On Nicholas Kristof and the Chinese Internet Censorship Loophole

New York Times op-ed columnist Nicholas Kristof published an article describing how he tested the limits of the Internet in China. He set up blogs on Chinese hosting platforms and found that politically sensitive words typed into posts were not deleted but replaced with asterisks. He took this as evidence that Chinese Internet police could not keep up with the explosive growth of blogs, and declared it a small victory for netizens.

The argument has a basic problem: everyone already knew this. The loophole was not a secret.

Internet service providers in China operate in a very competitive market. To attract users, most providers walk close to the edge of permissible behavior. One common practice is replacing censored keywords with asterisks rather than removing posts entirely—technically a violation of state censorship rules, but tolerated by Internet police under an implicit arrangement: authorities want a quiet life; providers want profit. Neither side forces a confrontation unless something creates a public embarrassment large enough to require a response.

The problem with Kristof’s article is that it created exactly that embarrassment—in the pages of the New York Times, an international audience. Once the loophole was publicized at that scale, the state had no choice but to react. The Internet police, now visibly humiliated, faced institutional pressure to clamp down. The people who paid the price were the very netizens Kristof claimed to be championing.

This is a recurring dynamic in Western coverage of civil society workarounds in authoritarian states. Publicizing the loophole for the sake of a compelling newspaper column trades a quiet, sustainable freedom for a brief moment of attention—and the people on the ground absorb the consequences while the columnist moves on to the next story.

Whether it is net censorship, informal markets, or bureaucratic tolerance of small transgressions, the operating logic is the same: these arrangements function because both sides pretend not to notice. Public disclosure shatters that pretense. The question is not whether the government is wrong to censor—of course it is—but whether the disclosure actually helps the people inside the system or primarily serves the commentator’s narrative.