Policy testing has been good for China

IN MAY 1919 John Dewey, an American philosopher, begins a lecture tour of China. “We’re going to see more dangerous adventures in life here,” he predicted. The celebration of learning through his work and social experimentation was enthusiastically embraced by the country’s brave reformers and dangerous revolutionaries. At least one of his speeches was attended by a young school teacher named Mao Zedong. “Everything through testing,” Dewey announced during his visit. Chairman Mao will later repeat this line as ruler of China.

Experiments on the scattered bases occupied by Chinese communists before 1949 were inevitable, Sebastian Hellman noted in his book, The Red Swan: How Unorthodox Policymaking Facilitated China’s Rise. The communists lacked manpower or administrative reach to impose a uniform policy. Instead they introduced new measures, such as land reclamation at model villages or “experimental points”, before they spread across the “surface” of their territory. The purpose was to learn without doing anything excessively catastrophic. This “model experience”, Mao writes, is “much closer to reality and richer than the decisions and directives issued by our leadership organs”.

After Mao, Chinese leaders adopted a similar “point-to-surface” approach. In fact, according to a recent study by Shawda Wang of the University of Chicago and David Young of Harvard University, the federal government has begun more than 630 such experiments since 1980. It has experimented with carbon trading, fisheries insurance, business licensing and fiscal devolution. A report by China’s Planning Agency last month outlined pilot schemes that cover everything from cross-border e-commerce and housing provident funds to green electricity and recyclable packaging.

These trials are not mere formalities. The results can go both ways. According to Messrs. Wang and Young, about 46% of experimental policies are never implemented nationwide. A failed trial can still provide useful lessons for future reform. Failure, as Mao once said, is the mother of success: “falling into the pit” can “gain your intellect.”

China has actually gained a lot by using this method. This is a “huge improvement” in a “rejection world” where all central policies are implemented without any scrutiny, Mr Wang argued. According to Mr. Hellman, the point-to-surface strategy is one of the reasons why Communist China has survived and advanced, and even other socialist regimes have stagnated or collapsed. Such unexpected results are sometimes described as “black swans”. In the case of China, he argues, red seems to be the more appropriate color.

Despite this long and celebrated history, China is surprisingly bad in policy testing. Its trials are not as clear as they could be, skewing the conclusions drawn by its leaders. One problem is their location. According to China’s planning agency, “sites should be fairly representative.” But contrary to this correct advice, according to Messrs. Wang and Young, 80% of the experiments have taken place in more affluent areas than the average since the 1980s. Another bias is financial. When local authorities examine a field of policy, such as education or agriculture, they spend 5% more in that area than in similar counties that do not participate in the test.

Tests can also be skewed by less measurable factors. Some local officials, for example, put more effort into this pilot practice than others. This is especially true of aspiring young cadres who are more likely to be promoted, as they are still far from retirement age. To measure this extra effort, Messrs. Wang and Yang created an intelligent proxy. They compare the language used by the local government to describe the test. Leaders who have more space for propaganda distinguish their language from the boilerplates used by their upward stagnant opponents elsewhere.

Excessive labor, high costs, and unusual prosperity can all make policy test results skewed. Some of these biases may be familiar to experienced Beijing policymakers. But then national leaders do not act as if they are aware of them. They favor the successful test regardless of the actual source of that success. The richer an experimental site, the more likely it is that the policy will be adopted nationwide. The host county is more likely to enjoy such a financial loss during the trial if such support is likely, as a sudden drop in interest rates increases the value of the land. The central government does not seem to exclude the virtues of an innovative policy from the idiosyncrasies of the places that drive it.

A duck dressed like a swan

It has national consequences. When the new policies spread across the country’s surface, the areas that are most closely related to the experimental “points” that benefit the most are judged by their subsequent economic growth. Since test sites tend to be richer than average, test-based policies can “systematically adapt” China’s richest parts, argues Messrs. Wang and Yang. This is not a result that Mao or Dewey would have welcomed: discrimination by experiment.

How can China reform this engine of reform, its experiments bring it closer to reality? Researchers suggest a useful place to start another interesting calculation. They point out that local officials are 22% more likely to be promoted if they take a successful test. To improve this strategy, therefore, Chinese leaders need to fix the politics involved in it. According to Mr Hellman, in recent years, under Xi Jinping, experiments have become “coercive and deceptive”. Local administrators slightly “enjoy the position and they are afraid to make policy and ideological mistakes along the way”. If local policy makers fear falling into the pit, there will be no gain in intelligence. 3

For more expert analysis of the biggest stories in economics, business and markets, sign up for our weekly newsletter Money Talks.

This article was published in the print edition of the section on finance and economics, entitled “The Red Swan Over China”.

Leave a Reply

Your email address will not be published.