The Impact of Automation on Corporate Decision-making

Research co-authored by Wharton’s Pinar Yildirim shows how automation is a strategic asset that can reshape the power dynamics within an organization.

Last year, the corporate world adopted a new term: the flattening. This phrase refers to how tech companies, which rapidly hired droves of middle managers during the pandemic boom, are now eliminating this layer through widespread job cuts.

Recent research by Mustafa Dogan, Alexandre Jacquillat, and Wharton’s Pinar Yildirim, published in the Journal of Economics & Management Strategy, also touches on this recent phenomenon. Using theoretical modeling, the study explores how automation impacts the structure of decision-making within organizations, specifically examining the trade-offs between centralization and decentralization. It challenges the conventional wisdom that technology will democratize organizations and empower lower-level managers by decentralizing authority.

The research shows, firstly, that automation can change how decisions are made in companies. Centralized firms, typically characterized by top-down decision-making, are more likely to automate tasks within divisions that face uncertainty, such as new product development groups. Doing so reduces a top manager’s reliance on the knowledge of mid-level managers in situations facing uncertainty, streamlines processes, and enhances top-level control. “Automating divisions that involve uncertainty can diminish the need for managers’ localized expertise, empowering executives to break free from their dependencies,” Yildirim explained.

In contrast, decentralized firms, where decision-making is more distributed across the managerial hierarchy, tend to allocate automation resources to more stable, business-as-usual tasks. For these firms, it is better to use automation to support stable, ongoing operations in existing product divisions. This helps to shield these operations from the negative effects of biased decision-making by mid-managers of other, uncertain divisions, helping to improve the overall financial performance of the firm.

Moreover, if companies can allocate more resources to automation, over time, they tend to centralize decision-making. Automation reduces the variability in operations, so lower-level managers don’t need to make as many decisions. This shifts more control to top executives, making the firm’s structure more centralized, regardless of their original setup. Yildirim noted: “This is reducing the strategic role of mid-level managers, who shift towards more operational tasks because lower-level ones are automated.”

Impact on Innovation

A second finding from the paper suggests that, as automation resources become more available over time, the gap between the innovative potential of centralized and decentralized firms could widen. Centralized firms may become increasingly resistant to change, while decentralized firms may become more agile or better able to adapt to new market conditions. “For decentralized firms, automating routine tasks in stable divisions allows managers to focus on adapting to changes and innovating, enhancing the firm’s agility and responsiveness,” Yildirim added. This divergence could have significant implications for the competitive dynamics in various industries.

Thirdly, automation deployment, as it changes the role of mid-managers, will also change the communication and the degree of disagreements in firms. As more tasks are automated in divisions facing uncertainty, the quality of communication between executives and managers may deteriorate. This can lead to a less informed decision-making process, ironically undermining the very efficiency gains that automation is supposed to deliver. Yildirim noted: “Automation makes communication from managers to top executives less informative. This means managing people can become harder when using technology strategically.”

Maintaining Alignment

The final insight from the study is that the strategic use of automation can also substitute for something else — traditional financial contracts used to manage conflicts within organizations. As automation reduces the need for managerial input, firms may find less need to align managers’ incentives with those of the organization through financial means. Instead, automation can standardize processes and reduce opportunities for bias or misalignment. “It’s a cost-effective way to maintain alignment and reduce internal conflicts,” said Yildirim.

The implications of the findings from the study are profound. The authors suggest that, as firms find it easier to access automation resources, they should also anticipate changes to their managerial hierarchy: top-down decision-making becomes more pronounced. This shift can reduce the role of mid-level managers, relegating them to more operational tasks and diminishing their strategic influence in the organization. In essence, the authors propose that automation is not only a tool used for efficiency; it can also become a strategic asset that can reshape the power dynamics within an organization.