Skip to content Skip to navigation

Networks and manufacturing firms in Africa

Please note that prior to September 2017, the Center on Global Poverty and Development was known as the Stanford Center for International Development (SCID).

Dozens of uniformed Ethiopian workers stitch garments in a large clothing factory.

<p>Factory workers stitch new garments in a large clothing factory in Ethiopia</p>

May 22 2015

Posted In:

Research Spotlights, SCID News

By Sam Zuckerman

What is the Issue?

Business experts often use the phrase “best practices” to refer to management methods with proven records of success. And that raises an obvious, but critical question: why doesn’t every business adopt recognized best practices? The reality is, best practices aren’t universal. Researchers have consistently found that management methods differ greatly from one company to another, even among companies in the same country and similar business lines. These differences are particularly striking in Africa and other developing regions, where variation in management quality is typically wider than in developed countries. Inconsistent management quality is an obstacle to development. For economies to advance and for countries to be competitive in a globalized market, it is essential that management standards rise.

In principle, best management practices might spread from company to company along their managers’ peer networks. A businessperson’s peer network is potentially a valuable resource. Through contacts with other businesspeople, a manager may get information about customers or suppliers, meet potential business partners, find out about technology, and learn how to better supervise a work force. Business peer relationships may be especially important in Africa, where formal training or outside expertise might not be easy to get. But the fact that African management quality varies so widely indicates that best management practices aren’t diffusing effectively along business peer networks.

Much research has been done on social networks outside the business arena, confirming that behavior often spreads widely among peer groups. But much of this research has looked at how people act when they are young and highly impressionable. Fewer studies of social networks have focused on businesspeople, and remarkably little is known about transmission of management practices through peer relationships.

Business managers work in a world far different from that of adolescents. They face pressure to produce profits, sometimes amid intense competition, and that may color their relationships with other managers. Researchers don’t know much about what information is shared among business peers, nor do they know what prompts businesspeople to act on information they receive. Do practices in some management areas, such as finance or labor management, spread more easily through networks than practices in other areas? What network features increase the chances that managers will take up practices they hear about from peers? On the other hand, what features might keep managers from making changes based on what they learn from other businesspeople?


Sub-Saharan Africa faces severe development challenges and greatly needs to expand business management know-how. Could policymakers use business peer networks to spread best practices and boost management quality? To answer this question, it is important to know to what extent management techniques are spreading through business networks and what changes might increase diffusion of best practices.

Marcel Fafchamps, an economist at Stanford University’s Freeman Spogli Institute for International Studies, and Simon Quinn, an economist at Oxford University’s Centre for the Study of African Economies, designed a novel experiment to investigate the spread of management practices through African business networks. They devised a way to manipulate the networks of manufacturing company managers by introducing new peer contacts. They then examined whether their subjects picked up management techniques from these new peers. Fafchamps and Quinn describe their study as one of the first to vary connections in business networks experimentally in order to measure the effects of peer relationships.

To build new peer relationships among African business managers, Fafchamps and Quinn organized business proposal competitions in Ethiopia, Tanzania, and Zambia in 2011, similar to the popular U.S. television show Shark Tank. Aspiring entrepreneurs got opportunities to pitch business plans to panels of managers, with cash prizes awarded for best ideas. For each competition, the researchers recruited more than 100 senior managers from small and medium-size food processing and manufacturing companies to serve as judges. Some were randomly assigned to judging committees, typically five or six members, which met three times to choose prize winners. Others served as solo judges and did not interact with panel members. The competition provided ample opportunity for judges to communicate. In addition to deliberating on prize winners, the judges were seated together at award ceremonies, where they were able to socialize. The aim: create an environment that encouraged judges to share experiences and opinions about management strategies, then determine whether the judges joined each other’s business networks. To complete the experiment, Fafchamps and Quinn investigated whether contact with other judges prompted managers to change the way they did business.

As part of the experiment, the researchers distributed four factsheets to committee and solo judges on such topics as labor or exports. Two-thirds of the judges received two randomly selected factsheets, while the remaining judges received none. The purpose was to randomly vary information provided to participants as an additional way for researchers to test how information traveled through networks.


Fafchamps and Quinn’s experiment created new business links. Participants remembered their fellow judges and, in some cases, communicated with them after the experiment. On average, each committee judge later spoke with one fellow committee member, sometimes specifically about management questions. But the contest experience did not change how participants perceived their business networks. Moreover, relationships formed with other judges brought about only limited transmission of management practices. If managers sat on committees in which a fellow judge had a business bank checking account, they had about a 4 percentage point higher probability of having a bank checking account for their own businesses by the experiment’s end. And if a fellow judge ran a business registered for the national value-added tax, the probability of VAT registration of their own business increased by about 7 percentage points. However, during the experiment, enforcement of VAT regulations was tightening in all three countries under study, which makes this a special case. Managers may have been looking to learn from peers about a business practice that was a big concern in the business community at the time.

When the researchers considered business size, they found more evidence of diffusion. Managers of the smallest businesses raised their chances of hiring an outside auditor if they served with judges whose businesses did so, while larger enterprises were more likely to advertise, introduce new products, or change production processes. Yet, the researchers saw no signs of transmission of management methods in client relations, labor, or innovation. On the whole, the study find no widespread evidence that business practices spread through peer networks.

But what about the limited spread of management practices that did take place? How did it happen? Fafchamps and Quinn saw two possibilities: One was that simply placing businesspeople on a committee together may have been enough to get them to exchange management information after the contest. Alternatively, they may have needed an additional nudge to boost the chances they would speak. The factsheets distributed to judges offer evidence. Those judges who randomly received the same factsheet were about 5 percentage points more likely to have spoken after the contest, and were more likely to have talked about labor management and innovation. These results suggest that the spread of business practices observed in the study stemmed more from conversations stimulated by the factsheets than from the business managers’ independent decisions to seek out peers whose expertise might be useful. In other words, once the judges had met, diffusion occurred primarily through relationships created by the researchers, not through relationships the judges chose on their own.

Policy Implications

Researchers in a range of social contexts have found evidence that behavior spreads through peer networks. But Fafchamps and Quinn’s study suggests that results in other areas can’t explain how management practices diffuse among competing businesses. Their business competition experiment introduced managers to potential additions to their peer networks. Yet the managers generally did not take advantage of these fresh contacts to adopt new management practices. Fafchamps and Quinn offer three possible explanations: First, entrepreneurs may not want to help real or potential competitors run their businesses better. Second, through peer relationships, managers may hear entrepreneurial horror stories that frighten them from making changes, for example, anecdotes about businesses that failed when they tried to boost exports. Third, businesspeople may simply be set in their ways and not see a need to learn from other managers. Reasons such as these might prevent business peer networks from transmitting best practices from one manager to another, explaining the wide variation in management quality in Africa and other developing regions.

These results help explain why some managers stick with inferior business practices. Despite some evidence of diffusion, managers’ peer networks do not appear to be good instruments for spreading best practices and reducing variations in management quality from one business to another. Policymakers seeking to promote development by improving management standards might have more success with other tools, such as training and technology. Peer networks may nonetheless be useful for diffusing information about recent regulatory changes.