Networks and Manufacturing Firms in Africa: Results from a Randomized Field Experiment

Addis Ababa, Dare Salam and Lusaka
337 firm managers
2010 - 2012
Target group:
  • Small and medium enterprises
  • Large enterprises
  • Managers
Outcome of interest:
  • Firm formality
  • Social cohesion
  • Soft skills
  • Take-up of program/social service/healthy behavior
Intervention type:
  • Information
  • Social networks

Management practices differ substantially across firms. Small firms in emerging economies, in particular, do not always employ best business practices, often causing them to fall behind in terms of productivity. Researchers conducted a randomized evaluation in Ethiopia, Tanzania, and Zambia to measure whether generating social connections between manufacturing firm managers could lead to the diffusion and adoption of best business practices. These social interactions led to the spread of best practices related to business formalization, but did not increase the implementation of other practices that may boost business productivity.

Policy issue

Management practices differ substantially across firms. Small firms in emerging economies, in particular, do not always employ best business practices, often causing them to fall behind in terms of productivity. In developing countries, where the transaction costs of learning from the market may be high, networking amongst firm managers may be a valuable source of information sharing about topics such as customers and suppliers, investment opportunities, access to technologies, business partners, regulations, and labor policies. However, open questions remain regarding why best management practices do not diffuse from firm to firm. Do social networks amongst managers lead to the diffusion of better management practices? If so, what kinds of management practices are spread across firms? Can researchers and policymakers promote these social networks in order to encourage the spread of best business practices and to increase firms’ productivity overall?

Context of the evaluation

From 2011-2012, Ethiopia, Tanzania, and Zambia ranked 129, 104, and 91, respectively, out of 140 countries in terms of business sophistication, according to the World Economic Forum’s Global Competitiveness Report—a measure of the quality of a country’s overall business networks and of individual firms’ operations and strategies. In these low-income contexts, business competition typically is concentrated in large, industrial urban centers and amongst small and medium firms. The average firm in this study, for example, had six permanent employees.

In recent years, business competitions—where participants propose business plans and models to win investment money—have grown in popularity in parts of Africa, often with entrepreneurs and managers serving as judges. These types of competitions provide an opportunity for firm managers to interact and learn best practices from one another.

Details of the intervention

Researchers conducted a randomized evaluation to measure the effect of social connections between manufacturing firm managers in Ethiopia, Tanzania, and Zambia on the adoption of best business practices.

In each country, researchers ran a business plan competition—called the Aspire Business Ideas Competition. Young entrepreneurs from five leading industries—food process, garments, leather, metal, and wood—presented proposals for new enterprises to win USD$1000 in investment money. Managers of small- and medium-size manufacturing firms served as competition judges. Researchers randomly assigned them to one of the two following groups:

  1. Committee judges: These judges assessed proposals together with other managers in a committee. Committees comprised between 3-6 judges and judged around 12 applicants. Each committee prompted managers to work together and socialize with other managers, giving them an opportunity to share best business practices on an informal basis. At the end of the competition, a prize ceremony gave committee judges the chance to engage socially.
  2. Non-committee judges: These judges assessed proposals individually and without conferring with other managers. This group served as a comparison scenario in which there was no orchestrated social interaction amongst firm managers.

In addition, at the conclusion of the competition, two-thirds of all managers randomly received two factsheets about either labor, innovation, or export practices, or about the research implementing group. This allowed researchers to assess whether new social connections occurred as a result of managers randomly receiving the same factsheet—thus having the same type of background business information to talk about—or as a result of managers self-selecting who to socialize with.

Overall, researchers assigned 239 managers to committees, leaving 106 as non-committee judges. Researchers measured whether serving as competition judges led to the creation of social networks between managers, and consequently whether these social networks improved firms’ business practices, using individual and firm-level survey data collected in late 2011 and early 2012.

Results and policy lessons

The business competition gave managers the opportunity to increase their social connections with other managers. In turn, managers’ social interactions led to the diffusion of best practices related to business formalization (VAT registration and opening a bank account), but did not increase the implementation of other types of best practices.

Social connections between managers: Committees created new social links amongst managers: 38.2 percent of managers assigned as committee judges recalled a peer who was actually on their committee, while 5.6 percent erroneously remembered a peer who was not on their committee. Managers from the same committee were 15.4 percentage points more likely to speak with each other after the competition than managers in different committees (who were only 2 percent likely to speak to each other following the competition) —implying that approximately one new social link was created per manager within a committee. Sharing a committee also increased the likelihood that participants discussed specific management practices including export strategies, labor management, and innovation.

Firms’ business practices: Having a VAT-registered committee peer increased the likelihood of VAT registration by about 7 percentage points–an 81 percent increase from an original VAT adoption rate of 8.1 percent. However, all three countries were undergoing major VAT reforms around the time of the competition—and only about 8 percent of manufacturing firms were registered for the VAT a year prior to the competition—suggesting firms may have been particularly interested in learning about this topic from their peers.

Having a committee peer whose business already had a bank account also increased the likelihood of committee firms opening a bank account by about 4 percentage points (a 10 percent increase over an original rate of 42.2 percent of comparison firms).

Researchers also found evidence suggesting that the diffusion of VAT registration practices did not occur through simple imitation of peers: firms were more likely to register if they had peers who had done so, but were no less likely to register if they had more peers who had not done so. Outside of VAT and bank account adoption, however, exposure to peers did not increase firms’ likelihood of adopting other business practices pertaining to labor management, relations with clients and supporters, or innovation-related matters.

While the mechanisms linking better business practices to managers’ exposure to networks are not fully clear, researchers found suggestive evidence that the diffusion of best practices occurred through conversations that researches randomly induced by distributing factsheets, rather than through managers self-selecting who they spoke with. Managers from the same committee who randomly received the same factsheet were 6.7 percentage points more likely to speak to each other than managers from the same committee with different factsheets.

Overall, these results imply that externally-crafted social connections can prompt the diffusion of best-practices. However, social connections may not be enough to completely narrow the gap in best management practices between firms. These results may be explained by the high costs that come with changing a firm’s business practice, or because managers may be reluctant to incorporate and share advise, particularly with possible competitors. Furthermore, social networks may also disincentive firms from adopting good practices if they hear that such practices were not effective for another business. As a result, inferior business practices and productivity differentials between firms may persist even when managers have access to business networks.

The researchers applied the insights from this study in subsequent randomized evaluations of management placement programs for young professionals and job fairs in Ethiopia.1 2 The research design of this study also inspired another research project that studies the effect of business networks on firm performance in China.3 Furthermore, in a subsequent analysis, the researchers of this study also studied the impact of the cash prizes of the Aspire competition on winners’ entrepreneurial success.4

Abebe, Girum, Marcel Fafchamps, Michael Koelle and Simon Quinn,“Learning Management Through Matching: A Field Experiment Using Mechanism Design”,NBER Working Paper No. 26035,July 2019.
Abebe, Girum, Stefano Caria, Marcel Fafchamps, Paolo Falco, Simon Franklin, Simon Quinn and Forhad Shilpi. 2018.“Job Fairs: Matching Firms and Workers in a Field Experiment in Ethiopia”.
Cai, Jing, and Adam Szeidl. “Interfirm Relationships and Business Performance”. The Quarterly Journal of Economics, Volume 133, Issue 3, August 2018, p. 1229–1282.
Marcel Fafchamps and Simon Quinn,“Aspire",Journal of Development Studies, October2017, 53(10): 1615-33.