Crime and punishment in contemporary South Africa III – From crime reporting to cases finalised

2 June 2020

INTRODUCTION

Once a crime is reported to the police, the next stage is detection and compilation of a docket for consideration by the National Prosecuting Authority. Data in this brief come from SAPS and NPA Annual Reports and are spotty, because there is no standard year to year format for information in either set of reports.

CRIME DETECTION

Table 1 sets out crime detection rates and total crimes detected.

Table 1: Crime detection rates

Just over a half of contact crime (murder, attempted murder, sexual offences, assault and robbery) are detected, compared with just over a quarter of other community reported crime. Police detected crime is ipso facto detected. This means that just under half of all crimes are detected, leading to about a million detected crimes a year.

THE NATIONAL PROSECUTING AUTHORITY

Between 2016/17 and 2018/19, the NPA received an annual average of 933, 933 new dockets. Virtually all of them were processed by the NPA, who then declined to prosecute in 491,986 cases (53% of the total), referred 323,105 (35% of the total) of the total back to the police for further investigation and initiated 116,306 prosecutions (12% of the total). The NPA also initiated 176 diversions (into a non-trial process) and 2,017 informal mediations.

Dockets referred back would either not be resubmitted, or resubmitted followed by a decision to prosecute or not to prosecute, or resubmitted followed a further referral back. Statistical evidence on the operation of the referral back process is not available, but it must over time unleash a further wave of prosecutions.

For 2016/17, we have data on the disposition of new dockets in regional and district courts.

Table 2: Disposition of new dockets 2016/17

Since regional courts deal with more serious cases, it is not surprising that they show a greater inclination to institute prosecutions and refer dockets back for further investigation and a lower inclination to decline to prosecute than district courts.

COURT ENROLLMENT STATISTICS

Court enrollment statistics illuminate little except churn. In 2014/15, 908,634 cases were enrolled in the courts and 435,175 were removed from the rolls. The distribution of the reasons for removal is set out in Table 3.

Table 3

That a case is removed from a court roll does not mean that it cannot be reinstated in the same or another court.

FINALISED CASES

NPA Annual Reports indicate that 481,442 cases were finalized per year on average between 2014/15 and 2018/19. Of these, 164, 881 (34% of the total) were settled using alternative dispute resolution. ADR is not used in high court cases, and settled only 8% of cases in the regional courts. It is much more widely used in district courts. Cases going to trial usually end in convictions.

Table 4: Cases finalised and outcomes

That finalized cases far exceed prosecutions initiated right away from new dockets suggests that dockets referred to the police for further investigation usually result in prosecution down the line.

FROM REPORTED CRIME TO ARREST: EVIDENCE FROM THE VOCS

Table 5 sets arrest rates (arrests divided by crimes reported to the police in 2014/15 and 2016/17 to 2018/19) for five categories of crime:

Table 5

These estimates compare with the ratio of finalized cases to crimes reported by the police of 23% derived from SAPS and NPA data. Not every appearance in court of an accused is preceded by an arrest. A simple summons to court can suffice.

The next brief will provide, as far as possible, estimates of the stocks and flows in the model set out in the first brief, leading to an assessment of the efficiency of the components of the criminal justice system.

By Charles Simkins, Head of Research, HSF