Yesterday, Rebecca MacKinnon of RConversation fame, testified before the US Congress on China’s censorship policies, especially with regard to domestic internet companies, and how those policies affect US investors.
Overall, her talk was nuanced and, I felt, quite accurate, but I do fear that given her audience and the preconceptions they’re likely to harbor, she may have given slightly the wrong impression.
Much of her talk focuses on self-censorship, or as the Chinese government prefers to call it, “self-discipline”. Ms. MacKinnon’s actual text is pretty fair, I think, but Baidu in particular certainly doesn’t come out of it well. There are two chief issues at play here, or to put it another way, two crimes of which Baidu stands accused: misleading advertising practices and overly vigorous censorship practices.
The advertising issue is essentially twofold. First, some people contend that what, exactly, is paid content and what is actual search results on Baidu is unclear. Second, some have suggested that the exodus of Google has led to absurd numbers of paid links being placed in search results.
When making a comparison with Google, it’s undeniable that it is less evident what’s an advertisement and what isn’t on Baidu, because there is no color underlay ((Actually, there are some advertisements on Baidu that appear at the top of search results with a color underlay as on Google, but for now we’ll ignore those as no one is complaining about them.)). However, all paid results seem to be marked with the characters “推广“. The characters are the same font size as the other text on the site, and they appear at the end of each listing, in gray. It’s easy to miss if you aren’t looking for it, but I can’t see anyone being tricked by that more than once or twice, especially as a native reader of Chinese. ((In my own experience, I find non-native speakers tend to “tune out” anything that isn’t clearly the core content and so probably would be more likely to ignore the 推广 notification because its placement makes it clear it is information about the webpage rather than information from the webpage, and non-native readers are likely to read as little as possible, for time purposes if nothing else.))
As for the absurd number of paid results, that appears to depend very much on what you search for. In her testimony, Rebecca MacKinnon cited a search for the term ”games“ (”游戏“） that returned ten paid results before the first real result. I am inclined to agree this is excessive. One would expect that on a search engine, after searching, one could see search results without having to scroll down at all. That seems reasonable, especially given today’s large screen sizes. On the other hand, I am now complaining about having to scroll down — perhaps this is a bit nit-picky. In any event, other search terms fare much better. A search for “console games” (“电视游戏”) returned no paid results. Neither did searches for “soccer” (“足球”), “World Cup” (“世界杯”), or even, rather surprisingly, “clothes” (“衣服”). Other searches turn up some, but not ten, paid results; ten appears to be the maximum possible number of paid results, and while that is a bit much, it doesn’t seem to be the norm, even with search terms you would expect to draw hordes of advertisers.
The other accusation leveled at Baidu specifically, and the one I think may be unfair, is that Baidu is more trigger happy than other domestic companies when it comes to censorship. To support this, MacKinnon compiles the results of several studies, including one of her own. For her methods, I suggest reading her paper, which is fascinating, but the short version is that her team found after some fairly rigorous testing that Baidu ranked third among 15 blog and BBS service providers in terms of how gung-ho they were about censorship. She declines to name any names specifically in the paper for fear of government repercussions against the companies that censor least, but in her testimony she fingers Baidu at third and also says that Tianya (a company Google has invested in, interestingly enough) is in first.
It’s difficult to measure scientifically how much a company censors, and it probably changes over time according to government directives; under the circumstances MacKinnon’s exhaustive research is probably about the best we can hope for unless someone chooses to leak an awful lot of secret documents from the government and major internet companies. Additionally, MacKinnon points to research from human rights groups that suggests Baidu’s search results are significantly more censored than Microsoft’s or Google’s circa 2006-2008. As of this writing, neither of these sites would load within China, so there’s no way for me to analyze their methodology, but it does stand to reason that Baidu would censor more than American companies. Why? I suspect it is mostly related to the level of scrutiny they’re under as a major company whose services are used by a huge slice of the population. MacKinnon’s research covered blogging and BBS services offered by Baidu, Blogbus, BlogCN, iFeng, Mop, MSN Live, MySpace, Netease, QZone, Sina, Sohu, Tianya, Tom, Yahoo! China, and YCool. Among that group, Baidu’s services stand out as some of the most widely known and used. After all, not everyone in China posts on Mop, but a huge percentage of the market searches with Baidu, and not just on computers, on cell phones and other mobile devices as well. And when compared to Google or Microsoft, Baidu has a lot more to lose than either of those companies, neither of which was drawing substantial revenue from its Chinese search engine. It seems, then, that Google can afford to be cavalier in their approach to censorship. If they lose their license to operate in China, they’ve got a big wad of Western markets’ cash to dry their tears with. But if Baidu loses China, they’re dead. Of course they need to try harder to stay on the government’s good side.
Baidu’s executives, MacKinnon points out, have at times appeared enthusiastic about censoring; to support this she cites a recent “self-discipline award” accepted by Baidu’s CEO from the government, and his participation in singing some “red songs” with CCP up-and-comer Bo Xilai. She also mentions a leaked blog post from another Baidu exec that suggests they aren’t all that excited about censoring. The truth is hard to pin down, because Baidu obviously cannot state publicly that they oppose censorship without risking losing their license to operate in China. Nor can they really refuse to “sing red songs” or accept awards bestowed by the government without causing dangerous loss of face.
I do not mean to suggest that censorship is right, but I wonder, as does MacKinnon, what good could possibly come from “bucking the system” at this point. If Baidu refuses to censor their searches to the government’s satisfaction, they will certainly lose their license, and then what are Chinese people left with? The new People’s Daily-affiliated search engine?
Self-censorship is an evil of sorts, but in the current system, its a necessary one. If Chinese people are to have access to any kind of internet at all, there must be Chinese companies willing to engage with the government and play by its rules, at least in the short term. If everyone stopped self-censoring and lost their licenses, that would send a great message to the government — but at a tremendous cost to the Chinese people and their already limited freedom of speech.
What I hope, then, is not that Baidu stops censoring their results — at least, not yet — but I hope that they are using their clout, their increasingly commanding market share, and the power of consumer demand for unfiltered search results to influence government policy makers and lobby for a freer internet. I also hope that American policy makers and investors will pay attention to the details of MacKinnon’s nuanced testimony, and not jump to the conclusion that Baidu and other Chinese internet companies are happily complicit in government censorship. They may be complicit, but there’s little evidence they’re very happy. And more so than any disengaged company, they have a little leverage they can try to use to shove the government in the right direction. American investors should push their Chinese investments for that kind of engagement, not the satisfying-but-pointless self-destructive shutdown advocated by some to “send a message”.
That might help foreign companies look “socially responsible” but the Chinese government has shown it doesn’t really care that much about such messages, so the only real loser is, as always, the Chinese people.
For the record, I have no connection, financial or otherwise, with Baidu or any other Chinese internet company.