Part III: From R Gravity Models to Python

Part III is the econometrics core of the course. The running case is the Post-Soviet gravity project: bilateral trade flows are modeled using flow, economic mass is measured with gdp_o and gdp_d, trade costs are represented by distw, comlang_off, and contig, and institutional variables include wto_joint, EU_joint, and EAEU_joint.

The module has two goals. First, students learn why each gravity estimator exists. Second, they learn how to translate R-style gravity workflows into transparent Python implementations that can reproduce and extend the Post-Soviet paper without fabricating results.

Chp	Topic	Main Output
07	OLS gravity	OLS replication workflow
08	Fixed effects	Structural gravity translation
09	Bonus Vetus	BVU/BVW implementation plan
10	PPML	PML estimator comparison

This part contributes to the final publication-ready paper by turning the replication exercise into an econometric toolkit. Students should leave Part III able to explain, estimate, and interpret each model family before moving to replication tables in Part IV.