This thesis is a collection of three independent essays that applies microeconometrics techniques to empirically study topics in development and labour economics. The first chapter uses evidence from a natural experiment in Bangladesh, where households were treated to different types of transfer, food grains and cash, at different periods in time, to test the effect of these transfers on household consumption behaviour. Using the fixed effect instrumental variable model, the estimation results show that though in-kind transfers did cause households to consume more grain than they would have chosen under equal-value cash transfers, the impact on calorie consumption and children health status is minimal. Households that received cash were able to reallocate their funds more effectively, and chose to spend their extra income on clothing and children's non-food consumption, while at the same time spending no more on vices. The second chapter investigates the dynamics of living standards in Thailand. Income and earnings processes are first modelled after the statistical Galton-Markov process before being extended to follow a more structural permanent earnings model. Empirical estimations of income and earnings persistence in Thailand employ both constructed pseudo-panel data from Thailand's Labour Force Surveys and the Townsend Thai panel data. Galton-Markov estimates found conditional persistence to be low in Thailand. However, quantile regression estimates find that persistence is low at the bottom of the distribution but high at the top, indicating a divergence in earnings as time passes. A study of the covariance structure of earnings finds that total variation in the earnings process is predominantly driven by moderately persistent transitory components following the AR(1) process. The third chapter attempts to empirically fit the power-law distribution and study the dynamics of inequality, especially at the upper end, of the income and consumption distribution in Thailand. We find that using the popular but incorrect method based on the linear regression approach will lead to researchers drawing a wrong conclusion. Regression estimates of the power-law exponent, a, provide strong evidence of power-law fit in Thailand. However, from the implementation of the superior Clauset et al. method, the evidence in support of the power-law fit is much weaker. Estimates of a for both income and consumption suggest that there is low inequality at the top in Thailand but further inspection finds that there is a high level of persistent between-group inequality between the top and bottom ends of the distribution. In addition, following Battistin et al. (2009), we find weak support for Gibrat's law of proportional random growth as the income-generating process in Thailand.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:719883 |
| Date | January 2015 |
| Creators | Sivakul, Aganitpol |
| Contributors | Crawford, Ian |
| Publisher | University of Oxford |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | https://ora.ox.ac.uk/objects/uuid:617fabeb-e47b-4194-bfab-a7601c0edce1 |
Page generated in 0.2539 seconds