By Philippe Aghion
When there’s increased turbulence, delegating power improves sales and productivity, boosting a firm’s chance of survival.
Imagine a ship at sea, at risk of sinking in a tempest. Is it better to empower the crew to do whatever it takes to save the ship, or should every decision be made by the captain and top officers? Similarly, what should the optimal form of firm organisation be during a severe downturn? The need to make tough decisions – including layoffs – may favour firms that concentrate power at the top. However, the turbulence and fast-shifting conditions magnify the value of the information held by local managers.
The two views can be compelling. Indeed, in the depths of the Great Recession of 2009, a survey of executives by The Economist Intelligence Unit revealed that decision making had become more centralised in the C-suite. The rationale: to emphasise “projects that provide benefits across the enterprise rather than individual units”. But in another report three months earlier, the same publication argued that “companies have to deal with dramatically more uncertainty, complexity and ambiguity in the current recession. Success does not come from centralisation.”
So who should be in charge: the crew or the captain?
A new paper by Aghion et al* (2021) titled “Turbulence, Firm Decentralization, and Growth in Bad Times”, shows that large downturns tend to have less dramatic consequences for decentralised firms. When the market is rife with uncertainty, decisions often need to be made in the blink of an eye. As local managers tend to have the most accurate and up-to-date information, they are best able to respond to the evolving situation in a timely manner. Firms that empower them get the advantage.
The paper looks at two large microdata sets: one from ten OECD countries (including France and Japan) and the other leveraging administrative data on manufacturing plants in the United States. Collected in the context of the World Management Survey (a source of high-quality data on management and organisational design across firms around the world), the first set involved direct interviews with plant managers in medium-sized manufacturing firms (between 50 and 5,000 employees). Plant managers were asked how much capital they could invest without prior authorisation. They also shared how autonomous they were in terms of introducing new products, making decisions related to sales and marketing, and hiring permanent employees. The second dataset involved similar questions, courtesy of the US Census Bureau’s 2010 Management and Organizational Practices Survey. These datasets were combined with firm and plant performance data before and after the 2008 financial crisis.
The worse the tempest, the more reason to delegate
In the sectors hit hardest by the crisis, decentralised firms outperformed their centralised rivals on all counts: sales, productivity and, critically, survival. By “hit harder”, we mean those firms in industries, such as durable goods, that experienced the steepest shortfalls in sales and the largest increases in product churn. A marker of turbulence, product churn refers to the pace at which an industry drops existing products to replace them with new ones.
In the international sample, decentralised firms in hardest-hit industries saw an 8.2 percent fall in sales compared to 11.8 percent in centralised firms, a statistically significant difference of 3.6 percentage points. In the US sample, the difference was almost identical, at 3.5 percentage points.
Interestingly, this difference in economic performance was confined to the crisis period. It emerged in 2008 and, in both datasets, the curves converged after roughly five years. So, deciding for or against decentralisation is not necessarily clear cut in non-recessionary times – from a purely business perspective anyway. For instance, a centralised firm can take advantage of economies of scale. Or it can avoid the cannibalising of sales by a given business unit. But when things go wrong, decentralised firms do better, especially those in the toughest environments.
During the Great Recession, the sales of centralised firms shrank three times as much as those of their decentralised rivals. In the worst-hit industries, the total factor productivity increased significantly in decentralised firms, as did their odds of survival. Further analyses showed that the ability of local managers to decide outputs – sales and new products – was more critical than their ability to control inputs like labour and capital investment.
What it means for firms and policymakers
With the Covid-induced downturn already well underway and some economies showing early signs of recovery, is it is too late for firms to change their management style? Not at all. Usually, large crises provide a great window for firms to reorganise. In boom times, every firm should prioritise serving its customers. Downturns reduce the opportunity cost for a firm to review the way it is run. This is what Gilles Saint-Paul and Aghion called the “virtues of bad times”.
Indeed, the two datasets show that after the Great Recession, firms did not immediately adopt the new optimal (more decentralised) form of organisation. But eventually a significant and positive relationship emerged between decentralisation and the size of the negative economic shock. Firms have learned, it appears.
What can policymakers do with these findings? Obviously the role of the state is not to tell firms how they should run their business. But what the state can do is to foster an educated workforce. Because the more educated the workforce, the more attractive decentralised structures become, as reliable employees can be counted upon to make smart decisions – whether in a downturn or not. Conversely, an uneducated workforce favours more hierarchical structures, of the type that belong to past eras that benefitted early industrialists such as Henry Ford. Aside from putting enough thought behind their national education systems, governments could also induce training within firms, via grants or tax credits.
In addition, governments should implement pro-competition and pro-entry policies. Over time, the combination of education and training on one hand, and competition-fostering policies on the other, will promote the emergence of more efficient structures, very often translating to more decentralised firms. In fact, any policy encouraging innovation and creative destruction can provide the right soil for decentralisation to take hold.
A trend for the betterment of society – beyond pure economics
There is already a movement towards flatter organisations in the US and the United Kingdom, as multiple researchers, such as Rajan and Wulf, have documented. Garicano and three of Aghion’s co-authors on the paper discussed here have showed that increasingly better information technologies (such as Enterprise Resource Planning for plant managers) have allowed firms to become flatter, giving workers more autonomy and a wider span of control. The Machine That Changed the World – a classic published in 1991 – examines how lean manufacturing contributed to the trend towards decentralisation.
Beyond the unmistakable business advantages they provide during downturns, flatter organisations create better jobs. These jobs are associated with better pay, better training, better job security, better opportunity for promotion – and ultimately better social mobility. Of course, the movement towards flatter organisations also has the potential to benefit other stakeholders. For instance, more innovative products can come onto the market, as shown in other research to which Aghion contributed.
In a business context that increasingly mimics a turbulent sea, with storms constantly lurking on the horizon, firms have to be ready to respond to crises as they come. Today it is a health crisis, tomorrow it may be an environmental one. Firms that will delegate decision making to smart, committed and empowered managers are most likely to reach their ideal port of destination.
*The paper is co-authored with Nicholas Bloom of Stanford University, Centre for Economic Performance, NBER and CEPR; Brian Lucking of Stanford University; Raffaella Sadun of Harvard University, Centre for Economic Performance, NBER and CEPR; and John Van Reenen of MIT, Centre for Economic Performance, NBER and CEPR.
Philippe Aghion is a Professor of Economics at INSEAD. He is also a Professor at the College de France, a visiting professor at the London School of Economics and a fellow of the Econometric Society and of the American Academy of Arts and Sciences.
Frontpage December 20, 2019