Economics 6054: Assignment 1
Ricardian Trade and Gains
Professor Philip Luck
1. This question asks you to discuss some recent empirical work on estimating the (reduced-form)
gains from trade, and how this work relates to theoretical work on the gains from trade.
(a) State how large the estimated gains from openness are in Frankel and Romer (1999), and
Feyrer (2009a) and Feyrer (2009b). Discuss whether you think the estimates in the two Feyrer
(2009) papers are smaller or larger than the true average treatment eect of openness to
(b) Assuming that the estimates in Feyrer (2009b) are unbiased estimates of the average treatment
eect of an extra unit of international trade on a countrys real income, do these estimates
make quantitative sense in the context of standard models of market distortions?
(c) Do the estimates in Feyrer(2009b) line up with the predictions for the size of the gains from
trade in the Eaton and Kortum (2002) model?
(d) Discuss an amendment to Feyrer (2009b) that would explore the extent to which the theoretical
predictions about the size of the gains from trade in Eaton and Kortum (2002) t the data. Be
clear about what regression youre proposing, why youre proposing it, and what the estimates
would tell us.
2. This question asks you to replicate and extend results from Costinot, Donaldson and Komunjer(2010)
as well as comment the theoretical framework and empirical strategy.
(a) Using the dataset CDK data.dta and the template code les CDK code.do replicate the
results found in table 3 using ivregress and dummy variables.
(b) Instead of using xed eects absorb constants using the command reghfe. How much quicker
is does this estimation run?
(c) Cluster your standard errors by industry, report what happens to the standard errors.
(d) Export tables 3, estimated as in CDK and with regfe and clustered standard errors as .tex
les and compile them in latex. Submit your do le, latex code and compiled PDF with both
(e) Brie y explain how developing a theoretical model helped guild the empirical exercise in CDK.
Be sure to mention the construction of the dependent variable as well the interpretation of
the coecient of interest.
(f) Describe what you see to be the most serious criticisms of this paper.
(g) Brie y describe the instrumental variables strategy using in CDK and suggest an potential
(h) Are there additional theoretical restrictions in the CDK model that are not being tested in
3. Consider extension of the model described by Costinot, Donaldson and Komunjer (2012).
(a) Describe an extension of the model in Costinot, Donaldson and Komunjer (2010) that would
add Heckscher-Ohlin features to it.
(b) Now outline an empirical paper that would use this extension to make as useful a contribution
to the empirical H-O literature as possible.
(c) Be sure to state exactly what regression(s) or other empirical tests/exercises you could
proposing, how they follow from the model, and what the estimates would tell us.
(d) State any attractive features of this approach you can think of, relative to existing empirical
work on the H-O model.
TO GET THIS OR ANY OTHER ASSIGNMENT DONE FOR YOU FROM SCRATCH, PLACE A NEW ORDER HERE