Maria Socorro Gochoco-Bautista: There is a difference between structural transformation in which the share of agriculture in total output declines over time, and structural transformation in which the share of agricultural sector employment to total employment declines over time as a country develops. The authors state that they would like to focus on the output side of structural adjustment (i.e., what happens to the sectoral shares of output rather than the sectoral shares of labor).I think that, as the focus of the paper is pro-poor growth, it is more important to talk about people and employment rather than output. Indeed, the authors define growth as pro-poor “if and only if poor people benefit in absolute terms, as reflected in an appropriate measure of poverty.” The paper by Loayza and Raddatz (2006) that the authors themselves cite argues that the labor intensity of the growing sector rather than which economic sector in particular is growing is more important.This leads to my suspicion that what makes growth pro-poor is the ability of growth to generate good jobs, and to ultimately reduce the share of agricultural employment to total employment as the economy undergoes structural transformation and develops. If this is the case, then I am afraid that the paper is unable to shed light on how to generate pro-poor growth, even though it is able to point to factors that may generate growth in the economy in general. Of course, because the demand for labor is derived from the demand for output, one would expect the demand for labor to increase in fast-growing sectors and reduce poverty. In Asia, however, we have generally seen that although output growth has been robust, the share of employment in agriculture relative to non-agriculture remains high, and we know that the quality jobs are not in agriculture and that the majority of the poor are in rural areas. Perhaps the employment elasticity of growth in the non-agricultural sector may be low especially if production there is relatively capital intensive. The authors do not give a measure of the labor intensity of the fast-growing non-agricultural sector relative to the agricultural sector.Where to focus reform proposals is important because this has a bearing on policy implications. It is important to take into account where the distortions in the economy are in the context of institutional limitations and which act to prevent growth from being pro-poor. For example, there are barriers to job creation and entry into high-paying jobs in the non-agricultural or industrial sector that act as a tax on the most progressive sector of the economy. This limits the ability of the industrial sector to absorb surplus labor from the rural sector and retards the shift from agricultural to non-agricultural growth. Evidence for these distortions exists. In urban centers, for example, the unemployment rate among those with tertiary education is higher than in those with only primary or secondary levels of education, and the employment rate of the less educated is higher in sectors where regulations are not binding. Such barriers in labor markets may account for gaps in labor productivity between agricultural and non-agricultural sectors. For pro-poor growth, I think that it is important to focus on the labor market and employment generation aspects of structural transformation.In the regression equation, agricultural growth and non-agricultural growth are both entered as regressors, but the two are related to each other. Growth in the agricultural sector makes growth in the non-agricultural sector possible by opening up non-farm employment opportunities, a point the authors make themselves. But because the focus of the paper is on output growth rather than employment growth potential, it is difficult to know with certainty whether growth in any sector is pro-poor.