This is a regular chart used by Mark Perry, updated for the most recent data. It is one of my favorites.
Mark has a lot of great analysis of this data, too long to excerpt, so I refer you back to the link to see his comments. But I want to add one of my own.
Let me restate the numbers in this chart one other way to give a pretty stark view of utility of raising the minimum wage as a poverty program. Let's look at the first lowest vs. the second lowest quintile.
Raising current workers in the poorest quintile to second quintile earnings ($28,970 to $35,858) would increase the first quintile's average income by $2,962. On the other hand, keeping first income wages flat but raising first quintile household's average earners to second quintile number of earners (0.43 to 0.91) increases the first quintile's average income by $13,905.
By this analysis, increasing first quintile hours worked (and % employment) has 4.7 times more leverage than fiddling with wage rates.
So now consider raising the minimum wage, which will raise some wages but reduce unskilled employment -- it is exactly the wrong thing to do, even before we consider that the majority of minimum wage earners who benefit from such an increase and keep their jobs are not in the lowest quintile but are 2nd and third earners in rich and middle class households.