BlumRangel

Trade Reform and Rising Inequality in Developing Countries: A Roy-Model Illusion? by Blum and Rangel: 

Trade Reform and Rising Inequality in Developing Countries: A Roy-Model Illusion? by Blum and Rangel Comments by Wolfgang Keller University of Colorado National Bureau of Economic Research Centre for Economic Policy Research

Impact of trade liberalization on income distribution in LDCs: 

Impact of trade liberalization on income distribution in LDCs Simplest 2 x 2 factor abundance trade model: LDCs are relatively abundant in low-skilled labor LDCs will export low-skill intensive products Low-skilled labor benefits relative to high-skilled labor (Stolper-Samuelson theorem) => Inequality in LDCs should fall after trade liberalization

Slide3: 

Feenstra-Hanson /97 Feenstra-Hanson 1997

A Revisionist View on LDC Inequality: 

A Revisionist View on LDC Inequality Authors say that inequality in Brazil and Mexico did actually decrease, as predicted by the model Key for getting this right is To control for changes in the dist’ns of low-skilled and high-skilled workers in terms of observables and unobservables To look at individual-level, not average wages

A Wage Decomposition: 

A Wage Decomposition How did the skilled-to-unskilled wage, ws- wu, change over time ? If skill is identified with education, and we knew that ws = b1*(educ)+error1 wu = b2*(educ)+error2 ws- wu = (error1-error2) + (b1-b2)educu +b1(educs-educu)

Brazil: no argument there: 

Brazil: no argument there Was there indeed a major trade liberalization in Brazil between 1989 and 1996? If so, may be focus the paper on Mexico

Mexico: 

Mexico

Some concerns: 

Some concerns Are the results driven by unobserved characteristics affecting selection, or by observed characteristics, such as age? Add the time interactions to OLS columns Does the time variable really identify only occupational choice? Need to defend identification more proactively How good is the fit of the selection model? R2 moves from 0.26 to 0.27 with 4 additional parameters

Other questions: 

Other questions Selection finding: Why is it that workers with more advantageous characteristics were more likely to be white collar in 1991, versus in 1987? If indeed conditional on occupational choice, inequality in Mexico declined after trade liberalization – which is the important result: conditional, or unconditional? Is occupational switching always possible? Costlessly? Skill formation policies: Should one emphasize LESS years of education, and MORE…what? Flexibility? It seems unhelpful, almost tautological, to define “skilled” in terms of “occupying a good occupation”