You are interested in comparing the earnings power of immigrants and native-born workers. You start collecting data in 1990 and survey 100 native workers and 100 immigrant workers who arrived in the US in the previous year. You re-survey these individuals in 2000. However, you are only able to find 60 members of the original immigrant sample in 2000. The remaining members of the immigrant community tell you that everyone dreams of returning home but only those who are successful in the US and are able to save up enough money for the journey. In 1990, you calculate a ratio of immigrant to native earnings of 0.75; by 2000, this ratio increased to 0.90.

Given this pattern of return migration, will the measured convergence between immigrants and natives be too high, too low, or unchanged? Why? Describe a method you could use to calculate the true degree of convergence between immigrants and natives with your available data.