Earlier this week, writer Amanda Marcotte wrote a piece for Slate about a recent court case in which a woman who had suffered domestic violence and sexual assault was arrested and jailed for missing scheduled meetings with prosecutors.

Calling the problem of non-cooperation “endemic” in domestic violence (DV) cases, Marcotte applauds the prosecutor’s decision. It is, she says, “a sad, unavoidable truth” that granting DV complainants autonomy in deciding how and whether to cooperate with prosecutors “means putting some very bad men back out on the streets,” and we, as a society, have to “decide what’s more important to us.” As Marcotte presents it, the issue is straightforward: In order to get tough on crime, we have to be willing to get tough on victims.

Many feminists have published cogent rebuttals to Marcotte’s piece in the two days since it appeared. The blogger stavvers wrote an eloquent post declaring that any approach to sexual violence prosecution that does not unconditionally affirm the choices of survivors “is just rape culture, rebranded.” A Tumblr post by Nerd Grrrl Island drew out the victim-blaming implications in the piece. Others, notably Lauren Chief Elk, took to Twitter to push back on what they saw as Marcotte’s blanket support for the criminal justice system.

All of these criticisms have merit, and all raise important issues that Marcotte should grapple with. But the bottom line is that Marcotte’s essay fails to make a coherent case for the policies she embraces.

In response to her critics, Marcotte has said that her concern in this piece was not with the validity of criticisms of the criminal justice system or with the morality of re-victimizing the victims of crimes of coercion. Her focus, she says, is specifically and exclusively on the conflict between respecting victims and successfully prosecuting offenders.

But even granting that narrow focus, even assuming that the larger issues can be disentangled from her specific argument, Marcotte’s position—that coercing testimony from survivors of violence means more victims testifying, which means more offenders jailed, which means less DV and sexual assault—is essentially undefended. It’s an abstract thought experiment, not a policy analysis, and as it turns out, the conflict she sees is largely nonexistent in the real world.

A prosecutor’s decision to jail a complainant in order to compel testimony is not made in a vacuum, and it’s one that carries criminal justice costs of its own. The state may succeed in forcing testimony by such tactics, or it may not. That testimony may be helpful to the prosecution, or it may not. It may result in a guilty verdict, or it may not. The idea that jailing a reluctant complainant will invariably improve the chances of convicting the offender is unsupported by evidence—in fact, such a move may well have the opposite effect, by inducing a complainant to flee, or recant, or testify unconvincingly.

And even if the testimony is obtained, and is effective, the practice of compelling testimony may dissuade other victims (or that victim in a subsequent case, possibly with the same perpetrator) from coming forward. That risk is highest among survivors with reason to fear the criminal justice system—the poor, the homeless, the undocumented, the addicted, sex workers, LGBT people, people with disabilities, and so on. These communities are already disproportionately victimized by sexual assault and domestic violence, and coercive prosecutorial tactics that victimize them again are completely antithetical to curbing such violence.

It must also be recognized that rapists and abusers understand the power of the state and how it can be wielded against victims, and they use this knowledge to facilitate their crimes. Perpetrators of domestic violence frequently threaten to turn their victims over to immigration officials, child protective services, or other authorities. The more widespread the judicial coercion of victim testimony becomes, the more the threat of incarceration by frustrated prosecutors will be deployed by abusers to perpetuate their own violence.

For all of these reasons, prosecutors and judges rarely resort to arresting assault and DV survivors in order to compel their appearance in court. (One law journal article on the subject found anecdotally that prosecutorial coercion of uncooperative victims’ testimony was used far more often to justify dropping charges than to secure convictions.) The practice is strongly discouraged by the U.S. Department of Justice, and agencies participating in DV projects funded under the Violence Against Women Act are urged to embrace “procedures that provide victims with the opportunity to make an informed choice about whether to testify” instead.

Similarly, a model framework for domestic violence prosecution recently promulgated by domestic violence experts and adopted in St. Paul, Minnesota, declares that the use of coerced complainant testimony is frequently “counterproductive, increasing risk to victims and decreasing offender accountability overall.” The framework imposes nearly a dozen stringent restrictions on the use of the tactic, declaring that it should only be used where necessary to prevent “serious or lethal harm” to the victim herself.

In her article, Marcotte frames the prosecution of domestic violence as a zero-sum game, in which solicitude for the rights and concerns of the victim must be weighed against the state’s interest in punishing perpetrators. In reality, the opposite is true: The project of holding rapists and abusers accountable for their crimes is advanced, not impeded, by respect for victims.

When police, prosecutors, and judges act with respect for the autonomy of those whose autonomy has been violated by domestic violence and sexual assault, survivors are more likely to report, more willing to cooperate with the legal system, and more able to advocate for their own interest in securing justice (however justice may be defined). This fact is one of feminist legal theory’s great contributions to the fight against sexual violence, and it is one of the great engines behind the victories we have begun to see in curtailing such crimes.

Marcotte writes that there is “no silver bullet … that will get all victims to see things the way prosecutors want them to.” But seeing things the ways prosecutors want us to is not an obligation that DV and sexual assault victims—or any of us—have. Most progressives understand that, as do most feminists, many prosecutors, and the Department of Justice itself. It’s a shame that Marcotte does not.