What I have been reading

  1. Angus Deaton and Nancy Cartwright on understanding and misunderstanding RCTs. A nice summary of the RCT trend in development economics and its limitations.
  2. Book review by Margaret McMillan on the making of Africa.
  3. Redding et al’s summary of recent developments in quantitative spatial economics. This area seems methodologically developed, and the approaches to model the general equilibrium can be applied to wider contexts beyond international trade.
Advertisements

How to read structural modeling papers in economics?

For the purpose of a research project, I have been reading a lot of literature on locational equilibrium sorting in public economics. While the topic is fascinating, it is easy to get lost in piles of papers without understanding how they unify under the same overarching theme. Since reading structural papers is likely to be a challenge for many economics PhD students, I thought it might be useful to share my thoughts on how to do it effectively.

The purpose of reading others’ papers is not to produce a thorough summary of their work but to critically assess the status quo of the literature and what you can contribute. I have found the following steps useful for achieving this goal.

Step 1: Bird’s Eye View

Start with the review papers. If the topic is well researched, it should have a summary paper (look for Journal of Economic Perspectives/Journal of Economic Literature/handbook chapters). Focus on the fundamental assumptions made in each class of models. Make a list of papers for further reading based on the bibliography of the review paper.

Step 2: Divide and Conquer

For each paper on the “to read” list, read its bare bones and understand the key message. Clearly outline the model assumptions (and whether they are made as abstraction or due to data limitation), data availability, and empirical approach.

Step 3: Say it in Your Words

After you feel you have a good understanding of the class of structural models, try to synthesize the papers by describing them in writing. Focus on how they are linked to each other, and critically assess the pros and cons of each approach.

Step 4: Make the Link to Your Research

By the end of step 3, you should have a fairly clear understanding of which approach (if any) is best suited to your own research, and how your research contributes to the existing literature.

 

My Thoughts on Presenting Preliminary Research

A few weeks ago I presented a new research project to a few faculty members. I was looking for feedback on the appropriate structural model that can be used to explain commuting and residential choice patterns in a household survey. The following is what I learned from the presentation.

  1. Be clear about what feedback you are looking for. Say it right at the beginning. When your work is in preliminary stage, there are usually concerns from all aspects — data, identification, theoretical framework, context, etc. Try to focus on the aspect that matters the most to you for now. This will make your presentation more structured and enable your audience to give more useful feedback.
  2. Know what you have to cover in your presentation and what you can skip. For example, if your purpose is to identify the right theoretical model for your research question, do not dwell on the data for too long. Address questions that are relevant, and leave the questions that stray too far from your theme “to future discussions”.
  3. When you are presenting a model, be clear about the assumptions. Which assumptions are fundamental to the workings of your model and the interpretation of your results? Which assumptions are necessary due to data limitations? Which assumptions are an abstraction and can be refined? Thinking over these questions also helps you to understand different models better.
  4. Do not put unnecessary information on your presentation slides. Slides are a form of visual aid — they make your speech more effective instead of replacing you the speaker. If you find yourself staring at a slide with too many equations thinking “it’s probably gonna be fine, I’ll just use it as reference”, then you probably should make it more concise.
  5. Anticipate your questions as much as you can. I usually make draft slides a couple days before the actual presentation, and go over the slides from an outsider’s perspective (or whoever will be at your presentation, if you know them well). If a particular line seems confusing, I revise the wording on the slide or think of alternative ways to present the same idea.

Hope this is useful.

Weekly NBER Digest 3/6/16

This is the second post in my weekly NBER digest series.

  1. Bertrand and Duflo summarize the field experiments on discrimination.Dynamics of discrimination and ways to undermine discrimination seem to be promising future research areas.
  2. Dinkleman and Mariotti investigate how circular migration from Malawi to South Africa helps to improve the human capital in origin communities. Using spatial variation in migration costs and two policy instruments, a removal of migrant quota and a ban on migration, they find that after twenty years of the shocks, “human capital is 4.8%-6.9% higher among cohorts who were eligible for schooling in communities with the easiest access in migrant jobs.”
  3. A new working paper by Hummels, Munch, and Xiang reviews the existing literature on the labor market impacts of offshoring.

Weekly NBER Digest 2/20/16

This is the second post in my weekly NBER digest series.

1.What can we say about optimal trade policy using heterogeneous firms theory?

Costinot and coauthors use the classic Melitz (2003) model of heterogeneous firms trade theory to derive optimal tax levels at the micro (firm) level. They find that optimal import taxes discriminate against the most profitable foreign exporters, while optimal export taxes are uniform across domestic exporters.

Relative to another recent paper by Costiinot, Donaldson, Vogel and Werning (2015), the assumption of monopolistic competition (rather than perfect competition) in this paper leads to conclusions that are exactly the opposite. More generally, this paper is part of the trend in international trade research to connect traditional macro theories to micro data regularities.

2. How to evaluate the impact of international competition on firm performance?

This is not a new topic, but De Loecker and Van Biesebroeck highlights two aspects that are not well addressed in previous research. First, the impact of international trade on market power and productive efficiency should be studied in an integrated framework. Second, trade liberalization has the potential to increase competition by enlarging the relevant market, but this effect is not well understood. The discussion on the relevant market definition in the trade context is especially insightful.

Weekly NBER Digest 2/14/16

I decided to start a series of weekly blog posts on the new NBER working papers on development economics, labor economics, and international trade that I find interesting. In the past few months, I have experienced the excitement of finding an interesting research question, going through the empirical methodology to answer the question, cleaning data, and then figuring out there is not enough variation to answer my question (due to the contextual nature of the question). Now I am opening myself up to new ideas, and my NBER digests will serve this purpose as well.

1. How does taxation affect growth through corruption?

This working paper builds an endogenous growth model to examine the relationship between taxation, corruption, and economic growth. Taxes have disincentive effects on entrepreneurs, but also provide them with public infrastructure. Political corruption governs how efficient tax revenues are translated into infrastructure. The model predicts an inversed-U relationship between taxation and growth, which is consistent with data from the Longitudinal Business Database (LBD) at the US Census Bureau.

This paper is an example of combining macro modeling with micro empirical analysis to address an interesting question.

2. Are trade policies no longer important?

This working paper by Goldberg and Pavcnik describes the declining research interest in assessing the impact of trade policies and reasons for this decline, and suggests future areas of research. A lot of useful insights. As an example,

The variation in trade policy across cross-sectional units and time is only helpful for identifying the effects of trade policy in the presence of some type of friction and/or heterogeneity in exposure to policy change. … the main limitation of relying on differential exposure of economic agents to trade policy to identify its causal effects is … that this approach by its nature will generally reveal only the relative and not absolute effects of a policy change. The latter require a theoretical framework within which the relative effects can be interpreted.

My Two Cents on Randomized Controlled Trials

Randomized control trials (RCTs) have been at the forefront of development economic research in recent years. How well do these inform us of policy alternatives to reduce poverty?

On the bright side, RCTs allow us to identify the causal impact of policy interventions, and a lot of studies provide evidence that some simple nudging can make a big difference on behavior (see Esther Duflo’s work on encouraging Kenyan farmers to use fertilizers). However, there are also a few caveats in interpreting RCT results:

Publication Bias: only significant results — either positive or negative — get published. Are we learning about the truth or the truth we WANT to know? For instance, microfinance has been applauded as an innovative and effective way to increase savings and investments, encourage entrepreneurship, and reduce poverty. But a recent working paper has found zero effects of access to microfinance on long term development outcomes.

Pre-analysis plan vs. manual selection after study is initiated: Here is a philosophical discussion in the Journal of Economic Perspectives by Ben Olken.

Heterogeneous treatment effects: the magnitude of the effects of policy varies a great deal. External validity is often a concern. Here is a thought-provoking paper by Eva Vivalt, the founder of AidGrade, a database on impact evaluations.

Experimental arms race: Are we simply adding more technical details into the same experiments without shedding light on fundamental channels of how they change behavior? Here is an article by David McKenzie on the tpoic. More specifically, Rachel Glennerster writes about what this implies for RCTs involving governments.

Your thoughts and comments are welcome.