Surviving the transition from classes to research in your PhD

Readers of this blog might have noticed that I have not published any post since November last year. This is because I spent the past eight months struggling and surviving the transition from a student to a researcher of economics. During this transition, I have experienced many episodes of self doubt and breakdowns, and my confidence plunged so much that I did not think I would put down any meaningful thoughts. Fortunately, I was able to learn from my struggles and successfully defended my PhD prospectus a month ago (I’m a PhD candidate now!). In this post I hope to share with you what I have learned from my experience.

  • Once you have identified a research topic (not even a specific question), go talk to faculty members with relevant expertise immediately. This sounds daunting, but you should do it because 1) they will be able to point out strengths/weaknesses of your research ideas and save you time, and 2) it will teach you how to communicate with others about your ideas, which you will need to learn eventually.
  • Find at least one faculty member whom you can communicate with on a regular basis. Conducting innovative research is no small task, and you will need guidance at the beginning of this journey. Talking with faculty regularly can also help you feel that you are constantly making progress, which is more important than you think.
  • Time management is crucial to a successful transition from a student to a researcher. Without a class schedule, you can easily develop habits that reduce your productivity. What I have found to be most useful is to develop a routine: set a fixed amount of time where you go to the office and work on your research, but also leave time for social activities and rest. Make sure you have some time to rest everyday so you have something to look forward to when research is not going well (that happens a lot!).
  • Put yourself out there. Do not take negative feedback personally; accept it and improve yourself based on it. When I presented my research for the first time in my second year, I felt so bad about receiving “negative” feedback that I burst into tears in the middle of the presentation. I decided the only solution to this fear of presentation is to present more. Therefore, I presented two more times in the first semester of my third year, trying to do just a little bit better each time. Now I am proud to say that I can present my research ideas clearly and peacefully.
  • Sharpen your communication skills, in writing and in person. Good interpersonal communication skills allow you to make a stronger impression, so others are more likely to remember your research and give meaningful feedback on it. Good writing (in academic papers and in daily email correspondence) will make others understand your goals and help you achieve them. Personally, I always take clear, on-point emails and as a sign that the other person appreciates my time. This makes me want to communicate with them and help them.
  • Do not work in isolation. This is very, very, very important. Because research is highly risky and things don’t turn out the way you expect 99% of the time, you need support along the way. Make sure your office is close to your friends’, and talk to them regularly. At the minimum, they will be able to share your frustration in research. NEVER sacrifice your social life for a marginal increase in “time devoted to research” (which, as we all know, is likely to be devoted to social media).
  • Be more supportive and less judgmental of others. Research is hard, and we all know it. Instead of trashing others’ research, try to understand it and offer your colleagues constructive feedback. If you don’t understand it, maybe you can offer advice on how it can be more understandable.

P.S. I am determined to publish at least one post per week from today on. Stay tuned.

Advertisements

Development economics in a developed country: how do poor Americans save?

This recent episode of NPR’s planet money talks about the financial lives of poor Americans. A few practices mentioned in the article, such as group lending, private lenders, and high interest rates for loans, are strikingly similar with what poor people do in developing countries in Africa and Southeast Asia.

This resemblance leads me to think whether ideas and methodologies in development economics can be (more extensively) applied in developed country settings. Classic models in development economics such as health-based poverty trap and credit constraint can be easily applied to study the causes of poverty in a developed country setting. Comparative studies of poor individuals in developed vs. developing countries can shed light on the impact of institutions, governance, and infrastructure on addressing poverty.

The data mentioned in this episode are pretty amazing — 235 poor households across America tracked over a year with high frequency financial diaries. I bet interesting research based on this data is on the way.

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.

Technological Adoption Analysis with Panel Data

I just submitted my final paper for the panel data class. Sharing it here. Comments and feedback are welcome!

I. Introduction

This paper provides a critical review on the econometric approaches to model technology adoption decisions in developing countries. These decisions include the choice of whether or not to adopt a particular technology (e.g. high yielding variety seeds) and the amount of inputs depending on the technology used.

The developing country setting presents two additional challenges to identifying the determinants to technology adoption. First, imperfect access to credit and insurance introduces correlation between lagged productivity shocks and current input choices, thus violating the strict exogeneity assumption that is commonly maintained in panel data models. Second, the prevalence of informal networks highlights the importance of incorporating learning and externality into the analysis.

Following Foster and Rosenzweig (1995), suppose we are interested in what factors determine the adoption of high yielding variety (HYV) seeds of farmers in developing countries. There are two broad sources of uncertainty that drives differences in different technology adoption behavior. First, farmers may know the returns to HYV seeds but not the optimal levels of inputs. Therefore, a farmer needs to experiment with different levels of input choices once she decides to use HYV seeds. Second, there may be uncertainty in the profitability of this new technology. This source of uncertainty can be especially relevant when the technology is new (Conley and Udry 2010). Although the two sources of uncertainty may co-exist, we focus on only one at a time given the complication of the problem.

II. Input Choice as Technology Adoption

In this section, we assume that returns to technology adoption depends on how close actual input levels are to optimal input levels, i.e. use a target input model. Foster and Rosenzweig (1995) uses this framework to examine how farmers HYV adoption decisions depend on own and neighbors’ experience. In their framework, expected profits of farmer j at time t is
1
where $\eta_h$ is yield using HYV varieties, $\eta_{ha}$ is the loss associated with using less suitable land as more HYVs are used, $A_j$ is the total amount of land, $H_{jt}$ is the amount of land using HYVs, $\sigma_{\theta jt}^{2}$ is the updated variance of the mean input level, and $\sigma_{u}^2$ is the variance of the error term in target input use (relative to the mean optimal input). The updating of the variance term depends on learning from own and neighbors’ experience. This will be the focus of section IV.

In the empirical analysis, the authors estimate the profit function adding education of the farmers as an additional covariate:
2
where $S_{jt}$ is the cumulative number of parcels planted by farmer j up to time t, $\bar{S}_{-jt}$ is the average of the cumulative experience of neighboring farmers, $\rho$’s are precision terms of own and neighbors’ experience as signals of optimal input levels. Two approaches are used for estimation.

The first approach uses IV and fixed effects to estimate a first-order reduced-form approximation of equation (2). Instrumental variables are used to address correlation between 1) contemporaneous profit shocks and production decisions, and 2) lagged profit shocks and contemporaneous adoption (potentially because of credit constraints). Fixed effects are used to eliminate individual level heterogeneity $\mu_i$. If we maintain the assumption that input decisions are predetermined, the IV approach address the concern that strict exogeneity is violated. Note that predeterminedness implies that the profit shocks in first differences exhibit first-order autocorrelation but are uncorrelated at all other lags. This seems a reasonable assumption if we believe the profit shocks are unanticipated and are not persistent over time. Because Foster and Rosenzweig (1995) do not describe the nature of the profit shocks, it is difficult to evaluate the validity of the predeterminedness assumption.

The second approach uses nonlinear IV fixed effects to obtain the structural estimates of the profit function. Equation (2) is differenced over time and estimated using standard nonlinear IV procedure. This approach is subject to the same concern as the first approach.

III. Discrete Technology Adoption Decisions

In this section, the outcome variable equals one if the individual adopts technology in period t. Because technology adoption contributes to accumulated experience, adoption in the current period may induce changes in the returns to technology in the next periods in a complicated way.

Foster and Rosenzweig (1995) examine HYV adoption using reduced-form predictions from the structural model. But without solving the decision rules, they are unable to estimate the structural parameters. To address this limitation, we might use nonlinear panel data models with stronger distributional assumptions of the error terms (e.g. logistic distribution) and use conditional maximum likelihood estimators. This, however, rules out serial correlation in the error terms and might be unrealistic. An alternative approach is Manski’s conditional maximum score estimation. This approach achieves identification from “switchers”, but observing enough individuals switching from adopting versus not adopting a specific technology might be challenging as there are often fixed costs involved in a new technology and hence persistence in adoption decisions.

Suri (2011) provides an alternative framework to examine why farmers make different adoption decisions. She uses the information on the correlation between productivity differences and productivity of a technology among farmers who use both technology to project the different productivity levels for farmers who use only one technology. More specifically, she assumes profits for farmer i with productivity

3

She estimates the following equation for yields:
4
Based on the primitives of the model,

5

The identifying assumption is mean independence of the composite error $(\tau_i+\epsilon_{it})$ and the comparative advantage component $\theta_i$, and the histories of the regressors. Translated into assumptions on what drives the hybrid switching behavior, this assumes the unobserved time-varying variables that drive the switching should not be correlated with yields.Chamberlain (1982) correlated random effects approach is used for estimation. Dependence of the observed $\theta_i$’s on the endogenous input $h_{it}$ is accounted for using the linear projection of $\theta_i$ on the full history of inputs and their interactions. Structural parameters are recovered from reduced-form estimates.

The correlated random effects approach reduces the threshold for identification, and it seems reasonable to assume that individual-level heterogeneity are uncorrelated with productivity shocks once the history of input decisions are controlled for in Suri’s setting. Moreover, the focus of Suri (2011) is to identify the \emph{cross sectional} heterogeneity in productivity and its consequence on hybrid seed adoption. It is unclear whether this focus warrants the use of CRE models.

IV.Learning in Technology Adoption

Recent literature on technology adoption highlights the importance of learning from own experience and the experiences of informal network members.

Conley and Udry (2010) collect data on social interactions and address the unobserved variable problem when studying learning effects in technology diffusion: pineapple planting. In their model, risk-neutral farmers each have a single plot, and maximize current expected profits by choosing discrete-valued input $x_{it}$ at time t. Pineapple output realized 5 periods after input decision is
6
where $\epsilon$’s are unobserved productivity shocks iid distributed with mean 0 and variance 1, $\omega_{it}$ captures spatially and serially correlated shocks to marginal product that is only observed by the farmer (not the econometrician). Farmers do not know the function $f$ but learn about it with a learning rule.

Identification uses the specific timing of plantings to identify opportunities for information transmission. Variation in planting decisions generate a sequence of dates where new information may be revealed to the farmer. Conditional on measures of growing conditions, Conley and Udry isolate events when new productivity information is revealed to the farmer. They then investigate whether new information is associated with changes in farmer’s input use that is consistent with social learning. A logistic regression is used to estimate how farmers’ input decisions respond to actions and outcomes of other farmers in their information networks (data collected by the authors).

The baseline regression model is
7
where $M_{it}$ is an index of good news on input levels constructed from inputs and profits five years ago and now. The identification assumption is that conditional on measures of changes in growing conditions $\Gamma_{it}$ and other farm level characteristics, the information measure $M_{it}$ is uncorrelated with unobserved determinants in growing conditions and therefore input use. A significant, positive $\beta_1$ is evidence for social learning.

An important limitation of this approach is that it completely ignores the endogenous formation on informal networks and the potential dynamic changes in informal networks. To study the learning effects in technology adoption, we need a better understanding about the formation of informal networks and the nature of learning to evaluate whether the identification assumptions are realistic.

References:

Chamberlain, G. (1982). “Multivariate Regression Models for Panel Data,” Journal of Econometrics 18: 5-46.

Conley, T. and Udry, C. (2010). “Learning about a New Technology: Pineapple in Ghana,” American Economic Review 100(1): 35-69.

Foster, A. and Rosenzweig, M. (1995). “Learning by Doing and Learning from Others: Human Capital and Technological Change in Agriculture,” Journal of Political Economy 103: 1176-1209.

Suri, T. (2011). “Selection and Comparative Advantage in Technology Adoption,” Econometrica 79(1): 159-209.

Julie Cullen on the Unintended Consequences of Education Reform

Julie Cullen at UCSD visited Duke this week and gave a talk this morning on the unintended consequences of education reform. Her talk is a general summary of the lessons drawn from three papers on 1) the impact of fiscal incentive on student disability rates, 2) school gaming under a performance accountability system, 3) strategic high school choice under Texas’ top ten percent plan.

These papers ask different specific questions, but they all address broadly the mechanisms through which education reforms can change individual behavior and create unintended consequences. The intuition is that whenever eligibility to funding/benefits are conditional on some manipulable measures, those measures will be manipulated. This same logic is applied to examine how fiscal incentives, performance standards, and policies aimed at expanding college access have distorted incentives of schools.

Although I am not very interested in the economics of education, I can imagine the same approach being used to assess other types of government policies, e.g. welfare program participation.

Using mobile phones to collect (otherwise ignored) data!

While I was doing a literature review on migration, I saw an interesting new article on inferring internal migration patterns from mobile phone usage data in Rwanda. The author Joshua Blumenstock, a professor in the information school at the University of Washington, has quite a few interesting projects using mobile phone data for policy evaluation in developing countries.

He also has a few courses on data analysis and visualization, including an introduction to data science and another covering core methods in data science, both of which might be appealing for a broad audience.

New Directions for Migration Research

If you are interested in economic research on migration, I highly recommend this presentation at the Center of African Economies 2015 conference. A great deal of fresh, puzzling recently discovered data patterns and exciting directions for future research from a group of young promising economists. The following are some quick notes. I encourage you to watch the whole presentation for your own inspiration.

Michael Clemens on Skilled Migration

– Future research questions: 1) look at surveys with details about migrants: what do they learn? How well do they assimilate into the local labor markets? 2) How does participation in the global labor market affect education systems in the origin country? 3) How should we model human capital externality (this is hard!)?

– Incorrect language can mislead our discussion on migration. “Brain drain”? Think about calling women’s labor force participation rate “family abandonment rate”. How would that feel?

Melanie Morten on Migration in Sub-Saharan Africa

– Facts: much migration in Sub-Saharan Africa in rural-urban, rural-rural, and urban-urban; there are large regional differences in migration flows.

– Apart from migration across sectors, we should think about labor flows within sector across areas as well.

– Potential reasons for the lack of migration in the presence of large wage gaps: 1) selection; 2) people care more than wages (amenities/quality of life, cost of living, congestion, etc), which is the compensating differentials argument; 3) it’s costly to migrate.

– There’s evidence for all three channels mentioned above. This points to the need for further understanding of the costs of migration and potential roles of policy to foster a better match between skill and location.

– Policies that are space based (e.g. subsidies) might prevent people from migration by making it easier to stay in low-productivity places. (I found this particularly interesting).

Clement Imbert on Seasonal Migration

– Seasonal migration is especially important for risk coping in certain areas with agricultural lean seasons (India, Bangladesh, etc); but the contexts are similar.

– Data: NSS data, REDS data, RICE survey.

– Two sectors: construction (spot market) and manufacturing (network/referral matters).

– Need to consider the general equilibrium effects of seasonal migration (on urban wages, amenities, etc).

– Little data and work done on seasonal migration in Africa.

Q&A:

1. How to study forced migration due to climate change or conflicts?

Morten: This is challenging to model because migration is not a choice. Current research has focused on the impact of large migrant inflows on the local economy. See Ran Abramitzky‘s work. In the case of climate change, which is a permanent shock, political economy might matter and the distributional impacts are important. See Esteban Rossi-Hansberg‘s recent work (he presented at Duke two weeks ago and it was fascinating!).

2. Are migrant non-migrant wage differences only seen in developing countries?

Morten: No. It seems people value amenities differently. In developing countries, the benefits can be similar with those in developed countries, but the costs can be very high. See my paper on nonparametric estimation of migration costs (my comment: gravity model in trade?). Thinking about differences between developing and developed countries can be useful for policy intervention.

Clemens: There is a broader literature on “Why don’t people make profitable investments?” There’s a lot to explore in developed countries — why do people stay in the Appalachia when they can move to Miami?

3. How do property rights affect models on migration?

Morten: This is a first-order consideration. See paper in AER: as people get formal property rights, they migrate more. Part of the lack of migration might be due to property rights.

4. What’s the role of information on migration?

Morten: Experiments. Need to find high quality data on wages that people can trust. There are many open questions and possibilities to extend existing models.

Imbert: Information seems to be more easily available for lower-skilled jobs.