The Howard League issued a press release a couple of weeks ago referring to a 'postcode lottery' in the sentencing practice of magistrates' courts. The press release refers to the extent to which immediate custody is used by magistrates in different areas. Overall, the figures appear to indicate a decline in the use of immediate custody between 2001 and 2011, albeit with increases in certain areas. The differences in the use of immediate custody in 2011 ranged from 1.5% of sentences imposed (in Warwickshire) to 6.5% (in Northamptonshire).
The 'postcode lottery' suggestion is eye-catching, and it captured media attention. The whole point of a lottery is its randomness, and while that may be appealing in the context of a raffle, we might troubled by it in other contexts such as the allocation of healthcare resources, or indeed in justice. To state the obvious, we don't like the idea of a lottery in sentencing because lotteries, unlike sentencing, have nothing to do with giving people what they deserve.
Below, I want to sketch out a few thoughts about the extent to which differentials in sentencing are undesirable in principle; thoughts which do not depend on the extent to which there actually is differential sentencing in practice. The data published in the Howard League press release are certainly interesting, although it is not clear whether and in what ways they factor in types of offence and offender history (this is not to suggest that they do not, merely that it is not clear on the face of the press release). Both offence type and offender history would be relevant to sentencing decisions generally, and specifically to the decision whether to impose immediate custody. Without more, it is difficult to determine the actual extent of differential sentencing practice.
Let's assume, for the sake of argument on the principles, two offences of criminal damage, identical in all material respects, one committed by Alf in Birmingham, the other by Charlie in Devon, with Alf and Charlie convicted by magistrates. Let's assume also that Alf and Charlie are themselves identical in all respects which might be material to sentencing. Let's then assume that Alf gets a harsher sentence. Does this point towards a sentencing postcode lottery? (If it helps, imagine a litany of Alf-Charlie type pairs undergoing corresponding experiences in Birmingham and Devon. Does that make the lottery argument any stronger?).
If our reaction to different sentences for the same offence is negative, this may be grounded in respect for the rule of law (and its emphasis on equal treatment) and the powerful pull of desert as the principal basis for sentencing. Alf and Charlie have done the same thing – they deserve the same sentence.
And yet different sentences may not be irrational. For one thing, there are purposes other than desert which inform sentencing, both as a matter of law (see section 142 of the Criminal Justice Act 2003), and in principle. So, let's say that there has been a spate of burglaries in Birmingham which has affected the quality of life not just of the primary victims, but also of members of the wider community who are now fearful of becoming victims. It may be defensible, for example on the basis of deterrence, to give Alf a more severe sentence than Charlie, if the community context in Devon is different.
Another potentially relevant factor is proportionality. Proportionality involves giving a sentence which is deserved by the offender in that it 'fits' the crime, both in relation to the crime's 'absolute' seriousness (this is sometimes referred to as cardinal proportionality) and its seriousness relative to other offences (referred to as ordinal proportionality). The key difficulty associated with proportionality is establishing, in practice 'how much' a crime is worth, in absolute and relative terms. One (non-trivial) question might be whether proportionality requires the same outcome across different contexts and communities. If not, then we might suggest that the context in Birmingham makes the offence 'worse' in Birmingham than in Devon, such that proportionality could call for a different sentence, independently of the non-desert based justifications referred to above.
And of course all of this is situated in magistrates' courts. Much is made of the role of magistrates' courts as local forums for the delivery of local justice. In that context, sentencing differentials might not merely to be expected, but desirable. The localism rationale has certainly come under significant pressure in recent years, with the closure of smaller courts, the consolidation of business in larger court centres, the growing role of the professional District Judge (Magistrates' Court) (DJ(MC) – one of my favourite legal system abbreviations), and shifts towards connecting magistrates to a particular area rather than just to a specific bench. If localism continues to have any purchase, one way in which we might see it at work is through sentencing practice, bounded by relevant guidelines while reflecting local priorities.
Where analysis of sentencing practice suggests that differentials do exist, the rule of law and the demand for like cases to be treated alike call for some kind of explanation. I am making the relatively modest claims here that plausible explanations may be available under certain conditions, and that those explanations may be based in a variety of principles which inform sentencing, and in the role of magistrates' courts.