Explore the vast range of titles available.

In this dissertation, I study the behavior and the factors that impact the performance of firms in developing countries. Chapter 1 and 3 investigate the determinants of patterns of trade in Myanmar, a country which over the past decade has been undergoing an extraordinary transition, from military control and diplomatic isolation to political and economic liberalization. Chapter 2 studies how firms upgrade the quality of their output to increase sales abroad. Specifically, in Chapter 1, I investigate the hypothesis that if matching frictions in international trade are important, a seller’s ability to connect with buyers could explain a substantial part of exporters heterogeneity in size. I do so in Myanmar’s bean export market. Despite beans having all the attributes of a commodity, there is significant transaction price dispersion across both exporters and foreign buyers. Empirical patterns are consistent with foreign buyers facing search costs to find exporters. I estimate a model of search and auctions, where foreign buyers first search for a set of exporters, and then run a competitive bidding process between exporters within that set. In the model, exporters are described by two parameters: a visibility parameter that impacts their likelihood of being found by foreign buyers and a cost parameter that drives the level of their price quotes and thus their market share with each foreign buyer. Visibility explains an important part of the firm size distribution. On the buyer side, searching for an additional exporter has an estimated cost of about $2,000. Moving to a centralized market would lead to a five percent decrease in transaction prices. Chapter 2 looks at the relationship between firms’ output quality and their organizational structure. Using data on the production and transaction chain that makes up Peruvian fishmeal manufacturing, we establish three results. First, firms integrate existing suppliers when the quality premium rises for exogenous reasons. Second, suppliers change their behavior to better maintain input quality when vertically integrated. Third, firms produce a higher share of high-quality output when supplier availability constraints shift them into using integrated suppliers. Overall, our results indicate that quality upgrading is an important motive for integrating suppliers facing a quantity-quality trade-off, as classical theories of the firm predict. Chapter 3 quantifies the impact of import license liberalization in Myanmar’s unique political economy environment. By contrast to previous literature on the issue, we find that liberalization did not lead to substantial entry in the sectors populated by firms connected to the party in power. We document two facts that rationalize these findings. First, connected firms tend to import products subject to important economies of scale, which provide opportunities for rent-seeking and act as a “natural” barrier to entry for small firms. Second, we show that a subset of the products liberalized de jure were not liberalized de facto. Products not liberalized de facto are more likely to be sectors where connected firms are present and where economies of scale are less important. This last result suggests that institutional arrangements were made to protect connected firms in the sectors where they faced higher potential competition.

A distinct feature of MNCs is a three-tier organizational structure: foreign managers (FMs) supervise domestic managers (DMs) who supervise production workers. Language barriers between FMs and DMs could impede transfers of management knowledge. We develop a model in which DMs learn general management by communicating with FMs, but communication effort is non-contractible. These conditions generate sub-optimal communication within the MNC. If communication is complementary with language skills, the planner could raise welfare by subsidizing foreign language acquisition. We experimentally assess the validity of the general skills and the complementarity assumptions in Myanmar, a setting where FMs and DMs communicate in English. The first experiment examines the general skills assumption by asking prospective employers at domestic firms to rate hypothetical job candidates. They value candidates with both higher English proficiency and MNC experience, a premium driven, in part, by frequent interactions with FMs. The second experiment examines the complementarity assumption by providing English training to a random sample of DMs working at MNCs. At endline, treated DMs have higher English proficiency, communicate more frequently with their FMs, are more involved in firm management, and perform better in simulated management tasks. Organizational barriers within MNCs can thus hinder knowledge transfers.

A distinct feature of MNCs is a three-tier organizational structure: foreign managers (FMs) supervise domestic managers (DMs) who supervise production workers. Language barriers between FMs and DMs could impede transfers of management knowledge. We develop a model in which DMs learn general management by communicating with FMs, but communication effort is non-contractible. These conditions generate sub-optimal communication within the MNC. If communication is complementary with language skills, the planner could raise welfare by subsidizing foreign language acquisition. We experimentally assess the validity of the general skills and the complementarity assumptions in Myanmar, a setting where FMs and DMs communicate in English. The first experiment examines the general skills assumption by asking prospective employers at domestic firms to rate hypothetical job candidates. They value candidates with both higher English proficiency and MNC experience, a premium driven, in part, by frequent interactions with FMs. The second experiment examines the complementarity assumption by providing English training to a random sample of DMs working at MNCs. At endline, treated DMs have higher English proficiency, communicate more frequently with their FMs, are more involved in firm management, and perform better in simulated management tasks. Organizational barriers within MNCs can thus hinder knowledge transfers and lead to an under-investment in English relative to the social optimum.

We study the relationship between exporters' organizational structure and output quality. If only input quantity is observable, theory predicts that vertical integration may be necessary to incentivize suppliers to increase input quality. Using data on suppliers' behavior, supplier ownership, supply transactions, and manufacturers' output by quality grade and exports from the Peruvian fishmeal industry, we show the following. After integrating with the plant being supplied and losing access to alternative pay-per-kilo buyers, suppliers take more quality-increasing and less quantity-increasing actions. Integration consequently causally increases output quality, and manufacturers integrate suppliers when facing high relative demand for high quality grades.

Working Paper No. 23949 We study the relationship between exporters' organizational structure and output quality. If only input quantity is observable, theory predicts that vertical integration may be necessary to incentivize suppliers to increase input quality. Using data on suppliers' behavior, supplier ownership, supply transactions, and manufacturers' output by quality grade and exports from the Peruvian fishmeal industry, we show the following. After integrating with the plant being supplied and losing access to alternative pay-per-kilo buyers, suppliers take more quality-increasing and less quantity-increasing actions. Integration consequently causally increases output quality, and manufacturers integrate suppliers when facing high relative demand for high quality grades.