Skip to main content

Punishment, Proportionality, and Aggregation


Criminal theorists struggle to account for the “totality principle”—the idea that no matter how many small crimes you commit, your punishment should not exceed that for a more serious offense. Andrew Ashworth, for instance, argues that “overall proportionality” should be preserved but that this is a “pragmatic” solution. This paper argues that a retributivist can accept overall proportionality without abandoning her retributivism. I offer two lines of defense. The first is to show that the unit that we are aggregating may be more complex than first appears. The second relies on insights from the headaches versus lives literature, and how we should rescue one person dying before preventing one million people from headaches. The harms that we impose on others, and the harms that defendants suffer because of punishment, do not exist on one single scale. Hence, the true problem, I maintain, is not how culpability functions, but how our particular mode of punishment does. When we only have one primary mechanism to punish (particularly once we start aggregating)—put people in prison—then aggregating offenses appears as though it must quickly lead to numerous acts of littering being punished more than arson. With a more nuanced punitive response, we can punish minor offenses every time without surpassing the kinds of punishment we impose on more significant offenses.

This is a preview of subscription content, access via your institution.


  1. 1.

    Shelly Kagan, “Do I Make a Difference?”, Philosophy and Public Affairs 30 (2011): 105–141; Larry Alexander and Kimberly Kessler Ferzan, Reflections on Crime and Culpability: Problems and Puzzles (Cambridge, Cambridge University Press 2018), pp. 56–59.

  2. 2.

    Jeff McMahan, “Defence Against Parfit’s Torturers”, in Jeff McMahan, Tim Campbell, James Goodrich, and Ketan Ramakrishnan, eds., Principles and Persons: The Legacy of Derek Parfit (Oxford: Oxford University Press, forthcoming).

  3. 3.

    Alistair Norcross, “Comparing Claims: Headaches and Human Lives,” Philosophy and Public Affairs 26 (1997): 135–167; Alex Voorhoeve, “How Should We Aggregate Competing Claims?”, Ethics 125 (2014): 64–87.

  4. 4.

    D.A. Thomas, Principles of Sentencing: The Sentencing Policy of the Court of Appeal Criminal Division 2d ed. (Heinemann: London 1970, 1979), p. 56. Elsewhere, I have argued for noncomparative desert. Alexander and Ferzan, Reflections on Crime and Culpability, ch. 10. That is the idea that if A, B, and C all commit the same crime, it is not a failure of retributivism if A is punished fully but B and C are not (even for potentially problematic reasons). But this problem is different. Here, the question is how A’s commission of several offenses should be punished in comparison to A’s hypothetical commission of another offense. That is, we need a scale of which offenses are more serious than each other, even if we think that no individual would have a desert-based complaint if someone else did not get what she deserved.

  5. 5.

    Jesper Ryberg, “Retributivism, Multiple Offending, and Overall Proportionality”, in Jesper Ryberg, Julian Roberts, and Jan W. De Keijser, eds., Sentencing Multiple Crimes (Oxford: Oxford University Press 2018), p. 15.

  6. 6.

    Michael Tonry, “Solving the Multiple-Offense Paradox”, in Jesper Ryberg, Julian Roberts, and Jan W. De Keijser, eds., Sentencing Multiple Crimes (Oxford: Oxford University Press 2018), p. 243.

  7. 7.

    18 U.S.C. § 703 (2012); 50 C.F.R. § 10.13; 16 U.S.C. § 707 (2012).

  8. 8.

    Andrew Ashworth, Sentencing and Criminal Justice, 4th ed. (Cambridge: Cambridge University Press 2005), p. 250.

  9. 9.

    Id. at 254.

  10. 10.

    Justifying a sentence in practice implicates myriad concerns. In Virginia, if a twenty-five-old woman is being sentenced for fifteen acts of shoplifting, her age and gender have a more significant impact on the unavailability of alternative sentencing than does the number of her criminal acts. See Virginia Sentencing Commission worksheet for larceny, available at (With respect to this offense, being a woman increases one’s likelihood of reoffending.).

  11. 11.

    Jesper Ryberg, “Retributivism, Multiple Offending, and Overall Proportionality”, pp. 13–14.

  12. 12.

    Tonry, “Solving the Multiple Offense Paradox”.

  13. 13.

    See also Jacob Bronsther, “Long-Term Incarceration and the Moral Limits of Punishment”, Cardozo Law Review 41 (2020): 2369–2433.

  14. 14.

    Victor Tadros suggested to me that prioritarianism can likewise yield a point at which we should temper or cease punishment.

  15. 15.

    Tonry, “Solving the Multiple Offense Paradox”, p. 254.

  16. 16.

    Ryberg, “Retributivism, Multiple Offending, and Overall Proportionality”, p. 14.

  17. 17.

    Id. at 24–27.

  18. 18.

    Victor Tadros, “Localized Restricted Aggregation”, in David Sobel, Peter Vallentyne, and Steven Wall, eds., Oxford Studies in Political Philosophy, Vol. 5, (Oxford: Oxford University Press 2019), pp. 175–76.

  19. 19.

    Dale Dorsey, “Headaches, Lives, and Value”, Utilitas 21 (2009): 36–58.

  20. 20.

    One of the first theorists to discuss the puzzle, Leo Katz, argued that when crimes occur together, we don’t individuate the acts. Hence, five shots fired in quick successions looks more like one attempt than five. Leo Katz, “Before and After: Temporal Anomalies in Legal Doctrine”, University of Pennsylvania Law Review 151 (2003): 863–885. In contrast, Larry Alexander and I partially justified these intuitions by the way that the culpable choices are made—crimes that occur all at once do not evince the same qualitative decision making as those decisions that occur over time. Larry Alexander and Kimberly Kessler Ferzan, with Stephen J. Morse, Crime and Culpability: A Theory of Criminal Law (Cambridge: Cambridge University Press 2009), pp. 254–57; Alexander and Ferzan, Reflections on Crime and Culpabilty, pp. 134–35.

  21. 21.

    Commonwealth v. Donovan, 478 N.E.2d 727 (Mass. 1985).

  22. 22.

    Id. at 735.

  23. 23.

    Michael S. Moore, Act and Crime: The Philosophy of Action and its Implications for Criminal Law (Oxford: Oxford University Press 1993), pp. 385–86.

  24. 24.

    Alexander and Ferzan, Crime and Culpability, pp. 250–54.

  25. 25.

    United States Sentencing Guidelines, ch. 3, pt. D, “Introductory Commentary.”.

  26. 26.

    Alexander and Ferzan, Reflections on Crime and Culpability, p. 137 n.6; cf. Christopher Bennett, “Retributivism and Totality: Can Bulk Discounts for Multiple Offending Fit the Crime?” in Jesper Ryberg, Julian Roberts, and Jan W. De Keijser eds., Sentencing Multiple Crimes (Oxford: Oxford University Press 2018), p. 66 (“moralistic retributivism opens up the possibility that such discretion is necessary to correct for distortions introduced by the form of legality”).

  27. 27.

    Consider California’s struggles with its double jeopardy test in this regard. Compare People v. Sexton, 37 Cal.App.5th 496, 507–08 (2019) (“[i]f one’s intent is to inflict harm on an individual who manages to free themself, then a subsidiary goal of the infliction of harm is to regain control of the individual. In such a case, however, the overarching objective throughout remains inflicting harm on the victim, and any subsidiary objectives—such as regaining physical control or shielding the harm from one’s children—necessarily depend on the overarching objective and are therefore not independent.”) with People v. Vasquez, 44 Cal.App.5th 732 (Ct. App. 2020) (finding mayhem to be independent of murder because the biting of the victim’s fingers in the course of the killing was retaliatory and not directed at the killing).

  28. 28.

    Accord Mitchell N. Berman, “Proportionality, Constraint, and Culpability” (this issue) (suggesting “culpability in relation to wrongdoing”).

  29. 29.

    Gabe Mendlow’s question—what are we punishing for—demonstrates precisely how difficult this question becomes. Gabriel S. Mendlow, “The Elusive Object of Punishment”, Legal Theory 25 (2019): 105–31.

  30. 30.

    See generally Alexander and Ferzan, Crime and Culpability.

  31. 31.

    Alexander and Ferzan, Reflections on Crime and Culpability, p. 136.

  32. 32.

    Of course, the case is entirely different if multiple other people are risked each time the defendant aims at her goal. If Clod’s BB gun risks hits Sarah and Joe who are studying on the quad, and then Liz and Lauren who are leaving for class, and so forth, then Clod’s wrongful choices should fully aggregate.

  33. 33.

    Alexander and Ferzan, Reflections on Crime and Culpability, ch. 2.

  34. 34.

    Larry Temkin calls this the “Disperse Additional Burdens View.” Larry S. Temkin, Rethinking the Good: Moral Ideals and the Nature of Practical Reasoning (Oxford: Oxford University Press 2012), pp. 67–68.

  35. 35.

    Interesting the United States Sentencing Guidelines do simply aggregate and then add additional penalties for the number of victims. Hack receives a base offense level of seven for theft, an addition of 30 for stealing over $550,000,000, and an addition of 2 for having ten or more victims. § 2B1.1.

  36. 36.

    There are questions here about one rich person versus lots of poor people. For the argument that this distributive principle is independent of maximin and similar concerns, see Temkin, Rethinking the Good, p. 71.

  37. 37.

    For further exploration of this voluntary failure by my husband, see Kimberly Kessler Ferzan, “On Jeffrie Murphy’s ‘Involuntary Acts and Criminal Liability’”, Ethics 125 (2015): 1119–1122.

  38. 38.

    The difficulty with any example outside of the criminal law is that a response might be appropriate even if it is not “deserved.” If I found that my nerves are constantly shot, my scholarship is suffering, and I am psychologically on edge all the time because of the sneezing, I could divorce my husband. But that would not be because he deserved it as a response to his “wrongdoing” but because of my right to exit the relationship. I thank Adam Kolber and Jae Lee for conversation on this point.

  39. 39.

    Some readers have objected to this oversimplification. To be sure, we have alternatives to pursuing criminal penalties, from drug courts to deferring prosecution. And, we have options such as probation or home detention. But the primary mode of punishment, particularly when we think someone deserves lots of it, is to put the person in prison. Incarceration is our primary means of stigmatizing offenders.

  40. 40.

    In his recent article discussing low level offenders, Doug Husak suggests that perhaps they ought to be treated like those who cut in line, and subject only to social sanction, because using the criminal law is disproportionate to their offense. Douglas Husak, “Criminal Law at the Margins”, Criminal Law and Philosophy 14 (2020): 381–393. He believes the current managerial model for such offenders, which tracks first time offenders but only punishes repeat offenders, is aimed more at deterrence. I believe, although I am not completely certain, that Husak and I agree about these cases. If low level offenses should be subject only to social sanction, then the argument that in aggregation, they should receive prison time is likely motivated not by retributive, but by deterrence rationales.

    In conversation, Husak has also suggested to me that my problem will disappear if we decriminalize a number of offenses. To be sure, we do not intuit the same sort of caps at the high end of the spectrum–many arsons or rapes may be as serious as a murder. Nevertheless, I strongly suspect that there will be offenses that we believe warrant some punitive sanction, but not a significant one. Husak and I may disagree on where to draw the line on this. Moreover, some of the arguments against criminalization are because we do not think the acts are culpable, wrong, and/or deserving of a state sanction, and some of the arguments against criminalization are because we live in a nonideal world where the state does a poor job of administering low level sanctions. But if my problem only disappears because we are eliminating the criminal sanction for pragmatic reasons, then it hardly offers a principled theoretical account for retributivists who live in more ideal times.

  41. 41.

    Even when we have lesser punishments, when we aggregate them, we tend to escalate to incarceration.

  42. 42.

    Dale Dorsey, “Headaches, Lives, and Value”; see also Temkin, Rethinking the Good, pp. 54–56 (analogizing Jim Griffin’s view to a position where no matter how many C papers one writes (at 2 points a piece), one will never deserve an A (at 4)).

  43. 43.

    Larry Alexander suggested to me that the aggregation question can be solved by making a Taurek-like move, but I think this is insufficient to meet the challenge. As Larry Temkin has noted, these aggregation questions can arise also intrapersonally such that one would rather suffer a minor annoyance for a very long time period than a significant harm for a much shorter time.

  44. 44.

    This leads to a different aspect of the project Adam Kolber has pursued. Kolber has argued that given experiential differences, a given punishment may be experienced differently. As we see, the same point may apply to victim, an implication Kolber recognizes. Adam J. Kolber, “The Experiential Future of the Law”, Emory Law Journal 60 (2011), pp. 622–47.

  45. 45.

    Youngjae Lee, “Multiple Offenders and the Question of Desert,” in Jesper Ryberg, Julian Roberts, and Jan W. De Keijser eds., Sentencing Multiple Crimes (Oxford: Oxford University Press 2018), p. 115.

  46. 46.

    Id. at 122.

  47. 47.


  48. 48.

    Id. at 125.


For written comments on this paper, I thank Larry Alexander, Mitch Berman, Adam Kolber, and Michael Moore. I thank Genevieve Tung at Penn and Michael Klepper at UVA for research assistance. This paper substantially benefitted from presentation at the Proportionality in the Criminal Law symposium sponsored by Georgetown Institute for the Study of Markets and Ethics and the Political Turn(s) in Criminal Law online workshop.

Author information



Corresponding author

Correspondence to Kimberly Kessler Ferzan.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Ferzan, K.K. Punishment, Proportionality, and Aggregation. Criminal Law, Philosophy 15, 481–494 (2021).

Download citation


  • Overall proportionality
  • Multiple offenses
  • Totality principle
  • Retributivism
  • Proportionality
  • Aggregation