Do low levels of trust limit the spread of ideas and knowledge among small-scale firms in African cities? A new study provides micro-level descriptive evidence on the spatial patterns of economic activity among small business owners in one of the fastest-developing cities in southern Africa: Lusaka, Zambia. Innovations for Poverty Action worked with researchers to collect new survey data to investigate the relation between knowledge sharing, trust, and business agglomeration within the city. Preliminary results suggest business agglomeration is positively correlated with both trust and knowledge sharing.

Policy Issue 

Small and medium sized enterprises (SMEs) make up a large portion of the economies of many developing countries and have the potential to help drive economic growth. Since these businesses are smaller, they may lack certain benefits that larger firms can enjoy—for example, ordering in bulk at discounted rates, or taking on large profitable orders. The “clustering” of businesses together increases the opportunity for interactions between firms and may relax the economic constraints of small medium enterprises (SMEs). Clusters of firms may work together to buy or sell to businesses higher or lower up in the production process. As cities in Africa become more urbanized, there is more and more potential for clusters to help drive growth. However, imperfect formal institutions and weak informal arrangements at the local level, such as the lack of reciprocal trust, may discourage business interactions and prevent the potential benefits of clusters to be fully realized. This project aims to investigate the relation between knowledge sharing, trust, and business agglomeration within cities.

Note: The data collection described here is the first step of a longer-term research project.

Context of the Evaluation 

The research project was conducted in Lusaka, Zambia, one of the fastest-growing cities in sub-Saharan Africa. The United Nations estimates that the number of urban dwellers in Zambia will quadruple by 2050 to almost 26 million individuals. In Africa as a whole, the number of urban dwellers is expected to almost triple to around 1.3 billion people.1 In 2016, the Zambian government was in the process of developing a National Urbanization Plan – motivated in part by the acknowledgement that “by ensuring density diversity and innovation, cities can boost economic activities.”2

Details of the Intervention 

Note: This study was not a randomized evaluation.

Innovations for Poverty Action worked with researchers to conduct a Census of Urban Entrepreneurs, which maps all of the economic activity in the city of Lusaka. Between May and September 2016, IPA collected data from 47,428 firms across all industrial sectors on: 1) location, 2) industrial classification 3) number of employees and 4) location description (e.g., type of site, electrification). These data provide unique first-hand evidence on the distribution, size, and characteristics of economic activities in a fast-urbanizing environment, which supplements the official datasets currently available.

IPA then administered a short survey to all businesses with less than 20 employees belonging to the following industries: manufacturing, mining, and construction. 2216 out of 3803 eligible business owners participated in the survey, which collected information on their business practices and history, beliefs and levels of trust, collaborative behavior with other businesses, and demographics.

The research team measured participants’ levels of trust by asking the extent to which they trusted their neighbors and strangers. The study used three measures of knowledge sharing: business cooperation, teaching business skills to others, and talking about business with other business owners. Participants’ propensity to cooperate with other businesses was measured by their self-reported history of engaging in cooperative behavior with other businesses in the same sector.

Results and Policy Lessons 

Trust among small-scale manufacturers: The average level of trust with respect to both neighbors and strangers in this population was particularly low when compared with Africa as a whole.3  With an average level of trust in neighbors of 2.41 (on a scale of 1-4, with four meaning “trust completely”), interviewed entrepreneurs had a lower level of trust than Zambia as whole in 2007 (2.66) and also many other countries in Africa, such as South Africa in 2006 (2.87) and Ethiopia in 2007 (3.12). However, there was some degree of variation within Lusaka.

Trust and business density: Data showed a positive correlation between businesses being located close to other businesses in the same sector and trust in neighbors.4 In areas where similar businesses located near one another (e.g., welders), their owners displayed greater trust levels in neighbors, but not in strangers. This may be either because more trusting business owners decided to move next to each other, or because proximity to similar businesses fostered reciprocal trust, or a combination of the two. The correlation was not affected by other characteristics of the owner, such as gender and experience, by characteristics of the business, such as size, or by features of the geographical administrative area where the business is located, such as their level of economic activity.

Knowledge sharing and business density: The proportion of cooperative activities undertaken by the owner increased with the density of similar businesses. Same-sector business agglomeration was also positively correlated with the likelihood that the owner learned the business from another entrepreneur and greater communication between owners (on business and personal matters).

Sources

[1] https://esa.un.org/unpd/wup/Publications/Files/WUP2014-Highlights.pdf

[2] http://www.urbangateway.org/icnup/sites/default/files/ICNUP%20Zambia.pdf

[3] World Values Survey, 2005-2009 wave: retrieved from http://www.worldvaluessurvey.org/WVSDocumentationWV5.jsp

[4] In the main specification, density was measured as the (log) number of businesses in the same sector within the same Census Supervisory Area (CSA) divided by the CSA area.