This dissertation consists of three essays that explore the interplay between economic shocks and local context. The first chapter studies the role of Korea's unique social fractionalization in explaining a 60% growth in agricultural productivity during the 1970s. Social fractionalization along dimensions like ethnicity or class can stunt economic development. This paper investigates how fractionalization affects a group's ability to respond to economic shocks by studying rural South Korea in the 1970s. Social groups in rural Korea were defined by one dominant characteristic: extended kin network identified by family name. Some villages displayed high homogeneity, with up to 90% of households sharing the same family name. This unique social context offers a reliable measure of social fractionalization that is otherwise difficult to measure. I combine this cross-sectional variation with the time variation in market access created by the construction of a new bridge, the Namhae bridge, in 1973. I find that homogeneous villages displayed higher agricultural investments and productivity growth than heterogeneous villages following the bridge construction. Homogeneous villages capitalized on the opportunities created by the bridge by providing complementary local public goods more effectively than heterogeneous villages. This paper highlights the critical role of social homogeneity in enhancing a community's ability to capitalize on new opportunities in the face of external shocks, such as improved market access. In the second chapter, I investigate the impact of expanding access to education on the quality of pre-existing schools, using India's Rashtriya Madhyamik Shiksha Abhiyan (RMSA) national secondary school construction program as a case study. I use a difference-in-differences strategy to assess the effects of new school construction on existing schools' enrollment, facilities, and the number of students taking and passing the nationally administered exam at grade ten and twelve. The analysis utilizes school-level data from the Secondary Education Management Information System (SEMIS) database, which tracks all secondary schools in India between 2009 and 2020. I find that while opening new schools increases the number of students taking and passing exams, it does not necessarily lead to improvements in the quality of education, as measured by the proportions of students taking and passing exams. This suggests that constructing new schools alone may not be enough to enhance educational quality through increased competition, and other factors or interventions may be necessary. This paper highlights the potential trade-off between expanding access to education and maintaining the quality of education provided by pre-existing schools. In the third chapter, I empirically test ``learning-by-exporting'' by examining the long-term performance of Indian IT firms after their exposure to the temporary export shock created by the Y2K bug. The Y2K bug was a computer software glitch that misinterpreted the year 2000 as 1900 due to the widespread use of two-digit date representations in computer systems at the time. As the new millennium approached, concerns grew that this programming oversight would lead to widespread system failures and potential global disruptions. This uncertainty prompted extensive efforts to identify and fix potential issues before 2000, resulting in a surge in outsourcing to Indian IT companies. However, by the year 2000, concerns proved to be largely overstated as few major problems materialized. This event represented an unexpected yet short-lived export boom for Indian IT firms. I use two exposure measures: IT firms with a history of exporting before 2000 and the Bartik instrument, which leverages regional variation in IT sector employment share. Overall, I find that exposure to the Y2K export shock had a positive and statistically significant effect on the long-term performance of Indian IT firms, supporting the concept of ``learning-by-exporting.'' Event study analysis shows that firms exposed to the Y2K shock experienced improvements in sales, export share, and total compensation paid to employees. However, the 2SLS results using the Bartik instrument reveal statistically insignificant findings, suggesting a potential weak instrument problem. This highlights the need for more accurate measures of Y2K shock exposure and further exploration of alternative estimation strategies.