Popular wisdom and those who defend the death penalty say that the most heinous crimes should be more harshly punished. But, as Lincoln Caplan points out in a recent New York Times editorial, this is simply not the case. Death sentences are far more random than that, as shown by a study of murder cases in Connecticut from 1973 to 2007.

The Connecticut study, conducted by John Donohue, a Stanford law professor, completely dispels this erroneous reasoning. It analyzed all murder cases in Connecticut over a 34-year period and found that inmates on death row are indistinguishable from equally violent offenders who escape that penalty. It shows that the process in Connecticut—similar to those in other death-penalty states—is utterly arbitrary and discriminatory.

The study revealed that, far from being blind, Lady Justice metes out harsher punishments based not on the egregiousness of crimes but more often on race and geography. These findings echo those of sociologists who have studied the death penalty, such as Scott Philips (as “discovered” in Contexts, Winter 2011). Philips examined how the victim’s social status affected whether a defendant was sentenced to death in Texas from 1992 to 1999. Results showed that if the victim was “high status” (e.g. white, no criminal record, college educated), defendants were six times more likely to be sentenced to death. Black defendants, though, whose victims tend to be of lower social status, were still more likely than others to be sentenced to death.

In light of such evidence and with the death penalty on the decline (some states, such as Illinois in 2011, have abolished it altogether) Caplan argues it’s time for this “freakishly rare,” “capricious,” and “barbaric” form of punishment to go.