One of the questions to be answered in this discussion is whether there is an increase in the length of a sentence of imprisonment when it is wholly or partly suspended.34
This phenomenon, which may be entirely unconscious on the part of the sentencer, is referred to as sentence inflation, and occurs when the court increases the term of a sentence of imprisonment because it is to be suspended.35 When Tait examined
sentencing patterns in the Magistrates’ Court in Victoria, he found an inflation rate of about 50%: in other words, a four month unsuspended sentence appeared to correlate to a six month wholly suspended sentence,36 with other studies finding similar
results.37 In my interviews with judges and magistrates, the majority of respondents
34 As discussed in
[2.5.2], this approach has been endorsed by the Supreme Court in Canada, but does not seem to be formally permitted in Australia. See also [2.2.3] and [3.4.2], (Q13) for discussion.
35 Sentence inflation may be regarded as a form of net-widening. For discussion of net-widening, see
[1.5.4]. See also Legislative Council Standing Committee on Law and Justice, Parliament of New South Wales, Community Based Sentencing Options for Rural and Remote Areas and Disadvantaged Populations, Report 30, Sydney (2006), [5.86]-[5.93].
36 David Tait, 'The Invisible Sanction: Suspended Sentences in Victoria 1985-1991' (1995) 28
Australian and New Zealand Journal of Criminology 143, 153.
37 See Anthony Bottoms, 'The Suspended Sentence in England: 1967-1978' (1981) 21
British Journal of Criminology 1, 6-7 and Wendy Searle, Judy Paulin and Tony Waldegrave, 'The Activation of
171
suggested that it was inappropriate to extend the term of the sentence to reflect the fact of its suspension, but a sizeable minority indicated that they do in fact so from time to time, albeit only to the extent of ‘rounding up’ the sentence.38
One simple means of establishing the existence of any sentence inflation is to compare the lengths of sentences of imprisonment that are to be served immediately with those which are wholly suspended. If the latter are appreciably longer than the former, it would be reasonable to conclude that sentencers are ‘inflating’ the term of imprisonment for suspended sentences. As Table 4-3 above indicates, the median length of an unsuspended sentence in the Supreme Court is twice that of wholly suspended sentences, while the mean sentence is four times longer.39 The fact that
unsuspended sentences are considerably longer than wholly suspended sentences accordingly seems to fly in the face of any sentence inflation occurring. The figures for the Magistrates’ Court in Table 4-10 also contradict the suggestion of inflation, with the median sentence of unsuspended sentences 50% longer than for wholly suspended sentences (90 and 60 days respectively) and the mean sentence almost twice as long (141 and 80 days respectively).
The median length for partly suspended sentences in the Supreme Court was the same as for unsuspended sentences, while the mean sentence was shorter (15 vs 24 months). It is important to note, however, that the median and mean suspended portion of the sentence were six and seven months respectively, so that the portion of the partly suspended sentence to be served is shorter than the term of unsuspended sentences, again contradicting the suggestion of sentence inflation. In the Magistrates’ Court, the mean and median length of partly suspended sentences were much higher than for unsuspended sentences, perhaps suggesting some sentence inflation for partly suspended sentences. The median term of partly suspended sentences less the median suspended portion, was, however, at 90 days, exactly the same as the median term of unsuspended sentences, while the mean was less than the mean unsuspended sentence.
Another way of examining the existence of sentence inflation is by use of the so- called ‘gap and bulge test’ devised by Tait, which operates as follows:
If diversion occurs across the range of sentence lengths, the distributions should be fairly similar. If diversion is more frequent at the bottom end, then the distribution of suspended sentences would tend to be more concentrated near the bottom. If there is sentence inflation, a gap in immediate sentences at the bottom end may be matched by a bulge in suspended sentences further up the ladder.40
Suspended Sentences of Imprisonment in New Zealand' in Philip Spier (ed), Conviction and Sentencing of Offenders in New Zealand 1989 to 1997, Ministry of Justice, Wellington (1998) .
38 See
[3.4.2], (Q13).
39 This is consistent with earlier findings for the higher courts in South Australia and England where
unsuspended sentences in the higher courts were shorter than suspended sentences: see Catherine Dengate, The Use of Suspended Sentences in South Australia, Department of Correctional Services,
Adelaide (1978), 16 and Bottoms, n 37, 6-7.
40 Tait, n 36, 153. It should be noted that Tait also undertook multidimensional mapping. It was
172
Fig 4.5: Distribution of custodial sentences by length (months) – Supreme Court 11% 37% 24% 12% 12% 2% 1% 10% 13% 11% 34% 14% 13% 4% 1% 6% 15% 16% 10% 15% 10% 9% 11% 8% 0% 5% 10% 15% 20% 25% 30% 35% 40% Wholly suspended (n=246) 11% 37% 24% 12% 12% 2% 1% Partly suspended (n=105) 10% 13% 11% 34% 14% 13% 4% 1% Unsuspended (n=372) 6% 15% 16% 10% 15% 10% 9% 11% 8% 0>3 3>6 6>9 9>12 12>18 18>24 24>36 36>60 60+
Fig 4.6: Distribution of custodial sentences by length (months) – Magistrates’ Court 58% 33% 7% 1% 1% 35% 26% 9% 19% 39% 37% 11% 3% 7% 3% 11% 0% 10% 20% 30% 40% 50% 60% 70% Wholly suspended (n=1032) 58% 33% 7% 1% 1% Partly suspended (n=200) 11% 35% 26% 9% 19% Unsuspended (n=485) 39% 37% 11% 3% 7% 3% 0>3 3>6 6>9 9>12 12>18 18+
173
Figure 4.5 sets out the distribution of all custodial sentences in the Supreme Court in 2002-2004. As can be seen, there is a strong tendency to impose very short wholly suspended sentences, with almost half (48%) of all wholly suspended sentences imposed for less than six months, compared with 10% for partly suspended sentences and 21% for unsuspended sentences.41 Sentences of less than a year accounted for
84% of wholly suspended sentences, compared with only 47% of unsuspended sentences. Tait suggests that such a concentration represents more frequent diversion from the ‘bottom end’, which may include not only very short unsuspended sentences but also non-custodial orders. There is no significant gap in the use of unsuspended sentences at the lower end, such sentences in fact being fairly evenly distributed, nor is there any bulge of wholly suspended sentences at the higher end of the graph. In fact, the ‘bulge’ for wholly suspended sentences appears at the lower end, with sentences of three to less than six months (37% of wholly suspended sentences), while sentences of 12 months or longer account for less than 4% of such sentences.42
Figure 4.6 sets out the distribution for custodial sentences in the Magistrates’ Court. As can be seen, wholly suspended sentences are again heavily concentrated at the bottom end, with sentences of less than three months accounting for 58% of all wholly suspended sentences. Unsuspended sentences are, however, also concentrated at the lower end (39% of sentences under three months), suggesting there is no ‘gap’ being filled by wholly suspended sentences. Interestingly, sentences of 3>6 months are fairly evenly distributed across the three groups, accounting for 33% to 37% of all custodial sentences.
These data suggest that there is no evidence of sentence inflation occurring in either the Supreme or Magistrates’ Court, but the Supreme Court figures in particular may support a theory of net-widening (or penalty escalation), that is, that wholly suspended sentences are being used instead of non-custodial orders, and not as an alternative to an unsuspended sentence. The SAC inferred from data demonstrating that suspended sentences were concentrated amongst shorter sentences that ‘it appears that suspended sentences sometimes result in net-widening in Victoria’,43 but
also suggested that ‘[r]ather than showing net-widening, these trends could suggest that courts are quite properly suspending short prison sentences, in recognition of the
41 The use of short sentences generally appears to be quite high in Tasmania. Data for 2002 indicate
that sentences of six months or less accounted for 33% of unsuspended sentences imposed in the Supreme Court, compared with only 1% of such sentences in the District and Supreme Court in NSW: see Jason Keane, Patrizia Poletti and Hugh Donnelly, Common Offences and the Use of Imprisonment in the District and Supreme Courts in 2002, Sentencing Trends and Issues, No 30, Judicial
Commission of New South Wales, Sydney (2004).
42 It is interesting to note that the ‘bulge’ for partly suspended sentences appears much later, at the
12>18 month mark. Due to the comparatively small numbers of such sentences, however, and the lack of clear gap or bulge for unsuspended sentences, it is difficult to draw any clear conclusions about the interaction of partly suspended and unsuspended sentences in this regard.
43 SAC Final Report, n 30, [3.33]. The figures for the higher courts indicated that 39% of suspended
sentences were less than 12 months, compared with 12% of sentences to be served immediately: [3.35]. In the Magistrates’ Court, 78% of sentences were less than six months, compared to 68% of sentences to be served immediately, a finding which was said to ‘be consistent with some penalty escalation at the point of initial sentencing’: [3.34]
174
negative effects of such sentences on offenders’.44 The same observations apply to
the Tasmanian data: the data may suggest either net-widening or correct use of suspended sentences.