“To err is human,” Deng Yuwen, the former editor of Study Times, wrote in the South China Morning Post today. “And a leader who tolerates no checks on his power is even more likely to err, because power can make one arrogant and impervious to other views.” As the National People’s Congress opens tomorrow, one crucial focus will be a proposed change to China’s Constitution removing term limits for the presidency. As many commentators have warned this week, such a change would pave the way for Xi Jinping to rule for an indefinite period, and would undo a key political reform measure introduced by Deng Xiaoping. Potentially vesting great power in the person of Xi Jinping, as opposed to collective leadership, it would also raise serious questions about whether Xi’s leadership can be sufficiently adaptive, or whether it will, as Deng Yuwen said, be “even more likely to err.” There has long been reason to wonder, in fact, whether Xi Jinping isn’t sealing things up so tight that the system is gasping for air. In late 2015, one year before his elevation as “the core,” Xi Jinping put his foot down on “improper discussion of [the policies] of the central Party,” or wangyi zhongyang (妄议中央), a phrase that was added to the Party’s updated Disciplinary Regulations. Don’t chatter about the Party’s business, he seemed to say: Just follow my lead. Visiting key Party media a few months later, Xi gave a speech in which he stressed that the media “must be surnamed Party” — meaning that they must do the Party’s bidding. In the same speech he mingled oil and water by stating that “positive propaganda,” which has traditionally signalled suppression of critical news coverage, is at one with “supervision by public opinion,” the phrase that has long been synonymous with more probing, and even investigative, coverage. “Supervision by public opinion and positive propaganda are unified,” Xi said. Yikes. So where, then, does China’s Mr. Personality get his information? Some technologists argue that new technologies have already presented authoritarian states like China with a solution, making possible “a big-data dictatorship.” Those in power can count on endless streams of information, from social media sentiments to credit information, all harnessed through artificial intelligence. “Given that many dictatorships collapse as a result of poor information,” Mark Leonard, director of the European Council on Foreign Relations, wrote on Project Syndicate this week, “digital technologies could become an even more powerful prophylactic against bad decision-making than term limits.” In fact, the Chinese Communist Party has long dealt with the delicate balance between information controls it sees as imperative, and the need to ensure restricted flow of valued intelligence. We can see this tension in one of the Party’s oldest systems of official intelligence, the insider news briefs and journals known as “internal references,” or neibu cankao (内部参考). Cheng Li, a senior fellow at Brookings, noted in his book The Power of Ideas that these internal references are produced by a range of actors, from think tanks and government departments. They can sometimes circulate only at the most senior levels — for example, among Politburo members — and at other times may be available to provincial or county-level leaders. During the heyday of the commercial press in China, from the mid-1990s through around the Beijing Olympics, internal references were often tip sheets used by enterprising reporters, clueing them in to potential stories. A journalist for a major Party-affiliated newspaper in Beijing once told me that his paper produced internal reference documents only on an irregular basis, whenever there were critical stories that editors or their Party superiors determined could not be made public. When they did produce an internal brief on a story, this could circulate at quite senior levels in the capital, sometimes even getting an endorsement from a member of the Standing Committee. Other media may produce internal references on a more regular basis. Systematic study of the internal reference system would present obvious challenges. But it would be fascinating to discover how the system might have changed, or is now changing, in light of the explosion of big data. My guess is that internal references remain as much or more relevant today. Despite the faith of technologists, there are plenty of facts and observations that must still derive from good old-fashioned reporting. In any case, the news media in China remain very much involved in the manufacture of internal references — a reminder of how the Party regards media as arms of intelligence gathering as well as propaganda. Earlier this month, a commentary in The Beijing News voiced concern over the abuse of two local news journalists in Hebei province who had been investigating a pollution case. According to media reports, journalists from the Internal Reference team of Hebei Television, the province’s official television network, had gone to Quzhou County (曲周) in southern Hebei to follow up on reports of pollution by a local enterprise. The journalists were mobbed and seriously beaten by assailants who also stole their filming equipment, wallets and mobile phones. At some point, one of the journalists was bound by the assailants, who threatened to throw them down a well. A subsequent investigation showed that the enterprise in question had identified itself as a goat breeding cooperative, but in fact was manufacturing industrial plastic sprays, and flouting environmental laws. The commentary in The Beijing News portrayed the case as a worrisome disruption by local thugs (with the possible involvement or negligence of county officials) of a necessary process of supervision by the media — not as an agent of the public interest per se, but as an indispensable arm of the provincial leadership. “The undercover journalists had a special mission,” the paper wrote, “to represent the Hebei Committee of the Chinese Communist Party in conducting an undercover investigation, performing a supervision role.” The paper continued in its affirmation of the role of Internal Reference-related investigations:

Undeniably, if it had not been for the investigation by these journalists, this illegal workshop would have continued to illegally pollute. . . . Protect the journalist’s right to report, and allow them to carry out their work with ease, with the courage to expose problems in society, and they will uphold the social and public interest. If the opposite occurs, and the environment for reporting grows serious, and certain people can use violence to prevent reporting, beating and threatening journalists with impunity so that they feel a general chill, then this will necessarily mean the loss of the front line in news supervision (新闻监督), and the greatest loss will be the social and public interest.

It is fascinating to note what this passage reveals about mainstream official views of the role of the “journalist.” There is clearly a righteous tone about the value of their work — and yet, there was never any real expectation of exposure in this case, not public exposure anyway. The point was internal exposure to provincial leaders, who apparently tasked the reporters with the investigation. In this case, the “social and public interest” is to be served only in secret. In fact, the “general chill” of which the writer speaks has already happened to the media in the broader sense. Nothing makes this point more clearly than Xi Jinping’s conflation of “supervision” and “positive propaganda.” In the face of severe restrictions on news reporting, one might imagine the internal reference system offering some relief in the form of premium restricted access content. But one substantial weakness of the system is that it maps right on top of the Party-state bureaucracy, meaning that provincial officials, for example, who have direct control over outfits like Hebei Television, have an incentive themselves to remove or tone down reporting of potentially damaging stories. By contrast, one reason that the investigative reporting carried out (largely) by commercial newspapers from the end of the 1990s was so effective was that it utilized “cross-regional reporting,” or yidi jiandu (异地监督), meaning that papers from one administrative region would report on malfeasance in another. A paper in Guangzhou could report on corruption at the city level in a neighboring province without immediate fear of reprisal from its Party bosses in Guangdong. Xi Jinping does not seem to have much interest in this sort of intersectional supervision. He has brought to heel China’s once unruly (by today’s standards) press. His leadership has been marked across the board by greater centralization and a top-down approach. We might suppose that information is not getting bottlenecked, that it is filtering up to Xi, and to others who make the key policy decisions — that they are not “arrogant and impervious.” But the reality is that we have no way of knowing. No one does. And that has to tell us something.