The last three decades have witnessed an explosion of theoretical work on the organization of firms (Robert Gibbons and John Roberts forthcoming). In parallel, there has been a massive increase in access to microdata which has revealed huge dispersions in productivity. For example, within narrow industries like cement, oak flooring, and block-ice the total factor productivity of plants at the ninetieth percentile is about twice that of those at the tenth percentile (Lucia Foster, John Haltiwanger, and Chad Syversson 2008). Unfortunately, analyzing to what extent this heterogeneity in productivity is due to management and organizational practices, unmeasured inputs, or other technologies has been held back by a lack of data. National statistical agencies do not usually collect data on the internal organization of companies, nor do firms report this in their accounts. Recently, however, social scientists have been starting to fill this gap by working closely with small numbers of individual firms (e.g., the “Insider Econometrics” approach described in Kathryn Shaw 2009) or covering wide cross-sections of firms (e.g., Nicholas Bloom, Raffaella Sadun, and John Van Reenen 2009). In this paper we describe some of the tools of this research, particularly Bloom and Van Reenen (2007)—henceforth BVR— for measuring management and organizational practices. 1

There is a widespread sense that over the last two decades firms have been decentralizing decisions to employees further down the managerial hierarchy. Economists have developed a range of theories to account for delegation, but there is less empirical evidence, especially across countries. This has limited the ability to understand the phenomenon of decentralization. To address the empirical lacuna we have developed a research program to measure the internal organization of firms - including their decentralization decisions - across a large range of industries and countries. In this paper we investigate whether greater product market competition increases decentralization. For example, tougher competition may make local manager's information more valuable, as delays to decisions become more costly. Since globalization and liberalization have increased the competitiveness of product markets, one explanation for the trend towards decentralization could be increased competition. Of course there are a range of other factors that may also be at play, including human capital, information and communication technology, culture and industrial composition. To tackle these issues we collected detailed information on the internal organization of firms across nations. The few datasets that exist are either from a single industry or (at best) across many firms in a single country. We analyze data on almost 4,000 firms across twelve countries in Europe, North America and Asia. We find that competition does indeed seem to foster greater decentralization.

Why Do Firms in Developing Countries Have Low Productivity? by Nicholas Bloom, Aprajit Mahajan, David McKenzie and John Roberts. Published in volume 100, issue 2, pages 619-23 of American Economic Review, May 2010