Foreign and Domestic Ownership of Rwanda’s Coffee Industry

Firms vary widely in their management practices and productivity, especially in low-income countries. In Acquisitions, Management, and Efficiency in Rwanda’s Coffee Industry (NBER Working Paper 30230), Rocco Macchiavello and Ameet Morjaria find that Rwandan coffee mills become more productive after they are acquired by a foreign firm, but not when they are acquired by a domestic one.

Coffee mills purchase coffee cherries — beans that have not yet had the outer pulp removed — from farmers and convert them into parchment coffee, an intermediate stage in bean processing. Managers are responsible for supervising employees in charge of each step of the process, developing relationships with farmers to source cherries, overseeing hiring of seasonal workers and agents, and paying farmers and workers.

As the industry grew and consolidated, foreign owners improved mill performance by better implementing good management practices, compared to domestic owners.

The number of Rwandan mills has grown rapidly, from a handful in 2002 to 310 in 2017. Until 2011, all mills were owned domestically, typically by their builders. Some owners were individuals who owned a stand-alone mill; others were groups that owned two or more mills. After 2011, foreign multinationals began acquiring mills. By 2017, over 50 percent of mills were under group ownership, including a sixth of mills owned by seven foreign groups.

The researchers conducted several mill-level surveys over the last decade to collect information on operational aspects: price and quantity of input, quantity and quality of output, production cost, employment, mill manager characteristics, and key managerial practices. They also interviewed all CEOs of group-owned mills to ask how they target mills for purchase, what alternative target mills they considered, and details of acquisitions that failed, enabling them to create multiple counterfactuals for comparison to the target mill. They combined surveys with administrative data on processing capacity, input procured, and ownership history to document and understand performance differences. By comparing mills that changed ownership type, the researchers find that foreign acquisition increases the likelihood that the mill is operational in any given year. While both foreign- and domestic-owned mills expand processing capacity and employment, foreign-owned mills are able to source more cherries from farmers. In comparison to domestic stand-alone mills, foreign-owned mills have 23 percent greater capacity utilization. Labor productivity, the ratio of output to labor, is similar for individual-owned and foreign-owned mills, but 25 percent lower for mills owned by domestic groups.

The researchers rule out the possibility that the performance differences between foreign and domestic groups is due to differences in processing technology across mills or differential access to sources of finance. They do find, however, that managers at foreign-owned mills are different: they are younger, more educated, and have higher ability. These manager characteristics can explain around a quarter of the differences in performance between foreign-owned and domestic-owned mills.

The researchers argue that the rest of the performance differences seem to be driven by managerial practices. However, they note both groups’ managers equally attempt management changes — suggesting lack of knowledge and lack of incentives are unlikely to be drivers of performance differences. Implementation differences could drive the performance differences.

Organizational capabilities post-acquisition are different across the two groups. Foreign firms deploy several complementary practices to support implementation: they grant managers more autonomy to implement changes, and they ensure that managers do not misuse their increased autonomy through a combination of monitoring (via IT infrastructure) and paying them wages above their market salary.

Managers of domestic-group mills report facing more resistance from farmers and workers when making management changes; the groups’ different pre-acquisition capabilities might explain this. Personal relationships in the community are important for domestic groups (which are barely mentioned by foreigners). The importance of these relationships in driving target selection is reflected in mill managers’ birthplaces — around 70 percent of managers of domestic-group mills were born in the same district as their mills, compared to only around 20 percent of managers of foreign-group mills.

These findings highlight the possibility that while domestic groups’ embeddedness in the local community creates opportunities to invest, it can also create pressure to maintain status-quo relational arrangements, hindering implementation of management changes.

—Whitney Zhang