Comments and Discussion

Ann Harrison:

Nancy Chau and Ravi Kanbur have written a very interesting paper exploring which countries adopt international labor standards and when the adoption occurs. This is an extremely important and current issue. While there is enormous interest in how globalization has affected the quality of our lives, there is simply not enough research on this question. Although much of the policy debate focuses on how globalization affects the labor force, the evidence in this area is slim. Chau and Kanbur's paper makes an important contribution by showing when countries are likely to adopt different types of international labor standards.

Chau and Kanbur's paper analyzes the determinants of the adoption of four kinds of ILO conventions: the right to organize, the abolition of forced labor, the elimination of discrimination, and the abolition of child labor. Somewhat surprisingly, the authors find that many factors do not matter, for example GDP per capita, openness to trade, education, political freedom, and urbanization. The factors that do matter include time (the probability of ratification increases over time), the type of legal system, and peer effects.

The results are provocative. Although standard measures of globalization, such as the share of trade in GDP, have no impact on the likelihood of adopting different labor standards, one measure does: whether similar countries have adopted the standard. To the extent that globalization implies that countries are more sensitive to what others do, this could imply a beneficial effect: countries are more likely to adopt labor standards when their neighbors do.

I have some suggestions for future revisions. One question of great policy interest is the relationship between openness to trade and measures of economic well-being. In particular many opponents of liberal trade policies have argued that trade leads to a "race to the bottom," as nations compete with each other by eroding labor standards and pay. The authors find no relationship between labor standards and trade. As their measure of openness to trade, the authors use the ratio of exports plus imports to GDP. The trade share is not, however, a policy measure. It is an outcome of trade policy and jointly determined with a number of other factors. Better measures of trade policy would include tariffs or quotas. The World Bank now publishes an average tariff measure that varies across countries and over time. It would also be interesting to see if there is any relationship between these other measures of globalization and the adoption of ILO conventions. Other interesting measures of globalization that could have effects on the adoption of labor standards would include capital flows or policies regarding the openness of the capital account. These types of data are compiled in the IMF's International Financial Statistics and its Annual Report on Exchange Arrangements and Exchange Restrictions.

The authors get the surprising result in some of their tables that higher GDP per capita is associated with a lower likelihood of adopting ILO conventions. This is puzzling. The authors may want to explore nonlinear effects of GDP per capita, and they might also try eliminating big data outliers, such as the United States. The United States has only adopted one out of the four conventions. The U.S. experience consequently implies that richer countries are less likely to adopt conventions, but it is probably a very unusual case.

The only robust determinant of adoption of ILO conventions, apart from legal status and time effects, are so-called peer effects. I worry that these peer effects may proxy for some omitted variables. It is unclear to me whether the authors allow for country-fixed effects, which could partially address this concern. I am also concerned that the coefficients on peer effects might reflect the fact that the time effect—which is significant and reflects the increasing likelihood of ratification over time—varies across different groups of countries. One possibility would be to introduce export status separately in the regression.

It is extremely interesting that legal systems are such critical determinants of the adoption of labor standards. The result that socialist countries are much more likely to adopt ILO conventions is a reassuring one...