Wonderland Sentencing: Therapeutic Perspectives on Pretrial Provisional Sentences

*Robert Kaufman ’57 Professor of Law, Brooklyn Law School. For helpful comments, I thank Kolette Bodenmiller, John Meixner, and Kate Skolnick, as well as participants at the 2024 CrimFest Conference at Cardozo School of Law and at a 2026 faculty workshop at Albany Law School.Show More

In Alice’s Adventures in Wonderland, the Knave of Hearts is accused of stealing tarts. The Queen of Hearts insists that they should “[s]entence [the Knave] first” and hear the verdict afterwards. Alice derides “[t]he idea of having the sentence first” because, of course, nothing would be more absurd. In this Essay, I defend that absurd idea, at least provisionally.

While we certainly should not punish someone prior to conviction, we should carefully consider the benefits of “wonderland sentencing,” in which those accused of crimes are given provisional sentences prior to the decision to proceed to trial. Doing so enables defendants to make better informed, fairer decisions about whether and under what circumstances to plead guilty. The proposal supports goals of therapeutic jurisprudence by (1) reducing prosecutorial leverage that many legal scholars consider both excessive and a cause of disproportional sentencing; (2) allowing defendants to be and feel heard by an impartial judge; and (3) providing defendants better information about their future sentencing liability, not only as it affects plea decisions, but also as it affects the very stressful period before their cases are resolved.

We give defendants very limited information prospectively about punishments for various crimes. Wonderland sentences would help correct our general failure to provide advance notice of the consequences of criminal conduct. While wonderland sentencing is just one of many therapeutic policy changes we could implement, given that approximately ninety-five percent of non-dismissed cases are resolved by plea bargains, wonderland sentencing has the potential to do a lot of good for a lot of people.

Introduction

In Alice’s Adventures in Wonderland, the Knave of Hearts is accused of stealing tarts.¹ 1.Lewis Carroll, Alice’s Adventures in Wonderland 165–66 (London, MacMillan & Co. 1866).Show More The Queen of Hearts insists that they should “[s]entence [the Knave] first” and hear the verdict afterwards.² 2.Id. at 187.Show More Alice derides “[t]he idea of having the sentence first” in an exchange scholars frequently cite to illustrate an extraordinarily absurd attempt at justice.³ 3.Id.; see, e.g., Charles H. Gustafson, Judicial Review of Jeopardy Tax Collection: Sentence First, Verdict Afterwards, 26 Case W. Rsrv. L. Rev. 315, 315–16, 315 n.* (1976); Robert S. Pasley, Sentence First—Verdict Afterwards: Dishonorable Discharges Without Trial by Court-Martial?, 41 Corn. L.Q. 545, 545–47 (1956).Show More In this Essay, I defend it (at least provisionally).

Of course, we shouldn’t punish anyone prior to conviction.⁴ 4.But cf. Saul Smilansky,Determinism and Prepunishment: The Radical Nature of Compatibilism, 67 Analysis 347, 347–48 (2007) (arguing that the compatibilist view of free will held by many philosophers leads to the conclusion that we ought to pre-punish).Show More But there is much to be said for what I call “wonderland sentencing,” in which those accused of crimes are told, prior to deciding whether to plead guilty, what their sentence would likely be if convicted at trial. Doing so could enable defendants to make better informed, fairer, and more socially advantageous decisions about whether and under what circumstances to plead guilty.

There are many possible variations on wonderland sentencing. My goal isn’t to settle the details of one particular approach. Rather, I lay out the benefits of wonderland sentencing in general terms and then consider several ways to fill in the details, keeping in mind that any particular approach will have to coordinate with the varied sentencing laws in existing jurisdictions.

The impetus for wonderland sentencing comes from what is typically the most important decision defendants must make about their cases: whether to plead guilty in exchange for a lighter sentence. The decision can be extraordinarily difficult because defendants have limited relevant information. For example, they don’t know exactly what evidence the prosecution would introduce or how convincing it would be to potential jurors. They also don’t know how well their own witnesses would testify on their behalf.

Most importantly, however, even if defendants knew all the information that would be revealed at trial, they would still lack critical information about sentence severity. In many jurisdictions, judges have wide sentencing discretion, leaving defendants uncertain about how long their sentences are likely to be.⁵ 5.See, e.g., United States v. Jackson, 835 F.2d 1195, 1197 (7th Cir. 1987) (recognizing the broad sentencing discretion typically afforded to judges in non-guideline jurisdictions).Show More In others, there may be sentencing guidelines that limit judicial discretion.⁶ 6.See, e.g., 18 U.S.C. § 3553(b)(1) (advising federal courts to ordinarily impose a sentence within the guideline range); Minn. Sent’g Guidelines & Comment. § 1.A (Minn. Sent’g Guidelines Comm’n 2024).Show More But even in those jurisdictions, judges still have significant sentencing discretion and are often called upon to interpret sentencing guidelines or the application of maxima and minima in ways that are hard for defense counsel to predict.⁷ 7.Cf. United States v. Booker, 543 U.S. 220, 245, 250–51 (2005) (holding the federal sentencing guidelines to be advisory and emphasizing the importance of judicial discretion).Show More

Even when extraordinarily harsh sentences are unlikely, they can still weigh on a defendant’s decision-making.⁸ 8.There is evidence in many domains, including mock criminal justice contexts, that vividly presented numbers can anchor people’s estimates and preferences. See, e.g., Stephanie A. Cardenas, Charged Up and Anchored Down: A Test of Two Pathways to Judgmental and Decisional Anchoring Biases in Plea Negotiations, 29 Psych. Pub. Pol’y & L. 435, 435–36 (2023) (discussing how the willingness to plead guilty or recommend guilty pleas of mock defendants and defense attorneys can be anchored by harsh potential sentences); Roland Imhoff & Christoph Nickolaus, Combined Anchoring: Prosecution and Defense Claims as Sequential Anchors in the Courtroom, 26 Legal & Criminological Psych. 215, 215, 223–24 (2021) (finding that law student experimental subjects were anchored in their mock sentencing determinations by the recommendations of both the prosecution and the defense).Show More For example, in 2024, Maine doctor Merideth Norris was found guilty of overprescribing painkillers.⁹ 9.Shawn P. Sullivan, Kennebunk Doctor Convicted on Multiple Counts of Illegally Distributing Opioids, Portsmouth Herald (June 24, 2024, at 11:23 ET), https://www.seacoasto‌nline.com/story/news/local/york-star/2024/06/24/kennebunk-maine-doctor-merideth-norris-c‌onvicted-of-unlawfully-distributing-opioids/74191342007/ [https://perma.cc/K6LG-CPLC].Show More She was convicted on fifteen counts of unlawfully distributing controlled substances and faced a maximum of twenty years’ incarceration on each count.¹⁰ 10.Id.Show More Predictably, Dr. Norris was said to face up to three hundred years’ incarceration.¹¹ 11.Id.Show More When she was ultimately sentenced, however, she received no prison time.¹² 12.Emily Allen, Kennebunk Doctor Will Not Serve Prison Time for Overprescribing Opioids, Portland Press Herald (May 16, 2025), https://www.pressherald.com/2025/05/15/ken‌nebunkport-doctor-will-not-serve-jail-time-for-overprescribing-opioids [https://perma.cc/7G‌4M-4C9R].Show More

In reality, Dr. Norris was never going to spend three hundred years incarcerated, and almost certainly, her attorney gave her more realistic estimates. But defendants can be traumatized simply by hearing that they face such extreme sentences. Federal judges are required to notify defendants of the maximum sentence they face under the law, but not the sentence the judge expects to assign or the sentence she considers the highest she would be likely to assign if the case were to proceed to a verdict at trial.¹³ 13.See Fed. R. Crim. P. 11(b)(1)(H).Show More Knowledge of theoretical sentencing maxima may be particularly influential and misleading when clients distrust their often court-appointed attorneys. Moreover, even with attorney estimates, defendants may face wide plausible sentencing ranges.

Prosecutors pressure defendants to plead guilty not only by threatening long sentences but also by making strategic use of uncertainty. Rational choice theory suggests that the uncertainty of sentences will drive risk-averse defendants to plead guilty to avoid trial. The more risk-averse the defendant, the more we expect risk to drive guilty pleas. Moreover, given that risk tolerance is associated with criminal and other antisocial behaviors,¹⁴ 14.SeeTim Friehe & Hannah Schildberg-Hörisch, Predicting Norm Enforcement: The Individual and Joint Predictive Power of Economic Preferences, Personality, and Self-Control, 45 Eur. J.L. & Econ. 127, 137–38 (2018); see also Thomas Epper et al., Preferences Predict Who Commits Crime Among Young Men, PNAS, Jan. 31, 2022, at 1, 5 (finding a statistically significant positive correlation between risk tolerance and criminality for property crimes, though not for violent, drug, or sexual crimes).Show More greater uncertainty at sentencing is likely to put more pressure on prosocial defendants to plead guilty than antisocial defendants—precisely the opposite of the direction we seek. Since prosocial defendants are less likely to be guilty than antisocial defendants as a general rule, uncertainty at sentencing has the perverse effect of pushing innocent defendants to plead guilty with greater force than it pushes guilty defendants to do the same.¹⁵ 15.See Robert E. Scott & William J. Stuntz, Plea Bargaining as Contract, 101 Yale L.J. 1909, 1948–49 (1992).Show More Wonderland sentencing provides a way to make sentences fairer and reduce erroneous convictions by allowing defendants to make less risky, better informed decisions about whether to plead guilty.

Imagine if we lived in a world of wonderland sentencing, where defendants made plea bargaining decisions with relatively accurate potential sentence information. In such a world, I suspect, a proposal to eliminate wonderland sentencing and hide such information from defendants would strike us as about as absurd as the Queen of Hearts’s proposal. And unlike some prior proposals to give judges a larger role in plea processes,¹⁶ 16.See, e.g., Daniel S. McConkie, Judges as Framers of Plea Bargaining, 26 Stan. L. & Pol’y Rev. 61, 65 (2015).Show More wonderland sentencing is unlikely to offend rules against judicial participation in plea bargaining. In fact, there may be no legal impediments to the implementation of wonderland sentencing; judges could start the process today, and it probably already occurs from time to time.

In Part I, I discuss in more detail how wonderland sentencing could work. In Part II, I analyze wonderland sentencing from traditional retributivist and consequentialist perspectives. I also discuss how those perspectives could be modified to address what I call “therapeutic retributivism” and “therapeutic consequentialism.” I conclude that while it is too soon to decide that wonderland sentencing will necessarily improve the criminal legal system, we have more than enough reason to pilot studies to gather more information.

1.  Lewis Carroll, Alice’s Adventures in Wonderland 165–66 (London, MacMillan & Co. 1866). ↑
2.  Id. at 187. ↑
3.  Id.; see, e.g., Charles H. Gustafson, Judicial Review of Jeopardy Tax Collection: Sentence First, Verdict Afterwards, 26 Case W. Rsrv. L. Rev. 315, 315–16, 315 n.* (1976); Robert S. Pasley, Sentence First—Verdict Afterwards: Dishonorable Discharges Without Trial by Court-Martial?, 41 Corn. L.Q. 545, 545–47 (1956). ↑
4.  But cf. Saul Smilansky, Determinism and Prepunishment: The Radical Nature of Compatibilism, 67 Analysis 347, 347–48 (2007) (arguing that the compatibilist view of free will held by many philosophers leads to the conclusion that we ought to pre-punish). ↑
5.  See, e.g., United States v. Jackson, 835 F.2d 1195, 1197 (7th Cir. 1987) (recognizing the broad sentencing discretion typically afforded to judges in non-guideline jurisdictions). ↑
6.  See, e.g., 18 U.S.C. § 3553(b)(1) (advising federal courts to ordinarily impose a sentence within the guideline range); Minn. Sent’g Guidelines & Comment. § 1.A (Minn. Sent’g Guidelines Comm’n 2024). ↑
7.  Cf. United States v. Booker, 543 U.S. 220, 245, 250–51 (2005) (holding the federal sentencing guidelines to be advisory and emphasizing the importance of judicial discretion). ↑
8.  There is evidence in many domains, including mock criminal justice contexts, that vividly presented numbers can anchor people’s estimates and preferences. See, e.g., Stephanie A. Cardenas, Charged Up and Anchored Down: A Test of Two Pathways to Judgmental and Decisional Anchoring Biases in Plea Negotiations, 29 Psych. Pub. Pol’y & L. 435, 435–36 (2023) (discussing how the willingness to plead guilty or recommend guilty pleas of mock defendants and defense attorneys can be anchored by harsh potential sentences); Roland Imhoff & Christoph Nickolaus, Combined Anchoring: Prosecution and Defense Claims as Sequential Anchors in the Courtroom, 26 Legal & Criminological Psych. 215, 215, 223–24 (2021) (finding that law student experimental subjects were anchored in their mock sentencing determinations by the recommendations of both the prosecution and the defense). ↑
9.  Shawn P. Sullivan, Kennebunk Doctor Convicted on Multiple Counts of Illegally Distributing Opioids, Portsmouth Herald (June 24, 2024, at 11:23 ET), https://www.seacoasto‌nline.com/story/news/local/york-star/2024/06/24/kennebunk-maine-doctor-merideth-norris-c‌onvicted-of-unlawfully-distributing-opioids/74191342007/ [https://perma.cc/K6LG-CPLC]. ↑
10.  Id. ↑
11.  Id. ↑
12.  Emily Allen, Kennebunk Doctor Will Not Serve Prison Time for Overprescribing Opioids, Portland Press Herald (May 16, 2025), https://www.pressherald.com/2025/05/15/ken‌nebunkport-doctor-will-not-serve-jail-time-for-overprescribing-opioids [https://perma.cc/7G‌4M-4C9R]. ↑
13.  See Fed. R. Crim. P. 11(b)(1)(H). ↑
14.  See Tim Friehe & Hannah Schildberg-Hörisch, Predicting Norm Enforcement: The Individual and Joint Predictive Power of Economic Preferences, Personality, and Self-Control, 45 Eur. J.L. & Econ. 127, 137–38 (2018); see also Thomas Epper et al., Preferences Predict Who Commits Crime Among Young Men, PNAS, Jan. 31, 2022, at 1, 5 (finding a statistically significant positive correlation between risk tolerance and criminality for property crimes, though not for violent, drug, or sexual crimes). ↑
15.  See Robert E. Scott & William J. Stuntz, Plea Bargaining as Contract, 101 Yale L.J. 1909, 1948–49 (1992). ↑

16.  See, e.g., Daniel S. McConkie, Judges as Framers of Plea Bargaining, 26 Stan. L. & Pol’y Rev. 61, 65 (2015). ↑