Search

How do mental models guide behaviour? Although a great deal of attention has been paid to mental models in psychology, relatively little is understood about how they influence incentivised decision-making or their relevance for policy. In this dissertation, I explore this question over the course of four papers. In the first paper, I synthesize the literature on mental models and aim to situate it in standard decision theory. In paper two, I look at how people use mental models of social groups to determine strategies of action under incentivized conditions. In paper three, I look at how unemployed young people's mental model of their agency within the world influences their decisions to search for work. In paper four, I look at how mental models of technology influence vaccine hesitancy.

The research compiled in this thesis aims to explore a few core questions related to the perceptions, management and measurement of climate risk within the context of smallholders in Kenya. It makes use of a unique data set comprising of nearly 200,000 adoptions of a pioneering weather index insurance that was bundled with a hybrid seed product and available over several seasons. The first empirical chapter econometrically explores the role of basis risk in the continued uptake of the bundled index insurance product. The findings conclude that localised experience of both positive and negative basis risk compound an uncertainty of the product and consequently stifle demand. This presents evidence to suggest that smallholders do indeed view these products risk mitigating mechanisms rather than investments under balanced reciprocity. The second empirical study, using the same dataset, explores the effect that 'experimentation' and 'learning', by repeat users of the bundled index insurance product, has on another adaptive agricultural behaviour (changes to sowing timings). This work provides suggestive evidence of 'learning' that consequently leads to improved sowing decisions and lower insurance pay-outs for repeated users. This offers some valuable policy insights for the structuring of insurance premiums within these products. The final research project used a sample of the same database of farmers to conduct a large scale field experiment which sought to collect risk preference data in a novel way (via mobile phone). This data is analysed within a pioneering Bayesian modelling framework to track how risk preferences update given changing environmental contexts. The study identifies a notable reduction in risk aversion post-harvest as well as evidence of increased levels of localised risk aversion in more challenging climates and following weather shocks. This work supports an evolutionary model of risk preferences and may also offer applied insights for development practitioners pursuing the effective delivery of risk mitigation mechanisms.

This dissertation consists of three empirical essays on firms, productivity and innovation in low-income countries. Chapter 1 examines whether changing management practices to encourage bottom-up innovation leads to beneficial information flows from production workers on lower hierarchical levels to higher-level management with the power to implement ideas. In a three-month field experiment with 1,600 workers in a Bangladeshi garment factory, I find that providing structured peer-to-peer lunches for workers to discuss ideas significantly improves the quantity and relevance of ideas which are implementable in the firm, driven by changes in workers' ease of submissions and their aspirations. Additional encouragement from trained supervisors, however, comes at the cost of suppressing voice on critical topics about supervisors, which works through a perceived deteriorated voice climate. Chapter 2 (with Christopher Woodruff, Atonu Rabbani and Rocco Macchiavello) explores the impact of a low-cost group-based management training programme on productivity and working conditions in 25 Bangladeshi garment factories, with significantly positive effects on workers. Practices aimed at improving communication on the production floor are found to be more effective than technical and production-related practices, and lead to positive effects on human-resource related outcomes. Chapter 3 (with Ferrán Vega and Christopher Woodruff) analyses detailed line-level data in a Bangladeshi garment factory over a period of five years to examine the role of low-level autonomy in the process of adaptation to a negative productivity shock, caused by switching to a 'faster fashion' production model imposed by industry demands. We find that performance differences can be traced back to varying degrees of autonomy exercised by production managers in deviating from pre-determined plans.

This thesis studies firm development and labour markets in South Asia from three different perspectives - gender, skills and infrastructure. Chapter 1 investigates how female career advancement in the workplace impacts on bargaining in the household. It exploits the design of a promotion programme for women in Bangladesh's garment industry, a sector where women continue to be underrepresented even in low-level management positions. The results suggest that women's promotions lead to higher bargaining power at home not just for the promoted women, but also for the female subordinates that they supervise. Chapter 2 - written with Christopher Woodruff - turns to analyse the effects of the same promotion programme on factory outcomes. It reports on the importance of selecting managers based on their initial characteristics as compared to training candidates in the same skills. The evidence suggests that formal diagnostics lead factories to choose candidates with higher initial skills, which are predictive of official promotions and performance. Training results in higher promotion rates but has only marginal effects on performance. Soft skills, such as confidence and leadership, are particularly relevant for the success of the new female managers. Lastly, Chapter 3 documents heterogeneous short-term effects of infrastructure improvements on firms in India. It finds large and significant negative impacts on incumbent manufacturing firms close to improved highways, which are driven by initially less profitable, younger and smaller firms. This appears to provide a snapshot of the process of reallocation of resources towards more successful firms when areas become better connected.

This thesis examines the effects of a large and ever-present development sector on the labour market of aid-dependent low-income countries, using Sierra Leone as a case study. The thread that connects the substantive chapters of this thesis is the positioning of the development sector as a third sector in the labour market. Alongside the public and private sectors, the development sector is also an employer and skills developer in host-countries. Embedding the development sector as a sector in itself, makes apparent a channel of aid transmission in the domestic economy, via the labour market - a channel that is often overlooked in both academic and policy discourses. This in turn brings to the debate the short-term and the long-term effects on the labour market of host-countries. The thesis is novel as it applies analytical methods that are primarily used in the economics discipline, to the types of questions which sit within critical development studies. The main findings are that the presence of a large development sector affects occupational choice of graduates and the skills available in the labour market. These effects interplay with: (i) cognitive ability of graduates – higher ability graduates are more likely to choose employment with a donor organisation, non-governmental organisation or international non-governmental organisation; (ii) patronage in the labour market – in low trust environments where patronage exist in the labour market the development sector emerges as a more transparent recruiter and thus attracts talent; and (iii) skills training programmes at vocational and university levels have been directly affected by donor policies and actions. This process of change in the labour market and skills composition is important in itself; but there are also potential implications for sustainable home-grown growth and continued aid dependence.

This thesis consists of a short introduction, an overview of the context and data used, three self-contained analytical chapters on migration and human capital in developing countries, and overall conclusions. The first paper documents detailed patterns of internal mobility and estimates the predictors of migration by gender and reasons for moving using a Linear Probability Model and a Multinomial Logit Model, respectively. My main finding is that young migrants are a very heterogeneous segment of the population and that “favourable selection” only holds for those who move for studies. The second paper explores the impact of internal migration on cognitive and psychosocial skills by estimating a 2-Stage Least Squares (2SLS) model, using weather shocks as instruments for migration. My key finding is that migration affects both cognitive and psychosocial skills, but these effects differ across migrants depending on their reason for moving: those that move for studies have higher cognitive and psychosocial skills than non migrants do, whereas those that moved for family formation have lower cognitive skills than non-migrants do. Lastly, the third paper focuses on whether age at migration has an impact on cognitive and psychosocial skills. I use sibling pairs to estimate a household fixed effects-2SLS model, using weather shocks as instruments for migration. My main finding is that younger migrants perform better than older migrants, although this effect can be offset or even dominated by input responses as a result of migration. In conclusion, this thesis provides evidence of the heterogeneous character of young migrants, as well as supports the claims that migration matters for skills formation and that the earlier it takes place, the better.

This thesis investigates the role of the state in promoting economic diversification in countries that are over-reliant on the extraction of non-renewable resources. The objective of this thesis is particularly timely given the current context of commodity price bust, which has revealed the economic vulnerability of several commodity-dependent countries. Economic diversification has consequently been brought back as a policy priority in such countries. However, various resource-dependent countries have attempted to diversify their economies but have failed to do so throughout history. As such, this study addresses the question of what needs to be done to diversify such economies, by focusing on the role of industrial policy. The ambition is that this thesis and its theoretical foundations can enrich our understanding of the underlying causes of export diversification and structural change in resource-dependent countries. The first contribution of this thesis is to reframe theoretical debates related to natural resources and economic development by elaborating a taxonomy of diversification strategies and by reconceiving the way we measure extractive resource endowment through the elaboration of a multidimensional indicator. Three main diversification pathways are discussed in this thesis, namely financial diversification, vertical diversification, and horizontal diversification. This study also demonstrates the contextual determinants that influence the urgency to follow diversification strategies over others. The second part of this thesis has adopted the analytical account of specific historical cases as the main heuristic for disentangling dynamics of economic diversification in natural resource-dependent economies. It compares the experiences of Chile and Malaysia in diversifying their economies since the 1950s in order to contribute to existing discussions on viable policy interventions for economic development and diversification. This research also contributes to reframing existing narratives on the development of Chile and Malaysia by challenging several misconceptions associated with both countries, notably regarding the role of free market forces. This thesis finds that the role of industrial policy goes beyond facilitating diversification, as it also shapes the direction of such diversification. In both case studies, government interventions that successfully promoted new sectors not only fixed market failures but also shaped the accumulation of productive capabilities, both within and beyond commodity value chains. It is therefore found that neoclassical and path-dependent approaches to diversification are not suitable in explaining the process of structural change of resource-dependent economies, in contrast to more dynamic approaches to comparative advantage and capabilities accumulation.

This thesis analyses the role of the Indian Ocean in contributing to a distinctly integrated maritime region, and assesses some of the implications this has for regional trade and industrialisation in the era of high imperialism between 1870 to 1914. The thesis also presents a collection of original trade data and commodity price statistics from British, French, and Dutch colonial archives, which form the basis of analysis. First, this thesis looks at the relative factors driving market integration in the Indian Ocean. Results show that market integration was primarily driven by port city integration with other port cities across the Indian Ocean, rather than with or between cities of the colonial hinterland. It also finds that steamships connecting port cities overseas contributed more to regional integration compared to railroads connecting cities of the colonial interior over land, and that the size or political function of a port city was not as much of a factor driving regional integration as simply whether or not a city was accessible to the Indian Ocean. Second, this thesis looks at trade costs within the Indian Ocean compared to trade from the Indian Ocean to other maritime regions either within the global periphery, or from the Indian Ocean to the global core of Europe. Findings show that trade costs were low for short--distance routes within the Indian Ocean, as well as long--distance routes from the Indian Ocean to Europe. However trade costs were the highest for trade from the Indian Ocean to other regions of the global periphery such as the Eastern Mediterranean, thus establishing that the Indian Ocean was a distinctly integrated maritime region within the global periphery. However in relation to the global core of Europe, the Indian Ocean was also highly integrated as part of a larger process of globalisation taking place during the period. Third, considering previous analysis on regional market integration and trade costs within and between the Indian Ocean and other regions, this thesis looks at some of the implications that an integrated Indian Ocean market had on trade growth and the export of large-scale industrial manufactures from the primary entrepot cities of Bombay and Calcutta. Comparative analysis of the intensive and extensive margins of trade for both cities demonstrates that trade ties across the Indian Ocean served as a major destination for more highly differentiated, higher value manufactured goods from both cities. Additionally, even though the vast majority of exports from both cities went to Europe, this was primarily in staple commodities rather than in manufactured goods. Thus an integrated Indian Ocean market provided an important destination for emerging manufacturing industries at the region's primary entrepot cities.

This thesis addresses Indian household behaviour towards savings and explores the effects of economic growth on poverty and inequality. In the second chapter, using data from the Indian Household Consumer Expenditure survey, I show that vulnerable Indian households do not treat gold jewellery as a luxury good, but as a substitute to for savings. Households in rural locations, with irregular income, with illiterate heads and with casual or agricultural labour as the main source of income have higher expenditures on gold. This can be explained by the Permanent Income Hypothesis. Households mitigate future uncertainty by accumulating gold to smooth out future income shocks. I test two alternative hypotheses for high levels of spending on gold in India. I find limited support for an alternative hypothesis that high levels of gold expenditure simply reflect expenditures on dowries, a form of cultural preferences stemming from religion. I find no evidence of a second alternative hypothesis, namely that intra-household bargaining power, measured by the by wife's relative education level, influences gold purchases. Finally, using Engel curves, I examine whether these results are due to anomalies in the Indian data by examining expenditures on food, a normal good. In the third chapter, I deal with non-random sampling of the Indian Household Consumer Expenditure Survey, by adjusting the median annual per capita consumption. While comparing absolute and relative methodologies for calculating poverty rates in India, I show that the World Bank methodology produces higher absolute poverty rates than the relative 60% of the median annual consumption methodology; spatial analysis used for robustness, confirms that neighbouring states present very different poverty rates, emphasising the effects of local legal and welfare systems in every Indian state. I then use a logit maximum likelihood model to show how 5-Year Indian GDP growth, preceding the 2009-2010 Household Consumer Expenditure Survey used for this chapter, has had a lagged effect on enhancing the chances of poverty in Indian households. This is more prominent in the case where the relative methodology is used for estimating poverty. Furthermore, I show how households in Indian states that have experienced Services-led GDP growth in particular, are more prone to be poor, both in the case of 5-year and 3-year growth, where the absolute methodology is used to draw poverty lines. Where the relative methodology is used, shorter-term Services-led growth seems to benefit Urban households only, in relation to poverty. In the fourth chapter I use the economic growth observed in India in the 2000s, to assess its impact on inequality amongst urban households. I also deal with non-random sampling in Round 66 of the Household Consumer Expenditure surveys, by assigning probability weights. I find that inequality has decreased overall from 1994-95 to 2009-10, while the difference in inequality between the bottom and top of the expenditure distribution amongst urban population has increased. Using intra-decile analysis, I find that inequality within the bottom 10% of the urban population has decreased, while at the same time inequality within the lower middle class has increased. I test the effect of economic growth on inequality, within the Marginal Productivity Theory model; I find no support at State level. I find that States with high percentage of casual labour, as part of their urban population, are more likely to experience in increase in inequality, while there is also some support around growth in the Services sector to have a positive effect on increasing inequality, within the 5-year time frame. Finally, I find no support for Social Security and Welfare government spending to have an impact on inequality, at State level.

Throughout the period of neoliberal globalisation the ability of nation states to shape economic outcomes has been significantly diminished. Not only the number of policy tools at their disposal has been reduced, but the pressures of global integration have decreased the effectiveness of those remaining. This dissertation discusses how the state can increase the effectiveness of these remaining policy tools and promote more desirable economic outcomes by shaping corporate governance arrangements and by making use of state-owned enterprises (SOEs). Chapter 2 provides a brief analysis of the global economy since the end of World War II (WWII). It examines the major ideological underpinnings and economic features of two distinct periods – the embedded liberal and the neoliberal. In addition, this chapter also discusses the impact of neoliberal globalisation on national economic sovereignty. Chapter 3 critically examines the Anglo-Saxon model of corporate governance, which is considered as the ‘best practice’ by the neoliberal doctrine with regards to this matter. It is argued that the theories that legitimise this model are fundamentally flawed, and an alternative framework for analysing corporate governance is proposed. Chapter 4 examines the evolution of corporate governance arrangements in three advanced economies – the United States, Japan and Germany – since the end of WWII. In addition, this chapter also addresses the relationship between corporate governance arrangements, national competitiveness, and the effectiveness of policy tools. Chapter 5 analyses the ‘rise’ of SOEs in the early post-WWII period and their subsequent ‘fall’, from empirical and theoretical standpoints. Some important recent developments regarding SOEs are also discussed. Chapter 6 is devoted to examining the Singaporean development experience with special reference to SOEs. The peculiarities of Singaporean SOEs and the role played by them at different stages of Singapore’s economic development are analysed. In addition, different aspects concerning the relevance of SOEs in the contemporary Singaporean economy are assessed.

Credit is a vital source of investment financing in most economies. Debt-based contracts allow cash-constrained firms to borrow against future profits without ceding control of their firm, using a relatively simple fixed repayment structure. However, excessive debt can lead to more fragile firms, and a misallocation of investment. While the optimal capital structure for a typical firm is likely to contain a combination of both debt and equity, most tax systems systematically discriminate against equity financing by subsidising debt. In Chapter 1, I investigate the effect of a policy change in Belgium that mitigated the classical tax discrimination between debt and equity financing through allowing firms to reduce their taxable income by an amount based on the equity in their capital structure. I find that this policy led to significantly lower leverage ratios, with the effect being greater for larger firms, as would have been expected given the relative incentives for large firms created by the policy change. Analysis of heterogeneous effects reveals that the decrease in leverage was most significant for those firms who were most exposed to risk in the pre-policy period, which provides favourable evidence for the effect of such policies on reducing leverage. The predominance of debt-based financing is also evident in developing countries, which may benefit from a greater availability of equity-based financing products. In Chapter 2, I present results from work with two of the largest microfinance institutions in Pakistan to investigate the effects of 'microequity' contracts on microenterprises. I implement an artefactual field experiment with Pakistani microenterprise owners, designed using a simple model of investment choice. Results reveal that equity-financed microenterprise owners chose investment options with a greater expected profit than those under debt financing, with heterogeneity analysis suggesting a larger effect for more risk-averse individuals. In the second part of Chapter 2, I present results from a field survey that provides insights on the reasons for which most microfinance institutions do not actually offer microequity products, despite their potential benefits. Results indicate the practical implementation challenges related to costly state verification, adverse selection into income-sharing contracts and moral hazard caused by inappropriatelytailored profit-sharing ratios. Chapter 3 investigates the demand for a shared-ownership microfinance contract, using a field experiment with the same participants as Chapter 2. The contract is designed to help growth-oriented microenterprise owners who have reached their borrowing limit to expand their business with the purchase of a fixed asset. High take-up was observed for the shared-ownership product in general, with a strong preference for a more flexible repayment schedule, which was randomly offered to half of the treatment group. Analysis of heterogenous demand provides some evidence that the preference for flexible contracts is particularly strong among the most risk-averse microenterprise owners, who may benefit more from the implicit insurance inherent in shared-ownership contracts.

This thesis examines the role of the state in shaping economic outcomes in rapidly urbanising developing countries. Increasing the effectiveness of the state is a major challenge facing most developing countries today. Chapter 1 focusses on one important factor that constrains the state's ability to provide public goods to all citizens: citizens' physical remoteness from their administrators. Using rich data on village India, and a spatial regression discontinuity design, this chapter shows that greater distance to administration reduces a village's access to public goods and worsens economic outcomes. Villages that are more remote from their administrators have fewer roads, schools, health centers and less irrigation. In turn, their residents have lower incomes, fewer assets, less literacy, and are more likely to be employed in agriculture. At least for roads, these effects are not driven by the higher cost of construction in remote villages, but higher cost of monitoring road quality. These results suggest that reducing the distance between the state and its citizens, and changing incentives for bureaucrats, can help to mitigate the large spatial disparities in living standards observed within many developing countries. In addition to hosting administrative headquarters, cities are also large goods and labour markets, and have been a major driver of recent employment growth in de- veloping countries. Chapter 2 uses newly matched and geocoded data from India's Economic and Population Censuses to study how growing cities affect the economic structure in neighbouring villages. Instrumenting for employment growth in cities using a Bartik (shift-share) instrument, the chapter finds that 10% higher growth in manufacturing employment in the nearest city reduces the growth in manufacturing employment in the village by 1.8%, though it does not affect employment growth in other rural non-farm sectors. The effects are stronger for villages located close to larger towns and cities. Further, the effects are stronger for villages that are less suited for manufacturing activities, such as more remote villages and villages that lack paved roads. Finally, Chapter 3 examines the effect of property titles on housing investment deci- sions of low-income urban households, using data from Delhi's resettlement localities. It finds that more secure property titles are not associated with a higher probability of engaging in housing investment, with higher levels of housing investment, or indeed with measures of housing quality. The results are robust to using matching estimators. The null results can be explained by the by the limited differences in perceived evic- tion risk between titled and non-titled households, and by low reliance on the formal financial sector by all households. These results suggest that urban property titling programs in developing countries can be more successful if they also expand access to complementary inputs, such as the formal financial sector.

Mobile money services have rapidly expanded across East Africa, with 66% of people using an account actively. Understanding the impact of these services is therefore an important area of research, which the first two chapters of my thesis contribute to. Chapter One of my thesis uses survey data from Tanzania to understand how mobile money services affect the risk sharing capabilities of households. I look at the aftermath of rainfall shocks at the village level, and examine how consumption has been affected for users of mobile money services, non-users who reside in villages with other users, and households that live in villages without access to mobile money services. I find that after a village-based rainfall shock, only users of mobile money services are able to smooth their consumption. Non-users, regardless of whether they have access to mobile money or not, experience a 6% fall in their consumption. This raises questions of who benefits from the introduction of new financial services. Chapter Two looks at integrating mobile money services into a microfinance loan product and the impact this has on the businesses of women in Uganda. In a field experiment, I randomly assigned women applying for a microfinance loan to receive the loan on a mobile money account or as cash. I find that women who receive their loan on a mobile money account invest 11% more of the loan into their business and experience 15% higher profits. These impacts are greatest for women who experienced high pressure to share money with family and friends at baseline. This study suggests that the manner in which a loan is disbursed has important implications for whether women are able to control how the loan is used. The third chapter of my thesis looks at a different topic: role models and how they impact behaviour. I study the impact of being randomly assigned to see a movie featuring a potential role model, Queen of Katwe, or a placebo movie, on the exam results of secondary school students in Uganda. Seeing Queen of Katwe results in lower secondary school students scoring 0.11 standard deviations higher in their maths exam, mainly because they are less likely to fail maths, and upper secondary school students scoring 0.13 standard deviations higher in their finishing exams. Impacts are largest for female students. Further research is needed to uncover the mechanisms by which a role model can raise exam performance, with both changing beliefs and raising aspirations potential channels.

This document summarizes my research on the production and consumption behavior of households in emerging markets, notably in relation to infrastructure, which emerging markets have often lacked until recently. As academic, policy, and business attention turns to the opportunities, challenges, and features of emerging markets, I look to contribute to the academic debate as well as to policy and managerial thinking. There are two essays. The first essay focuses on production, and explores drivers of innovation adoption among farming households. I was motivated by watching the work of NGOs in rural India teaching agricultural techniques to farmers. Some farmers would adopt the techniques while others would not. What drives this heterogeneity in responses to innovation? Conversations with NGO workers revealed conflicting theories, and existing literature points to factors - often psychological traits or cultural attitudes - that would be difficult to implement for an NGO on the ground or as policy. Using the NGO's unique dataset of their past innovation disseminations, I show that infrastructure is a substantial factor behind innovation adoption that operates by freeing up slack time for households to innovate. The second essay focuses on consumption, and explores drivers of education demand among poor rural households. Demand for education is of unique interest in emerging markets because it can lift the household out of poverty through higher future earnings. What motivates a household to pursue education? Existing literature has focused on the economic returns - if future earnings are higher, more will be spent on education. This essay proposes an infrastructure driver for education demand in emerging markets - poor rural households pursue education when they are more frequently exposed to urban areas. I show that exposure to cities via highway construction increases household demand for education.

Between 2000 and 2015 the Millennium Development Goals were the focus of much global attention and activity. They were selected in light of astounding poverty, with over 1 billion people at the time living on less than $1 per day. In a sense the MDGs were morally undeniable. The focus of this study is MDG2, universal primary education. It sets out to establish how and why MDG2 came to be selected. Whilst its selection seems obvious, for years developing countries complained about the short-sightedness of prioritising primary over secondary and tertiary education (Klees 2008). A task force commissioned by the World Bank and UNESCO at the time showed that the Bank’s rate of return analysis on primary education was flawed. It argued that developing countries need highly educated people to be economic and social entrepreneurs, develop good governance, strong institutions and infrastructure. In this way MDG2’s selection is problematic. Using case study method, first the literature is examined. Three hypotheses are generated: one based on a rational synoptic theory, one on critical theory and one on world society theory. A range of data are used to establish findings and test hypotheses. The study then considers implications of the findings for theory and the policy process. The findings show that priorities promoting more equal opportunities, such as MDG2, were gradually preferred. Whereas priorities promoting more equal outcomes, such as elimination of trade barriers, were gradually excluded. The study finds no evidence that the General Assembly ever voted on the list of 8 MDGs. Rather, the MDGs were selected by elite policy actors, addressing multiple interests. The study considers the assertion that marginalization of the poor does not happen because people harbor ill will toward them, rather because “The poor have no friends among the global elite” (Pogge 2011, p. 62).

This dissertation comprises three essays on the allocation of labour and capital in a large developing country, Indonesia. In the first essay, I examine the impact of the 1998 East Asian recession on child schooling outcomes in Indonesia. Using panel data on 7-15 year-olds, I exploit the heterogeneous impact of the recession across urban communities as measured by the variation in rice price increases, under the assumption that communities where rice prices increased the most were those where real wages declined the most. I find that for the youngest children (aged 7-12 years) there is a large negative impact of higher rice prices on school attendance and no effect on labour market participation. For older children (aged 13-15 years), schooling enrolment does not respond to rice prices but labour market participation declines sharply in the worst-hit communities. I find no evidence of adverse long-term consequences on human capital formation. In the second essay, I test the hypothesis that there exists a significant earnings differential between similar workers in the formal and informal sectors. Using panel data on salaried and self-employed individuals, I find that after controlling for firm size and individual-specific heterogeneity, there is no formal sector earnings premium, except in the public sector. The results are robust to the presence of unpaid family workers, measurement error, and non-random attrition in the survey. This questions the commonly held belief that labour markets in developing countries are segmented because of legal institutions that protect high formal sector earnings. In the third essay, I estimate the effect of a large exchange rate depreciation on the performance of importers. The ability to manage volatility in the cost of imported inputs is likely to depend on a firm's access to external sources of finance as well its ability to hedge against exchange rate movements. Using data from a census on Indonesian firms, I find that while domestic importers face lower value-added due to a rise in their costs of production, foreign-owned importers fare better: they are more likely to sustain higher value-added, hire more labour and use more materials than domestic owned firms. This suggests another channel through which FDI can add value to a firm in a developing country, particularly with the increasing importance of trade in intermediate goods.

The global financial and food crisis of 2007 and 2008 was followed by a surge in foreign interest and investment in African agriculture. Renewed global interest in African agriculture was also accompanied by an increase in international efforts to influence domestic agricultural policies, including in Ghana. In the context of an increasingly globalised food regime and integrated commodity markets, this thesis answers the question: to what extent do international actors impact domestic agricultural policies in Ghana? Policy 'impact' is understood as the marked influence that international actors have on policy goals and the resources, institutions, and knowledge used for achieving them. This thesis compares case studies of cocoa and rice policy over two different periods in Ghana's recent history (1983-1995 and 2003-2012) to understand how international actors use their power and resources to impact agricultural policies. The comparison of cocoa and rice policy is used to address two gaps in existing literature by examining how the impact of international actors relates to: 1) the political economy for a specific crop; and 2) the interaction between actors at international, national and local levels. Findings from the comparative analysis are then used to test existing theories for how international actors influence government policy in Africa more generally. In particular, findings provide new insights into how the impact of international actors on African agricultural policies is strongly associated with the effect of policy decisions on the longer-term political economy for a particular crop.

Garment sector trade unions have proved largely powerless to combat hypermobile transnational capital’s systematic extraction of surplus value from the newly industrialized Global South. Optimized conditions for accumulation coupled with the 2005 phase-out of the Multi-Fibre Agreement (MFA) have meant a radical geographic reconfiguration of the globalised garment industry heavily in favour of capital over labour. The thesis approaches the global garment sector from multiple vantage points across the world with the goal of uncovering the obstacles to workers' organisation, examine workers' strategies of resistance, and analyse the changing composition of labour and capital within the clothing commodity chain. The thesis highlights five distinct but interconnected case studies including a transnational workers campaign from a garment factory in Honduras; a history and present-day feasibility of establishing a transnational collective bargaining from El Salvador to Turkey to Cambodia; the prospects for a countermovement in the organizing strategies at the bottom of the clothing commodity and supply chain in Bangalore; the growth of a 'full package' denim manufacturer in changing the relationship between 'buyers' and 'suppliers' on the outskirts of Bangalore; and finally a continuation of this analysis the case of a strike at a monopoly footwear supplier in China. The central research question is: How do workers build power and establish workers' rights in the globally hypermobile garment sector? Ultimately, what is demonstrated within this thesis is that the actions of garment workers shaped and circumscribed the actions of capital in the sector, and as capital transformed new landscapes for accumulation new vistas for opposition begin to emerge.

Against the backdrop of the brain drain-brain gain debate, this thesis explores certain facets of the return migration phenomenon. Drawing on several theories, the decision to return among high skilled migrants is likely to be influenced by the prospect of using their overseas-acquired knowledge to secure a better livelihood back home. While ample consideration is given to motivations to return, the choice of employer and issues adjusting to the work and social surroundings, the main objective of the research is to understand migrants' transfer of overseas-acquired knowledge upon their return to India. Given the interdisciplinary nature of the topic, the scope of the thesis is focused on returnees working in two organizations in New Delhi—Fortis Escorts Hospital and Research Centre (EHIRC) and Tata Consultancy Services' Government Industry Solutions Unit (GISU). Adopting a mixed methods approach, survey data and case interviews are analyzed to address the core research question: “How and in which ways do returnees transfer their newly acquired knowledge, skills and experiences in employing organizations?” A key hypothesis is that returnees' social ties affect the extent and nature of knowledge transfers and thus confer intended benefits and may lead to unintended consequences for their organizations. The analyses pit McPherson's (2001) principle of homophily in social networks against Granovetter's (1973) weak ties hypothesis to grasp the role of returnees in knowledge transfers within EHIRC and GISU. Results drawn from data collected on returnees, non-migrants and transnationals strongly confirm that social ties—strong, intermediate or weak—affect the transfer of knowledge to stakeholders in their organizations. The contribution of this thesis to the existing body of research is to shed light on both the potential and limitations of returnees as a conduit for transferring knowledge, upgrading skills and relaying insights to non-migrants, teams or units in the workplace.

Informal sectors in developing countries provide a substantial pool of jobs for some of the world's poorest people. Self-employment comprises a large portion of the job opportunities available to individuals working in these sectors. This thesis is concerned with the factors that drive people to become self-employed and determine their welfare as an entrepreneur, with a special emphasis on differences between women and men. In Chapter 1, we explain the Ghanaian context to which this thesis relates and outline the contribution of each main chapter and the common themes. In Chapters 2 and 3, we examine the trade-off between domestic work, such as caring for children and household chores, and market work. In Chapter 2, we consider the extent to which individuals are able to substitute between these two tasks to adjust to short-run variation in domestic productivity brought about by outages in electricity. We find that self-employed workers adjust non-monotonically to changes in domestic productivity, initially increasing their levels of domestic work to preserve consumption levels, but then substituting towards market work when power outages become more severe. We show that this relationship is heterogeneous by sex, and build a model of time allocation to demonstrate the theoretical mechanisms behind these results. In Chapter 3 we examine whether the factors that drive occupational selection differ by sex. It is often argued that women choose jobs in self-employment because this allows them to balance income-generation with childcare and other domestic work. We test the plausibility of this claim and its implications for labour market outcomes. First, we use a simple model of occupational choice to clarify our ideas about which notions of 'job flexibility' are important for the Ghanaian context. Second, we examine whether differential selection forces between women and men may explain the raw sex earnings gaps that appear to persist in various sectors, using a multinomial logit model to adjust for non-random occupational selection. We find that controlling for selection substantially widens the earnings gap amongst the self-employed, but shrinks it for the wage-employed. Third, we interrogate our selection equations and show that domestic obligations increase women's likelihood of entering low-input self-employment jobs more than men. We assess the importance of endogeneity using a maximum simulated likelihood estimator to couch the idea that selection on observables can be used as a guide for selection on unobservables, focussing on the discrete choice made over occupation. In Chapter 4, we turn to theory to try and resolve some of the empirical puzzles that remain from Chapter 3. In particular, we attempt to reconcile the fact that female participation in self-employment is so high even when the average differences in potential earnings are large. To do this, we construct a search model, which allows for individual heterogeneity and participation in both self- and wage-employment, as well as discrimination against female workers in the wage sector. We numerically solve and simulate this model, using calibrations from the existing literature, to explain a set of stylised facts generated from a longitudinal dataset of workers in urban Ghana. We show that wage sector discrimination leads to average earnings gaps in \emph{all} sectors of the economy, even if the underlying ability distribution is the same for both sexes. We also conduct a series of experiments to examine how women and men may be affected differently by government policy. Finally, in Chapter 5 we connect our main findings to policy and make some suggestions for future work.

This thesis is about imperfections in urban labour markets of three developing countries. I study how physical living conditions place constraints on labour force participation, and increase risks associated with unemployment. In Chapter One I test for the impact of high search costs on labour market outcomes of job seekers. I use a randomized trial of transport subsidies among youth living far away from the centre of the city in Addis Ababa, Ethiopia. Lowering transport costs increases the intensity of job search and leads to better employment outcomes. Weekly phone call data shows that treatment works to stop job search activity from declining over time. I show that the results are consistent with a dynamic model of job search with cash constraints and monetary search costs. Income from temporary work is used to smooth consumption and pay for the costs of search. I find that subsidies reduce participation in temporary work. Chapter Two looks at the links between poor housing conditions in slums and market labour supply. I test for the effect of free government housing in South Africa on households, using four waves of panel data and a natural experiment due to the allocation of new housing according to proximity from housing projects. I then use planned but cancelled projects to control for non-random selection of housing project sites. I find that government housing leads to large increases in household incomes from wage work, and increases in the labour supply of female household members. I argue that these results are due to reduced burdens of work in the home of improved housing, especially for women. In Chapter Three we look at how labour markets respond to large but temporary economic shocks caused by typhoons in the Philippines. We use quarterly aggregate, repeated-cross sectional and panel data to demonstrate robust evidence of downward wage flexibility. Lay-offs do not occur when storms hits, but hours per worker fall. We explain these results with a model of implicit contracts under which risk is shared between workers and firms through wage cuts, but workers are insured against lay-offs so that adjustments in labour demand occur through reductions in hours per worker. Our results are particularly strong for workers in long term contractual relationships in the private sector.

This is a dissertation about the organisational structure of the World Bank, the professional practice of evaluation, and the meaning of the concept of aid effectiveness in practice. In international development, evaluation is a professional activity that determines and then reports on the impacts of aid projects and programmes to the clients of such efforts and to the public. 'Aid effectiveness' is a concept that refers to a standard of how aid projects and organisations should operate and the results, such as economic growth and poverty alleviation, which these efforts should deliver in order to work. The concept of 'aid effectiveness' has also been used in the debate about international development as a system and its reform. Given that aid policymakers and academic researchers often use the data contained in development organisations evaluations to determine the extent to which aid projects and programmes are 'effective', it is critical to analyse what these evaluations measure and what influences their ratings and judgments. Based on a case study of the World Bank, the analysis is primarily qualitative and draws on both interviews with evaluation professionals in the World Bank and content analyses of the logical framework, indicators, and language in the World Bank's evaluations at the project- and country-level. Building on the previous theoretical work in post-structuralism that considers how international development organisations 'produce' their work through certain terms and processes (Escobar, 1995; Crush, 1995), I assess how the professional practice of evaluation in the Bank 'produces' the results of aid at the project- and country-level, specifically in the evaluation reports that it makes publically available. The World Banks data and evaluation reports are a window through which to understand the impact of aid, and several factors that influence this 'window' are assessed, including the institutional role of evaluation, the professional practice of evaluation, and the required evaluation processes within the World Bank. The study has important implications for practitioners of international development, academic researchers, and evaluation professionals who endeavour to improve the aid system and often rely on the results of the World Bank's evaluations to inform their understanding of the impact of particular development efforts. By reshaping the discussion from one which considers if aid 'works' to one about the data and the process of making a judgment about the success of aid projects and programmes, I articulate what the role of evaluation is in practice and what the World Bank's resulting evaluative data do and do not reflect about the World Bank's work. The relationship between the 'scales' of aid is also analysed by comparing and contrasting the evaluation processes at the project-level and the country-level. I challenge the notion of a 'micro-macro paradox' (Mosley, 1986) between the successful results of the World Banks projects and the economic development in its client countries by articulating the actual meaning of this data in context, the unseen institutional forces that shape this data, and the difficulty of asserting a linear relationship between the results of projects and programmes on different scales of aid.

In this thesis I study how networks are formed and I analyse the strategies that well-connected individuals adopt in public good games on a network. In chapter one I study an artefactual field experiment in rural India which tests whether farmers can create efficient networks in a repeated link formation game, and whether group categorisation increases the frequency of in-group links and reduces network efficiency. I find that the efficiency of the networks formed in the experiment is significantly lower than the efficiency which could be achieved under selfish, rational play. When information about group membership is disclosed, in-group links are chosen more frequently, while the efficiency of network structure is not significantly affected. Using a job-referral network experiment in an urban area of Ethiopia, I investigate in chapter two whether individuals create new links with the least connected players in the network. In a first treatment, competition for job-referrals makes it in the player's interest to link with the least connected partners. In this treatment, links to the least connected players are significantly more likely than links to better connected individuals. In a second treatment, connections only affect the welfare of the new partner. Choosing the least connected player minimises inequality and maximises aggregate efficiency. This may motivate other-regarding players. In this treatment, however, links to least connected partners are not significantly more likely than links to other players. In chapter three I explore the characteristics that individuals value in the people they approach for advice. Using cross-sectional data on cocoa farmers in Ghanaian villages and a matched lottery experiment, I find an association between the difference in the aversion to risk of two farmers and the probability that one farmer is interested in the advice of the other farmer. In chapter four I study a one-shot public good game in rural India between farmers connected by a star network. Contributions by the centre of the star have a larger impact on aggregate payoffs than contributions by the spoke players. I use the strategy method to study whether the centre of the star contributes more than the average of the spokes. In selected sessions, I disclose participants' expectations about the choices of the centre of star. I find that the centre player contributes just as much as the average of the spokes, and that he is influenced by the expectations that other players hold about his decisions.

This work proposes a new theory of authoritarian regimes: responsive authoritarianism. Most existing theories of autocracies take as their point of departure elite politics or the state’s repressive apparatus to explain the rise and fall of regimes. I argue that, for many states, regimes also have to consider the consent of the governed when designing policies. Specifically, when regime legitimacy is low but the central leadership maintains a long time horizon, autocratic regimes are predicted to become more responsive to the needs of citizens. This theory is tested against a number of aspects of the Chinese fiscal system dealing with public goods provision during the period of 2002-2011 and generally finds in favor of the theory. Chapter 4 tests the fiscal transfer system, Chapter 5 tests the fiscal expenditure data, and Chapter 6 tests data on the results of the transfer and expenditure data: actual public goods provision. This theory has a number of implications that suggest that scholars begin to rethink how they conceptualize power dynamics within an authoritarian regime, in particular paying closer attention to the relationship between the ruler(s) and the ruled. It suggests that, at least in the political science literature, power be returned to the people.

This thesis consists of four independent chapters each of which addresses the relationship between clientelism and social policy in relation to welfare state development from different perspectives. The overarching research question examines whether the adoption of such policies leads to de-clientelisation, and, if so, to what extent. The research extensively draws upon both cross-national data and that from Thailand between 2000-2012 during which populist welfare policies have gained significant influence on political development. Chapter 1 employs a global dataset of developing countries to offer a comparative perspective on the subject and shows that political parties generally trade-off between social policy and their engagement in clientelism. The latter three chapters take Thailand as a case study empirically investigate clientelist mechanisms at different geographic levels. Focusing on the household level, Chapter 2 evaluates the role of patron-client relations in determining access to the Thailand Village Fund based on the Socio-Economic Household Surveys. The provincial level is examined in Chapter 3 which studies economic and political determinants of two types of provincial-level distributive transfers: social policy spending and discretionary spending. Chapter 4 examines the clientelist mechanism at the national level through an assessment of the electoral linkage dynamics by measuring changes in personal votes. The findings show some degree of resilience of clientelist relations as they intervene with social policy allocation, particularly at local level. The global trend contrasts with the case of Thailand where, as in-depth analyses of the latter three chapters have shown, clientelist relations often persist and convert into a new form, for example the southern model of welfare regimes.

The mainstream investment climate literature often fails to take account the methods that firms in developing countries adopt in order to mitigate the effects of a poor investment climate. A better understanding of these informal “coping strategies” may add to the body of knowledge on what is important, and what isn’t, when it comes to prioritising investment climate reforms in developing and transitional economies. Original research from Vietnam, a country which is growing rapidly and in the midst of its transition from plan to market, shows how firms have been able to adapt their business operations given an unstable and partially reformed institutional enabling environment. By comparing the behaviour of manufacturing enterprises across a number of differing local jurisdictions, we are able to discern just how firm level coping strategies adapt. We find evidence that entrepreneurs are able to use a variety of informal institutional mechanisms to invest and operate in an inhospitable business environment where private property rights are not well protected and develop “second best” response mechanisms. These mechanisms include establishing formal and informal networks and linkages, seeking patronage and protection, and by sharing ownership with potential expropriators. We also find evidence that in the face of weak property rights protection, firms adopt approaches to reduce the costs to the original investors if third party expropriation is attempted and are less likely to reinvest retained earnings. Where they do invest, it is principally in dissolvable and/or movable assets, and adopting a higher discount rate or risk adjusted time value of money for capital investments. Similarly, we find evidence of linkages between measures of firm confidence in the local investment climate, and the extent to which firms are willing to employ outside salaried management. Thus, the thesis provides a contribution to the growing literature reviewing the development of formal and informal investment climate institutions in transitional and developing economies. The principle research finding, namely that the establishment and use of informal or second best institutional arrangements can offset some of the costs and risks associated with an otherwise weak and unstable business environment, has important implications for policymakers when it comes to the prioritization of investment climate reforms in developing countries.

This thesis is a collection of essays on the relevance of property right reforms on the wellbeing of poor households in developing countries; specifically titling programmes in urban Peru. The first essay assesses the effects of titling on housing investment. The availability of a unique dataset permits us to trace households' investment behaviour. This allows us to investigate if tenure security induces households to make sizeable investments, the evolution of this effect over time, and whether heterogeneous expectations about future tenure security matter for the estimated impact. Evidence reveals significant effects, especially on large investment; however, the response on this type of investment may take several years to become effective. The second essay contributes to the debate on the sustainability of property rights reforms by emphasising the importance of strong registration systems. Policymakers have focused on the process of granting titles but the conditions needed to maintain the formality of future plot transactions have been left unattended. The analysis exploits an exogenous variation in legislation to examine the impact of a change in the registration process on the registration rate of plot transactions. Evidence suggests a large negative effect, implying that a weak registration system could threaten the reform. The third essay explores the impact of titling on the risk preferences of slum dwellers. The analysis provides evidence that titled dwellers reported lower values of the risk aversion measure than their non-treated counterparts. Results also suggest that tenure security can influence slum dwellers' preference formation process. Evidence shows that beneficiaries who were exposed to tenure security during their youth report on average lower values of the risk aversion indicator than individuals titled at an older age.

This dissertation contributes to the long-standing debate between those who argue that the enclosure of the commons was as a precondition to foster economic growth and those who defend common property regimes can be efficient and sustainable. Exploiting historical evidence from 19th century and early 20th century Spain, this research shows that the persistence of the commons in some Spanish regions was not detrimental to economic development, at least relative to the institutional arrangements they were replaced with. On the contrary, during the early stages of modern economic growth, the communal regime not only did not limit agricultural productivity growth, but indeed constituted a crucial part of the functioning of the rural economics in a number of ways. On the one hand, these collective resources complemented rural incomes and, subsequently, sustained households' consumption capacity. The reduction in life expectancy and heights in the provinces where privatisation was more intense, as well as the negative effect on literacy levels, strongly supports that the privatisation of the commons deteriorated the living standards of a relatively large part of the population. On the other hand, the communal regime also significantly contributed to financing the municipal budget. Deprived from this important source of revenue, local councils became unable to adequately fund local public goods and ended up increasing local taxes. Lastly, the social networks developed around the use and management of these collective resources facilitated the diffusion of information and the building of mutual knowledge and trust, thus constituting a vital ingredient of the social glue that hold these rural communities together. All things considered, the persistence of the commons in some regions provided peasants with cooperation mechanisms different from the market and made the transition to modern economic growth more socially sustainable.

The first essay [“Twin Deficits and Free Lunches: Macroeconomic Outcomes In Anticipation of Foreign Aid”] concerns itself with situations in which private agents anticipate a future windfall (free lunch) that will help service the debt resulting from a present fiscal expansion (implemented via a temporary tax cut). Such expectations of a windfall can arise in the context of natural resource discoveries or, more interestingly, due to perceptions by agents in “too important to fail” countries that will be bailed out through higher foreign aid or debt relief. We employ an overlapping generations model featuring credit constraints to study the real effects of such free lunch expectations in a small open economy, drawing contrasts with the standard tax and money finance closure rules. The model is solved analytically and shows that anticipated aid is equivalent to current aid when agents have perfect foresight, so that a temporary tax cut is seen as permanent. Accordingly, agents raise their consumption and indebtedness (at the expense of future generations) by an amount that is an increasing function of their “impatience” (subjective rates of time preference plus probability of death). A worsening of the current account obtains (twin deficits) across a range of plausible closure rules, including those featuring money finance. The introduction of credit constrained households (we study the variant where myopic agents spend their current disposable incomes) does not alter the basic result in the case of full aid finance, but does matter for mixed tax-aid regimes, in more complex settings where agent expectations and donor promises on aid diverge, and when governments face borrowing constraints so that the timing of aid delivery matters. The second essay [“The Role of Domestic Debt in Economic Growth: An Empirical Investigation For Developing Economies”] focuses on the remaining source of government financing, i.e. domestic debt, and the role it can play in mobilizing private savings, facilitating credit intermediation in higher risk settings (i.e. serving a “collateral” function on bank balance sheets), developing financial markets and supporting economic growth in general. To investigate this question empirically, we set up a new domestic debt database covering about 100 developing economies, going back three decades to 1975; explore Granger causality links between domestic debt and key macroeconomic and institutional variables; and estimate the growth impact of domestic debt using panel regressions, allowing for non-linear effects. Domestic debt, as a share of GDP is found to exert a significant positive impact on economic growth, with potential channels including domestic savings mobilization, provision of risk-insurance on banks’ balance sheets; and greater institutional accountability of the state to its citizens. Although this result countervails more established arguments against domestic debt (i.e. that it leads to crowding out and banks to become lazy), there is some evidence that above a ratio of 35 percent of bank deposits, domestic debt does begin to undermine economic growth. The growth payoff also depends on debt quality, with higher payoffs observed for positive interest-rate bearing marketable debt issued to nonbank sectors. The third and final essay [“Why Do Banks in Developing Economies Hold Domestic Government Securities?”] explores demand-side determinants of domestic debt, by focusing on commercial bank holdings of government paper, discriminating carefully between voluntary factors (such as mean-variance portfolio optimization) and statutory ones (cash reserve and capital adequacy requirements). The analysis is made possible by the construction of a dataset on government and private returns (real and nominal) for almost 600 banks from 70 emerging and low-income economies, spanning the (pre-Basel II) period 1995-2005. A battery of structural cross-section regressions indicates that banks’ portfolio decisions are at least as significantly influenced by mean-variance considerations as regulatory factors: the actual portfolio share of government securities (λ) responds intuitively, and sizably, to variations in the moments of the distributions for government and private returns as well as in the minimum-variance portfolio share (λ*). Higher cash reserve requirements tilt portfolios away from government securities toward riskier private lending, while higher capital adequacy requirements work the other way. The association between actual portfolios and the identified determinants is noticeably weaker at lower ends of the λ distribution, suggesting the domination of non-CAPM factors in those contexts.

The use of group lending for poverty alleviation is a widespread feature of modern microfinance. The structure of joint-liability credit - if one member defaults the others are held financially responsible - produces a natural tension between a borrower's social and economic interests. This study integrates theory from economics, sociology, and behavioural experiments to address the question, "How do social and economic mechanisms interact to shape a microcredit group's financial behaviour?" The empirical analysis involves an original dataset from a microfinance institution in Sierra Leone. The total dataset includes 7,025 joint-liability borrowers involved in 47,931 repayment transactions from 2005 to 2011. The empirical methods used are diverse: ethnographic fieldwork, GPS spatial analysis, social affiliation survey design, and multilevel statistical analysis of loan performance data. The original work is structured as three distinct papers. In the first paper, I examine social collateral, the formal use of a borrower's relationships as security against loan default. How does a group's spatial structure affect the efficacy of social collateral? Spatial concentration improves a group's economic performance up to a certain level after which the effect reverses and performance declines. The relationship is driven by a social trade-off between ability and willingness to enforce the loan. Further, groups that consist of multiple spatial fragments produce worse performance. Spatially fragmented groups are prone to splitting into social factions. In the second paper, I question what drives the self-selection process of microcredit group formation. The results show that group leaders prefer members with pre-existing social ties, who are spatially proximate, and have matching business types. The preference for socio-spatial factors is likely motivated by reducing the risk of strategic default by group members. In the third paper, I explore how economic cooperation in small groups evolves over years of repeated interaction. Despite the selective retention of better performing groups, average cooperation rates consistently decline, in terms of contribution and effort. Further, variance across groups continues to increase over 30 months of repeated interaction, suggesting that convergence to a stable cooperation rate has not occurred. Given that group lending exhibits many of the factors found to promote cooperation in laboratory experiments, it is surprising to find such a marked decline in this field setting. Overall, this thesis contributes to economic sociology by dissecting the difficult trade-offs between social and economic motives in group lending and offers policy implications for microfinance institutions regarding group formation heuristics, contract design, and loan management.

This thesis consists of a short introduction and three self-contained analytical chapters. Chapter 1 focuses on the question of learning gaps and divergence in achievement across countries. I use unique child-level panel data from Ethiopia, India, Peru and Vietnam to ask at what ages do gaps between different populations emerge, how they increase or decline over time, and what the proximate determinants of this divergence are. I document that learning gaps between the four countries are already evident at the age of 5 years and grow throughout the age trajectory of children, preserving country ranks from 5 to 15 years of age. At primary school age, the divergence between Vietnam and the other countries is largely accounted for by substantially greater learning gains per year of schooling. Chapter 2 focuses on learning differences between private and government school students in India. I present the first value-added models of learning production in private and government schools in this context, using panel data from Andhra Pradesh. I examine the heterogeneity in private school value-added across different subjects, urban and rural areas, medium of instruction, and across age groups. Further, I also estimate private school effects on children's self-efficacy and agency. I find modest or insignificant causal effects of attending private schools in most test domains other than English and on children's academic self-concept and agency. Results on comparable test domains and age groups correspond closely with, and further extend, estimates from a parallel experimental evaluation. Chapter 3 uses panel data from the state of Andhra Pradesh in India to estimate the impact of the introduction of a national midday meal program on anthropometric z-scores of primary school students, and investigates whether the program ameliorated the deterioration of health in young children caused by a severe drought. Correcting for self-selection into the program using a non-linearity in how age affects the probability of enrollment, we find that the program acted as a safety net for children, providing large and significant health gains for children whose families suffered from drought.

This thesis consists of four separate chapters which address different aspects of inequality and income determination. The first three chapters are country-level studies which examine (1) how incomes are shaped by spatial price differences, (2) the factor income composition, and (3) enterprise size. The final chapter analyses how income inequality changed at the global level. The first chapter investigates the implications of regional price differences for earnings differentials and inequality in Germany. I combine a district-level price index with administrative earnings data from social security records. Prices have a strong equalising effect on district average wages in West Germany, but a weaker effect in East Germany and at the national level. The change in overall inequality as a result of regional price differences is small (although significant in many cases), because inequality is mostly explained by differences within rather than between districts. The second chapter is motivated by the rapid increase in top income shares in the United States since the 1980s. Using data derived from tax filings, I show that this pattern is very similar after controlling for changes in tax unit size. Over the same period as top income shares increased, the composition of these incomes changed dramatically, with the labour share rising. Using a non-parametric copula framework, I show that incomes from labour and capital have become more closely associated at the top. This association is asymmetric such that top wage earners are more likely to also receive high capital incomes, compared with top capital income recipients receiving high wages. In the third chapter, I investigate the positive cross-sectional relationship between enterprise size and earnings using panel data from Ghana. I find evidence for a significant firm size effect in matched firm-worker data and a labour force panel, even after controlling for individual fixed effects. The size effect in self-employment is stronger in the cross-section, but it is driven by individual time-invariant characteristics. The final chapter studies the global interpersonal income distribution using a newly constructed and improved database of national household surveys between 1988 and 2008. The chapter finds that the global Gini remains high and approximately unchanged at around 0.7. However, this hides a substantial change in the global distribution from a twin-peaked distribution in 1988 into a single-peaked one now. Furthermore, the regional composition of the global distribution changed, as China graduated from the bottom ranks. As a result of the growth in Asia, the poorest quantiles of the global distribution are now largely from Sub-Saharan Africa. By exploiting the panel dimension of the dataset, the analysis shows which decile-groups within countries have benefitted most over this 20-year period. In addition, the chapter presents a preliminary assessment of how estimates of global inequality are affected by the likely underreporting of top incomes in surveys.

This thesis presents three core chapters examining different aspects of the relationship between natural resources and economic development. While addressing different questions they share several features in common: a concern with causal inference; overcoming the challenges of endogeneity between resource abundance and other characteristics of developing countries; and the use of new and novel datasets with spatially identified units of analysis. The work contributes to a rich and growing empirical literature seeking to deepen our understanding of the underlying mechanisms affecting the fortunes of resource-abundant countries. In the introductory chapter I discuss the extensive literature on this topic and in particular focus on the new generation of well-identified within-country studies, seeking to understand the empirical relationship between resources and economic development. Countries typically welcome the news of a resource discovery with joy and indeed, resource discoveries hold great economic potential. But what determines whether a country is resource rich or not? Is it more than just a chance finding, or good geology? In Chapter 2, entitled Institutions and the Location of Oil Exploration I present an investigation into this question. I examine the relationship between governance and choices of where to drill for oil. This work utilises a new dataset on exploration wells and looks at the distribution of drilling close to national borders. This allows me to identify estimates for the effect of differences in governance between neighbours. Two times out of three, investors choose to drill on the side of borders that are better governed, all other things being equal. This suggests that resource-wealth itself may be contingent on factors beyond geology, and indeed may be endogenous to the process of development. In Chapter 3, entitled The Local Effects of Resource Extraction, I turn my attention to the local economic consequences of industrial mining in Indonesia. I present a simple three-sector general equilibrium model to generate predictions for the local labour market, akin to the Corden-Neary Dutch disease model of the macroeconomy. I test the predicted effects in response to an exogenous resource sector shock by looking at mine opening or mine expansion events across three hundred mines. I test the predictions of the model, first by estimating the economic footprint from industrial mining; found to be an average of fifteen kilometre radius. I then examine the response of reported labour market activity from households surveyed in nearby communities. Here I find no evidence for a shift of local labour into the mining sector. I do find however a notable movement of labour from the traded sectors (agriculture and manufacturing) to the non-traded service sector, with a strong effect for foreign-owned mines versus domestic ones. Chapter 4, entitled Disentangling the Effects of Resource Extraction: Local Government and Investment Multipliers, examines the oil and gas boom in Indonesia from 1999-2009. Here I deploy a variety of identification strategies to attempt to disentangle the regional effects of the boom, measured in terms of district GDP. I estimate effects arising from transfers of revenue to local government. Using an instrumental variable approach I isolate the fiscal channel from resource projects. I find a positive and significant effect of increased local government revenues on district GDP over the boom decade. I then examine the spillovers from resource projects, isolating them from fiscal transfers. For districts neighbouring resource rich districts I find evidence for a modest positive effect arising from project investments, rather than fiscal transfers. In Chapter 5 I present concluding thoughts and discuss a future research agenda. I also summarise the burgeoning landscape of resource data available for within country and spatially identified studies and offer some thoughts on how this might evolve.

This dissertation attempts to integrate the household and children more fluidly into measures of well-being in the past. In part one, I develop a Monte Carlo simulation to test some of the assumptions of Allen’s welfare ratio methodology. These included his assumptions that family size was constant over time, that there were no female-headed households and that women and children did not participate in the labour force. After all of the adjustments, it appears that Allen’s welfare ratios underestimate the welfare ratios of a demographically representative group of families, especially if women and children’s labour force participation is included. However, the predicted distributions also highlight the struggles of agricultural labourers, who are given separate consideration. Even the average agricultural labourers’ family with women and children working would have had to rely of self- provisioning, gleaning, poor relief or the extension of the working year to make ends meet at the poorest point in their family life cycle. Part two adjusts Floud et al.’s estimates of calorie availability in the English economy from 1700 to 1909 for the costs of digestion, pregnancy and lactation. Taken together, these three additional costs reduced the amount calories available by around 15 per cent in 1700 but only by 5 per cent in 1909 because of the changing composition of the English diet. Part three presents a new adaptive framework for studying changes in children’s growth patterns over time and a new methodology, longitudinal growth studies, for measuring gender disparities in health in the past. An adaptive framework for understanding growth provides a more parsimonious explanation for the vast catch-up growth achieved by slave children in the antebellum American South. The slave children were only able to achieve this catch-up growth because they were programmed for a tall height trajectory by relatively good conditions in utero. Finally, impoverished girls experienced greater catch-up growth than boys in two schools in late-nineteenth century Boston, USA and early-twentieth century London, suggesting that girls were deprived relative to boys before entering these institutions.

Some 5 million people were classified as internally displaced in Colombia at the end of 2012, which represented about 10 per cent of the population and the highest number in the world at the time. Colombia differs from other countries with high levels of displacement in that it is comparatively politically stable, has effective national institutions, a relatively strong formal economy, and can by no means be described as a ‘failed’ or ‘failing’ state. The displacement literature tends to characterise the phenomenon as a humanitarian crisis and a side effect of the long-running civil war. But Colombians continue to be displaced in very large numbers despite the formal demobilization of the paramilitaries in 2006 and the diminished military capacity and engagement of the guerrillas since about the same period: the same groups that are widely regarded as being the main perpetrators of displacement. This thesis contends that displacement of the civilian population in Colombia is frequently not a consequence of violence, but rather the primary objective, where violence plays a facilitatory role. Moreover, the thesis asserts that these massive levels of displacement are substantively linked to predominantly economically-motivated logics and are regionally specific. By examining an agricultural commodity that has significantly expanded relatively recently in Colombia - African palm oil - this research examines if and how expanded cultivation may be linked to displacement. Using a political economy framework of analysis combined with empirical fieldwork, it explores the ‘localised displacement logics’ whereby land is coercively acquired by powerful local groups. The thesis concludes that the abandonment and dispossession of land from poor and marginalised groups constitutes part of an ongoing process of capitalist expansion and statebuilding in Colombia. Contrary to assertions that it is the intra-state conflict that constitutes the central obstacle to development, Colombia’s current trajectory of capitalist development may actually be a central obstacle to sustainable peace and not lead to an end to displacement.

Powerful individuals often influence the delivery of government services for their own purposes. Officials may prefer inherently to direct assistance to their own relatives and social contacts (nepotistic preferences). Alternatively, they may use government services strategically in exchange for favors (patronage) or to gain voter support (clientelism). Most existing literature examines these three phenomena separately, or does not distinguish the motivations for politicians’ influence on distribution. Causal identification has also been a problem in the empirical literature. In the first chapter of this thesis, I develop a theoretical model of interaction between three levels of actors: politicians, local patrons, and households. The model allows for politicians and patrons to influence government services for nepotistic, clientelistic, and patronage purposes. In chapters 2-4, I test the predictions of the model using two novel household survey datasets I collected along with my collaborators in rural Punjab, Pakistan. Chapter 2 tests the theoretical predictions for the interaction of politicians, patrons and voters. Chapters 3 and 4 provide quasi-experimental evidence on the causal effect of links with politicians on assistance. I find evidence that politicians exert dramatic influence on the targeting of government assistance in this setting. Consistent with the theoretical model, the most assistance goes to a small “inner circle” of their closest contacts. Politicians assist this “inner circle” based on their inherent preferences, regardless of electoral pressure. When politicians face electoral pressure, they also deliver assistance to a wider group, in particular members of the same clan. In contrast, local patrons do not appear to have significant independent influence over the targeting of the government assistance programs I study, but they do provide other types of assistance to households. Their behavior is more consistent with the idea that they are motivated by inherent preferences for assisting their contacts. The results have implications for the interpretation of empirical literature on nepotism, clientelism, and patronage. They can also inform the policies of donor agencies and civil society organizations who aim to engage or pressure governments to reduce corruption and improve public spending.

I examine the importance of human capital for economic development in Thailand during the period of high economic growth and structural transformation (1985-2000), using labour force survey data. The three main chapters attempt to estimate the effects of education, as a measure of human capital, on three major outcomes in the Thai labour market, namely (i) earnings; (ii) sector of employment; and (iii) earnings inequality. I address the endogeneity problem of education using an education policy shift—the change in the compulsory schooling law—that produces exogenous variation in education. The three main chapters adopt distinct modelling frameworks. The details of each of the main chapters are as follows. The third chapter investigates how education increases earnings and the probability of being in the non-agricultural sector. As the education policy shift influences educational attainment in a discontinuous way, a regression discontinuity (RD) framework is adopted to identify the average returns to education and the effect of education on the sector of employment. It is important to emphasise that the RD technique constrains the effects of education on the two outcomes to be linear and to be applicable only to sub-populations. My results confirm significant effects of education on both earnings and the sectoral sorting process. In addition, there are heterogeneous effects of education by gender. The fourth chapter is an extension of the previous chapter. I allow the returns to education to be heterogeneous across education levels and sectors of employment, while attempting to estimate the returns for the entire population. I use a control function (CF) approach and a double selection correction to estimate the sectoral earnings process, while jointly accounting for the choice of education and the selection into sectors and paid employment. I find that the returns to education are non-linear and higher in the non-agricultural sector especially for medium and highly educated workers. This suggests that human capital plays a crucial role in facilitating a structural transformation towards the non-agricultural sector. In the final chapter, I study how the increased primary education completion rate affects earnings inequality. While there exists a burgeoning literature on the average returns to education, less attention has been devoted to estimating the effects of education on the distribution of earnings. I identify the effects of primary education completion on earnings at different points of the distribution, and thus earnings inequality, using a recently developed approach, called regression discontinuity distributional treatment effects. My results suggest that the increased primary education completion rate reduces earnings inequality as the returns to primary education are larger for the poor than the rich.

This thesis consists of two parts: Part I (Chapters 2 and 3) critically assesses a set of methodological tools that are widely used in the literature and that are applied to the empirical analysis in Part II (Chapters 4 and 5). Using a randomised experiment, the first chapter compares pen-and-paper interviewing (PAPI) with computer-assisted personal interviewing (CAPI). We observe a large error count in PAPI, which is likely to introduce sample bias. We examine the effect of PAPI consumption measurement error on poverty analysis and compare both applications in terms of interview length, costs and respondents’ perceptions. Next, we formalise an unproven source of ordinary least squares estimation bias in standard linear-in-means peer effects models. Deriving a formula for the magnitude of the bias, we discuss its underlying parameters. We show when the bias is aggravated in models adding cluster fixed effects and how it affects inference and interpretation of estimation results. We reveal that two-stage least squares (2SLS) estimation strategies eliminate the bias and provide illustrative simulations. The results may explain some counter-intuitive findings in the social interaction literature. We then use the linear-in-means model to estimate endogenous peer effects on the awareness of a community-based development programme of vulnerable groups in rural Tanzania. We denote the geographically nearest neighbours set as the relevant peer group in this context and employ a popular 2SLS estimation strategy on a unique spatial household dataset, collected using CAPI, to identify significant average and heterogeneous endogenous peer effects. The final chapter investigates social network effects in decentralised food aid (free food and food for work) allocation processes in Ethiopia, in the aftermath of a serious drought. We find that food aid is responsive to need, as well as being targeted at households with less access to informal support. However, we also find strong correlations with political connections, especially in the immediate aftermath of the drought.

The thesis consists of a short introduction and three self-contained analytical chapters on land policy in developing countries. Chapter 1 examines the agricultural investment choices of small-scale farmers in Ethiopia whose land will be expropriated to provide space for a large factory. I use data from a survey of households conducted before expropriation occurred, but after the policy was announced. I identify the anticipation effects of land expropriation using variation in whether households own plots located inside or outside the proposed project boundary. Households facing immediate expropriation hedge against future income risk by using less fertilizer on their plots, and and growing less risky crops. These households are more likely to grow sorghum (a safe crop) and less likely to grow wheat (a relatively riskier crop). Households also respond to the threat of expropriation by reducing long-term investments in soil quality. Using two-stage least squares I show that subjective beliefs about the likelihood of expropriation act as a channel through which the threat of expropriation affects investment decisions. The results are robust to a number of other specifications, including some that account for unobservable geographic variation in plot characteristics. Chapter 2 explores the consequences of land expropriation for small-scale farmers in Ethiopia. Expropriation of farmland is used by all levels of government in Ethiopia as a tool for providing new land for industrial investors, commercial agriculture and expanding cities. Farmers usually receive a cash payment in exchange for their land based on a fixed formula to establish the price of land. I evaluate the impact of such a policy on a group of small-scale farmers and assess the extent to which they make the transition to new livelihoods. On average, households lose 70% of their land and receive compensation payments that are about 5 times the value of their annual consumption expenditure. Using data collected before and after the intervention I examine the impact of expropriation and compensation on household consumption, productive assets, livestock holdings, savings and labour market participation. Households in the treatment group increase their consumption, start more businesses and participate more in non-farm activities than households that do not lose farmland. These households also reallocate livestock portfolios away from oxen and towards small ruminants and cattle, reflecting a shift away from growing crops. However, these shifts to new livelihoods are relatively small compared to the amount of compensation kept as savings: with the exception of a few households, most of the compensation payment is left in commercial banks earning a negative real return. Chapter 3 focuses on the recent increase in large-scale agricultural land deals across Africa and the nature of the contracts reached by governments and foreign investors. In recent years, multi-national firms and foreign governments have entered into long term contracts with host countries in which large tracts of land are purchased or leased for commercial agricultural production in exchange for promises of infrastructure development, job creation and rural infrastructure improvement. The profitability of these projects is uncertain, especially at a time of increased agricultural commodity price volatility in world markets. Based on stylized facts about land deals I present a theoretical model of land contracts reached by host governments and foreign investors that explains the policy tradeoff between investment timelines, revenue generation and uncertainty. When agricultural projects require fixed infrastructure investment and yield uncertain payoffs, firms benefit from being able to complete the fixed investment in stages. If firms can learn more about payoffs by holding off on investment, they effectively hold an option to abandon the project. The value of this option provides a channel by which uncertainty affects the terms of the land contract. When host governments determine the terms of the contract by setting an income tax, a royalty rate and an investment timeline, the value of this option will affect government's optimal policy choice. In particular, I find that if governments benefit a great deal from investment spillovers the optimal contract will be designed to encourage firms not to abandon a project. But, if governments benefit relatively little from investment spillovers, governments will choose contract parameters to extract the value of the firm's option to abandon the project. I end by examining the effect of increasing uncertainty on the government's optimal policy choice.

This thesis consists of five stand-alone chapters empirically evaluating questions relating to the life cycle of natural resource extraction. We use three different data sets to shed light on the local impacts of natural resource booms and busts. In chapter 2 to 4 we use the household budget survey of Kazakhstan to explore the impacts of the oil boom on the local population. In the second chapter, we explore the distributional effects of the oil boom and show that average household income increased and income inequality decreased. In the third chapter we study how the increase in average income was perceived by the local population and find that households' satisfaction with income decreased. In the fourth chapter we study how the boom affected households' expenditure and show that the likelihood that households pay tuition fees for tertiary education increased. In chapter 5, we explore the long-term impacts of a negative labour demand shock following the coal mine closures in the UK. To do this we construct a new data set containing the location of all active coal mines since 1981 and link it to the UK census. We find that the dramatic lay off of miners since 1981 was associated with a persistent reduction in female labour force participation in the affected districts. In chapter 6, we study the determinants of drilling costs and their impact on the real price of oil using a new global data set on the number of exploration wells drilled and costs of drilling. To do this, we propose a structural model of the upstream sector in the oil and gas industry. The model allows us to decompose the variation in the reduced form errors of the estimated VAR into three structural shocks, and estimate the dynamic responses of the variables in the system to these shocks. We confirm that the upstream sector of the oil and gas industry is subject to increasing costs. But we do not find that the real oil price is permanently affected by shocks to costs of drilling.

This thesis shows how the Process Analysis Method (PAM) can be applied to assess technologies used to mitigate arsenic from drinking water in rural India, using a set of sustainability indicators. Stakeholder perspectives, gathered from a fieldwork survey of 933 households in West Bengal in 2012, played a significant role in this assessment. This research found that the ‘Most Important’ issues as specified by the technology users are cost, trust, distance from their home to the clean water source (an indicator of convenience), and understanding the health effects of arsenic. It was also found that none of the ten technologies evaluated are economically viable, as many do not charge user-fees, which creates reliance upon donations to meet recurring costs. Utilisation of a technology is strongly related to sociocultural capital, but in many cases, features that contribute to sociocultural value, like regular testing of the treated water, are not included in the financial budget. It is suggested that increased awareness might change attitudes to arsenic-rich waste and its disposal protocols. This waste is often currently discarded in an uncontrolled manner in the local environment, giving rise to the possibility of point-source recontamination. All technologies proved to have difficulties in dealing with waste, except the Tipot and Dug wells which produce no waste. Of the methods considered, the BESU technology scored highest, but still only with 47-62% of the maximum scores achievable within each domain. This explains the widespread failure of mitigation projects across the region. The indicators and metrics show where improvements can be made. A model scheme based on these findings is outlined which could be applied with the objective of increasing utilisation and improving sustainability. It can be concluded that a product stewardship approach should be taken in regard to design, implementation and operation of the technologies, including the creation of a regulated toxic waste collection and disposal industry.

This project approaches ongoing debates over the impact of increased Chinese engagement in African countries through the lens of production and industrialisation. Emerging market FDI into Africa is growing rapidly, and an increasing proportion of this investment is into manufacturing and productive sectors. This trend is led by the commercial expansion of private Chinese manufacturing firms across the continent. The goal of this project is to examine the differentiated impacts on African industrialisation attempts of this phenomenon. It takes as its case study industrial zone development projects in Nigeria, namely, the two official economic and trade cooperation zones being developed as large-scale FDI projects by Chinese firms, with Chinese and Nigerian government support, in Lagos and Ogun states. Analytically, four dimensions of this process are identified for study: the home country context, the host country context, the zone structures and institutions, and the firms themselves. Special attention is paid to the interface between foreign actors and the particular political economy of Nigerian manufacturing, as well as the at times substantial gaps between policy and practice in terms of industrial planning, investment and production. The thesis argues that SEZ projects in general, including the Chinese ETCZs, are industrial policy tools that operate on particular assumptions regarding the organisation of global production. As such, they incentivise the insertion of export-oriented firms into established global networks supplying international markets. However, a closer examination of industrial policy in China, the production environment in Nigeria and the behaviour of internationalising firms reveals that these assumptions are not always accurate. Thus, the SEZ institution as it is currently conceived in Nigeria is ill-suited to lend support to the trend towards Chinese relocation of producer firms, as well as to the reality of Nigerian production—both of which are predicated on domestic and regional markets as the primary driver of African industrialisation and productive sector growth.