The demand for sentence discounts: some empirical evidence

Sentence discounts are now routinely used by Australian courts to encourage guilty pleas. In this article, the authors examine three populations of not on bail defendants who went to trial and were convicted in New South Wales in 2004 for the offences of aggravated robbery, burglary and murder respectively, with the objective of estimating the percentage reduction in sentence quantum that would have induced them to plead guilty. Since conviction (acquittal) probabilities following a trial are likely to be uniformly distributed between 0 and 1, the expected mean probability of conviction (acquittal) for a defendant pleading not guilty was 0.5. The average reductions in the prison sentence corresponding to this probability were: 21%, 23% and 27% respectively. The maximum (minimum) values were: 39% (1.3%), 40% (1.9%) and 39% (1.5%). This range of values reflects the wide dispersion of actual prison sentences handed down by the courts. The distribution of actual sentence discounts offered by the judges in exchange for a guilty plea is not available, consequently the authors cannot comment on why these defendants chose a trial.