Interpreting U.S. criminal case prosecution and conviction case-share statistics requires careful consideration of base population of cases and disposition of cases through means other than case dismissal, plea-bargained conviction, or trial. The base population of cases in local jurisdictions depends strongly on charging practices and case screening. In some U.S. local jurisdictions, charges are automatically filed upon arrest of a suspect. Hence prosecutors make initial decisions about whether to prosecute after charges are filed. In other U.S. local jurisdictions, prosecutors make initial decisions about whether to prosecute before charges are filed. In some jurisdictions, police decide whether arrested persons’ cases are referred for prosecution. Case dismissal rates are sensitive to the extent and timing of case screening relative to the points of case counting. That effect is starkly apparent in dismissal rates in cases of intimate-partner violence: 21% in jurisdictions where prosecutors screen cases prior to the defendant’s first court appearance vs. 49% in jurisdictions where prosecutors screen cases after the defendant’s first appearance.^ Such differences in dismissal rates cannot be reasonably understood without consideration of charging practices and case screening.

The U.S. lacks a national compilation of local jurisdictions’ charging practices. The Court Statistics Project of the National Center for State Courts tracks felony caseloads in state trial courts. These statistics indicate only felony cases filed in state courts of general jurisdiction. However, the caseload statistics include data by local court jurisdictions on the unit of counting cases in that court jurisdiction.^ Case counting practices vary in the procedural point of counting. The different points of case counting are citation, warrant, summons, complaint, accusation, information, advisement, first appearance, arraignment, indictment, and docket. Some courts have multiple points of case counting. Case counting also varies by the allocation of multiple defendants, multiple charges, and multiple incidents to cases. While charging practices vary greatly by local jurisdictions, they are relatively stable over time. For comparisons over time in a fixed set of jurisdictions, differences in charging practices affect aggregate statistics through changes in the distribution of case activity across jurisdictions over time.

Under-appreciated means of disposing criminal cases are numerically much more significance than results of criminal trials. About two-thirds of all felony cases are resolved through plea-bargained convictions. About of quarter of felony cases are dismissed either by prosecutors through initial screenings after case filing or through subsequent decisions (nolle prosequi) or by the judge at some point in case processing. The criminal justice system thus works primarily through plea bargaining and dropping charges (case dismissal). Only about 4% of U.S. felony cases go to trial.^ Other dispositions of cases account for 8% of cases, a share twice as large as the share of felony cases that go to trial. Other dispositions of various types are variously named prosecutorial diversion, deferred prosecution, conditional discharge, deferred adjudication, continuance, and adjournment in contemplation of dismissal. Study of cases in three upstate cities in New York State suggests that more than a third of misdemeanor domestic-violence cases are disposed through adjournment in contemplation of dismissal.^ Restraining orders, which commonly are imposed after arrest for domestic violence, are probably imposed along with adjournments in contemplation of dismissal. The U.S. Bureau of Justice Statistic’s State Court Processing Statistics calls these case dispositions “special assignments.”^ These special assignments in total are less special than criminal trials. They deserve more attention in study of court statistics and in analysis of the functioning of criminal justice systems.