2018 Theses Doctoral

Essays on Development Economics

de Rochambeau, Golvine

After two civil wars and a health epidemic destroyed most of Liberia’s economy, paths to the country’s recovery and growth have seldom been studied in the economics literature. This dissertation contains three chapters on topics around a particular factor of growth: the private sector. The first two chapters address the freight transport sector in Liberia. Understanding how and at what cost commodities are transported intra-nationally is key to promote sustainable growth. The third chapter explores the tender process, and how it can be made more efficient.
In Chapter 1, I study the introduction of a new technology in Liberia’s transport sector. While standard economic theory predicts that the introduction of a monitoring device should increase the productivity of the worker, I show that this is not necessarily the case. I use an experiment where I introduce a monitoring technology on trucks to measure the impact of the technology on the worker. I find three main results. First, the technology had an overall positive effect on the drivers who received the technology. Second, the managers choose to install the technology only on the drivers they trust the least. Third, the technology had an adverse effect on the drivers at the margin. These results are consistent with a theory of intrinsic motivation. The key insight of the model is that monitoring workers crowds out their intrinsic motivation to work hard. If a driver is intrinsically motivated, monitoring him may have adverse effects on the effort he provides. This chapter is the first evidence from the field of the adverse effect of a monitoring device. The results provide evidence that a blind application of monitoring devices to the entire worker-base may produce suboptimal effects and that managers seem on average to be able to identify the workers who are not intrinsically motivated.
In Chapter 2, I propose an estimation of Liberia’s intra-national trade costs, and shed light on why these estimates are so high. Using data on Liberia’s commodity prices, I estimate how trade costs vary with distance. I find that estimated trade costs per distance are significantly bigger in Liberia than they are in other countries. I then use data from a survey of transport companies to estimate travel times. Trade costs per travel time — and not per distance – are still bigger in Liberia than they are in other countries, but to a lower extent than trade costs per distance. I conclude that the speed of vehicles, which captures road-specific characteristics such as road quality, explains a significant share of the difference in trade costs between Liberia and other countries.
Chapter 3 evaluates the effect of a training program that teaches firms how to apply to tenders. We show that firms who took such a training increased the number of contracts won as well the quality of those contracts. However the impact of this training is heterogenous across firms. In particular, we explore this heterogeneity along expected discrimination. Firms who believe that they are being discriminated against often have lower chances of winning tenders. These firms also accumulate less experience in applying to contracts, and could therefore benefit more from the training. Results show that the effect of the training is a non-monotonic function of expected discrimination. For firms in the low ranges of expected discrimination, the effect of the training counter-balances the effect of higher expected discrimination. In higher ranges of expected discrimination, the effect of higher expected discrimination dominates and more discriminated firms are less likely to provide additional effort in response to the training. We conclude that the firms who benefit from the program are not the most discriminated firms but the ones in the middle of the distribution.