Some thoughts on multiple sentences and the totality principle: can we get it right?

Author:Manson, Allan


When it comes to numbers and criminal justice, there is no one better in Canada than Professor Anthony Doob. (1) Over the years, I have been fortunate to have engaged in hundreds of conversations with him about myriad criminal justice topics, issues, and data. One problem that we have discussed only tangentially is the perplexing question of imposing sentences for multiple offences in a way that makes both pragmatic and theoretical sense. As I recall, we briefly talked about this problem after a seminar given a few years ago by Professor Kevin Reitz at the Centre of Criminology in Toronto. That seminar resulted in a chapter in Roberts and von Hirsch "Previous Convictions at Sentencing" (Reitz 2010). As well, it generated an energetic conversation that ended with all of us shaking our heads.

Most Western jurisdictions have developed some version of totality which operates to ensure that an aggregate sentence for multiple convictions is not excessive. That is, a sentencing judge does not simply decide on the appropriate sentence for each offence and then add them up. Instead, she must stop to consider whether that aggregate would be "unduly harsh." If so, the individual sentences need to be adjusted in some way. In Canada, this is statutorily recognized in section 718.2(c) of the Code, which provides that "where consecutive sentences are imposed, the combined sentence should not be unduly long or harsh." Fantastic ... a principle that reflects restraint but leaves a series of questions. At the pragmatic end, what is the aggregate sentence that is not "unduly long or harsh"? And how does a judge configure the various individual sentences to reach that mythical and elusive total? Does the judge use a combination of concurrent and consecutive sentences or calculate sentence adjustments to reach the right result? These situations are common, so we know that judges somehow manage to answer them. We also know that the total is almost always less than the aggregate sum. Hence, the epithet of the "bulk discount." But can the "total" be explained in a way that conforms with sentencing theory and principles, especially in a jurisdiction which has embraced proportionality?

This is one of those situations in which an academic can legitimately say that there is good news and bad news. The good news is that, over the past few years, some of the best thinkers in the English-speaking sentencing world have turned their attention to this problem, including in alphabetical order, Andrew Ashworth (2010), Anthony Bottoms (1998), Nils Jareborg (1998), Kevin Reitz (2010), and Martin Wasik (2012). Even the UK Sentencing Council (2012) has addressed this issue. The bad news is nobody has cracked it.

Sentencing scholars have raised several important questions: How do judges determine and construct appropriate total sentences? Does the concept of overall culpability advance the debate? Can we learn anything from how a jurisdiction treats the seemingly related issue of previous convictions? Does the "one transaction" concept have any traction in determining when sentences should be concurrent? How important is the question of presentation? Can a guideline or range for the category of offence provide a cap for multiple convictions? Do we need better empirical data about what judges are doing to find better answers?

In this brief paper, I cannot address all of these questions. However, thinking about them has led me to conclude that the only approach to multiple sentencing which can even come close to being called principled would be one that distinguishes between different categories of multiple sentence situations. That is, different situations may require different approaches. Here, I argue that we can distinguish between high end cases and non-high end cases in ways that lead to appropriate and fair responses, or at least ones that we can abide without feeling that we have cast principles completely aside.


In a sentencing system that claims to be based on proportionality, Reitz (2010) has examined the relative treatment of the sum of serial sentences imposed over time compared to the total sentence imposed for the same offences by a single judge at one time. In most systems, some aggravating effect is given to previous convictions [the so-called recidivist premium] and some mitigation, sometimes substantial, can result from the total sentence for multiple offences [the so-called bulk discount]. As a result of this differential treatment, Reitz (2010) argues that proportionality is undermined. We can attempt a principled reason for this disparity, but it only makes sense in some cases. For example, one can claim that the recidivist is eventually more culpable and thus warrants an increased sentence. Conversely, one can also argue that the offender sentenced for multiple offences deserves a break because he has not yet had the personal benefit of the deterrent effect of prosecution and conviction. However, we are still faced with the fact that the sum of sentences for 10 break and enter offences imposed over a period of years will be substantially greater than the total sentence for the man who pleads guilty to the same 10 offences at one time. At the end of the day, Reitz (2010) offers, facetiously, a "wild card" approach, "forged of unmanageable complexity" and "lacking in compelling moral force" (154) as a very questionable answer to the proportionality problem. I view this as a concession that there is no all-purpose principle which will work but that some categorized approach may be possible.

Wasik (2012), in his recent thoughtful article, examines many facets of the totality conundrum and also addresses proportionality. He argues that, so long as the individual elements of a total sentence are properly presented--that is, offence by offence--a semblance of proportionality is preserved because each one will display its individual proportionality even if the total is configured into an aggregate that is less than the sum of the elements. Yet, this use of internal and external totals, as Wasik calls them, is really a way to redefine proportionality to suit the totality purpose. It does not really preserve proportionality. However, it leads his attention to presentation, which is really a concern that, in some cases, there is a need to accurately express a specific sentence in distinct terms. This affects how a total sentence should be configured. There have always been two general mechanical options for complying with the notion that the total sentence should not...

To continue reading