Research Journal #6

random
Back
  1. Successfully read paper in 3 out of 7 days. This is not satisfactory, but already the largest number of days I have ever read paper in a week! It's much much more rewarding than I think it would be.
    • Wealth Inequality in the United States since 1913: This is the first inequality paper in ECON I ever read. The work is rather descriptive, aiming to provide a more accurate measure of wealth inequality in US. Measurement of individual wealth at a broader scale is in itself a challenging task to do, invovling the valuation of different types of assets. I learnt some useful descriptives which can be used to measure inequality.
    • Robert Fogel & Douglas North Nobel Prize contribution overview: It's really encouraging to see that there are scholars in economic history who are very successful. The type of questions they asked just grabbed my interest instantly. Robert Fogel, for example, challenged the established opinion at his time such that slavery was an outlandish method of production for the old capitalism system. He argued and showed that slavery was economically efficient. Its abandonment is due to political reasons, but not an economic one. It's amazing to see that economics allows scholar to penetrate into these type of big, historical questions. Hope that I can integrate some of these flavors into my work as well.

2. "Selection bias" as the core concept in econometrics. To my first time in this year, I finally understand something important in econometrics. In most types of the causal question we want to ask, the reason we cannot answer them using observational data is that there's selection bias. In an idealized randomized control trials, there's no selection bias. Control group and treatment group would largely resemble each other. Thus, the treatment is the only variant. The difference of outcome between two groups is the treatment effect. In observational data, two groups are always inherently different other than the treatment. For example, If we want to study whether smoking is bad for health, the smoker group could be systematlically different from the non-smoker group in terms of their educational background, family income etc. which can all affect one's health. Thus, how to overcome this selection bias is the key objective of econometrics methods.

© Zhiwei (Berry) Wang.RSS