UC Berkeley

Low productivity in the private sector, especially amongst small businesses, and poor public sector service delivery are significant barriers to sustainable and equitable development in low-income countries. This dissertation studies these barriers through the lens of management challenges. It comprises of three chapters, where the first chapter examines the potential of SMS-based business trainings to address gaps in management skills amongst micro-entrepreneurs in Kenya, the second sheds light on the demand for SMS-based business trainings amongst these micro-entrepreneurs, and the third studies the effectiveness of a command and control management intervention implemented at scale in Punjab, Pakistan.

In my first chapter, titled "Short Messages Fall Short for Micro-Entrepreneurs: Experimental Evidence from Kenya", I study the effectiveness of SMS-based business management trainings for improving outcomes for micro-entrepreneurs. SMS-based trainings are becoming a popular tool to remotely support micro-entrepreneurs in low-capacity contexts due to their scalability and low costs. However, little evidence exists on the effectiveness of such trainings to improve business outcomes. In this study, I evaluate a field experiment in which access to an SMS-based training was randomized across 4,700 micro-entrepreneurs in Kenya. After three months, I find positive effects on knowledge and adoption of best business practices. Younger entrepreneurs see stronger effects on sales, profits and business survival, driven by higher engagement with training content, more time spent on business, and getting larger loans. Contrary to predictions elicited from social scientists, I find that these positive effects disappear twelve months after the intervention, as all engagement with content ended within the first five months. Findings from this study suggest that, despite the promise and wide-spread use, SMS-based trainings are unlikely to be effective for micro-entrepreneurs in the long run. Results highlight the importance of lack of engagement as a major challenge limiting the potential of remotely provided information-based support.

My second chapter, titled "Demand for SMS-Based Business Trainings Amongst Kenyan Micro-Entrepreneurs", studies the demand for SMS-based business management trainings in Kenya. I leverage two key components added to the field experiment from the first chapter to measure the demand; first, upon completion of the business management training, or prolonged disengagement from it, micro-entrepreneurs in the treatment group were given the opportunity to buy a second SMS training through Take-It-Or-Leave-It (TIOLI) offers where the asking price was randomized across three levels. Observing buying decisions across the randomized price levels allows me to study how demand changes with price, and also sheds light on correlations between individual and enterprise characteristics and demand. Second, I conduct an in-person demand elicitation activity with a select subset of the sample across the treatment and control groups, using a modified version of the Becker-DeGroot-Marschak (BDM) method. In the TIOLI sample, 70% of individuals chose to accept the additional training when it was offered for free, 68% accepted when the price was half the marginal cost faced by the service provider, and about 50% accepted when the price was double the marginal cost. In the BDM sample, the average willingness to pay for SMS trainings was five times the marginal cost, and almost a quarter of the respondents were willing to buy the training for ten times the marginal cost. Both methods of demand elicitation thus showed that micro-entrepreneurs were willing to pay a positive amount for SMS-based business management trainings. I also find correlational evidence suggesting demand for trainings was higher amongst individuals with more children in the household, those that recently applied for a loan, those with more knowledge of best practices, and those with higher education levels. Taken together with results from the first chapter, these findings suggest that engagement levels might not reflect true demand for SMS-based trainings, pointing towards possible behavioral explanations driving under utilization of the resource.

Finally, in my third chapter, titled "Command and Can’t Control: Assessing Centralized Accountability in the Public Sector", I study the potential and limitations of centralized management in the public sector, with Saad Gulzar, Juan Felipe Ladino, and Daniel Rogger. A long-established approach to management in government has been the transmission of information up a hierarchy, centralized decision-making by senior management, and corresponding centralized accountability; colloquially known as 'command and control'. We examine the effectiveness of a centralized management and accountability system implemented at scale in the public education sector bureaucracy of Punjab, Pakistan, for six years. The scheme automatically identified poorly performing schools and jurisdictions for the attention of central management. We find that flagging of schools and corresponding de facto punishments had no impact on school or student outcomes. We use detailed data on key elements of the education production function to show that command and control approaches to managing the general public sector do not induce bureaucratic action towards improvements in government performance.