Turbulence, Firm Decentralization, and Growth in Bad Times
GAICIL Insights
In a storm, the instinct is to grab the wheel at headquarters.1 This paper suggests handing it to the people on deck. Turbulence, Firm Decentralization, and Growth in Bad Times, by Philippe Aghion, Nicholas Bloom, Brian Lucking, Raffaella Sadun and John Van Reenen, pits two views against each other. The localist view holds that greater turbulence after macro shocks increases the value of local information, favoring decentralized firms. The centralist view holds that tough decisions favor centralized firms.
Executives deploying enterprise AI might start here: design it first as a tool for the plant manager. When the next storm comes, the firms that win may be the ones whose AI made local judgment sharper.
Two large micro datasets, the World Management Survey covering firms in 10 OECD countries and administrative data on US establishments, back the localist view. Firms that delegated more power from headquarters to local plant managers before the Great Recession outperformed centralized counterparts in sectors hardest hit by the crisis. Negative shocks hurt growth in centralized firms significantly more than in decentralized ones, measured by export shocks or by exogenous predictors such as product durability.
Turbulence drives the effect. It concentrates in industries with greater increases in turbulence, measured by product churn and stock market volatility. When conditions change rapidly, managers close to operations hold information headquarters lacks. The conclusion is robust to alternatives such as managerial fears of bankruptcy and changing coordination costs. Decentralization will be suboptimal in many environments, yet it appears beneficial for the average firm in bad times. Command-and-control fails exactly when it feels most necessary.
The macro implication is bold. The US having relatively more decentralized, and therefore flexible, firms may have helped it weather the global storm better than more centralized countries, such as those in Southern Europe. Organizational structure may contribute to national resilience. The authors see the paper as a first attempt to unravel the relationship between growth and firms' internal organization using micro data with observable measures of decentralization. A nation's org charts may be part of its shock absorbers.
The parallel with today's technology is close. AI enters as a fork in the road. AI reshapes how information flows within organizations. It can centralize decision-making by giving headquarters better data and analytics, or it can empower local managers with tools enhancing their ability to act on local information. In turbulent times, the paper implies, local information and the authority to use it are valuable. Firms adopting AI should ask whether their systems strengthen or weaken local managers' capacity to respond. A dashboard that lets the CEO override every plant manager may be a liability in a storm.
AI is also a source of turbulence. As it transforms markets, product churn and volatility may increase. Firms delegating authority to managers who understand local conditions may adapt better. AI tools that augment local decision-making, supplying information and analysis while preserving managers' authority, may be particularly valuable. The technology that creates the turbulence can also equip people to navigate it, if it is pointed at the edge.
Resilience requires resisting a reflex. Centralized control looks attractive in a crisis, yet decentralized firms fared better when shocks were large and turbulent. Organizations designing responses to AI's disruptions should be wary of excessive centralization, which may slow adaptation.
For policy, national resilience may partly depend on firms' organizational practices. Policies encouraging good management, including appropriate delegation, may strengthen economies' capacity to withstand shocks. With AI possibly making shocks more frequent, such resilience grows in value.
The measurement is the methodological prize. Measuring decentralization directly in large samples of firms and establishments allows systematic study of organizational structure. Similar measurement will be needed to study how AI affects the allocation of authority within firms.
The thread running through it all: organizational design matters for performance in bad times. Decentralized firms outperformed centralized ones in turbulent sectors during the Great Recession because local information became more valuable. As AI transforms both the information inside firms and the turbulence of markets, this research offers a framework for allocating authority to thrive.
Introduction
What is the optimal form of firm organization during “bad times”? The greater turbulence following macro shocks may benefit decentralized firms because the value of local information increases (the “localist” view). On the other hand, the need to make tough decisions may favor centralized firms (the “centralist” view). Using two large micro datasets on decentralization in firms in ten OECD countries (WMS) and US establishments (MOPS administrative data), we find that firms that delegated more power from the central headquarters to local plant managers prior to the Great Recession outperformed their centralized counterparts in sectors that were hardest hit by the subsequent crisis (as measured by export growth and product durability). Results based on measures of turbulence based on product churn and stock market volatility provide further support to the localist view. This conclusion is robust to alternative explanations such as managerial fears of bankruptcy and changing coordination costs. Although decentralization will be suboptimal in many environments, it does appear to be beneficial for the average firm during bad times. (JEL D22, G12, G32, G34, L23)
Takeaway
Are decentralized firms more resilient to large negative shocks? On the one hand, a shock like the Great Recession may reduce the congruence between the CEO and the plant manager, thus making centralized firms more resilient (the “centralist” view). On the other hand, recessions are associated with greater turbulence, making the plant manager’s local information more valuable, which would imply that decentralized firms will perform relatively better in unexpected downturns (the “localist” view). To empirically investigate these issues we collected new data on a panel of firms in 10 OECD countries (WMS), and plants in the US (MOPS) and exploited the negative shock of the Great Recession which reduced demand across industries and countries in heterogeneous ways. Using our pre-recession data on decentralization we find that negative shocks hurt growth in centralized firms and establishments significantly more than in their decentralized counterparts. This is true whether we use export shocks which vary at the industry by country (WMS) or establishment (MOPS) level, or exogenous predictors of these negative shocks like product durability. Further, as the localist view suggests, this effect is driven by the industries which experienced a greater increase in the turbulence (as measured by product churn and stock market volatility) that accompanied the crisis. Potentially, the fact that the US has relatively more decentralized (and therefore flexible) firms, meant that it could weather the global economic storm better than many more centralized countries (e.g. in Southern Europe).1
We see our paper as a first attempt to unravel the relationship between growth and the internal organization of firms using micro data with observable measures of decentralization. There are many directions to take the research.
Notes
- Philippe Aghion et al., Turbulence, Firm Decentralization, and Growth in Bad Times, 13 Am. Econ. J.: Applied Econ. 133 (2021), https://doi.org/10.1257/app.20180752. Takeaway source: Conclusion, reproduced from a free author or repository copy. ↩
Cite as: Philippe Aghion et al., Turbulence, Firm Decentralization, and Growth in Bad Times, 13 Am. Econ. J.: Applied Econ. 133 (2021), https://doi.org/10.1257/app.20180752
Header image: Meeting room Säynätsalo Town Hall.jpg by Antti Leppänen, CC BY 4.0, via Wikimedia Commons.
