Imagine the US in 2026 and imagine that the immigration reform opened a path to citizenship for the majority of immigrants.
Once they become citizens, on average Hispanics born outside the US experiment a one level increase in education and a marginal increase in the probability of reaching a higher level of income; all this based on an self-generated econometric model using more than 3 million census observations for Hispanics– details are on the methodology note in the end.
For the Mexicans, citizenship means that additional immigrants would be more probable of finishing high school (8.5 percent) and college (9.3 percent). Likewise, the number of Mexican immigrants without education would be reduced by 2.8 percent.
As for Hispanics (excluding Mexicans), citizenship would also mean better educational achievements. The big difference is that citizenship would only have a positive affect in college attendance. Hispanics (excluding Mexicans) would be 19.02 percent more probable of obtaining a BA.
Regarding income, the outcome is just as positive. First, we would observe a significant reduction in the number of Hispanics without an annual income: for the Mexicans the reduction would be of 5.93 percent, while for the Hispanics (excluding Mexicans) it would be of 3.92 percent.
Second, the greatest impact would be felt among those in the $20,000 to $24,999 annual income bracket. Mexicans would be 2.97 percent more probable of belonging to this income bracket and Hispanics (excluding Mexicans) 1.46 percent.
The large difference between owning or renting a house also corroborates how exceptionally positive citizenship would be for Hispanics. The probability of owning a house would increase 16.51 percent for the Mexicans and 14.46 percent for the Hispanics (excluding Mexicans) in the US.
The one issue in which we find divergent results is employment. For the Mexicans, citizenship would not increase significantly their probability of being employed, but for Hispanics (excluding Mexicans) it would. Specifically, the probability would increase in 3.08 percent.
Because income and education are categorical (discrete) variables, we used an Ordinal Logistic Regression. Also, because employment and house ownership are categorical (discrete) variables of two values, we used a Logistic Regression. Logistic Regressions provide certain intuition on the coefficients and significance levels; nonetheless they do not permit us to formulate robust conclusions on the issue investigated. Therefore, we utilized probabilistic tests (Clarify: prchange, prvalue and prtab), to interpret the statistical results. The independent variables include: citizenship (Mexican and Hispanics – excluding Mexicans), poverty, total income (when the dependent variable is education, employment or house ownership), education (when the dependent variable is total income, employment and house ownership), age and sex.
Click on the following link to access the data on citizenship: