2022, the values spike to around 2 to 3 standard deviations. The Aggregate demand bars range between -0.5 and 0 standard deviations between 2018 and 2020. In early 2020, they take on values of around -4 standard deviations but then become positive from mid-2020 until 2023, ranging from around 0 to 1.5 standard deviations. The US quarterly goods inflation line hovers between -1 and 0 standard deviations until 2020 before dipping to -2 standard deviations in 2020. It then rebounds and increases to 2 standard deviations by 2022 before declining back to around 0 in 2023. The note on the figure reads, The monthly series are normalized to have a mean of zero and a standard deviation of one over the sample period, January 2017 to September 2023. The source line reads, Source: Researchersʼ calculations using data from various sources.

The COVID-19 pandemic led to major disruptions in global supply chains. In The Causal Effects of Global Supply Chain Disruptions on Macroeconomic Outcomes: Evidence and Theory (NBER Working Paper 32098), Xiwen Bai, Jesús Fernández-Villaverde, Yiliang Li, and Francesco Zanetti analyze container ship data to measure these disruptions and investigate how they affected inflation during and after the pandemic.

They find that the drop in inflation at the onset of the pandemic was due to a drop in aggregate demand due to mobility restrictions, and that the subsequent rise was mainly due to adverse shocks to supply chains. By 2022, the main driver of inflation shifted from supply chain shocks to constraints on productive capacity, likely due to reduced levels of labor supply. In late 2022, inflation began to decline as a result of weakened demand, strengthened capacity, and supply chain recovery.

Port congestion initially drove COVID-related inflation but was supplanted by productivity shocks later in the pandemic.

Containerized seaborne trade accounts for 46 percent of all international trade. Large container ships operate on fixed itineraries, and even mild congestion can lead to substantial delays, costs, and trickle-down consequences. During the pandemic, wait times at some ports extended from only a few hours to two to three days.

Ports include both anchorages and berths. Vessels moor at berths to load and unload cargo; if a port is not congested, a vessel can moor directly at a berth upon arrival. When a port is congested, a vessel will moor first in an anchorage area; mooring patterns can be used to measure congestion. The researchers obtain data from January 2017 to September 2023 from the automatic identification system of the International Maritime Organization, which tracks all vessels larger than 300 gross tons at high frequency. They train a machine learning algorithm to identify areas with high densities of ships and then determine whether ships are at a berth or an anchorage. They define congestion as the fraction of ships that first moor at an anchorage when reaching port and compute the average congestion rate in each month, a ship-visit weighted average of congestion over the top 50 container ports worldwide.

They find that congestion was declining prior to the pandemic, and was around 25 percent from early 2019 to mid-2020. It rose to 37 percent in mid-2021 before declining again; it returned to normal levels in mid-2023.

The researchers develop a macroeconomic model of congestion in which producers and retailers must match to each other to trade. A supply chain shock increases transportation costs and makes it harder to match. Using this framework, they estimate the effects of aggregate demand, productive capacity, and supply chain shocks on GDP, personal consumption expenditure (PCE) prices, import prices, retail market tightness, unemployment, and the average congestion rate index. A negative one standard deviation supply chain shock leads to real GDP decline and an unemployment increase of around 0.2 percent. Retail market tightness initially increases but falls in the following quarter. PCE goods prices increase by up to 0.3 percent and import prices by up to 0.5 percent.

— Whitney Zhang