Read about the failures of post-war gender quotas. Received this from a friend? SUBSCRIBE [[link removed]] CRITICAL STATE Your weekly foreign policy fix.

If you read just one thing… ...read about regular old bias masquerading as algorithmic bias.

Many tech companies have to undertake the long process of developing, testing and deploying a complex algorithm in order to inject systematic [[link removed]] racial [[link removed]] bias [[link removed]] into their products, but the social media analysis company Dataminr found a short-cut: just let analysts be racist and call it an algorithm. Dataminr, a Twitter partner company with a range of national security and law enforcement contracts, claims that it uses artificial intelligence to flag certain people to police as likely gang members based on their social media posts. In reality, sources inside the company told [[link removed]] an Intercept reporter, its employees trawl social media for anyone they think seems like a gang member and send alerts to police based on those hunches. With no internal definition of what kinds of social media posts might indicate a likely gang member, the process relies on the guesses of a mostly white group of analysts to determine who looks like they’re in a gang. It's about the results you’d expect from that description, with the company telling cops to watch out for particular Black middle schoolers based on a Twitter hashtag.

Distributed censorship

One of the major differences between the US and Chinese economic systems is that US corporations pay under-qualified employees to conduct social media surveillance on the government’s behalf, while Chinese corporations outsource social media surveillance on the government’s behalf to users who will do it for free. Censorship [[link removed]] on Weibo, China’s answer to Twitter, is prevalent, and done substantially by nationalistic users who report content they see as “trying to divide China.”

State censorship of the site has grown since 2011 when activists began to use the site to organize protests against government policies.

Some of the monitoring of the site is done directly by the government, but increasingly censorship comes in the form of “account bombing,” in which hundreds or thousands of pro-government users report a certain account until the company deletes it.

FORWARD TO A FRIEND [[link removed]] From campaign rhetoric to foreign policy

Donald Trump and Joe Biden have turned the US presidential race into a contest of who can offer the loudest anti-China rhetoric, but will that matter for US policy after the election? A new article [[link removed]] in Legislative Studies Quarterly suggests that it will. By comparing anti-China TV ads during 2010 congressional races to bills co-sponsored in the subsequent Congress, researchers found that legislators who get elected on an anti-China platform tend to try and deliver while in office.

In 2010, 34 congressional candidates ran ads attacking their opponents for being “soft” on China, with more ads coming from Democrats than from Republicans.

Challengers who attacked incumbents over China policy and were then elected were more likely to support anti-China security legislation than other members of Congress. On the flip side, incumbents who won reelection despite being attacked for their position on China police were more likely to support anti-China legislation in all issue areas than their colleagues.

FORWARD TO A FRIEND [[link removed]] DEEP DIVE No John, it really is the institution: Part II

Last week on Deep Dive, we looked at a paper that investigated the source of institutional strength in revolutionary democracies. The authors found that democratic institutions grew stronger in the Reconstruction-era American South in the areas where federal soldiers were deployed. Black communities were better able to build institutions to access power in a democratic system that could withstand eventual violent white backlash when they were protected in the early years by a strong-armed force.

This week’s paper tracks a different evolutionary pattern. Revolutionary democracies are fragile, but the opposite is true of revolutionary autocracies. Non-revolutionary authoritarian governments last just 15 years on average, while revolutionary authoritarian governments average nearly 40 years in power. The question is, why does the revolutionary origin of an authoritarian government matter so much in determining its durability? In a new article [[link removed]] in the journal World Politics, political scientists Jean Lachapelle, Steven Levitsky, Lucan Way, and Adam Casey locate the source of durable autocratic regimes not in the safety of the regimes’ early years but in the chaos of that time.

The existing theories of why governments last for long periods have little to say about regime origin, Lachapelle et al. write. Regime origin doesn’t determine a country’s economic growth or natural resource allocations, both of which tend to help authoritarians stay in power. Strong ruling party institutions — another indicator of authoritarian longevity — are more likely the result of a revolution than the cause of one. There is, then, something particular about revolutions that produce long-lasting autocratic rule.

For Lachapelle et al., that particular thing is the disorder of the revolution itself. The period of upheaval brought on by the revolution, they argue, creates a situation in which budding revolutionaries have to achieve supremacy quickly in order to ever take control of the state in the first place. When creating order from chaos, there are powerful incentives to create institutions that lend themselves to autocracy. If the institutions created are powerful enough to take control of the revolution, then it stands to reason that they’ll be powerful enough to remain in place for a long time.

To test their theory, Lachapelle at al. looked at all the autocratic regimes since 1900 and measured their likelihood to fall for each additional year they stayed in power. By controlling for economic growth, prosperity, population and oil and gas production, they set aside all the popular explanations of authoritarian longevity that do not relate to revolutionary origins. Holding those variables equal, growing out of a revolution indicates that an authoritarian government is 74% less likely to fall in any given year than a non-revolutionary authoritarian government.

With the effect of revolutions established, the authors then dug into the question of whether revolutionaries actually produced more effective authoritarian institutions early on in their rule than non-revolutionaries. By measuring civil society strength, military size and party control over the military and government, they checked how quick and effective autocrats were at seizing control. Under revolutionary autocracies, civil society was weaker, militaries larger, and parties stronger, in control of both the military and the government than under non-revolutionary autocracies.

To say that the state makes war and war makes the state is hardly a new observation in political science. Yet, Lachapelle et al. detail how particular forms of ambitious, ideological struggle incentivize particular forms of statehood. An autocracy that brought its form of order from chaos will likely outlast one that simply transitioned from another form of state order.

FORWARD TO A FRIEND [[link removed]] SHOW US THE RECEIPTS

Shirin Jaafari profiled [[link removed]] Iranian human rights lawyer Nasrin Sotoudeh, on the occasion of the release of a new documentary about her life. Sotoudeh, a leading advocate for women, LGBTQ people, religious minorities, and prisoners of conscience in Iran, is serving a 38-year prison sentence herself after representing women who removed their headscarves during a protest. The documentary was shot in secret before Sotoudeh’s most recent arrest, and captures footage of her arguments before Iranian courts and of her daily life outside of her work. Sotoudeh is currently having heart problems in prison, and her husband reports that the Iranian state is denying her medical care.

Critical State’s own Laicie Heeley is back [[link removed]] with another season of “Things That Go Boom,” the Inkstick Media and PRX podcast about current issues in national security. The new season starts Nov. 9, and the preview reminds us that, though it feels like the US presidential election is the end of the world, it is actually just a scene-setter for how the world confronts a range of other problems. And some of those problems, from climate change to nuclear war, could actually be the end of the world.

Halima Gikandi spoke [[link removed]]to Sudan watchers about the recent deal struck to remove Sudan from the US government’s list of state sponsors of terrorism. The designation subjects Sudan’s government and people to harsh financial sanctions that make it difficult to bring money into the country. As part of the deal, the new Sudanese government paid $335 million to American victims of terrorism sponsored by the previous Sudanese regime of Omar al-Bashir. Final removal from the list will require an act of Congress, but many Sudanese are already excited about the prospect of renewed economic opportunity.

FORWARD TO A FRIEND [[link removed]] WELL PLAYED

Clearly, the first three lightning strikes weren’t enough of an indication that just standing a 70-foot metal missile with a nuclear warhead inside in the middle of a field and hoping for the best was an unsustainable [[link removed]] strategy.

Fort Bragg’s Twitter presence died of thirst [[link removed]] last week, but the OnlyFans model who prompted the terminal display of horniness isn’t mourning [[link removed]].

The occasional Critical State series, “What’s Going On With These United Nations Training Videos?” continues this week, with some puzzling [[link removed]] animation choices from the Office of the High Commissioner for Human Rights.

There’s a debate in the US about whether giving police military equipment leads them to do more violent, risky and escalatory things. There’s plenty of quantitative data suggesting that it does [[link removed]] , [[link removed]] but people still wonder whether just having certain capabilities will lead cops to make foolhardy decisions they otherwise would have avoided. Enter this [[link removed]] video, from Japan, about what happens when you even give police power tools, much less armored personnel carriers.

When reached for comment about General Richard and his wall decorations [[link removed]], Ayatollah Khamenei referred reporters to his spokeswoman, Mariah Carey. [[link removed]]

Three years of law school and six figures of student loans were all worth it for a lawyer to get to say [[link removed]], in federal court, “I understand juggalo, but then they do boogaloos?”

There’s no shot of the whole parliament for comparison, but it’s hard to imagine that Ethiopian Prime Minister Abiy Ahmed is wrong here [[link removed]].

FORWARD TO A FRIEND [[link removed]] Follow The World: DONATE TO THE WORLD [[link removed]] Follow Inkstick: DONATE TO INKSTICK [[link removed]]

Critical State is written by Sam Ratner and is a collaboration between The World and Inkstick Media.

The World is a weekday public radio show and podcast on global issues, news and insights from PRX, BBC, and GBH.

With an online magazine and podcast featuring a diversity of expert voices, Inkstick Media is “foreign policy for the rest of us.”

Critical State is made possible in part by the Carnegie Corporation of New York.

Preferences [link removed] | Web Version [link removed] Unsubscribe [link removed]