Job Talk Questions Flashcards
Why static model?
- Most of the trade papers that quantify welfare effects using structural gravity models employ a static framework.
- These models have been designed to capture the source of comparative advantage (utilizing the trade movement) rather than how the economic variables (capital, population, technique of production, and tastes) evolve.
- In structural gravity models there’s no accumulation equation or state variable because the purpose is to capture the sources of trade movement at a given point in time.
- I believe that given the scope of the research question in this paper, which is to quantify the benefits against the costs, the static model suffices.
- I use the static model as a shortcut to the steady-state of the dynamic model.
- You can think of a setup where waste by-product in period t which is used in recycling in t+1 and that is an input to manufacturing in t+2 to study the evolution of waste itself and how that would change with policies. Estimating this model is now possible with the Exiobase3 database that has multi-region input-output tables with waste flows within and across economies.
How do the differences in elasticities determine the specializing commodity of a country?
The rich are technologically better while the poor have cheaper labour. For a commodity that is more sensitive to trade costs, the cost considerations swamp the technological considerations during production. As lower-income countries have lower costs of production (due to cheap labour), they tend to specialize in the generation of low-value waste. Conversely, the rich tend to specialize in the production of manufactured goods.
Are your trade elasticity estimates credible? They seem large.
- 2SLS procedure to tackle classical measurement error.
- However, SW (2014) show that the trade barrier measure is underestimated, which means that the trade elasticities are overestimated.
- They propose a simulated method of moments approach, which would be hard in my framework because I’m solving for the parameters in 3 sectors simultaneously.
- I use the Robson and Whitlock (1964) modified measure of trade barriers—less biased but as efficient.
- My estimate for manufactured goods is close to EK’s median estimate.
- However, the lack of estimates for waste flows, prevents me from making a similar comparison.
Is the size of welfare gains from waste trade reasonable?
- The size of welfare gains is partially an artefact of the modelling assumptions: one sector, multiple sectors, market structure, etc.
- Fally and Sayre (2018): Accounting for features of primary commodities such as low elasticities of demand and supply and concentration of resources in select countries create large welfare gains (three times larger for the median country).
- Etkes and Zimring (2015) use a natural experiment—Ghaza blockade 2007-10—to study the welfare effects of trade. They find much larger estimates (15-27%).
- What’s important is to compare waste trade gains to overall gains.
Are the environmental cost estimates reasonable?
- I’m using existing estimates of SMCs combined with model predictions on how waste disposal changes with a change in policy.
- The assumptions governing changes in real incomes are the same ones governing changes in waste volumes. So these estimates are partially an artefact of the assumptions and studies have shown could be much larger.
- However, what matters more is the size of environmental costs relative to the welfare gains. And I show that though the environmental costs are smaller compared to welfare gains, in the case of low-quality waste trade the costs swamp the benefits for lower-income countries.
If you had to reject your paper at a journal, what grounds would you use for rejection?
- The biggest concern in my paper is that it’s hard to find credible estimates on the cost of pollution externalities.
- The way I tackle this is by using the Ban amendment as a revealed preference to back-calculate the cost of pollution externalities or using existing estimates in the literature to quantify the environmental costs. But even with existing estimates, the information is limited. However, the size of my environmental costs hinge on these figures, so if they’re meaningfully off from the true values, the results from my paper could be questioned.
- However, I do several robustness checks to provide wide bands on these estimates and even with that, the main qualitative conclusions continue to hold.
- So, I believe that I did the best that anyone with this data could do.
- One way to capture the social marginal cost of waste is by quantifying the value to a country from entering an international waste trade agreement. In my paper with George Deltas, this is what we intend to do. The data on adoption of the Basel Convention provides variation at both time and country-level and so facilitates stronger identification of the value to a country from entering the agreement. The estimates from this work can be fed back into my job market paper.
Why don’t you use a game-theoretic model where two countries with a bilateral agreement negotiate prices for waste trade?
- First of all, even though information on several multilateral/bilateral agreements among countries is available, it is incomplete and lacks data on prices of different types of waste, so it would be hard to estimate a game-theoretic model.
- Given the goal in this paper, which is to quantify the welfare effects across countries, a gravity model is suitable because the welfare estimates depend only on a couple of sufficient statistics–trade elasticities and share of expenditure on domestic goods–which can easily be computed with existing trade data. This is an advantage of using a gravity setup in that data requirements are not severe and much less compared to also a computable general equilibrium model.
- However, if the ideal data were available, I believe that we could exploit the natural experiment to get at more credible estimates.
Do your findings hold for waste other than industrial waste?
- Industrial waste accounts for 94-97% of waste generation and the data I have is mainly industrial in nature.
- Municipal (mixed) waste is part of the low-value waste in my sample.
- Consumers in rich countries are producing more low-value plastics/paper waste compositionally compared to lower-income countries. They also produce more waste per capita. Demand across countries should be similar.
Does waste trade only happen from rich to poor?
It happens even from the poor to the rich countries. However, I find that there’s heterogeneity in the type of waste going from the rich to poor and from poor to rich. Rich countries (technologically better) mainly import high-value waste like metals while lower-income countries with cheaper labour import low-value (mixed) waste that requires manual sorting.
Can a unilateral regulation on top of the China Ban make things better?
- Rich stop exporting–lower-income countries should become worse-off because waste exports are like a transfer from rich to poor.
- My results show that the poor disproportionately gain from the waste trade because they care more about the inputs to manufacturing than the environmental externality. The rich, in contrast, care about the environment are more sophisticated exporters and in a way, their decision to trade is endogenizing the SMC from waste. So if there’s any unilateral policy from the rich, it would make the poor worse off.
- Ban amendment makes countries worse-off.
Why does waste trade create relatively larger welfare gains than regular trade?
- Because it’s also creating externalities, so for trade to happen waste trade has to create much higher gains.
- Only one recycled good while manufacturing substitutes are available.
One contribution of my entire work that’s a highlight I would want to tell anyone?
- Lots of questions that study the effect of trade on the environment through air pollution. My contribution is asking how waste trade affects the environment.
- No standard way in the literature to model the flow of a commodity that’s a byproduct of manufacturing and input to recycling. On top of that, you need to capture the source of its comparative advantage.
Why is low-value waste recycling rate closer to overall recycling rate?
The US is disproportionately recycling a larger fraction of high-value waste than other richer countries, like Germany and Sweden, which also recycle a lot of low-value waste. This is a way to account for the general pattern across countries that a larger fraction of high-value waste is recycled than low-value waste.
Do you obtain two different values for externality parameters?
Yes, the calibrated values would be different for the two types of waste. However, in dollar terms, the value of the externality from both types of waste would be the same because the parameters are chosen so that each additional tonne of either type of disposal amounts to the same value for a country.
Why Logit and IHS transformations?
- Can’t do PPML because the dependent variable is ratio and not count.
- Can do log(ratio), but will lose zeros and adding 1 is ad-hoc.
- Logit and IHS are variance-stabilizing transformations.
- Logit transforms to a scale of (-inf, +inf).
- IHS prevents loss of zeros obs.
