Is Productivity About to Skyrocket?
The answer has big implications for the US economy and American workers.
Is Productivity About to Skyrocket?
vectorfusionart/Adobe Stock
Why Companies Don’t Run at Full Steam
The answer comes from looking at the individual-level rather than economy-wide averages.
- By
- June 03, 2025
- CBR - Economics
Manufacturers, for the most part, don’t run their operations at full capacity. Machines sit idle. Processes are slowed. Macroeconomists have long studied the effects of these costly choices, asking, among other things, whether this slack capacity in the system makes recessions last longer.
But economists often focus on averages that mask important differences in how and why companies choose to adjust their capacity, write the Central Bank of Sweden’s Niklas Amberg, Stockholm School of Economics’ Richard Friberg, and Chicago Booth’s Chad Syverson. When they zoomed in to see what might influence the decisions of individual companies, they find that two specific features—uncertainty about future demand and the size of product markups—explain nearly half the variation in capacity utilization.
The study empirically validates a long-standing theory, the newsvendor model, which articulates the trade-off between running at full capacity versus maintaining some flexibility to increase output in the event that demand suddenly spikes. This model is often used in operations studies. For example, Booth’s John R. Birge applied it to explain why inventory is most volatile at the most and least profitable companies. It’s also been used—by supply chain software company Deposco’s Yanyang Zhao (a recent graduate of Booth’s PhD program), Birge, Booth’s Levi DeValve, and Robert Inman of General Motors—to demonstrate why centralized warehouse inventory levels are crucial to improving distribution, particularly when there’s risk of disruption. And work by DeValve and North Carolina State University PhD student Jabari Myles generalizes the newsven-dor model so that it can help address a new group of challenges in settings that involve multiple points from which orders can be fulfilled.
Related reading
Amberg, Friberg, and Syverson also employed the newsvendor model, testing specific factors that affect the optimal amount of capacity. They looked first at demand. While some demand is general and stable, they focused on the portion that is unpredictable. Using quarterly survey data from Sweden’s National Institute of Economic Research, as reported by nearly 1,000 companies between 2021 and 2024, they mapped fluctuations in capacity utilization. They then paired this with answers to a key question on the NIER survey, which asked companies to describe the future of their business situation. Companies could choose one of four answers: easy to predict, fairly easy to predict, fairly difficult to predict, or difficult to predict.
Capacity was tightly linked to uncertainty about the future, the researchers find. Companies that described the future as fairly difficult or difficult to predict had lower capacity utilization—averaging 6 and nearly 10 percentage points lower, respectively—than those that said the future was easy to predict.
The research also finds a strong relationship between capacity and product markups. Using an annual survey that collects detailed accounting data for every company in Sweden, the researchers computed markups for the same 1,000 companies they studied in the NIER data. Among companies that charged high markups, those that were uncertain about the future reduced capacity by about 8 percentage points compared with those that felt certain about the future. The relationship didn’t carry over to companies charging low markups.
This finding makes intuitive sense, Syverson points out: Companies with high markups stand to make more money if they can meet spikes in demand. They are more willing to absorb the costs of idle capacity because they expect to eventually earn these losses back.
He notes that the research does not contain clear prescriptions for businesses. If anything, it suggests that the vast majority of companies are already doing what might be expected.
But the results do offer a reminder that regardless of what aggregate statistics show, companies could be responding to their individual situations in different ways and on different scales—a nuance that gets washed out by averages. For anyone trying to make high-level economic predictions or recommendations, Syverson says, “variation at the micro level can really matter for what happens in these macro movements, and so it’s important to get under the averages.”
- Niklas Amberg, Richard Friberg, and Chad Syverson, “Have We Got News For You: Firm-Level Evidence on the Optimal Choice of Expected Capacity Utilization,” Working paper, April 2025.
- John R. Birge, “The Relationships Among Inventory Volatility, Profitability, and Capital Structure,” Responsible and Sustainable Operations, July 2024.
- Levi DeValve and Jabari Myles, “Approximation Algorithms for Dynamic Inventory Management on Networks,” Management Science, October 2024.
- Yanyang Zhao, John R. Birge, Levi DeValve, and Robert Inman, “Managing Multi-Tier Inventory Networks with Expediting Under Normal and Disrupted Modes,” Working paper, May 2023.
More from Chicago Booth Review
More from Chicago Booth
Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.