1.2 Time Series Trends

The 2003 census was carried out with methodology elements that differed from those of previous prison censuses (discussed more fully in Section 17). These changes inevitably introduce potential for discontinuity to many of the time series trends, as fluctuations on indicators may reflect changes to data collection methods rather than actual changes to the composition of the inmate population.

That said, growth of inmate numbers is perhaps the most accurately tracked variable, and this year’s total reflects a 47 percent growth since 1991 (not counting the home detainees). There has been a greater proportional increase in remand numbers over sentenced, and of course, since its introduction in 1999, a steady increase in home detention numbers.

Other noted trends include:

• Proportionally fewer inmates serving shorter (less than 12 months) sentences, and more inmates serving longer sentences.
• A gradual increase in the proportion of inmates who have more than 10 previous custodial sentences.
• A flattening of the inmate age profile, with fewer under thirty years, and more over thirty.
• A drop in the proportion of inmates having their first custodial sentence at less than 17 years old.