Lights and GDP relationship - What does the computer tell us? (Revised and resubmitted in Empirical Economics ) [PDF] [Online Appendix] [Data & Code]
The relationship between nighttime lights and GDP varies from country to country. However, which factors drive variations in the lights-GDP relationship across countries remains unclear. This paper examines the significance of approximately 600 potential drivers of uncertainty in the relationship between night lights and GDP worldwide. I employ three novel modern statistical techniques to select variables within a high-dimensional context: LASSO, minimax concave penalty, and spike-and-slab regression. Institutional quality emerges as the most important factor in explaining the difference between luminosity data and GDP across countries.
Conference/seminar: Vietnam Symposium in International Business Conference (Vietnam), Asian and Australasian Society of Labour Economics (University of Tokyo, Japan).
Does energy consumption lead to economic growth over time? Evidence from semi-parametric estimation for a panel of Asia Pacific countries. (Under review) [PDF] (with Russell Smyth, Mita Bhattacharya, and Xibin Zhang) [Data & Code]
Energy consumption changes over time and across countries because of how countries use energy in their various stages of growth. This paper extends the empirical literature on estimating energy demand elasticities in two directions. First, we employ nighttime light data as an instrumental variable for energy consumption. Second, we consider a semi-parametric model to examine country-specific and year-specific energy elasticities, allowing the coefficients to vary with continuous and categorical variables. We apply these strategies to explore the effect of energy consumption on the gross domestic product in the Asia-Pacific region between 1992 and 2018. We find that while the elasticities of energy consumption are time-varying and heterogeneous across countries, their magnitude is generally between 0.1 and 0.2 and exhibits an inverted U shape. We also find that the coefficients of energy elasticities are downward biased without accounting for endogeneity.
Measurements of true income: Do PWT, UQICD and WDI behave similarly? [PDF]
Internationally comparable income is obtained using purchasing power parity (PPP) adjustments over space and time across countries. The method used to adjust can lead to differences in the PPP adjusted income. National accounts GDP in local currency can be converted to PPP adjusted income, and some authors in the poverty measurement literature have suggested using income obtained from household surveys. In their recent QJE paper, Pinkovskiy and Sala-i-Martin (2016a) (P&Sa) proposed a combination method whereby optimum weights are given to GDP and surveys’ based income. They used World Development Indicators (WDI)’s GDP income for their study. This thesis extends their work in three directions: First, the robustness of the findings is tested by using four alternative PPP adjusted GDP income to WDI, they are PWT 7.1, PWT 8.1, PWT 9.0 and UQICDv2.1.1. This also extends the recent NBER paper (P&Sb) which evaluated WDI income against PWT 7.1, PWT 8.0 and PWT 8.1. Second, the robustness of the findings is tested by running the models by income and geographical country groups. Third, the robustness of the findings is tested by using an alternative econometric estimation approach to obtain the optimum weights.
Conference/seminar: The Vietnam Economist Annual Meeting (VEAM 2016)
Temperature shocks and human development