Comments on the Model
1. Authors use the AIDS to derive the expenditure share in good j for individual h, which is given by equation (10); s_j (p,x_h )= α_j+∑_(k=1)^J▒〖γ_jk lnp_k+β_j ln(x_h/(a(p))) 〗
Here a(p) is the cost of the subsistence basket of goods which is independent from nonhomotheticities, that is, it is assumed that a(p) does not vary with the level of income. Though this is a plausible assumption within a country, across countries it is more like to vary given the differences in the level economic prosperity and tastes of countries. This is important as βj is interpreted as the income elasticities of the good and so βj’s for various countries for the same good are directly comparable, that is, income elasticities of the same …show more content…
Here T_ni^s cannot be directly estimated from the available data, so authors use the proxy which is given by τ_ni^s=d_ni^(ρ^s ) Π_j g_j^(〖-δ〗_j^s ) ϵ ̃_j^s, dni is distance ρ^s is the elasticity between distance and trade costs in sector s, the g’s are other gravity variables such as common border, common language, and ϵ ̃_j^s is an unobserved component of the trade cost between i and n in sector n. Here ϵ ̃_j^s could be potentially correlated with Ω_n. Suppose in country n there are unobservable trade barriers such subsidy or other support in sector s, so, Ω_n is likely to be affected by these type of unobservable component of trade cost, that leads to the before mentioned correlation. When author estimate the equation (27), the ϵ ̃_j^s is consumed in error terms which causes the endogeneity in estimation the equation statistically, and so sectoral income elasticities parameters β_i^s are inconsistently estimated with incorrect statistical significant, which is reported in table 1 in the …show more content…
The model does not consider the change in the income distribution due to greater openness. Authors mention at the beginning of the paper that the Stolper-Samuelson theorem performs poorly in terms of empirical evidence, however, suggest no alternative hypothesis regarding the effect of trade on income distribution and employment situation of the different factors of production. Most of the within country opposition to trade is due to job and income loss in import competing industries. Thus, the gains from trade estimated in this paper are overestimated due to partial equilibrium nature of analysis. For the complete measure of gains from trade, a general equilibrium approach is
1. Authors use the AIDS to derive the expenditure share in good j for individual h, which is given by equation (10); s_j (p,x_h )= α_j+∑_(k=1)^J▒〖γ_jk lnp_k+β_j ln(x_h/(a(p))) 〗
Here a(p) is the cost of the subsistence basket of goods which is independent from nonhomotheticities, that is, it is assumed that a(p) does not vary with the level of income. Though this is a plausible assumption within a country, across countries it is more like to vary given the differences in the level economic prosperity and tastes of countries. This is important as βj is interpreted as the income elasticities of the good and so βj’s for various countries for the same good are directly comparable, that is, income elasticities of the same …show more content…
Here T_ni^s cannot be directly estimated from the available data, so authors use the proxy which is given by τ_ni^s=d_ni^(ρ^s ) Π_j g_j^(〖-δ〗_j^s ) ϵ ̃_j^s, dni is distance ρ^s is the elasticity between distance and trade costs in sector s, the g’s are other gravity variables such as common border, common language, and ϵ ̃_j^s is an unobserved component of the trade cost between i and n in sector n. Here ϵ ̃_j^s could be potentially correlated with Ω_n. Suppose in country n there are unobservable trade barriers such subsidy or other support in sector s, so, Ω_n is likely to be affected by these type of unobservable component of trade cost, that leads to the before mentioned correlation. When author estimate the equation (27), the ϵ ̃_j^s is consumed in error terms which causes the endogeneity in estimation the equation statistically, and so sectoral income elasticities parameters β_i^s are inconsistently estimated with incorrect statistical significant, which is reported in table 1 in the …show more content…
The model does not consider the change in the income distribution due to greater openness. Authors mention at the beginning of the paper that the Stolper-Samuelson theorem performs poorly in terms of empirical evidence, however, suggest no alternative hypothesis regarding the effect of trade on income distribution and employment situation of the different factors of production. Most of the within country opposition to trade is due to job and income loss in import competing industries. Thus, the gains from trade estimated in this paper are overestimated due to partial equilibrium nature of analysis. For the complete measure of gains from trade, a general equilibrium approach is