A Tale of Two Blue Cities

What Atlanta and Athens Reveal About Who Bears the Risk of Progressive Compassion

Five years ago yesterday, apparently, I was arguing with a then-friend about whether a convicted child sex offender should be entitled to state-funded “sex-reassignment” surgery. I know the date because Timehop reminded me.

The timing was interesting because about two weeks before, I was on a Zoom panel talking about the period when my views on gender identity began to change. That conversation already produced Woman Enough to Exploit, Not Woman Enough to Protect, but there was another moment I hadn’t finished thinking about: realizing that the transgender-identified people I knew personally weren’t necessarily all that different from the radical trans-activists online. As an example, I recalled a usually level-headed transgender-identified female friend nearly cursing me out over my objection to the Adree Edmo case.

Then Timehop handed me the screenshots. My memory had embellished the profanity a little, but not the substance.

Adree Edmo was an Idaho prisoner serving a ten-year sentence for sexually abusing a 15-year-old boy. Edmo sued the state for sex-reassignment surgery, arguing that denying the procedure constituted inadequate medical care under the Eighth Amendment. The federal courts ultimately agreed, and Edmo received the surgery in July 2020 before later being transferred to a women’s prison. After years of additional litigation, a federal judge in 2025 also awarded Edmo’s attorneys nearly $2.44 million in fees and another $45,000 in expenses.

At the time, the story had been shared in the Facebook group of a local “gender-expansive” organization I was affiliated with, where people I actually knew were celebrating the outcome. Apparently, I reached out privately to a trans-identified friend I was particularly close to because I genuinely wanted to understand why.

Reading the conversation five years later, her reasoning was actually more considered than the one profane line I remembered. She told me she had initially been “heart reacting to the headline,” but supported the decision more broadly because of what it could mean for other “transgender prisoners.” She believed inmates should have access to good medical and mental-health care and worried about establishing a legal principle under which some prisoners could be deemed undeserving of it.

“If you’re drawing the line,” she asked, “who gets to decide what that line is?”

Then came the part I did remember: “ultimately I don’t care fuck all about a pedophile but I care about inmates’ rights and trans rights at large.”

Her reasoning was not as quite as ridiculous as I remembered. Prisoners retain constitutional rights, and rules established around an unsympathetic prisoner can affect more sympathetic ones later. Five years later, I think she asked an important policy question: Who gets to draw the line?

Government does. Legislatures, courts, agencies and administrators constantly draw lines between competing rights and interests. What strikes me now isn’t that my former friend cared about Edmo’s rights, but who disappeared from the analysis once Edmo became the vulnerable person who needed defending.

The women in prison were inmates, too. Once a male prisoner convicted of sexually abusing a child was transferred into a women’s correctional institution, female prisoners became stakeholders in that decision. Yet their interests never appeared after my then-friend’s concern for “inmates’ rights and trans rights at large.”

I recognize that mode of thinking because I once shared it: identify the marginalized person, identify what they are being denied, and assume expanding access is the compassionate position. Public administration requires us to follow a policy outward—to its costs, tradeoffs and foreseeable risks, including those imposed on people other than its intended beneficiary.

Five years ago, I was asking what Edmo should receive. Today, I’m asking something else: Who bears the risk? Sometimes what we call compassion is really a transfer of risk—and that risk tends to travel downward.

Who gets to escape the consequences?

The people most exposed to the consequences of public disorder are rarely those with the greatest ability to shape public policy. A well-paid professional can support permissive policies and drive home. A woman dependent upon public transportation cannot simply opt out of the train station.

I confronted a version of this question in June, after 66-year-old Margaret Swan was randomly stabbed to death aboard a MARTA train. In Representation and Protection, I asked what public institutions owe people who cannot simply opt out when public systems become unsafe.

Atlanta received another grim example this weekend. Twenty-nine-year-old Ramon Harris was fatally stabbed on the BeltLine near Ponce City Market; police arrested Jonathan Hamby, a convicted felon Mayor Andre Dickens said had 30 prior arrest cycles and a history of mental illness. Hamby had cycled through arrests, probation, mental-health evaluations and treatment requirements. His family says they had spent years trying to get effective help and had been “desperately fighting a broken system to prevent this exact tragedy.”

We cannot know which intervention would have prevented Harris’s death. But we do know the risk was not invisible. Government had encountered Hamby repeatedly, recognized both his mental illness and dangerous behavior, and still failed to interrupt the cycle before another member of the public encountered him.

The victim in that case was male, so the policy failure I’m interested in isn’t exclusively female. But public disorder does not distribute risk equally by class or sex. Women face an additional increased possibility: that an encounter with a violent or unpredictable man can become sexual. Physical differences between the sexes matter whether progressive politics finds them convenient or not.

That is why another developing story in Athens caught my attention.

Athens, again

In Athens, police are investigating the possible sexual assault of a 20-year-old University of Georgia student who awoke near City Hall after a period she could not remember, with an unknown man reportedly masturbating behind her. Police listed an unnamed homeless person as a suspect, although no arrest had been announced as of Tuesday morning. Two other homeless men helped her get home—a useful reminder that homelessness itself is not synonymous with dangerousness.

The story comes amid an ongoing debate over how Athens manages its public spaces. Earlier this month, the Athens-Clarke County Commission rejected, 6–4, an ordinance that would have prohibited camping, sleeping and lying on sidewalks. Critics argued that the measure effectively criminalized homelessness. After the student’s report, State Rep. Houston Gaines blamed local officials for allowing homelessness and encampments in public spaces to grow.

I don’t think “criminalize homelessness” versus “do nothing” is a serious policy framework. Athens-Clarke County has committed millions in federal funding to supportive services, outreach, encampment services and low-barrier shelter…and good on them for doing so. Homelessness is itself a form of vulnerability and government should pursue ways to get people off sidewalks and into stable housing.

But compassion for one vulnerable population cannot require pretending another population has no vulnerabilities of its own. What obligation does government have to the people who use those public spaces while those interventions are being pursued? A city can owe something to the person sleeping on its sidewalk and to the woman who needs to walk down that same sidewalk. Compassion should require government to take both seriously, not declare one person vulnerable and quietly assign the other the risk.

That question is too often treated as evidence that someone lacks compassion. I increasingly think the opposite is true. A policy is not compassionate merely because its intentions are compassionate toward the person at its center. The people living with its consequences count, too.

Compassion for whom?

“Vulnerable population” is an important concept in public policy. But I’ve become increasingly interested in who gets recognized as vulnerable in the first place. Progressive politics has developed an elaborate vocabulary for describing the vulnerability of prisoners, homeless people, migrants, addicts, people with mental illness and transgender-identified people. Many genuinely are vulnerable and deserve humane treatment, but vulnerability can become a hierarchy rather than a description and the people whose interests conflict with the designated vulnerable population can fade from view.

Women present an especially uncomfortable problem for this politics because taking our vulnerability seriously sometimes requires acknowledging things contemporary progressivism would rather make negotiable: sex, physical differences between men and women, patterns of male violence, and the legitimacy of female boundaries. A woman in prison cannot identify her way out of the physical asymmetry between herself and a male prisoner. A woman waiting for a train cannot make male violence less sexed by adopting more inclusive language. A college student walking through a public space cannot make herself equally capable of defending against a larger male stranger because acknowledging that difference has become politically unfashionable.

This is part of what I was getting at in Woman Enough to Exploit, Not Woman Enough to Protect. Women—and Black women especially—can be extremely useful to progressive politics as symbols of vulnerability. Our histories can establish the moral urgency of somebody else’s claim, but when protecting actual women requires a boundary that conflicts with another progressive priority, our vulnerability suddenly becomes much harder to see.

Women aren’t the only people who disappear this way. The elderly woman who depends on public transportation, the low-wage worker who cannot Uber home, and the prisoner who cannot choose her cellblock all have vulnerabilities of their own. Yet politically, they share a disadvantage: their vulnerability does not necessarily advance the policy being sold as compassionate.

This isn’t an argument against compassion. It is an argument against compassion that sends someone else the bill. There is nothing particularly compassionate about improving one person’s circumstances by quietly transferring foreseeable danger to someone with less ability to escape it. If a policy protects the person whose vulnerability is politically legible while making another vulnerable person less safe, we should at least be honest that compassion has not been expanded, merely rationed.

And the ability to escape that rationing is itself unequal. Wealthier people can buy distance from public disorder through cars, safer neighborhoods, controlled-access buildings or simply the ability to leave. The woman taking MARTA because she cannot afford another option cannot. The prisoner cannot. The student living where she attends school cannot always. Policies made in the language of compassion can therefore produce a peculiar inversion: the people with the fewest options are asked to absorb risks created by people with considerably more options.

A politics does not become compassionate simply by identifying someone to have compassion for. The test is what happens to everyone else once that compassion becomes policy.

When does “unintended” stop meaning anything?

This is where I’ve found myself entertaining a darker thought. Sometimes I wonder whether any of this is actually by design.

I don’t mean that politicians are sitting around plotting to get women assaulted or strangers stabbed. I have no evidence for that and public administration gives us a more useful standard anyway…foreseeability.

The first time a policy produces an unexpected consequence, perhaps policymakers genuinely didn’t anticipate it. But then there is data, another incident, testimony, another victim…at what point does continuing to make the same tradeoff become a policy choice?

Hamby’s history illustrates the distinction. Intent matters morally, but it cannot be government’s only measure of responsibility. A harm need not be desired for the risk that produced it to have been repeatedly accepted. There is a difference between intending an outcome and repeatedly deciding that its foreseeable risk is an acceptable cost.

The question I should have asked five years ago

Looking at those screenshots now, I know my friend wasn’t evil. She was doing something I once did myself: beginning with the person society had designated vulnerable and reasoning outward from there.

Five years ago, she asked who gets to draw the line. Government does, inevitably. The better question is whether the people drawing it are willing to account for everyone on both sides.

That’s the question I didn’t know how to ask then. Not simply Who deserves compassion? Not even Who has rights? Who bears the risk? Because when government has been warned about a foreseeable consequence again and again, “nobody intended this” becomes an increasingly inadequate defense. Eventually, the question is not only who receives compassion, but who is repeatedly asked to absorb its risks.