Changes in the headcount rate are the standard metric for gauging how public transfers and taxes affect poverty. An alternative strategy, one theoretically more appealing and complete, is to rely on distribution-sensitive indexes [Sen (1976, 1981)]. How would policies measured impacts change if such an approach were to be used? This study provides new empirical evidence based on Luxembourg Income Study data for seventeen countries covering various years between 1969 and 1997. Poverty is measured using three indexes from the class developed by Foster, Greer, and Thorbecke (1984), one of which is the headcount rate. Estimates of the policy impacts are obtained by computing index values with before- and after-policy income. Evidence is also provided on the determinants of cross-country differences in index values and policy effectiveness, and on the extent to which variations in the different indexes are correlated with those in the United Nations Human Development Index.