Just before I left the Virginia faculty, Fred and I were debating this hypothetical for a chapter he was drafting:
Suppose someone, whom we can call Harvey, has been accused by four different women of sexual assault or other acts of serious sexual misconduct. As is common in such instances, there were no witnesses to any of the alleged acts, nor is there any physical evidence. But each of the women presents a largely believable account… Harvey offers a forceful denial…Footnote 1
Schauer posits that each case has an 80 percent likelihood of being true.Footnote 2 Assuming these accusations are independent, Schauer contends that if we aggregate the probabilities, the likelihood that Harvey committed at least one of these acts is over 99.9 percent.Footnote 3 He then throws down the gauntlet: “Thus, under circumstances in which it is virtually certain that Harvey has committed at least one act of sexual assault – that he is a sexual assaulter – he will be acquitted in four serial criminal prosecutions, and will accordingly not be punished. Whether this outcome is problematic or troubling, in criminal trials or in other contexts, is the principal theme of this chapter.”Footnote 4
I confess that when Schauer first posed this hypothetical, I was mildly horrified by the prospect that we ought to just go ahead, do some multiplication, and feel comfortable concluding that “Harvey committed one crime.” You won’t be surprised to hear that I couldn’t dissuade him. In the chapter, Schauer deploys this hypothetical to make two arguments. First, the arguments against punishing Harvey for one (albeit undetermined) offense are far weaker than they first appear.Footnote 5 Second, because Schauer views punishment for Harvey within the criminal law as a nonstarter (practically and potentially normatively), Schauer asserts that his insights should have particular sway outside of criminal (and civil) proceedings.Footnote 6 Schauer thus condemns Harvey in the shadow of criminal law.
Schauer stopped too soon. Though provocatively raising the prospect of conviction, Schauer ultimately settles for something far less. But we can’t settle, and we need to take on the criminal law directly. My aim in this paper is first to surface precisely what makes this puzzle such a compelling one for criminal law – namely, that sexual assault cases create justice deficits – and, second, to ask the hard questions that this deficit demands, in the way that Schauer would have. Specifically, we should revisit whether we have the burden of proof right, whether we can alter statutes, or, at the very least, whether we need to rethink our evidentiary rules. In other words, I aim to deploy Schauer’s methodology to his own moral quandary.
At the end of the day, I offer both a bold and a timid answer. Boldly, I was wrong to think that Schauer went too far; he didn’t go far enough – on Schauer’s own terms, we should be convicting Harvey of all four offenses. More timidly, the Harvey case is too easy because of a unique fact set where we need not give up any of our deeply held commitments to reach conviction. But further extensions come with real costs. Though there may be some opportunities to tinker at the margins, we ultimately cannot normatively be bold and act consistently with what justice demands for defendants. This means that we face the tragic choice of which injustice we are prepared to live with, likely opting for the course that minimizes harm as best we can. But even if this is a timid and pessimistic place to end, it cannot diminish the greatness of Schauer’s enduring and indispensable contribution in teaching us how to sharpen the questions and clarify the stakes.
I. Recasting the Problem
“Harvey” is, of course, a thinly veiled reference to Harvey Weinstein. Although the hypothetical raises questions about aggregating across offenses, it implicitly raises a different question. Indeed, it is telling that when Schauer raised aggregation in prior work,Footnote 7 it lacked the important feature present here – that the problem that leads to the need for aggregation is within the context of sexual violence and believing women.Footnote 8 What 80 percent likely means, when likelihood rests solely on the testimony of both sides, is that the jury forms a credence that each woman is telling the truth to 80 percent. Each woman is credible, but not credible enough.
Let’s be realistic – today, 80 percent is incredibly optimistic. But what the hypothetical pushes us to confront, I think, is that even in a far more ideal world, that may be as good as it gets.
Let’s say we take seriously #BelieveWomen’s call not to discount claims of sexual violence.Footnote 9 A woman comes forward to report a rape. But in the criminal context, now consider what kind of case she faces. In acquaintance rape, the question is almost never whether sex occurred – it’s whether it was consensual or whether the defendant had the required mens rea. There are no witnesses. Often, there’s no physical evidence. So, she said/he said.Footnote 10 We understand it is a costly signal so the probability she is telling the truth is quite high.Footnote 11 Still, even if we take her testimony seriously, it may still be difficult for the state to overcome reasonable doubt.
This isn’t unlikely. Indeed, in their empirical study, David Bryden and Sonja Lengnick interviewed veteran criminal lawyers in Minnesota about crimes that routinely become credibility contests.Footnote 12 Many pointed to nonsexual assault – cases where one person claims to be attacked, the other denies they started it, and there’s little or no corroboration.Footnote 13 One defense attorney described these cases as a “race to the courthouse” to determine who gets to be the “victim.”Footnote 14 As a prosecutor in Washington, D.C., my own personal experience was that these cases were called “criss-cross” and were declined; you just couldn’t prove them.Footnote 15 Credibility contests are hard, and reasonable doubt is harder still.
This means proof beyond a reasonable doubt can pose an insurmountable hurdle to justice in sexual assault cases. Tort law isn’t going to fill this gap, practically or normatively. Practically, defendants without deep pockets are judgment-proof. Normatively, it is hard to believe that tort law should be the only venue for vindication of the profound rights violation at stake.
We’ve got a problem. The state has a duty to respond to wrongdoing. As Antony Duff argues about rape and murder:
Given the seriousness of such wrongs, and the way in which they violate basic conditions of civic life, it is important that we take formal notice of them as wrongs, and seek to call their perpetrators to account: if we allowed such wrongs to be committed with impunity, we would be betraying the basic values to which we purport, as a polity, to be committed.Footnote 16
That’s the real puzzle Schauer presents. It isn’t about whether we should feel comfortable doing math in or out of criminal law. It is that criminal law has a justice deficit. No woman is believed enough – in individual cases or in the aggregate. Testimonial justice can only take us so far, and in a more ideal world, in cases with one victim and those with multiple, defendants will walk.
If credible allegations repeatedly fall short, we’re not talking about individual error – we’re talking about systemic design. But I suggest we not settle for an answer outside of criminal law before we look long and hard at what criminal law can, and should, do.Footnote 17 Schauer went straight to the evidence puzzle, but there’s more depth to the question and more breadth to the potential answers. We should attack the problem along three Schauerian lines: burdens, statutes, and evidence.
II. The Burden
The demandingness of the beyond-a-reasonable-doubt standard is the obstacle to conviction in these she-said/he-said cases. Is the standard too demanding?
This is not a question Schauer answered; indeed, it wasn’t one he asked. Schauer was typically more interested in showing the flexibility of setting standards outside of the criminal law. For instance, in The Proof, he simply sets forth a brief foundational introduction to Blackstone’s ratio before moving on to whether that is how we should think about university disciplinary proceedings.Footnote 18
Additionally, Schauer’s interaction with the criminal law’s burden was, I think, one of frustration. Criminal law casts a large shadow, and its concepts are often deployed outside the criminal context, including whether to appoint a Supreme Court justice or to hire a babysitter. Time and again, Schauer chastised us for reverting to the distraction of omnipresent criminal law mantras.Footnote 19
But even if this wasn’t his go-to area of analysis, Schauer certainly would have recognized the potential here. Normatively, this is the natural pressure point. As Schauer notes, “Setting a burden of proof is inescapably an exercise in determining what, substantively, is at stake….”Footnote 20 Hence, our analysis crucially depends on how we think about a citizen’s relationship with the state,Footnote 21 what factors are permissibly traded off and how,Footnote 22 and, even among retributivists, how to weigh failing to punish the guilty in comparison to unintentionally punishing the innocent.Footnote 23
Though I cannot undertake a complete analysis of the myriad views, nor can I defend a position here, we can still reach some preliminary conclusions. First, if the state has a commitment to us as citizens that requires the beyond-a-reasonable-doubt standard, then it is not as if that commitment is less weighty when the crime is rape. Second, consequentialism is a Pandora’s box. If consequentialism is unleashed, it may amount to reduced standards for all sorts of crimes. (For consequentialism, it isn’t that clear why we’d need a crime.) Third, it will be difficult – expressively, practically, normatively – to defend “rape law exceptionalism” with a different burden of proof just for sex offenses.
Nevertheless, a case can be made for rape law exceptionalism. First, Schauer would have been the first to point to the interactive effects of the burden with the evidence. The burden of proof only works if the game is fair. As Michael Pardo and Larry Laudan have noted, if the quality of the evidence is systematically poor – or systematically skewed in favor of one side – then the burden of proof doesn’t just distribute errors,Footnote 24 it distorts them. It amplifies false negatives.Footnote 25 If we are trying to reach a numeric ideal beyond a reasonable doubt, then we should alter the standard to make up for evidentiary deficit. In contrast, if we think the standard embodies a deontological requirement between state and citizen, then the fact that the state’s evidence is flimsy doesn’t undermine the citizen’s legitimate demand for a stringent burden. This means that, under some understandings of the standard’s justification, there is a reasonable case for lowering the standard, not because we are giving up on what the balance should be but because we are readjusting for the real-world effects of what the balance is.
Here’s the second point. It might be that the state simply cannot fulfill its obligation to both sets of citizens in the nonideal world in which we live. Maybe rape will always be hard to prove and defendants will always have robust procedural rights, and the two will just conflict.Footnote 26 This isn’t then a matter of how to end injustice; it’s a matter of how to minimize it. But tradeoffs aren’t trumps, and so, the fact that we might wrong some innocent defendants is not a decisive answer to our desire to be more responsive to sexual violence.
Schauer would have recognized the academic puzzle of rethinking criminal law’s burden of proof, and he might have been tempted to think that there should be a lower burden in sexual assault cases, but he definitely wouldn’t have written that paper. Schauer sought to have real influence, not to shout in the wind. And, here, however unfair and insurmountable the burden is when it comes to sexual assault cases, the Supreme Court isn’t likely to reconsider the reasonable doubt requirement anytime soon; it’s a political nonstarter. So whether there is a reform that can end injustice or one that can just minimize its effects, we have to look beyond the burden; we should look at the statutes.
III. The Statute
If you can’t change how much you have to prove, change what you have to prove. That is, circumvent the reasonable doubt standard. Bill Stuntz taught us that criminal law does this all the time.Footnote 27 Stuntz observes that criminal law has pathological politics: Legislatures inappropriately empower prosecutors. If ABC is the moral wrong but C is hard to prove, then make the crime AB and rely on prosecutorial discretion to weed out those cases where C does not exist.Footnote 28 Or make the offense ABE where E is an easier to prove proxy for C.Footnote 29
Schauer makes a distinct, but related, observation about criminal statutes. He rejects the claim that if the moral wrong is ABC, then that is what we need to criminalize. Schauer suggests that we might criminalize XYZ instead.
To Schauer, criminalization should take into account probabilities and inferences. Take possession of extraordinary amounts of cocaine.Footnote 30 Should we punish this person – convicted of possession—not as a big-time drug possessor but as what they surely are, a big-time drug dealer? It is possible that the defendant might be stockpiling for personal use. But that is exceedingly unlikely. And, Schauer claims: What is the difference between the jury inferring from the evidence of extreme amounts to the intent to distribute as opposed to just making the crime possession but punishing it as the level of intent to distribute? After all, if the crime included the intent to distribute, the jury would use the fact that the defendant had tons of drugs to infer that he intended to sell them. Probabilistic proxies are everywhere.Footnote 31 As Schauer contends, “The view that the probabilistic and indirect nature of presumed offences renders them worrisome appears to be at odds with the probabilistic nature of evidence itself.”Footnote 32
This move by Schauer introduces some play in the joints; it doesn’t endorse pathological politics. Schauer’s thought is that there is nothing morally problematic about crimes that don’t directly reflect moral wrongs, so long as the requisite probabilistic relationship exists.
However, there remains the worry that some individuals will defy that prediction. For them, the offense will sweep too broadly. Here, Schauer’s answer – that the offender is on notice about this possibilityFootnote 33 – fails to fully come to terms with whether notice is realistic in the real world,Footnote 34 as well as how notice is supposed to bridge the proportionality gap between being a drug dealer and being a drug possessor.Footnote 35
Schauer’s answer? Error distribution:
And even if act-based decision-making is hard to fault in an ideal world, as a matter of non-ideal theory in a multi-decision-maker environment – one in which those who make the rules are not necessarily the same as those who are expected to follow them – maximum particularism or individuation may come at a cost, and the cost may at times be high.Footnote 36
Returning to sexual offenses, what does this move to presumed offenses get us? Here are two ways to go. First, we can look for other conduct that is evidentiary of the moral wrong. It might be causal, or it might correlate, but the idea would be that if we rely on evidentiary inferences at the proof level, there is nothing inherently objectionable about relying on them when drafting statutes. An attempt along these lines would be Ian Ayres and Kathy Baker’s separate crime of reckless sex, which focused on the failure to use a condom during a first-time sexual encounter, as a proxy for nonconsensual sex.Footnote 37 It won’t be perfect, but it can protect against error. However, presumed offenses may not cover that much ground.
Our second option is to embrace the Stuntzian insight. Take a bug and make it a feature. If the Supreme Court won’t consider the proof-beyond-a-reasonable-doubt standard directly, or we recognize that directly attacking that burden could do more harm than good, then we might minimize injustice or re-alter the balance by adjusting the statute such that it approximates guilt at a lower level. Maybe we think that it is easier to prove the lack of affirmative consent than to prove the lack of mental willingness, even if the latter is what consent really is. To be clear, I believe that most affirmative consent advocates truly believe that that is what consent consists of. But they need not. They could think that mental acts are the correct moral answer, but that we need a different legal answer in order to right the unfairly tilted burden.Footnote 38 And, so, whether we are morally or legally obligated to do so, this is still a mechanism to do indirectly what we cannot do directly – alter the burden by altering the statute.
Overcriminalization is easy to accomplish. The Supreme Court’s jurisprudence in this area provides few guardrails to prevent such manipulation.Footnote 39 And broad statutes can serve as an important counterweight to an otherwise too demanding burden of proof.
But you might not find that this ends injustice; it just shifts it. And it hands enormous power to prosecutorial discretion – which, in the U.S., we know can play out along racial lines. So we may be trading one structural injustice for another. Indeed, much of the debate over the revision to the Model Penal Code was expressly about this problem – in a world of overcriminalization, where sexual violence is underenforced, will we get more or less justice by adding to the criminal law?Footnote 40
There’s also a legitimacy problem. Rape is a crime that the public must recognize and condemn. We want a crime of rape that tries to track morality, not one that fakes it, indirectly aiming at conviction. If we redefine rape too far from its underlying moral wrongfulness, we risk undermining public trust – and the seriousness of the wrong itself. And if we just punish wrongs of rape as minor offenses – such as failing to get consent not to use a condom, for which the defendant just gets a slap on the wrist – we fail to condemn and punish as we ought. It is difficult to determine whether our statutory tinkering will result in more or less justice, and so it is difficult to take the leap of faith when we know we are departing from our deeply held principles. Perhaps that’s why Schauer left his burden and statute drafting tools in the toolbox and immediately jumped to evidence, to which we now turn.
IV. Evidence
Evidentiary rules indirectly alter the burden of proof. And unlike statutes and burdens, they are rarely transparent to the public or even the jury. Can we tinker more, with less political downside?
One thing to note is that with each passing layer of intervention, we lose justice. If we don’t change the burden, but alter the statutes, then our intervention will only affect those cases that fall within the proxy crime. And evidentiary rules implicate even fewer cases. Specifically, there has to be the sort of evidence where altering the evidentiary standard can make a difference. As we will see, that may be true for Harvey, but it won’t be true for the case where there isn’t evidence that the defendant committed other sexual assaults. Nevertheless, let’s see how much progress we can make.
For good or for ill, evidentiary rules play a role in shifting how hard it is to make a case.Footnote 41 For example, Larry Alexander insightfully suggests that when voluntary intoxication is inadmissible to prove that the defendant did not have the required mental state, this has the effect of creating a permissive but irrebuttable presumption, an effect that makes it much harder for the defendant to create reasonable doubt.Footnote 42 The statute gives, but the evidentiary rule takes away.Footnote 43
Sexual assault cases have two sorts of special evidentiary rules. First, with some exceptions, rape victims cannot be questioned about their sexual past.Footnote 44 The victim’s prior sexual behavior has minimal probative value, and protecting victims from grueling cross-examinations will encourage greater rape reporting. Second, in many jurisdictions, the usual bar on using character evidence to show the defendant’s propensity to commit the crime is removed. Typically, character evidence is impermissible – “once a drug dealer, always a drug dealer” because juries may place too much weight on past acts or seek to punish defendants because if they haven’t committed the crime at issue, they may still have committed some uncharged offense.Footnote 45 Yet, Congress enacted Federal Rule of Evidence 413 that allows prosecutors to argue propensity for sex offenses.Footnote 46
In some ways, Federal Rule of Evidence 413 seems to be the fix. Just let propensity in. But there are reasons to pause rather than to celebrate. First, these rules have been subject to significant scholarly criticism.Footnote 47 It just isn’t true that all sex offenders are likely to recidivate.Footnote 48 Second, these rules aren’t adopted everywhere – the Cosby prosecutor in Pennsylvania had to persevere without such an easy evidentiary fix. Any endorsement of these rules has to be earned, not taken as a baseline presumption.
Again, at this point, the cases we can impact have narrowed. We are now only able to influence cases in which defendants are accused, or who have been convicted, of multiple acts of misconduct. If we can’t get justice elsewhere, what about here? And that takes us back to Harvey, and how Schauer seeks to resolve the puzzle.
Schauer rejects that his puzzle is about punishing for character or thoughts – and he’s right. On Schauer’s setup, this isn’t a propensity argument. This is classic doctrine of chances. If all your wives die in the bathtub when you’ve got insurance on them, it’s unlikely they were all accidents.Footnote 49 So, too, if four allegations are each 80 percent likely, then according to the probabilities, one of those allegations is true.Footnote 50
There are two reasons to get off the boat here, one legitimate, the other misguided. First, though I will return to this objection briefly later, I am largely bracketing the substantial evidentiary debate that burdens of proof are about probability and that we can even approach this question as a matter of math.Footnote 51 Second, though it feels like there is an issue with statistical knowledge, there isn’t. The defendant has the ability to directly attack the factual basis of the statistics as they pertain to specific facts about him.Footnote 52
What does the math get us? As Sean Sullivan has been keen to point out, you can’t use the doctrine to show they were all bathtub murders, but to show that one was.Footnote 53 The doctrine of chances relies on independent probabilities to show that, in the aggregate, they can prove one act (and sometimes more). Schauer correctly grasps that he is only getting one crime by Harvey.
So, we do have one crime beyond a reasonable doubt. So, why not convict? Schauer notes we can, “there might be a plausible argument in favour of even criminal prosecution under the circumstances just described”Footnote 54 but then he blinks. He calls the section “Getting Realistic” and moves out of law entirely:
There may be many reasons to doubt the likelihood of the kind of aggregation discussed here ever becoming part of common law legal systems, whether in the context of criminal law or even civil liability. The necessity of commencing with a particular act before focusing on a potential perpetrator is simply too entrenched for that and raises issues going to the very operation of the sanctioning dimension of law. But the point of this exercise, even apart from academic (in the best sense, and not as a pejorative) interest, is that the issue has important implications in many non-formal or non-judicial administrative contexts.Footnote 55
Sort of. To be sure, we can aggregate outside of criminal law, but Schauer’s earlier focus – on questions of punishing for thoughts and character – are distinctly about criminal law. In the criminal context, he seems sorely tempted. He wasn’t writing a paper about whether we should sit next to Harvey at the lunch table.
So why not allow conviction for “one of the four?” That isn’t so crazy. After all, we don’t require jury agreement as to means.Footnote 56 And imagine you had two identical twins, both caught on camera robbing a different bank, both leaving DNA – but you can’t tell which one was in each bank.Footnote 57 Should they really escape justice? Maybe we need not be totally and completely wed to THIS ACT.Footnote 58 As Peter Westen and Eric Ow argue, Due Process should not be offended so long as there is jury agreement that if not this, then that:Footnote 59 if one twin didn’t rob Bank A then he must have done B and vice versa. Though Schauer raises criminal law, he ultimately refuses to cross the threshold and take a normative stand. Again, his thought is this is a practical nonstarter.
But I actually don’t think Harvey was a hard case for Schauer, given his own jurisprudential commitments. Indeed, as I revisit our conversation, I pushed him the wrong way. Instead of arguing against convicting for one offense, I should have urged him to argue for conviction for all four.
To start with, Schauer buys the math. So, let’s do the math. Admittedly, not every evidence theorist is going to agree. The Harvey hypothetical makes math seem far more palatable than it is in many other cases.Footnote 60 Compare Alon Harel and Ariel Porat’s original gambit – indicating a willingness to aggregate across a charge of pickpocketing and rape.Footnote 61 One might have thought such a case was a reductio, but they fearlessly assert that cases should be treated probabilistically and that, accordingly, we can be certain that one crime was committed.Footnote 62
Now, that we have one offense, why not all four? Of course, the question is what justifies the inference from one offense to others. Things are easy when we have a legitimate non-propensity basis such as identity, plan, or intent.Footnote 63 When there are legitimate evidentiary inferences that avoid running through propensity, we may be able to swiftly reason from one to all.Footnote 64
Harvey, however, doesn’t give us that. There’s nothing that allows us to leap from one certain act to four without running through the inference – Harvey committed the other crimes because Harvey is a rapist. Outside of the law, Schauer’s thought is that reasoning through propensity was perfectly legitimate – we do it all the time.Footnote 65 And, I suspect that Schauer would have little reticence in thinking (though perhaps not for putting it in print) that we’ve got enough to convict Harvey for all of the offenses.
So, what if I had pushed Schauer to go for all four: doctrine of chances plus propensity? Here’s a virtue. If we think that Harvey can be convicted of all of the offenses, then the criminal law objection to a single arbitrary conviction evaporates. Without inconsistency, we can fire on two cylinders. We can rely on the independence of each victim’s claims to allow us to aggregate probability, and we can rely on the dependent variable – Harvey – to allow us to reason from one assault to four.
While there are dangers nearby, this strikes me as an easy case for Schauer to step out of the criminal law’s shadow and fully endorse the conviction. Why? Because he believes we are entitled to believe that Harvey committed one act, because he believes that propensity is a legitimate evidentiary inference, and because he must have seen that, though burden tinkering could not be done directly, he had put his finger on precisely the kind of case where something has to give. So, I was wrong to think that Harvey shouldn’t be convicted of one offense, and Schauer was wrong to intimate that there should be the possibility of conviction for one. We both should have boldly embraced conviction for all offenses, at least on these facts where all the planets align.
Of course, it is easy to suggest that he should have crossed the Rubicon. This case has 80 percent probabilities, not the scant evidence required in actual cases where merely sufficient evidence to support a jury finding will do.Footnote 66 And, as I consider the hypothetical, all of the cases are being simultaneously adjudicated and somehow one court has jurisdiction over all the offenses.
Indeed, however clever it is for us to use aggregation to get Harvey for the first offense, that first step is a doozy – one that meets with significant objections. First, some theorists firmly reject that the evidentiary credences of juries are the same as objective probabilities. They don’t add (or multiply) up. Kevin Clermont gives the example of a close case in contract and a close case in tort. You don’t get to win one just because you almost won both. As he puts it: “A number of failures do not create success.”Footnote 67 Second, the expressive function of the criminal law is really, really weird if all we can say is that you may have committed some crime…maybe rape and maybe … pickpocketing? Third, procedurally, we truly want the jury to determine that the defendant committed this crime; how can you receive what you deserve if we can only point to some unidentified offense?Footnote 68
In the criminal context, it is deeply disturbing to think that the math yields that one offense has been committed. It is notable that even Harel and Porat thought that we would need external constraints to constrain abuse, including requiring that each offense reach at least a 0.5 threshold.Footnote 69 And, they note that the more similar the offense, the less we worry about undermining criminal law’s expressive function.Footnote 70 But the Harel/Porat 0.5 threshold appears arbitrary. If the math is right, why is it constraining abuse? If a teeny bit of smoke in lots of cases means there’s fire, why would it be wrong for the state to do this? We need a normative answer, premised on our political commitments – perhaps that without at least probable cause, the state has insufficient evidence to call us to account for a wrong.
Accordingly, we should expect reformers to balk at the prospect of allowing aggregation. But maybe we don’t need the math. Instead, we might simply rely on propensity.
In the real world, Harvey – and any case where there is evidence of prior misconduct – simply unravels into the simple question of whether we are prepared to endorse Federal Rule of Evidence 413. That is, Harvey is both the hard and the easy case. Easy because given how all the facts are construed, we may be more willing to allow both aggregation and propensity. As an “academic” exercise, Schauer should have embraced four convictions, not one. But Harvey also shows us that for this easy conviction to obtain, we will need to embrace aggregation, propensity, or both. And realistically, that is going to be propensity. Propensity does not threaten proof for one offense, and we know it is politically feasible – 413 is already on the books.
Should we endorse this rule? The justification isn’t empirical. Instead, our best justification is that it fixes the skewed evidentiary scales. It is a corrective to how the deck has been unfairly stacked from the get-go – how the unavailability of evidence has made proof beyond a reasonable doubt unattainable. Moreover, the argument is that it won’t tilt the scales too far in the other direction. Victims are extremely likely to be telling the truth.Footnote 71 Accordingly, we need not be as concerned that we have unfairly shifted error distribution.
And yet, endorsement of this rule comes with real loss. It has the capacity to construe offenders as monstrous and irredeemable, in ways that fuel rash punitive impulses. And it yields that innocent people will be convicted, most likely those with the least resources to mount a strong criminal defense.
And so, we are back where we started. To achieve justice for sexual misconduct, we must change the burden, the statute, or the evidence. And any change is likely to bring injustice to some innocent defendants. Someone’s got to lose.
A confession. Everything I have ever done and written cuts against these arguments. As a prosecutor and theorist, I have clung to proof beyond a reasonable doubt. With Larry Alexander, I have railed against proxy crimes.Footnote 72 As an advisor to the ALI, I sought revising the Model Penal Code’s sexual assault provisions to mirror morality and to avoid evidentiary shortcuts.Footnote 73
But my work on the #MeToo movement has led me to question whether we should just take the status quo for granted, hold these values constant, and let the chips fall where they may. Fred taught us to frame the problems precisely and to expose the real costs of our answers. He appreciated that we too quickly jump to conclusions and conflate distinct domains. And so, I’ve pushed his argument back into the criminal law where it belongs. And within criminal law, the stakes are stark and unforgiving: if we cannot get acquaintance rape cases over the finish line, then we’re not just managing error – we’re choosing what kind of injustice we’re willing to live with.
Acknowledgements
I thank the participants at Rules and Reasons: A Conference for Fred Schauer, at the University of Virginia, as well as the participants at the Consent, Truth, and Power conference at Balliol College, Oxford, for their helpful comments. Théo Fehr provided outstanding research assistance. I am deeply grateful for the opportunity to celebrate Fred. I miss his wisdom, kindness, and occasional snark more than I can say.
Open access
