Colin McCallin Nov. 8, 2018

A pattern has emerged: the use of statistics showing low ratios of criminal convictions to criminal reports to support criticism of the U.S. legal system. How, the implication goes, can we trust the law we live under when Department of Justice surveys indicate that only 22 out of 619 robbery reports end in felony convictions, and only 7 out of 310 sexual assault reports? (Source.) At first view those rates are unacceptable; they seem to announce disease within the country’s jurisprudence. But the issue gets more complicated the longer you think about it.

Why Is It Complicated?

First of all we have to consider that, if 22 out of 619 robbery reports end in felony convictions, a much larger number will end in lesser convictions in the process of plea-bargaining. In other cases perpetrators will have to abide by the terms of deferred judgments to avoid conviction altogether. These sorts of commutations are too useful not to have a place in our law; they give the justice system the flexibility it needs to prosecute a huge number of cases in limited time and with limited funding. To promote a vision of courts simply ignoring the majority of cases, declining again and again to hold criminals accountable, would be misleading.

Also, we have to determine whether the legal system of our country in fact operates unusually poorly or merely exhibits the shortcomings we should expect of any such system under similar conditions. To do this we might look at attrition rates—the rates at which criminal reports fail for any reason to generate convictions. Are these higher in the U.S. than elsewhere? And if so, at what stage in the legal process do reports here tend to fail? The evidence we have concerning these matters raises as many questions as it answers, unfortunately. But an analysis of sexual assault cases by Kathleen Daly and Brigitte Bouhours (found here), drawing on data from the early 2000s in five countries including the U.S., suggests at least that things here are not outrageously bad. Cases in South Australia in 2001-2005, for example, had the same attrition rate before trial (31%—whether by police inaction, plea-bargaining, court dismissal, or other causes) as U.S. cases from 2002 and a lesser chance at trial of ending in conviction (47% compared to 74%). Records in England and Wales, meanwhile, did not even include what would be necessary to calculate an attrition rate or an overall conviction rate. Canada’s trial rate was much higher than the U.S.’s; its overall conviction rate lower. In short, we are seeing a fragment of the whole picture, and the whole picture would probably show nothing conclusive either. However, the possibility presents itself that low ratios of convictions to reports are simply a feature of modern courts around the world and not necessarily cause for alarm.

Why Might Low Conviction Rates Be a Good Thing?

One could even argue it’s a good thing that we see so few criminal charges borne out. As we have said in posts before, a system of justice moving too fast or with too little impetus injures the public good. While there is certainly a point at which the virtue of moving slowly becomes a fault, and while attrition rates do deserve our attention, we have made it a principle of our law that, in a choice between letting guilty people go free and convicting innocents, we let guilty people go free. This means erring on the side of caution, holding prosecutors to high standards of proof, and, yes, resigning ourselves to the falling away of criminal reports that in other circumstances might end in convictions.