Essays in international trade and development economics

Kurniawati, Sandra

Vorschau

PDF
Kurniawati_Dissertation.pdf - Veröffentlichte Version
Download (3MB)

URN:	urn:nbn:de:bsz:180-madoc-707164
Dokumenttyp:	Dissertation
Erscheinungsjahr:	2025
Ort der Veröffentlichung:	Mannheim
Hochschule:	Universität Mannheim
Gutachter:	Fadinger, Harald
Datum der mündl. Prüfung:	2025
Sprache der Veröffentlichung:	Englisch
Einrichtung:	Fakultät für Rechtswissenschaft und Volkswirtschaftslehre > Internat. Handel, Makroökonomik, Entwicklungsökonomik (Fadinger 2014-2024) Außerfakultäre Einrichtungen > GESS - CDSE (VWL)
Fachgebiet:	330 Wirtschaft
Fachklassifikation:	JEL: F10, F63, L52, O25, Q17, R10,
Freie Schlagwörter (Englisch):	commodity booms , manufacturing sectors , Dutch disease , dynamic spatial equilibrium , agglomeration economies , dynamic externalities , industrial policy , local content requirement (LCR) , Indonesia
Abstract:	This dissertation consists of three chapters that examine questions in the fields of international trade and development economics. Specifically, I study how resource demand shocks in the global market affect economic outcomes in resource-exporting countries, and how industrial policy affects firms' outcomes in developing economies. Chapter 1: The Impact of Resource Demand Shocks on Manufacturing: Evidence from Indonesia. Rising global demand for vegetable oils has driven the expansion of oil palm plantations and of the vegetable oil industry in Indonesia. I use the palm oil boom in Indonesia as a natural experiment to examine the impact of the expansion of resource-based industries on other manufacturing sectors. In particular, I examine whether competition for intermediate inputs plays a role in propagating the effects of the commodity boom. First, to identify local shocks, I exploit regional variation in suitability for oil palm cultivation at the district level, obtained from Gehrke and Kubitza (2021). Second, to investigate potential crowding-out effects in input markets, I construct a measure of input similarity using sector-level input-output data. Then, I examine how exposure to the boom affects local employment and manufacturing outcomes, such as sales and labor productivity. I find that highly exposed regions tend to have higher growth in total employment, which is mainly driven by higher growth in agricultural employment. While the export boom led to growth in the vegetable oil industry, it reduced sales and labor productivity in other manufacturing sectors. In terms of employment, there was no significant effect on total manufacturing employment, but the growth of non-production workers in highly exposed regions was slower than that in less exposed regions. I also find that the negative effects of the boom were particularly pronounced among industries that use inputs similar to those of the vegetable oil industry. The findings suggest that competition between the commodity and manufacturing sectors occurred not only in labor markets but also in intermediate input markets. Chapter 2: Quantifying the Effects of Commodity Booms on Regional and Sectoral Outcomes Based on the empirical evidence from Chapter 1, I develop a dynamic spatial equilibrium model that builds on Desmet et al. (2018) and Conte et al. (2021) to quantify the aggregate and welfare effects of the commodity boom. My main contribution is to revisit the 'Dutch disease' hypothesis using a multi-sector spatial model that incorporates both static and dynamic externalities in the manufacturing sectors. The model also features sectoral linkages, internal migration, and both internal and international trade. I calibrate the model to the Indonesian economy, taking 2000 as the baseline year, and simulate the response to the commodity boom that the country experienced in the 2000–2011 period. I then use the calibrated model to quantify the effects of the commodity boom by shutting down the commodity export shocks in the 2000s. First, I find that while the commodity boom increased GDP during the boom period, it potentially reduced welfare. Second, the effects of the commodity boom varied across sectors. By the end of the boom period, the GDP shares of the commodity and service sectors increased (by 7.86 percentage points (p.p.) and 2.40 p.p., respectively), while manufacturing contracted. In particular, the non-food manufacturing share declined by 7.67 p.p. Third, the impact of the commodity boom was heterogeneous across regions. By the end of the boom period, regions outside Java experienced a 10.62% increase in GDP, while Java—the more industrialized region—experienced a 4.61% decline. Lastly, I show that agglomeration economies and dynamic externalities (or learning-by-doing) play important roles in amplifying the impact of the commodity boom. Agglomeration economies strengthen the positive impact of the commodity boom on aggregate GDP during the boom period. In contrast, dynamic externalities in manufacturing amplify the negative effects on aggregate GDP and on the regional GDP of the more industrialized areas after the boom period. Chapter 3: Firm Responses to Industrial Policy: Evidence from Local Content Requirement (LCR) Policy Governments in both developing and developed countries increasingly use industrial policy to promote domestic production and technological upgrading. One of the policy instruments is the local content requirement (LCR), which mandates that targeted sectors or firms source a certain proportion of inputs from domestic suppliers. Using detailed manufacturing plant-level data from Indonesia, I study the effects of the LCR policy in the telecommunications sector. First, I investigate how the LCR policy affects firms' input composition and cost structure. Second, I estimate the impacts of the policy on firm-level outcomes, such as sales, employment, and value-added. Lastly, I examine whether firms in upstream sectors benefit from the policy through production linkages. To identify causal effects, I exploit cross-plant variation in exposure to the LCR policy in the telecommunications sector. I measure plant-level exposure based on the share of LCR-targeted products in each plant's output. I use pre-policy data from 2006 to mitigate endogeneity concerns, as firms may have adjusted their product mix in response to the LCR policy. Then, I use a Two-Way Fixed Effects (TWFE) model and a dynamic extension of the TWFE model to examine both static and dynamic effects. While the policy is intended to promote local sourcing, the results suggest they have no significant impact on firm sales. Firms tend to experience higher labor costs and a decline in labor productivity. There are short-term positive effects on employment among upstream firms, but these effects dissipate in the longer run.