Testing the Divergent Mobility and the

This paper examines changes in individual earnings during positive and negative growth periods in three Latin American economies: Argentina, Mexico, and Venezuela. We ask whether those individuals who start in the best economic position to begin with are those who experience the largest earnings gains or the smallest earnings losses; this is the “divergent mobility ” hypothesis. We also compare periods of positive economic growth with those of negative economic growth, asking whether those groups of individuals tha t experience large positive earnings gains when the economy is growing are the same as those that experience large losses when there is a recession; this is the “symmetry of mobility ” hypothesis. Unconditional and conditional tests of these hypotheses are performed. Empirically, the divergent mobility hypothesis and the symmetry of mobility hypotheses are decisively rejected. Rather, in these three countries, the conditional mobility patterns are found to be overwhelmingly structural.