Biased Learning: Factor abundance and inequality in trade liberalization

Effects of policy learning on trade liberalization have been explained as a function of accountability and responsiveness of political regimes. I argue that apart from political regime type, the effects of policy learning depend on the country’s relative factor abundance and economic inequality. This claim is grounded on an extension of the selectorate theory where individuals that belong to two groups – capital owners and workers – vote on matters of redistribution and trade policy. The country’s factor abundance signals the groups’ preferences in trade policy but it is not sufficient to determine group preferences without any information of how trade policy will impact the domestic economy. It is by learning from other countries that beliefs on how trade policy will affect economic performance are updated and group preferences finally determined. To stay in office, incumbents need to follow one group’s preferences and buy the other’s via redistribution, and it is economic inequality that determines how costly it is to counter each group’s preferences. Using dyadic trade data from 82 countries and from 1950 to 2000 my analysis shows that inequality mitigates the effects of trade liberalization as a result of learning in labor abundant countries, and accentuates the same effects in a capital abundant country. The results also suggest that weaker beliefs on the benefits of trade liberalization are necessary to liberalize trade in capital abundant countries than in labor abundant countries.