the Decentralization of the Firm

Technology, Information, and the Decentralization of the Firm D Acemoglu, P Aghion, C Lelarge… – … Quarterly Journal of …, 2007

Decentralization of the firm: theory and evidence AA Christie, MP Joye, RL Watts – Journal of Corporate Finance, 2003

The organization of firms across countries N Bloom, R Sadun, J Van Reenen – The quarterly journal of …, 2012

Does product market competition lead firms to decentralize? N Bloom, R Sadun, J Van Reenen – 2010


Technology, Information, and the Decentralization of the Firm D Acemoglu, P Aghion, C Lelarge… – … Quarterly Journal of …, 2007

Recent increasing interest in the determinants of firms’ organizational choices
← motivated by the belief that new technologies are inducing firms to become less hierarchical and more decentralized
→ limited work on the determinants of the decentralization decisions of firms
→ how the allocation of authority within firms changes as the information structure in an industry evolves.

A simple model:
firms make choices on how to implement new technologies
Different organizational forms are distinguished by the amount of information they use in these decisions:

centralized control relies more on the information of the principal (Aghion&Tirole, 1997)) + Decentralized control delegates authority to a manager, who potentially possesses more information than is available in the public history
← but agency problem as the interests of the principal and the manager are not perfectly aligned
=> tradeoff between the superior knowledge of the manager and the agency costs of managerial delegation determines the optimal degree of decentralization

→ how learning from the experiences of other firms changes this tradeoff (the more a principal learns the implementation of new technologies, the less she needs to delegate control to the manager)

Using this basic framework, 3 sets of empirical predictions:
1. Firms closer to the technological frontier are more likely to choose decentralization, because they are dealing with new technologies about which there is only limited information in the public history.
2. Firms in more heterogeneous environments are more likely to be decentralized, because greater heterogeneity makes learning from the experiences of others more difficult.
3. Young firms, which have had a limited history in which to learn about their own specific needs, are also more likely to be decentralized than older firms.

two large data sets of French firms and establishments and one smaller set of British establishments in the 1990s
A range of empirical patterns consistent with these predictions:
firms closer to the technological frontier of their industry,
firms operating in more heterogeneous environments,
younger firms are more likely to choose decentralization.

+ the relationship between heterogeneity or distance to frontier and decentralization is significantly stronger in high-tech than in low-tech industries.

Measure of decentralization:
whether different units of the firm are organized into “profit centers.” → proxying decentralization by the extent of delayering or measures of managerial autonomy over investment and/or employment decisions

On the theoretical side, our paper is most closely related to the literature on the costs and benefits of delegation or decentralization in firms. A first strand of this literature, for example, Baron and Besanko (1992) and Melumad, Mookherjee, and Reichelstein (1995), investigates the conditions for delegated contracting to replicate the constrained efficient centralized contracting. As emphasized by Mookherjee (2006), however, the presence of complete contracts in these models implies that delegation can at best replicate the constrained efficient allocation that is also achievable by centralized contracting. A second strand emphasizes information processing and communication costs as determinants of centralization or decentralization in firms.2 Although we also stress the importance of learning, our focus is different, namely on how public information affects how much autonomy the principal would like to grant to the agent. Closer to our paper are the recent models emphasizing the trade-off between loss of control and better information under decentralization—in particular, Aghion and Tirole (1997), Baker, Gibbons, and Murphy (1999), Rajan and Zingales (2001), Dessein (2002), and Hart and Moore (2005).3 The main differences between these papers and ours are twofold: first, because there are no incentive effects of the form of the organization, our framework is significantly simpler and allows us to focus on the basic tradeoff between information and loss of control; second, we allow the principal to learn from other firms’ or from her own firm’s past experience, which is the source of all the comparative static results we investigate in the empirical work.

The main contribution of our paper is the empirical evidence we provide on the determinants of decentralization. Previous work in the literature focuses on the general move toward “flatter” organizations.5 Rajan and Wulf (2006) provide the most systematic statistical description of recent organizational trends, showing a strong movement toward flatter corporations in the United States between 1986 and 1999. Caroli and Van Reenen (2001) and Bresnahan, Brynjolfsson, and Hitt (2002) report a positive association between various measures of decentralization and organizational change on the one hand and information technology (and human capital) on the other. Baker and Hubbard (2003, 2004) document the effect of new technologies (on-board computers) on ownership patterns in the U.S. trucking industry. Other related papers include Colombo and Delmastro (2004), who present empirical models of decentralization in Italian manufacturing plants; Lerner and Merges (1998), who examine the allocation of control rights in biotechnology alliances; and the papers by Ichinowski, Prenushi, and Shaw (1997), Black and Lynch (2001), and Janod and Saint-Martin (2004), which examine the impact of human resource practices and firm reorganization on productivity. None of these papers investigate the relationship between decentralization (or organizational change) and distance to frontier or heterogeneity.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s