A free national technical assistance center exists because defense teams routinely do not know how to litigate abuse history. Michigan gives them less statutory material to work with than a growing number of states, its highest court left the central evidentiary question open in 1995, and in the stalking category the person being followed is often criminalized by the person following them, through a documented strategy the research says works on the audience.
A criminalized survivor is a person who experienced domestic violence, sexual assault, stalking, or trafficking, and who is then charged with a crime connected to that abuse. The National Defense Center for Criminalized Survivors, a project of the Battered Women’s Justice Project and formerly the National Clearinghouse for the Defense of Battered Women, provides free case-specific technical assistance to defense teams handling those cases nationwide. It does not represent anyone.
The existence of that center is the finding. A national help desk staffed for defense attorneys means the expertise is not resident in the defense bar. In Michigan the problem compounds, because the state has no statute directing courts to weigh documented victimization at sentencing, no resentencing mechanism for incarcerated survivors, and a Supreme Court decision that approved expert testimony on battering when the prosecution offers it while expressly declining to resolve whether a defendant may offer the same thing.
Stalking is the category where the machinery is most exposed, because the survivor is often criminalized without doing anything that requires explaining. The stalker performs the criminalization directly by filing first and reversing the roles, a documented perpetrator strategy known as DARVO. Michigan’s protective order system then supplies the enforcement: free filings, statewide venue, no jury, and a rule permitting a private party’s retained attorney to prosecute a criminal contempt carrying jail time.
What is a criminalized survivor?
A person who experienced abuse and is then charged with a crime connected to that experience. The four recurring categories are defensive force, coerced criminal activity, failure-to-protect charges, and custodial interference charges filed after a parent leaves with a child.
Does the National Defense Center take cases?
No. It provides free technical assistance to defense teams, helps locate qualified expert witnesses, supplies case law and sample litigation materials, and works at any stage from pretrial through appeal. It does not enter appearances.
Does Michigan have a survivor justice law?
No. As of August 25, 2026 there is no Michigan statute making documented victimization a defined mitigating factor at sentencing, and no resentencing pathway for people already incarcerated.
How does a stalking victim end up as the defendant?
Most often by reversal rather than by anything they did. DARVO describes a perpetrator denying the conduct, attacking the credibility of the person reporting it, and taking the victim role. Applied to a court, that means filing first. Miller and Smolter documented abusers obtaining unfounded protective orders that expose the person they abuse to arrest before that person has filed a legitimate petition of their own.
What is the exposure in a PPO contempt case?
Up to 93 days in jail and a $500 fine for criminal contempt under MCR 3.708(H)(5)(a). There is no right to a jury. The respondent is entitled to a lawyer, and to appointed counsel at public expense if the court determines it might impose jail and the respondent cannot afford one.
Can a Michigan defendant put on an expert about abuse?
The Court of Appeals allowed it in People v Wilson in 1992. The Supreme Court in People v Christel in 1995 addressed only the prosecution’s use and reserved the defense question. Trial courts are working from an appellate record that never squarely answered it.
What the Center Actually Does, and What Its Existence Proves
The National Defense Center for Criminalized Survivors is a program of the Battered Women’s Justice Project, headquartered in Saint Paul, Minnesota. Before its renaming it operated for decades as the National Clearinghouse for the Defense of Battered Women. It works on cases in every state. Consultation is free.
Its published description of its own function is unusually precise about scope. It identifies defense strategies, supplies relevant case law and sample litigation materials, locates qualified expert witnesses, and helps assemble support networks for people facing trial or working through an appeal. It also takes collect calls from incarcerated survivors. What it will not do is appear in court on anyone’s behalf.
That combination is worth sitting with. A national organization has determined that the highest-leverage intervention available to it is not representation but the technical education of the lawyer who already has the case. Organizations do not build help desks for problems that are being handled well.
The center’s own intake materials tell defendants to hand a printed page to their attorney. That is an organization designing around a known failure mode: the attorney does not know the resource exists, and the client is the only person in the room with an incentive to close that gap.
The center also states plainly what many advocacy organizations will not, which is that raising a defendant’s abuse history is not always useful, advisable, or permissible, and that whether to do it is the defense attorney’s call. That sentence is doing real work. Abuse history introduced badly can hand the prosecution a motive theory. It can open the door to the defendant’s own prior conduct. It can produce an expert who gets excluded and a jury that now suspects the defense went looking for an excuse.
The technical assistance model exists because the risk of doing this wrong is not theoretical. It is the ordinary outcome when a generalist criminal defense practice encounters a case type it sees once every several years.
Michigan’s Self-Defense Statute Measures the Wrong Interval
Michigan codified self-defense in 2006 through Public Act 309, now MCL 780.971 through 780.974. Under MCL 780.972, a person not engaged in the commission of a crime may use deadly force anywhere they have a legal right to be, with no duty to retreat, if they honestly and reasonably believe the force is necessary to prevent imminent death, imminent great bodily harm, or imminent sexual assault. MCL 780.951 supplies a rebuttable presumption of honest and reasonable belief in home invasion situations. MCL 780.973 preserves the common law of self-defense where the act does not reach.
Read as written, the statute is generous. It removes the retreat requirement. It extends beyond the dwelling. It applies to defense of another. On its face there is nothing hostile to a survivor in it.
The problem is the unit of measurement. Imminence is an instantaneous standard. It asks what a reasonable person would have believed about the next few seconds. The danger assessment a person under sustained coercive control is actually running is longitudinal. It is built from escalation patterns, from what happened the last four times a certain look preceded a certain act, from strangulation history, from what the abuser said would happen if the person left, and from the documented reality that the period around separation is when intimate partner homicide is most likely to occur.
A jury instructed on imminence is instructed to ignore most of that unless something makes it admissible. The two statutory conditions compound the problem. The person must not be engaged in the commission of a crime at the time, which is a live exclusion for survivors coerced into criminal activity by the same person they are defending against. And the person must have a legal right to be where they are, which becomes contested in cases involving contested residences, protective order boundaries, and disputed tenancies.
The statute asks what the person believed in the moment. The survivor’s evidence is about the pattern. Getting the pattern in front of the jury is the entire defense, and Michigan has never settled whether it comes in.
The Christel Asymmetry
In 1992 the Michigan Court of Appeals decided People v Wilson, 194 Mich App 599, holding that expert testimony on battering was admissible where domestic violence was a factor in the offense. The defendant in that case had shot her husband while he slept.
Three years later the Michigan Supreme Court decided People v Christel, 449 Mich 578. The posture matters. Christel was not a survivor-defendant case. The defendant was charged with criminal sexual conduct, the complainant was the person who had been abused, and the prosecution offered expert testimony to help the jury understand why she had delayed reporting and recanted. The Court held that such testimony is admissible when it is relevant and helpful to evaluating a complainant’s credibility and the expert is properly qualified. It also drew a firm line on scope, holding that the expert may explain the general characteristics of the pattern but “the expert cannot opine that the complainant was a battered woman.”
Then it stopped. The Court expressly refrained from approving or disapproving the use of that evidence when a criminal defendant offers it as exculpatory. In 2020 the Court of Appeals in People v Spaulding, 332 Mich App 638, adopted the broader term battered partner syndrome and reiterated that expert testimony is confined to generalities without opining on the facts of the case.
The resulting structure is asymmetric in a way that is easy to state and hard to justify. When the abused person is the state’s witness, the Michigan Supreme Court has told trial courts how to admit expert testimony explaining her behavior. When the abused person is the defendant, the highest court in the state has said nothing binding, and the trial judge is working from an unreviewed Court of Appeals decision from 1992 and a reservation of the question from 1995.
This is not a finding that Michigan bars survivor-defendant expert testimony. Wilson remains on the books and trial courts admit this evidence. The finding is narrower and more useful: the question has been open at the Michigan Supreme Court for thirty-one years, which means admissibility varies by courtroom, and a defense attorney litigating it is litigating without a controlling answer. That uncertainty is precisely what free technical assistance is for.
Stalking, DARVO, and the Advantage of Filing First
The National Defense Center names four populations it serves: victims of domestic and dating violence, sexual assault, stalking, and human trafficking who have been criminalized as a result of being abused. Stalking is the least discussed of the four, and its criminalization mechanism is the one that does not fit the pattern of the others.
In the other categories, the survivor does something. They use defensive force. They commit an offense under coercion. They fail to prevent something. The state then charges the thing they did. In the stalking category, the survivor frequently does nothing that requires explaining. The stalker performs the criminalization directly, by going to court first and describing the person they are stalking as the stalker.
Psychology has a name for the move. Jennifer Freyd introduced the acronym DARVO in 1997 to describe how perpetrators respond when confronted: they deny the conduct, attack the credibility of the person reporting it, and reverse the roles of victim and offender so that the perpetrator occupies the injured position and the person harmed becomes the accused. Harsey, Zurbriggen, and Freyd reported in 2017 that DARVO was commonly used by people who were confronted about their conduct, and that higher exposure to it correlated with increased self-blame in the person doing the confronting.
The 2020 follow-up is the finding that matters for a courtroom. Harsey and Freyd ran experimental vignette studies measuring how observers judged the same underlying incident with and without DARVO present. Participants exposed to DARVO rated the victim as less believable, more responsible for the violence, and more abusive. They rated the perpetrator as less abusive and less responsible. The tactic worked on the audience, not on the parties.
DARVO is not a misunderstanding the court falls into. It is a strategy aimed at the court. The reversal is performed for an audience whose entire function is to decide which of two accounts to believe, and the peer-reviewed evidence says the audience moves.
Where the reversal becomes a criminal charge
Miller and Smolter documented the operational version of this in 2011, in the research note that gave the field the terms paper abuse and procedural stalking. Among the tactics they recorded was one that is specific and consequential: abusers filing first for unfounded protective orders, which places the person they are abusing at risk of arrest for violating those orders before that person has filed a legitimate petition of their own.
Read that sequence carefully, because it is the whole conversion. The person being stalked has not yet gone to court. The stalker has. The order that issues names the stalker as protected and the survivor as restrained. It is entered into LEIN. From that moment forward, the survivor’s ordinary conduct, including conduct undertaken to document or escape the stalking, is enforceable against them by arrest. No new act is required. The legal status of both people has simply been inverted by paperwork, and the criminal exposure now attaches to the person who was being stalked.
Clutch Justice has documented the Michigan architecture that permits this at length, including the same-day ex parte process, the absence of any cross-county registry of serial filers, and the functional nonenforcement of the felony false-statement provision at MCL 750.422a. That analysis is available here and is not repeated in this piece. The question here is narrower and downstream. Once the reversal has succeeded, the survivor is a criminal defendant, and this is the point at which the criminalized survivor framework applies to them.
Criminal contempt for violating a personal protection order carries up to 93 days in jail and a $500 fine under MCR 3.708(H)(5)(a). There is no right to a jury trial under MCR 3.708(H)(1). Under MCR 3.708(G), the prosecuting attorney handles the criminal contempt unless the petitioner retains their own attorney for it. A person who successfully reversed the roles at the petition stage may therefore prosecute the person they were stalking, through counsel they selected, before a single judge, with no jury.
That structure is where DARVO does its heaviest work. A contempt hearing on a protective order violation is close to a pure credibility contest. There is often no physical evidence, no independent witness, and no jury to divide the judgment. One judge decides who is telling the truth, frequently in a hearing measured in minutes, on a record that contains a sworn petition from one side and a denial from the other. The Harsey and Freyd findings describe exactly that decision environment, and they describe an audience that shifts.
The survivor is also, at that point, arguing from the least persuasive position available in a courtroom. They are asking the court to disbelieve an order the court itself already issued. They are proving a negative. And the conduct they would ordinarily use to establish the underlying stalking, meaning their own documentation, is now the conduct they are being accused of.
The statutory condition that lets it work
Michigan’s stalking statute does not resist the inversion, because it is not written to. Under MCL 750.411h, a course of conduct is a pattern of two or more separate noncontinuous acts evidencing a continuity of purpose, and unconsented contact expressly includes following or appearing within sight, appearing at a workplace or residence, and contact by telephone, mail, or electronic means. Aggravated stalking under MCL 750.411i adds a restraining order violation or a credible threat and makes it a felony. Nothing in those elements records which person is afraid.
The sorting is left to a single clause: harassment does not include constitutionally protected activity or conduct that serves a legitimate purpose. That clause is doing all of the work, and it is applied by the same judge, in the same credibility-dependent hearing, without a jury. A statute that describes the conduct of both people identically is not the cause of the reversal. It is the condition that lets a reversal survive contact with the elements.
The 2020 experiments contained a second result that points directly at the remedy. Participants who had been educated about DARVO before reading the same vignettes rated the victim as more believable and less abusive, and rated the perpetrator as less believable. The tactic is effective on an uninformed audience and measurably less effective on an informed one. Its counter is not better evidence. It is a decision-maker, and a defense lawyer, who can name the pattern when they see it.
That is the argument for the technical assistance model stated in empirical terms. A free national center that trains defense teams to recognize and litigate this pattern is not a referral service. It is the only widely available intervention aimed at the variable the research identifies as decisive, and it reaches a criminalized stalking survivor only if that survivor has a lawyer who calls it.
Clutch is free to read. It isn’t free to produce. Court records, FOIA requests, research databases, document access, technology and hundreds of hours of investigation make this work possible. Clutch Confidential is how readers who value that work can help sustain it, and get more from Clutch in return.
$10/month ?Where Michigan Sits in the Statutory Wave
Survivor justice legislation does something narrower than popular framing suggests. It does not create an excuse or an acquittal pathway. It permits a court to treat documented victimization as a mitigating factor at sentencing when the abuse was a significant contributing factor to the offense, and it requires the defendant to prove that nexus. Judicial discretion is preserved. Evidentiary standards apply.
The Court of Appeals holds expert testimony on battering admissible where domestic violence was a factor in the commission of the offense. The defendant had shot her husband while he slept, a fact pattern that cannot be reconciled with a strict reading of imminence without evidence about the pattern.
The Court holds expert testimony on battering admissible when relevant and helpful to a jury’s evaluation of a complainant’s credibility, with the expert confined to general characteristics. The abused person in Christel was the state’s witness, not the defendant.
MCL 780.971 through 780.974 remove the duty to retreat anywhere a person has a legal right to be and condition deadly force on an honest and reasonable belief in imminent death, great bodily harm, or sexual assault. MCL 780.951 adds a rebuttable presumption for home invasion.
New York’s Domestic Violence Survivors Justice Act amends Penal Law 60.12 to permit alternative sentences and creates Criminal Procedure Law 440.47, a resentencing pathway for people already incarcerated on terms of eight years or more. Relief requires a proven nexus between the abuse and the offense.
Oklahoma enacts the Survivors’ Act in 2024 over a gubernatorial veto. Georgia’s Survivor Justice Act is signed in May 2025 and takes effect July 1, 2025. New Jersey’s is signed in January 2026. The American Legislative Exchange Council adopts a model Survivor Justice Act in January 2026.
A five-bill trafficking package would expand the criminal convictions that may be set aside on grounds of trafficking victimization, broaden juvenile expungement, and strengthen protections for minors. The House Judiciary Committee took testimony on January 21, 2026, including written support from the Attorney General’s office.
Michigan has a human trafficking commission operating since 2015 and a partial expungement framework for trafficking victimization. It has no statute directing courts to weigh documented abuse as a defined mitigating factor at sentencing, and no vehicle for a person already incarcerated to raise it.
The Population Nobody Is Counting
The research on how many incarcerated people are criminalized survivors is thin, and that thinness is itself the point. What exists is consistent. The Council on Criminal Justice has reported that between half and three-quarters of incarcerated women report experiencing sexual violence, and that more than seventy percent report intimate partner violence. Research compiled by survivor justice advocates puts the range of incarcerated women with a history of domestic or sexual violence between seventy and ninety-five percent. A Stanford analysis found that close to a quarter of women incarcerated for homicide were serving time for offenses tied directly to intimate partner violence.
Michigan supplies its own numbers. Analysis published by the R Street Institute in April 2026, drawing on Prison Policy Initiative and Vera data, put Michigan’s incarceration rate for women at 78 per 100,000 residents, noted that the number of women in Michigan prisons more than tripled between 1978 and 2017, and reported that Michigan incarcerated thirty percent more women between 2009 and 2015 while its male prison population declined. Michigan State Police crime data recorded more than 68,000 domestic violence incidents in 2024 producing over 74,000 victims, with roughly 80 domestic violence related homicides and manslaughters annually.
None of those figures establish how many people are incarcerated in Michigan for offenses their own victimization produced. No agency collects that. The absence of the data is a policy choice, and it is convenient for everyone whose budget depends on the answer staying unknown.
No Michigan statute makes documented victimization a defined mitigating factor. Four states have one.
No equivalent to New York’s Criminal Procedure Law 440.47. A person sentenced before any future reform would have no vehicle.
Wilson permits it. Christel reserved the question in 1995 and the Supreme Court has not returned to it. Admissibility varies by courtroom.
Generous on retreat, hostile on timeframe. The commission-of-a-crime condition excludes coerced defendants by operation.
A partial set-aside framework exists and HB 5009 through 5013 would expand it. Expungement is backward-looking relief that arrives after the conviction has done its work.
Free national technical assistance exists and reaches Michigan. It is only reachable through counsel, which makes appointed counsel the bottleneck.
Jail exposure of 93 days with no jury, and a private petitioner’s retained attorney may prosecute the criminal contempt. The right to appointed counsel exists in the rule and depends on the respondent invoking it.
Nothing in the intake, elements, or hearing structure is designed to detect a reversed filing. MCL 750.411h records conduct, not fear, and the sorting sits in one carve-out clause applied without a jury.
No Michigan agency tracks how many incarcerated people are serving sentences for offenses connected to their own victimization.
What the Counterargument Actually Says
Prosecutorial objections to survivor sentencing legislation are not frivolous and should not be waved off. The three that recur are the prospect of high-volume petitions with weak nexus showings, the emotional cost to a deceased person’s family of relitigating a resolved case, and the integrity of a process where the person best positioned to describe the abuse is also the person with the strongest incentive to describe it. R Street, which supports this legislation, treats all three as legitimate drafting inputs rather than obstacles.
The drafting answers are known. Nexus requirements do the heavy lifting: New York’s statute requires proof that the abuse was a significant contributing factor to the offense, which screens out generalized trauma histories. Judicial gatekeeping at the petition stage filters legally insufficient applications before a hearing. Notice to prosecutors and structured victim input are standard. Oklahoma and Georgia passed with these safeguards and with law enforcement at the table.
There is a second objection worth naming because it is rarely stated directly. If courts may consider abuse history as mitigation, some percentage of people who were not meaningfully coerced will benefit. That is true. It is also true of every mitigating factor a sentencing court considers, including age, cognitive impairment, and addiction. The question is not whether a mitigation framework can be gamed. It is whether the current arrangement, in which the information is excluded by default and admitted only when a lawyer knows how to fight for it, produces more accurate sentences. It does not.
What Is Documented, What Is Not, and What Remains Open
Documented: the National Defense Center’s scope, services, and cost structure, taken from its own published materials. The holdings and procedural posture of Wilson, Christel, and Spaulding. The text of MCL 780.951, 780.972, and 780.973, and of MCL 750.411h and 750.411i. The procedural terms of MCR 3.703, 3.706, and 3.708, including the fee structure, the statewide venue rule for adult respondents, the prohibition on mutual orders, the 93-day and $500 criminal contempt sanction, the absence of a jury right, the private prosecution provision, and the advice and appointment of counsel requirements. The enactment dates and mechanisms of the New York, Oklahoma, Georgia, and New Jersey statutes. The introduction date and committee status of Michigan House Bills 5009 through 5013.
Also documented: DARVO as a described and measured construct. Freyd introduced it in 1997. Harsey, Zurbriggen, and Freyd reported prevalence and self-blame findings in 2017. Harsey and Freyd published the credibility experiments and the education effect in 2020. Miller and Smolter documented the first-filing tactic in peer-reviewed research in 2011.
Analysis, not finding: the characterization of the Christel reservation as an asymmetry that disadvantages survivor-defendants. Christel is a criminal sexual conduct case and its silence on defense use is a limitation of its posture, not a ruling. Also analysis: the application of the DARVO credibility research to Michigan PPO contempt hearings specifically. The 2020 experiments used written vignettes and university student participants. They were not conducted on judges, and no study has replicated them in a courtroom. The argument here is that a contempt hearing shares the decision conditions those experiments modeled, which is a reasoned inference rather than a measured result.
Not established: that reversed or retaliatory protective order filings occur at any particular rate in Michigan. The tactic is documented in the research literature and the Michigan rules permit it procedurally. Michigan does not measure how often it happens, and this piece does not claim a frequency.
Unanswered: how many people in Michigan custody are serving sentences for conduct tied to their own victimization. How often Michigan trial courts admit or exclude defense-offered expert testimony on battering, and whether that varies by circuit. How frequently PPO contempt proceedings are prosecuted by a petitioner’s retained counsel rather than by a prosecuting attorney, and whether outcomes differ. How often respondents in those proceedings appear without counsel after being advised of the right to appointed counsel. None of those questions has a public answer, and most are answerable from records the state already holds.
Why This Matters Beyond Any One Case
The through line in every category the National Defense Center handles is that the state was present before the charge. Someone called police. Someone sought a protective order. Someone went to an emergency room, or a shelter, or a caseworker. The record of institutional contact usually exists, and it usually shows the system had information it did not act on.
Then the survivor acts, or is coerced into acting, or fails to stop something they had no power to stop, and the same system that produced no protection produces a charge with impressive speed. The state’s failure at the front end becomes the state’s evidence at the back end. The person who could not get anyone to intervene is now the person the intervention is aimed at.
Incarceration does not repair any part of that. It does not treat the injury, restore the household, or return the years. What survivor justice statutes do is considerably more modest than their critics allege and considerably more useful than their absence: they let a judge see the whole file before deciding what a sentence is for.
The stalking channel makes the same point faster, because there the state’s failure and the state’s charge can arrive as the same document. Michigan built a protective order system that is free to enter, filable in any county, enforceable on signature, and backed by jail. That design is defensible. The people it was written for often have hours rather than weeks. But a system optimized for speed does not ask which party is which, and a person who understands that will reach the courthouse first. The research says a reversal succeeds on an audience that has not been taught to expect one, and Michigan has built no part of its intake, its elements, or its hearing structure to expect one.
Michigan has not built the framework. Until it does, the difference between a survivor who gets a fair hearing and one who does not comes down to two accidents: whether they had a lawyer at all, and whether that lawyer had ever heard of the free number that would have taught them what they were looking at.
Williams, Rita. When the Survivor Is the Defendant: Michigan’s Missing Framework for Criminalized Survivors, Clutch Justice (Aug. 25, 2026), https://clutchjustice.com/2026/08/25/criminalized-survivors-michigan-defense-framework/.
Williams, R. (2026, August 25). When the survivor is the defendant: Michigan’s missing framework for criminalized survivors. Clutch Justice. https://clutchjustice.com/2026/08/25/criminalized-survivors-michigan-defense-framework/
Williams, Rita. “When the Survivor Is the Defendant: Michigan’s Missing Framework for Criminalized Survivors.” Clutch Justice, 25 Aug. 2026, clutchjustice.com/2026/08/25/criminalized-survivors-michigan-defense-framework/.
Williams, Rita. “When the Survivor Is the Defendant: Michigan’s Missing Framework for Criminalized Survivors.” Clutch Justice, August 25, 2026. https://clutchjustice.com/2026/08/25/criminalized-survivors-michigan-defense-framework/.
Continue Your Investigation
If this reporting raised more questions, use the Clutch Justice ecosystem to keep going.
