4% vs 1% is not huge. Strictly using welfare equates to 43K more blacks on welfare. There is no way to break it down as to how long each group is on welfare. Welfare is dynamic as people are leaving and entering all the time. There are some you make it a career choice but that is true for all races.
Most people make it out that 50% of the blacks in this country are on welfare. Even at 4%, blacks mirror the total amount of welfare recipients which is 4.1%.
Here is how it actually breaks down...
If 38.8% of the entire non-hispanic white population were on welfare, the roles would be much larger than 48M to cover that since 0.388*196,817,552 = 76,365,210 based on 2010 numbers.
When looking at the numbers the way I suggest (48M welfare recipients divvied up in percentages per capita), then the numbers work.
Blacks would be: 19,104,000 (39.8% of 48M) The black population of the US per the 2010 census is 37,685,848. That works out to ~50% of the entire black population on welfare (more likely food stamps) based on your numbers.
Whites would be: 18,624,000 (38.8% of 48M) The white non-Hispanic population per the 2010 census is 196,817,552. This works out to 9.463% of the entire white population receiving said benefits.
Hispanics would be: 7,536,000 (15.7% of 48M) The Hispanic population per the 2010 census is 50,477,594. This works out to 14.93% of the entire Hispanic population receiving said benefits.
Like it or not, this IS how the math works based on the numbers you started out with and the real census data broken down in the demographic categories you highlighted.
I get that the numbers aren't static, I get that there are legitimate reasons to use the benefits temporarily. Given the random normal distribution and opportunity for abuse, I'd say that the group most likely to abuse the system is the group that is already taking advantage of the system in much greater percentages of their demographic to begin with. There is no stigma associated with the behavior that has an impact within the ghetto culture anymore.