New Mexico has a large variation in income yet when mapped a certain way, the income variation gets erased. Using natural breaks in data, you perceive a large, evenly spaced set of data where each data grouping is even in the state.
However, this isn’t the most accurate way to apply this data to the map.
When you do the calculations and divide the data up based on the range, you end up with a lot less variation, and a large break in data. The lower income parts of the state are no longer divided into small groups, hiding the gap. Instead, the viewer will notice a large portion of the data falls below one limit.