How Human Rights Watch counters atrocities

Human Rights Watch is one of my favorite non-effective-altruist-identified nonprofits. I’ve often referenced their reports when I want an unbiased, objective source about what atrocities are happening in a particular country. So I was eager to read Righting Wrongs, which is a memoir by their former CEO, Kenneth Roth, about his experiences working there.¹

You might wonder whether Human Rights Watch is doing any good. Observably, many authoritarian dictators exist and innovate endlessly in terms of human cruelty. Is Human Rights Watch managing to accomplish anything other than issuing reports for me to reference in blog posts?

The effectiveness of Human Rights Watch

To be sure, Human Rights Watch will never be able to prevent all human rights abuses. But it can make committing human rights abuses somewhat more expensive on the margin. Remember your introductory microeconomics. If an action is more costly, people will tend to do less of it, or will tend to do it less intensely.

Human Rights Watch does sometimes have unambiguous victories:

Joseph Kabila [President of the Democratic Republic of the Congo] was one of my oddest interlocutors. Despite his corruption and other abuses, he cared about his human-rights image and was eager to please. For example, he surrendered more suspects to the ICC than any other national leader.

I first met him in 2009 in Goma. We had just published a report on rape by Congolese security forces in eastern Congo, and I wanted to discuss how he could stop it...

After a few minutes, a young man in blue jeans walked up and sat down in the presidential chair. It took me a moment to realize it was Kabila. We had a constructive conversation about the rape problem, during which I outlined three things that I hoped he would do to curb it: (1) Announce a “zero-tolerance” policy; (2) prosecute any rapist; and (3) dismiss commanders who tolerated rape. In a speech two days later, he announced a plan to pursue all three recommendations. Within a short period, two senior army officers had been convicted of rape, and two others were arrested. These initial steps sent the message that there would be consequences for using rape as a weapon of war, even if Kabila’s follow-through was, as typical for him, less than vigorous.

The oddest but perhaps most important meeting I had with Kabila was in 2015 in Kinshasa. His two constitutionally permitted terms as president had ended, but he resisted stepping down. He did not have the political clout to change the constitution to permit a third term, as some African leaders had done, so he was engaged in what the Congolese called “glissement”—sliding into a third term. Protests were mounting, and protesters were being shot. Diplomats had been pressing him to move on, but I suspected they had been doing so, well, diplomatically. Our strategy for the meeting was guided by our research, which revealed that Kabila was not necessarily intent on becoming president for life but was worried about his security and future should he step down.

Kabila received Ida and me cordially. I laid out two possible paths that I saw for him. He could allow elections to proceed and hand power to a successor in what would be the first democratic transition in the country’s history. He would be celebrated as the father of Congolese democracy and could look forward to many years as a respected statesman. I said this knowing that Kabila liked mainly to spend time on his ranch playing video games, but I tried to make the option sound attractive. The other path was to continue his glissement. There would be more protests, more security-force shooting of protesters. Ultimately, despised, he would be forced to resign. Prosecution and imprisonment would be real possibilities.

Kabila wound up leaving office and being replaced by someone who wasn't his puppet, but who also wasn't the person the Congolese people had voted for. As peaceful transitions of power go, Kabila leaving office is lacking—but it's significantly better than the Democratic Republic of the Congo had managed previously.

But in the absence of an authoritarian leader as cooperative as Kabila, Human Rights Watch’s work is even more a matter of partial victories and incomplete successes. A government stops using chemical weapons on civilians for five years, before it returns to doing so. A dictator doesn’t hold free and fair elections, but she also doesn’t jail opposition leaders; a dictator jails opposition leaders, but she doesn’t order them tortured. People are trapped in overcrowded and unsafe refugee camps, but they at least receive food aid shipments.

Do these victories matter, as small as they are? Well, if someone was using chemical weapons on me, I would sure appreciate a five-year break. A slight decrease in the number or intensity of human rights violations is measured in lives—and not dozens of lives but hundreds and thousands. Moral purity is a luxury for people whose primary connection to human rights abuses is sharing viral articles about them on Bluesky.

So how does Human Rights Watch make actions more costly?

Roth explains a standard therapy-culture read-too-much-Brene-Brown distinction between shame and guilt. Guilt is internal, coming from your own sense of right and wrong, your own feeling that you’ve violated moral standards. Shame is external, coming from the fear that other people will disapprove, the thought of a thousand judging eyes all watching you. When reading this, I thought I knew what would come next: shame is destructive and harmful, guilt is good and drives moral behavior, we must all overcome both our own shame and our desire to shame others to bring about a better world.

And then Roth elaborates:

Shame is different from guilt. Human-rights abusers may feel no personal remorse whatsoever, but they would prefer to avoid the opprobrium of others. Our job was to force them to behave better regardless of their character.

Human rights abusers probably don’t have consciences; if they did, they’d be pediatricians or public-interest lawyers or technical AI safety researchers. Appealing to their better nature is pointless. If they had a better nature, they wouldn’t be torturing people in the first place.

But human rights abusers can be shamed. Look at what human rights abusers do. Why would Russia and China bother to have troll farms or state propaganda outlets if they didn’t care about their reputation, both internationally and at home? That’s not to say that reputation is an overwhelming consideration that overrules everything else. But because human rights abusers care about their reputation, threatening their reputation can cause a decrease in the number of atrocities on the margin.

(Note that human rights abusers don't necessarily feel the emotion of shame when Human Rights Watch shames them—just that they’re motivated to avoid social disapproval.)

Why do human rights abusers care about their reputation? Human rights abusers like being powerful, and that doesn’t just mean that they like being able to order the horrifying deaths of children. They like all the trappings of power.

They like going to important meetings with other heads of state where they both give statements to the press about their countries’ strong personal relationship. They like being invited to fancy summits where they make speeches about the importance of international cooperation. They like attending parties with other powerful people, drinking champagne, eating caviar, and gossiping about powerful people who aren’t in the room.

And Human Rights Watch can take the trappings of power away from them—not 100%, but somewhat and on the margin. Instead of lobbing softballs about international development, journalists start asking questions like “do you have a comment about all the opposition leaders you imprisoned?” The other heads of state feel obligated to give some statement about how the human rights abuser shouldn’t be doing all those war crimes, which totally kills the mood at the press conference. Invitations to fancy summits dry up, because it’s just too embarrassing for other countries to explain to their people why this guy was invited. When they do attend fancy summits, other leaders start vaguespeeching about how, whatever some people seem to believe, you really shouldn’t commit genocides. The human rights abuser starts having to have awkward conversations at parties about why they’re using chemical weapons on civilians! They’re just sitting there, sipping their champagne and munching on their caviar, and people come out of nowhere to go “seriously, dude, children?” Rude!

I understand this completely. If I were a political leader, I too would radically change every aspect of my policies in order to avoid awkward conversations at parties.

Further, we often think of human rights abusers as absolute rulers whose every whim is obeyed without question. But that’s not actually how running a country works. Every authoritarian dictator has to appeal to key power bases: the army, the secret police, the people who staff her administration, whoever loans her money, and most of all whoever guards her while she sleeps. And while by definition authoritarian dictators tend not to listen to the opinions of the average person, they do need to maintain some level of legitimacy and approval among their people. If a dictator becomes unpopular enough, she risks a popular revolution. Sure, the army can shoot people in the street. Sometimes that’s not enough!

And sometimes human rights abuses are committed by liberal democracies. Shame can be a particularly powerful tool in countries whose government can be voted out of office if the citizenry disapproves of their behavior. Liberal democracies are, in fact, so easy to shame that Human Rights Watch often indirectly shames recalcitrant countries through shaming their liberal democratic allies: Saudi Arabia might be hard to shame directly, but the United States can be shamed about Saudi Arabia’s behavior.

And countries also have to appeal to the international community for practical reasons, not just invitations-to-fancy-summits reasons. Poorer countries are often reliant on wealthier countries for foreign aid and military assistance; every country relies on other countries for trade and the advancement of various diplomatic goals. Human Rights Watch is often hesitant to call for economic sanctions, since these often end up hurting the exact people whose Human Rights we are supposed to be Watching. But no one needs the latest fighter jets, and military aid can often be made contingent on good behavior. Similarly, trade deals can be made contingent on (say) removing slavery from the production of exported goods.

I want to emphasize that shame often directly attacks the material interests of human rights abusers. If the United States or the European Union is unhappy with you, that doesn’t just result being iced out at parties. It can mean you don’t get the aid or trade you were counting on to feed your people (or build your thirty-seventh palace, whatever). It can mean you don’t get the tanks and planes you need to combat the Islamic State of Wherethefuckever. It can, apparently, mean that Donald Trump orders your abduction or assassination (and then your country is a trashfire but what do you care about that, you’re still arrested or dead).

The major weakness of Human Rights Watch’s tactics, according to Kenneth Roth, is that they don’t work against Kim Jong Un. North Korea’s strategies for cementing support among key power bases within its own country have nothing to do with ruling well; committing more atrocities, to some extent, actually makes elites support the government more. And North Korea has been an international pariah for so long that they have no international reputation to lose. Some dictators are, indeed, immune to shame.

But it’s not obvious to me that this is the only weakness of Human Rights Watch’s politics.

Human Rights Watch has its most unambiguous successes in (a) liberal democracies that care somewhat about human rights and (b) small countries which have relatively weak economies and are dependent on the patronage of liberal democracies that care somewhat about human rights. It is notably less successful with, for example, Russia and China. This isn’t because Vladimir Putin or Xi Jinping are immune to shame or don’t care about their international reputations. It’s because they’re powerful enough that they can tell Human Rights Watch to fuck off.

The European Union, the United Kingdom, and the United States might reconsider their relationship with the Democratic Republic of the Congo over human rights. They’re not going to reconsider their relationship with China over human rights. It would cost them too much. And, for similar reasons, China needs the United States and the European Union less than the Congo does.

Indeed, when China tries to expand its sphere of influence into Africa, a major part of its pitch is “we don’t care about human rights, if you’re in our sphere of influence you can ignore Human Rights Watch.” Dictators often prefer to be clients of Russia and China, even if they’re getting a much worse deal. For example, Burkina Faso distanced itself from the West-- in spite of serious economic costs-- and cozied up to Russia. Relatedly, Burkina Faso has banned political parties and Russia doesn’t give a shit.

Introductory microeconomics works both ways. You can deter behavior by making it costlier on the margin. But as cutting a deal with you becomes more expensive, people start to look for some other source of military assistance, economic aid, trade deals, and fancy summit invites.

How Human Rights Watch Works, When It Does

To make public shaming work, Human Rights Watch needs to be a neutral, trustworthy outside observer. How do they do this?

First and most importantly, Human Rights Watch needs to tell the truth. Kenneth Roth hits this point over and over again. Human Rights Watch never puts out a report unless it’s certain the report is accurate. Its credibility is its most important asset. Any mistake, any exaggeration, harms not only the campaign it’s part of but every other campaign Human Rights Watch will ever engage in. “They’re making it up,” human rights abusers will say, “just like they did last time.”

In human rights work, it’s very tempting to shade the truth. People are dying, after all. If you make the story just a bit more convincing, sand off just a few of the rough edges, then people might listen and you’ll save hundreds of lives. But giving into temptation is short-sighted.

Roth writes:

These retorts were not terribly effective because we were extraordinarily careful about our fact-finding. From day one, researchers were told that their top priority is accuracy. I would rather have had a researcher come home empty-handed than to return with inaccurate information. We aspired for the sober rather than the sensational, for exactitude rather than exaggeration. Because we were so careful, we welcomed battles over the facts. A government’s obfuscation and spin went only so far in the face of a detailed recitation of evidence.

Second, Human Rights Watch must be even-handed. Human rights abusers love whataboutism. “You say I shouldn’t commit war crimes,” they say, “but what about my enemy, who is also committing war crimes?” “You say that I shouldn’t imprison dissidents,” they say, “but what about the United States, which has the highest prison population in the world and which tortured suspected terrorists?”

The best possible response to this question is “here’s our report about that, and we have a meeting with them scheduled next week. But we’re not talking about them right now. We’re talking about you.”

Human Rights Watch has two researchers assigned to Israel/Palestine.² I feel like these two researchers have the worst job in the world. Not only do they have to spend all day collecting information about atrocities, half of everyone is accusing them of being motivated by rank anti-Semitism and the half of everyone is accusing them of being motivated by purest Islamophobia. The only reason Human Rights Watch has any credibility in the region is that, whenever Israel tries to deflect attention to Hamas, Human Rights Watch can go “we also think Hamas shouldn’t do that”, and whenever Hamas tries to deflect attention to Israel, Human Rights Watch can go “we also think Israel shouldn’t do that.”

Ultimately, Human Rights Watch’s moral power comes from the sense that they’re fair: they hold everyone to exactly the same standards; no one they criticize is being unfairly targeted.

Third, Human Rights Watch must only recommend actions that are possible. This isn’t really a constraint on civil and political rights: it’s always possible to hold elections, stop bombing civilians, not commit genocide, let women drive cars, and so on. But some people have declared that healthcare is a human right, or climate change is a human rights violation, or global poverty is a human rights issue.

However, no matter how enthusiastically you shame poor governments, they can’t get money from nowhere. Many poor countries—even if they were run perfectly—would be unable to provide an American or European standard of living for their citizens. And climate change is a complex problem that every country on earth contributes to; no country can unilaterally decide to cap global warming at 1.5ºC. If you shame these countries, you’re blowing political capital for no reason.

Human Rights Watch does address so-called economic rights, but only unambiguous corruption or unambiguously wasteful vanity projects—situations where there’s a specific, concrete action a country could stop doing.

Fourth, Human Rights Watch must be uncompromising, yet persuasive.

Kenneth Roth is diplomatic with human rights abusers, in that he doesn’t raise his voice or call them names. But he is factual and blunt, not mealymouthed and diplomatic. He is willing to respond to a leader’s claim that no one is being tortured with “people are literally being tortured in the basement below us right now.” He tries to treat all world leaders as his equals. He writes: “Leaders may like groveling but do not respect it.” The power of shaming comes from saying exactly what you mean.

At the same time, often countries get away with human rights abuses because important power blocs (the military, elites, the citizenry, the international community) don’t think the human rights abuses are bad. When the U.S. government tortured suspected terrorists, or Uganda imprisons people for life for having gay sex, these policies are successful because they are popular. Simply repeating “this violates international human rights law!” does nothing—no one cares about international human rights law.

Instead, Human Rights Watch tries to make the case that people should care about human rights violations. For example, they might tell the vivid stories of specific individuals, like Aleksi Navalny or Jamal Kashoggi, instead of bland statistics. Absolutely no one wants their own human rights to be violated, so a useful tactic is convincing people that they should want others to be treated with the same respect they want for themselves. Often, Human Rights Watch argues that a government that violates your enemy’s rights might violate yours or those of someone you like. Sometimes Human Rights Watch manages to appeal to a key demographic, even if they don’t manage to convince everyone: for example, urban residents might care more about rights than rural residents; a country’s diaspora might be able to read reports which are censored in the home country.

I'm not sure that I think that, at current margins, Human Rights Watch significantly outperforms cash. But I do think, if you strongly value preventing human rights violations, you might be well-served by throwing some money Human Rights Watch's way. And I think Human Rights Watch is a useful case study for people aspiring to work in policymaking or politics. Naming and shaming isn't a good fit for every issue, by any means. But I think a lot of people would be well-served by understanding what naming and shaming is, when it works, and how to do it, so that they know when to use the tactic to advocate about global poverty, factory farming, AI risk, or biosecurity.

1. I recommend this book highly. Kenneth Roth has retired and has an entire career’s worth of scores to settle.

2. This isn’t unusual. Most countries only have one researcher assigned to them; sometimes a single researcher is shared by multiple relatively uneventful countries. The largest team is assigned to the United States, which has six because it is large, powerful, prone to committing human rights abuses, and unusually responsive to shaming.