University of Warwick institutional repository: http://go.warwick.ac.uk/wrap A Thesis Submitted for the Degree of PhD at the University of Warwick http://go.warwick.ac.uk/wrap/72955 This thesis is made available online and is protected by original copyright. Please scroll down to view the document itself. Please refer to the repository record for this item for information to help you to cite it. Our policy information is available from the repository home page. University of Warwick Doctoral Thesis Buyer-Seller Relations, Prices and Development: A Structural Approach Exploring the Garment Sector in Bangladesh Supervisor: Author: Prof. Christopher Woodruff Julia Cajal Grossi and Prof. Gregory Crawford A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy in the Department of Economics June 2015 Declaration of Authorship I, Julia Cajal Grossi, declare that this thesis titled, `Buyer-Seller Relations, Prices and Development: A Structural Approach Exploring the Garment Sector in Bangladesh' and the work presented in it are my own. I confirm that: This work was done wholly or mainly while in candidature for a research degree at this University. Where any part of this thesis has previously been submitted for a degree or any other qualification at this University or any other institution, this has been clearly stated. Where I have consulted the published work of others, this is always clearly at- tributed. Where I have quoted from the work of others, the source is always given. With the exception of such quotations, this thesis is entirely my own work. I have acknowledged all main sources of help (Please, see AppendixK with details on specific contributions to this thesis). Where the thesis is based on work done by myself jointly with others, I have made clear exactly what was done by others and what I have contributed myself. Signed: Date: i UNIVERSITY OF WARWICK Abstract Department of Economics Doctor of Philosophy Buyer-Seller Relations, Prices and Development: A Structural Approach Exploring the Garment Sector in Bangladesh by Julia Cajal Grossi This thesis aims at understanding how manufacturers' heterogeneity affects the configuration of trading relations and prices in a dynamic environment. The institutional context I study is that of the Ready Made Garment sector in Bangladesh over the 2005 - 2012 period. The research here represents a contribution to that goal in four dimensions. First, accessing customs records we constructed a dataset containing buyer - seller trade interactions at a disaggregated level, including volumes and unit prices of the traded goods and, for a subsample, prices and quantities of the inputs required for manufacturing them. This feature allows us to go a step further than most studies based on matched importer - exporter data and opens a fruitful research agenda. Second, using this dataset I offer a first characterisation of the dynamics of the relations between manufacturers and large international buyers in matters of (i) duration of the relations, (ii) evolution of volumes, prices, orders and profitability over time, (iii) heterogeneity of the manufacturers and (iv) the probability of trading links arising. I find that relations with large buyers tend to be exclusive, that higher prices are associated with longer lasting relations, which tend to grow over time and fail whenever the manufacturer starts dealing with another large player. Importantly, I present a characterisation of suppliers heterogeneity novel in the literature and show evidence on two salient facts: the higher the heterogeneity across suppliers faced by a buyer, the more persistent its relations are and the higher the markup the buyer is willing to pay .Third, I develop a dynamic discrete choice game of linking and bargaining that realises those patterns in the data. I implement an algorithm that computes Markov Perfect Equilibria to discuss aspects of computation, convergence and multiple equilibria in the game and I scan a large parameter space to characterise the mechanisms that drive the dynamics in the industry. Fourth, I present the structural approach developed by Lee and Fong (2013) for estimating network formation games with endogenous bargaining and discuss three aspects in which its application is not immediate in my setting. These are related to (i) the availability of prices in our data, (ii) the difficulties in recovering conditional choice probabilities from the data, and (iii) the construction of the distance score. These difficulties lead to a pseudo Monte Carlo exercise that compares (sixteen) alternative estimation procedures. This preliminary study suggests that restricting the objective function to the observed states, using an auxiliary parametric assumption on the conditional choice probabilities in unobserved states and exploiting the data on prices could be fertile paths to explore towards adapting Lee and Fong's approach to estimate structurally the parameters of my game with the data we constructed. Acknowledgements I express my most heartfelt gratitude to my supervisors, Christopher Woodruff and Gregory Crawford. It has been my privilege to work with them and without their invaluable guidance this PhD would have not been achievable. I cannot thank them enough for their patience, inspiration and advice. I have been very fortunate to have two informal advisors whose support kept me going during challenging times: I would like to deeply thank Guillermo Noguera for his im- mense generosity, words of wisdom and untiring help during my last year in the PhD and Carlos Noton whose motivation and mentoring in early stages of the programme taught me by example. Words fail me to describe the magnitude of their contribution. I have learnt a lot from Rocco Macchiavello, Robert Akerlof, Bhaskar Dutta, Sascha Becker and Wiji Arulampalam. I really appreciate the time and effort they devoted sharing their knowledge and offering advice to make me a better researcher. I am also indebted to Pablo Becker and Andres Carvajal for encouraging me to take this challenge. It was life changing and all worthwhile. The Economics Department at Warwick has been a fantastic environment to embark myself on this project and I am grateful to the Head of the Department, Abhinay Muthoo, the Faculty and Support Staff for making of this an outsanding experience. I would also like to thank my colleagues and friends Nicola Pavanini, Federica Liberini, Ana Rita Madeira, Natalia Gonzalez, Luciana Nicollier, Giuila Lotti, Andreas Menzel, Zeyyad Mandalinci and Yasin Acar for their support, their companionship and the many enriching discussions we had. I am especially grateful to Daniel Pearce for believing unwaveringly in me and for walking with me this last part of the journey. We made it. Finally, words are short to express my gratitude to my family for teaching me everything I needed to know to get here. It is to them that I would like to dedicate this thesis. iii Contents Declaration of Authorshipi Abstract ii Acknowledgements iii Contents iv List of Figures viii List of Tablesx 1 Introduction1 2 Buyer-Seller Relations in the Ready Made Garment Sector in Bangladesh8 2.1 Introduction...................................8 2.2 The Data.................................... 10 2.3 A Simplified Framework............................ 12 2.4 Measuring Heterogeneity............................ 14 2.5 Heterogeneity and Market Outcomes..................... 18 2.5.1 Persistency in Buyer - Seller Relations................ 18 2.5.2 Profitability and Price-Cost Margins................. 21 2.6 Discussion towards a Model of Network Formation............. 22 3 A Dynamic Game of Linking and Bargaining 25 3.1 Introduction................................... 25 3.2 Intuition and timing.............................. 28 Network Formation:...................... 29 Bargaining on Prices:...................... 30 3.3 The Game.................................... 30 3.3.1 Per-Period Profits........................... 30 3.3.2 The Network Formation Stage.................... 31 3.3.3 The Pricing Problem.......................... 33 3.3.4 The Dynamic Specification...................... 35 3.3.5 Markov Perfect Equilibria....................... 37 iv Contents v 3.4 Computation of Equilibria........................... 40 3.4.1 The computation algorithm...................... 42 3.4.2 Realising Equilibria of a Small Game................. 44 Size of the Game and Computation Time........... 45 Non-Convergence........................ 46 Smoothness of Convergence................... 46 Non-uniqueness......................... 47 3.5 Parameters and Market Outcomes...................... 49 4 The Structural Approach 54 4.1 Introduction................................... 54 4.2 The Econometric Approach.......................... 55 4.3 The Algorithm................................. 58 4.4 Data Restrictions and Operational Assumptions.............. 61 4.5 Results from a Preliminary Monte Carlo Exercise.............. 63 5 Conclusions 67 A Sources of Data 73 A.1 The Main Source of the Data and its Structure............... 73 A.2 Additional Sources of Data.......................... 74 A.2.1 UN Comtrade Data.......................... 74 A.2.2 Board of Exporters........................... 74 A.2.3 BOND Licenses............................. 74 A.2.4 Euromonitor Data........................... 75 A.2.5 World Bank Survey.......................... 75 A.2.6 Series of Prices of Cotton....................... 75 A.2.7 Exchange Rates............................. 75 A.2.8 Members List