Practice 2. International trade, globalization and economic geography

Lecturer: Carlos Llano

In this practice, students will continue working with the PC-LAB in gravity equation. Please focus on Part III, which includes several exercises for modelling interregional flows within Spain at the regional level (nuts 2).

Part III: The gravity model and the home bias at the regional level in Spain.

You can work in group, but you should deliver a brief individual report (2-3 pages) by the date of the final exam.

Read the code and explain (briefly) what is the purpose of the main lines of the code.

  1. Based on the explanations given in class, what is the difference between the following regressions: you can visit the “help” in Stata. Be brief.

Line 98: reg lyod lgdp_o lgdp_d intra_reg contigui ldist

Line 100: areg lyod lgdp_o lgdp_d intra_reg contigui ldist, a(year) cluster (dist)

Line 102: xi:reg lyod lgdp_o lgdp_d intra_reg contigui ldist i.reg_o i.reg_d i.year, cluster (dist)

Line 112: xtreg lyod lgdp_o lgdp_d intra_reg contigui ldist,re

Line 115: xtreg lyod lgdp_o lgdp_d intra_reg contigui ldist,fe

  1. Why when estimating Line 115, STATA omits the results for some variables? What alternative estimate can be used instead in a gravity equation if you are really interested in these dyadic variables? Clue: Have a look on “gravity 2.do” (line 34) of Part II and read the following note:

Note: Stata allows 2 ways to generate/include dummies:
  1. Option 1: As in “gravity 2.do” of Part II. First creates the dummy (lines 20,21) and then, they are included in the regression (lines 34) using: ced* cid*

  1. Option 2: As in “gravity_spain_nuts2.do” of Part III. You can create the dummies as in (line 102) directly in the regression, by using “i.[variable]”: i.reg_o i.reg_d i.year

  1. Be sure that you understand why “Option 2” includes(i.year), but you don’t have “year*” in “Option 1”.

Run the STATA program “gravity_spain_nuts2.do”. Maybe, you may have to “search” and “install” (ppml; outreg2). Then do:

  1. Use a descriptive (graph or a table) to illustrate the intensity of intra-regional/inter-regional trade 1st considering Spain as a whole.
  2. Based on the results obtained in the main routine, obtain the “home bias” (border effect) for you region(the one assigned to you in Practice 1). You can use the specification in line 102. You may want to explore the following options: 1) drop all observation for the other regions; 2) create an interaction between the “border effect” dummy and a dummy for your region.
  3. How is the border effect obtained for your region in comparison to the one for Spain as a whole? Do you have any explanation?