A survey researcher wants to estimate the mean income of a population. They calculate a standard error of $SE = 500$. If they want to double the precision (i.e., halve the standard error), how much must they increase the sample size $n$?