Centralise or de-centralise?

In my career, I have been working for two global banks and one thing that always baffled me was the constant reorganisations as bosses and “strategies” changed. One dimension that nobody ever could settle on was whether to organise the company based around a central sales and product unit or whether to de-centralise the company and regionalise sales and production.

Time and again, the higher-ups would decide to de-centralise the business to be able to serve customers better and understand the local needs of customers. This, of course would lead to local kings who treated their regions like their own little fiefdom and took less and less notice of the demands and requirements of the company overall. Then, a new boss came in and decided to centralise operations and sales in order to be able to cut costs and increase economies of scale. After a while, the regions would complain about unsuitable products (either perceived or real) and the pendulum swung back to de-centralisations.

In my experience, none of these measures really added value. Instead, they cost the companies a lot of money and created new jobs for some executives but in the end, I never thought the company or its shareholders got much out of it.

I don’t follow the literature on business strategy because in my experience, this is mostly unsound theorising combined with a lack of empirical foundations. So, it could be that the question if a company performs better when it is centralised or de-centralised has been addressed there. But in economics, I think a new paper by Philippe Aghion and his colleagues is a good first effort in trying to quantify the benefits of either organisational form.

The researchers looked at export-oriented firms in the United States and 10 developed countries and their performance during the financial crisis of 2008 and the Eurozone debt crisis of 2011 to 2012. They compared the sales growth of de-centralised firms vs. centralised firms in times of crisis and in normal times. And while this is only a first effort and needs to be checked with other crises, the result is quite interesting.

Apparently, in normal times, there is no significant difference in sales growth between centralised and de-centralised firms. In the United States and in other developed countries, de-centralised firms have slightly lower sales growth than centralised firms, but the differences are so small as to be meaningless. But when a crisis hits, de-centralised firms clearly outperform in terms of sales growth since de-centralised firms can adjust better to local needs and better understand the impact a crisis has on the local market, Thus, over a cycle, it seems as if de-centralised firms tend to have an advantage in being more robust to external shocks. And that is something to remember for investors as well as managers. 

Sales growth of de-centralised vs. centralised companies (Panel A: 10 OECD countries, Panel B: United States)

Source: Aghion et al. (2021).