ECO 2300F

October 2004

Daniel Trefler

Empirical Essay on the Heckscher-Ohlin-Vanek Theorem

Using the data supplied to you, discuss some feature of the HOV theorem or the empirical literature associated with that theorem. You have two data choices:

  1. Trefler, Daniel, “Trade Liberalization and the Theory of Endogenous Protection: An Econometric Analysis of U.S. Import Policy,” Journal of Political Economy, February 1993.
  2. Trefler, Daniel, “The Case of the Missing Trade and Other HOV Mysteries,” American Economic Review, December, 1995.

You will find links to the data on my website at .

The format of the paper should be as follows.

  1. The paper should contain about 4 double spaced pages of text (maximum of 1500 words). It should also contain two pages of neat, easily understood, uncrammed tables.
  2. In the introduction you will state the aim of the paper. This aim must be very well defined, very focused, and very narrow. You do not have space to do anything elaborate.
  3. In the theoretical section you should discuss what your approach will be for shedding light on the aim stated in the introduction. What are the strengths of your approach.
  4. Discuss your empirical results. Explain how the tables are pertinent to your aims. There is only one reason for having tables not 100% connected to your aims - you have uncovered an unexpected feature of the data that will fascinate the reader.
  5. Summarize your conclusions, noting any weaknesses of your approach.

In writing this paper I want you to get used to conducting independent research in a form (though not quality) suitable for journal publication. I will be looking for two things. Can you formulate a clear question? Can you confront that question with persuasive empirical work?

Tip: Start with a paper that is longer than 4 pages and edit out all but the strongest arguments.

Data Description for

“Trade Liberalization and the Theory of Endogenous Protection”

Data Documentation for

"Trade Liberalization and the Theory of Endogenous Protection:

An Econometric Study of U.S. Import Policy," Journal of Political

Economy, February 1993, volume 101(1), pp. 138-160

by Daniel Trefler

September 29, 1994

The data are copyright. Users interested in the data must send a

written request to the author. The address is

Department of Economics

University of Toronto

150 St. George Street

Toronto, Ontario

Canada M5S 1A1

The data appear in two accompanying files, N.DAT and M.DAT . N.DAT contains all the variables used in the NTB equation exceptimports. M.DAT contains all the variables used in the importequation except NTBs.

N.DAT: The columns correspond to the following variables described below.

NTB_N SCRCONC BCRCONC SCRCOMP BCRCOMP MEPS KSTOCK OCCUP5 OCCUP4 OCCUP2 OCCUP1 UNEMPLOY UNION NE GEOG2 TENURE X M80 VS79_83 IOCODE SIC4.

M.DAT: M K_SERV INV ENGSCI WHITECOL SKILL SEMSKILL UNSKILL CROPLAND PASTURE FOREST COAL PETRO MINERALS IOCODE SIC4.

The following defines the variables and provides some statistics

to verify that the data have been read properly by the user.

Notes: 1. For the NTB equation, "skilled" is omitted since it is

1 minus the sum of the other 4 occupation variables.

2. Delta(import penetration) equals M - M80.

3. The industry code is IOCODE and is a 6-digit code from

the Bureau of Economic Analysis Input-Output

classification. SIC4 is the numerically smallest SIC

code that maps into IOCODE. See Appendix B of "Input-

Output Accounts of the U.S. Economy, 1983" in "Survey

of Current Business" 69 (Feb. 1989).

Variable N Mean Std Dev

------

NTB equation

NTB_N 322 0.1105137 0.2387031 NTBs

SCRCONC 322 0.3781198 0.1899786 Seller concentration

BCRCONC 322 0.3754561 0.0703089 Buyer concentration

SCRCOMP 322 0.2700320 0.3548055 Seller number of firms

BCRCOMP 322 0.6670407 1.3184600 Buyer number of firms

MEPS 322 0.0302950 0.0597414 Scale

KSTOCK 322 0.3454289 0.2208742 Capital Stock

OCCUP5 322 0.0393972 0.0393853 Engineers, scientists

OCCUP4 322 0.2833637 0.0935765 White-collar

OCCUP2 322 0.3941035 0.1175837 Semiskilled

OCCUP1 322 0.0835156 0.0649416 Unskilled

UNEMPLOY 322 0.1211361 0.0587925 Unemployment

UNION 322 0.3457042 0.1291685 Union

NE 322 0.0490435 0.0878578 Employment size

GEOG2 322 0.6973935 0.1547467 Geographic conc.

TENURE 322 5.3801242 1.5672990 tenure

X 322 0.0906389 0.1238969 exports

M80 322 0.0829694 0.1263472 import penet., 1980

VS79_83 322 -0.1335873 0.2252051 industry growth

IOCODE 322 356630.17 160773.36 industry code

SIC4 322 3068.04 605.5315256 see note above

Import Equation

M 322 0.1076952 0.1517962 Import penetration

K_SERV 322 0.1108321 0.0331315 Physical capital

INV 322 0.0322772 0.0149108 Inventories

ENGSCI 322 0.0305586 0.0206737 Engineers, scientists

WHITECOL 322 0.1568722 0.0406919 White-collar

SKILL 322 0.1012067 0.0407723 Skilled

SEMSKILL 322 0.1165377 0.0395616 Semi-skilled

UNSKILL 322 0.0383608 0.0287589 Unskilled

CROPLAND 322 0.0235892 0.0644692 Cropland

PASTURE 322 0.0074905 0.0257857 Pasture

FOREST 322 0.000314535 0.0013469 Forest

COAL 322 0.0021271 0.0021443 Coal

PETRO 322 0.0321484 0.0462959 Petroleum

MINERALS 322 0.000995599 0.0023531 Minerals

------

Data Description for

“Trade Liberalization and the Theory of Endogenous Protection”

Some Additional Details

There are 13 variables in the data set. All data pertain to 1983. Except for the trade and trade restriction data, all figures are in millions of dollars. Their descriptions follow.

Observation

1) 4-digit SIC industry code

Trade Statistics

2) Imports M

3) Exports X

These are scaled by domestic consumption i.e., let y be industry output (item 13 below), let m be the level of imports, and let x be the level of exports. Then M = m/(y+mx) and X=x/(y+mx).

Direct+Indirect Factor Shares

(value of factors needed to produce the final output of the industry. When divided by final output these become the factor shares we discussed in class. For example, let ALAND,g be the element of A corresponding to land and industry g. Let yg be the output of industry g. Then yg is an element of item 13 below. Let wLAND be the price of land. Then wLAND*ALAND,g*yg is an element of item 9 below.

4) Semi-skilled labour

5) Skilled labour

6) White collar labour

7) Engineers and scientists

8) Physical capital(depreciation & rental charges on fixed capital & machinery)

9) Land (cropland and pasture)

10) Subsoil resources(coal, petro, and minerals)

Trade Restrictions

11) Average tariff rate

12) NTB coverage ratio

Miscellaneous

13) Final output (Value of shipments)

Data Description for “Trade Liberalization and the Theory of Endogenous Protection”

(Continued)

The following are the output of PROC CONTENTS and PROC MEANS. (Note that the label here differs slightly from what appears in the PROC CONTENTS output. The one below is the correct one.)

# Variable Type Len Pos Label

10 ENGSCI Num 8 68 Eng,Sci,Math,Archit,Surveyors

11 K_SERV Num 8 76 Deprec+Rentals, 1982 or 1983

12 LAND Num 8 84 Cropland + Pasture + Forests

6 M Num 8 36 Imports/(VS+Imports-Exports)

4 NTB_N Num 8 20 Kishore"s all NTBs

7 SEMSKILL Num 8 44 Oper,Assembler,Inspector,Transprt,Moving

1 SIC4 Char 4 0 smallest SIC4 code mapping into IOCODE

8 SKILL Num 8 52 Mechanic,Trades,Extract,Precisionprod

13 SUBSOIL Num 8 92 Petroleum + Coal + Minerals

5 TAR_N Num 8 28 Kishore"s tariffs

2 VS Num 8 4 Value of shipments

9 WHITECOL Num 8 60 CLERK+PROF+SERV1

3 X Num 8 12 Exports/(VS+Imports-Exports)

Variable N Mean Std Dev Minimum Maximum

VS 270 5706.55 13438.55 47.0000000 182591.75

X 270 0.0899200 0.1165143 0 0.8555000

NTB_N 270 0.0915674 0.2020039 0 1.0000000

TAR_N 270 0.0489067 0.0480479 0 0.4188000

M 270 0.1067352 0.1497854 0 0.8211000

SEMSKILL 270 619.3198585 1062.42 3.7189000 9966.73

SKILL 270 581.0928067 997.2018170 0.9738000 8136.14

WHITECOL 270 930.1051811 1776.43 2.5672000 16505.49

ENGSCI 270 214.6931115 481.0106131 0.9117000 3613.90

K_SERV 270 628.7241770 1234.97 1.8420000 12581.50

LAND 270 37.1451904 113.5567793 0 1127.31

SUBSOIL 270 505.8571611 5142.20 0 83989.03

Data Description for “The Case of the Missing Trade and Other Mysteries”

This directory contains the data used for the paper "The Case of the Missing Trade and Other Mysteries" by Daniel Trefler November 15, 1994. There are two files.

1. hov_pub.dat

This is an ascii version of the data set. It was written with

the following sas PUT statement:

PUT COUNTRY $1-12 FACTOR $14-25 " " AT V Y B YPC;

Thus, the character variable COUNTRY appears in positions 1-12, the character variables FACTOR appears in positions 14-25, a space appears in position 26, and the variables AT, V, Y, B, and YPC are formatted by the SAS "best" format and separated by spaces.

The variables are defined as follows:

at = factor content of trade, F=A*T

v = endowment

y = gdp, world bank, y=p*Q

b = trade balance, world bank b=p*T

ypc = gdp per capita, PWT

country = alphabetic country name

factor = alphabetic factor name

pop = population, wb

2. sample.sas

This is a SAS program that computes some of the statistics used in the paper as well as in my "International Factor Price Differences: Leontief was Right!", Journal of Political Economy, Dec. 1993.

To check that the data are properly read, the following are sample statistics:

Variable Label N Mean Std Dev

------

AT factor content of trade, F=A*T 297 -169046742 6642034856

V endowment 297 43992270693 276423695475

Y gdp, world bank, y=p*Q 297 258760169537 614560302529

B trade balance, world bank b=p*T 297 -2873808912 13345058955

POP population, wb 297 36644092.36 51387252.69

YPC gdp per capita, PWT 297 7694.61 3849.37

INDEXC country identifier 297 17.0000000 9.5379753