Mobility is typically defined in terms of moving between income quintiles in the population -- relative, not absolute. If the income range of the quintiles is very compressed (i.e. everyone's income is close to the median) then a small change in absolute income can lead to a large change in income percentile. Compared to places with higher income variance, this tends to overweight the "mobility" of places with low income variance. At the very least it produces a misleading result relative to what most people would define as meaningful mobility.

The same problem applies to comparing mobility between countries. If you have someone with the same (PPP) wage in the lower quintiles in Denmark and the USA, a $20k increase in absolute income in Denmark will cause a big jump in the quintile (and therefore "mobile") and a $40k increase in absolute income in the USA won't even though it is twice the real gain! Simply due to the fact that Denmark has low income variance and USA has high income variance. Income mobility measures that don't adjust for variance tend to produce misleading results when there are significant differences in income variance in the sample population.

If Utah has a relatively low income variance (not a lot of poor, not a lot of rich) then it will lead to percentile based mobility measures that are less meaningful than States with higher income variance.