Back to Portfolio for the Future™

Banned in China: High Praise for a Statistics Textbook

May 13, 2015

chinabanThere was a time when the expression ”banned in Boston” was a selling point. Indeed, the expression may have its roots as far back as 1651, when the leaders of the capital city of the Massachusetts Bay Colony were scandalized by the work of one William Pynchon, the founder of Springfield, portrayed here. Pynchon criticized pivotal aspects of Puritan theology in his tome, The Meritorious Price of Our Redemption.

I can’t say whether the controversy, and the ban in Boston, helped sales of the book. But if we fast forward to the early 20th century, the “banned in Boston” headline certainly did help Clifford Odets’ play, Waiting for Lefty, and H.L. Mencken’s magazine, The American Mercury.

Perhaps sadly, those days are gone for Boston, killed by the Warren Court and its curtailment of the power of municipalities to ban much of anything.

What about “banned in China,” though? Is that a phrase with a future?

Sobering News

This question comes to mind because the latest news out of China, from an investment standpoint, is sobering. The HSBC purchasing managers’ index (PMI) showed contraction in April. The final number was 48.9, which was itself a downward re-adjustment from the preliminary number of 49.2.

More broadly, it has become apparent that the Chinese policy makers are torn. They want to preserve or strengthen the value of the yuan re: the U.S. dollar (and a long-run goal close kin to that may even be replacing the dollar as the world’s reserve currency); but they also want to stimulate growth with supportive monetary conditions, which at this moment seems to require quantitative easing, regardless of what is happening with the dollar.

This is the sort of conundrum that creates opportunities for macro trades.

Oddly enough, at this moment there seems to be a fair amount of fluttering in social media about a statistics textbook that was banned (four years ago) in the People’s Republic: Andrew Gelman’s book, Data Analysis Using Regression and Multilevel/Hierarchical Models.

Regular readers of this blog may recognize that name. Gelman is a regular participant in discussions of the meaning and value of a Bayesian approach to statistics/probability, in financial contexts as elsewhere. He is a Bayesian of an independent sort, sometimes challenging the verities of that tribe.

Path Specificity and Ideology

His blog has the provocative title, “Statistical Modeling, Causal Inference, and Social Science.” In January 2011, Gelman told the world via that website that he had received a message from the textbook’s publisher, that “due to various politically sensitive materials in the text, the China edition has not met with the approval of the publishing authorities in China, and as such PTP will not be able to proceed with the publication of this edition.”

This made Gelman feel delightfully subversive. His best first guess as to the reason his book was considered sensitive involved an integral feature of Bayesianism – the path-specific nature of its probabilistic reasoning. Your hypothesis of the probability of H on Tuesday (the prior) is supposed to be updated according to what happens or doesn’t happen Wednesday, thereby yielding the posterior probability of H.

In the context of Chinese communist ideology, Gelman suggested, it is likely that this suggests that the discredited pre-revolutionary traditions live on as “priors” continuing to have an impact upon the present. That may be sensitive enough, at least to get a timid publisher to pull the plug.

Thus, from 2011: the re-emergence of this banning in tweets in early May 2015 was fortuitous. But it does make me wonder: maybe the authorities there want to re-consider. After all, from one point of view the “priors” on China’s financial strength are very impressive. Only the day before yesterday it seemed a world-beater. Not to superficial media pundits, but to serious folk like Niall Ferguson and Zachary Karabell.

The worries are much more recent. Tempering these posteriors with the older priors might be exactly the approach to statistics the Chinese fiscal and monetary authorities (and, for that matter, outside investors contemplating China) want to take at the moment.