Thursday, January 2, 2014

China's Dirty Land Statistics

China's Ministry of Land Resources has announced that their latest land survey (which is actually four years old) discovered that the country has 11-percent more cropland than they have been reporting since 2008. They were really lucky to find this land, because agricultural officials are about to declare large swathes of cropland off-limits to production due to pollution with toxic heavy metals or vulnerability to erosion. And, by golly, when this poor-quality land is removed from the new total, China will still be above the minimum amount of land needed to maintain "food security." How about that?

The "new" land survey reports that China had 135.4 million hectares of cultivated land as of December 31, 2009. That exceeds the total reported for 2008 by 13.7 million hectares, or 11 percent. (The 2008 data were described as results of a Ministry of Land Resources survey.) It's also bigger than the total of 130 million hectares found in the country's 1996 agricultural census.

Actually, the survey found more of every kind of land. Somehow, in one year China added 17.9 million hectares of forests, 25.5 million hectares of grassland, 3 million hectares of orchards and tea plantations, while also devoting 1.8 million hectares to mining, industrial and residential uses and 5.4 million hectares for roads and rails. Not only is China an "economic miracle," the country also apparently has figured out how to create more land! What an amazing country (not!)

China's land area statistics,
then and now
  2008 2009  
land survey)
---------million hectares-------
Total of uses below 689.9 770.8 80.9
  cultivated cropland 121.7 135.4 13.7
   orchards            11.8 14.8 3.0
   forests 236.1 254.0 17.9
   grassland/pasture          261.8* 287.3 25.5
   other land for agricultural use 25.4 na** -25.4
   residential, industrial, mining use  26.9 28.7 1.8
   transport infrastructure       2.5 7.9 5.4
   water--reservoirs and irrigation                       3.6 42.7*** 39.0
Source: 2008 data published in China Statistical Yearbook; 2009 data from Ministry of Land Resources "2nd land survey".
*Yearbooks from 2008 to 2012 reported exactly 400 million hectares of grassland.
**No "other agricultural" land was reported by the 2009 survey.
*** "water" for 2009 may include fish ponds that were included in "other agricultural" in 2008.

A press conference was held to explain why the cultivated land total increased. The Minister of Land Resources explained that the second survey used new standards, was more complete, and used better methods. Moreover, farmers used to underreport land when it was subject to tax, but now they report their previously hidden land so they can get subsidies or rent it out.

A careful comparison of these figures with data on land area published in the past suggests statistical subterfuge. First, the Ministry appears to have kept the 2nd land survey results secret for over three years. Why did it take the Ministry of Land Resources four years to report data for December 2009? The Ministry's 2009 "Land Resources Communique" (published April 2010) reported that the second land survey was progressing well and that all rural units had reported their data--surely the survey was finished in 2010 sometime. The 2008 land survey results had been published the year following the survey.

Officials were quick to emphasize that China cannot loosen its protection of agricultural land despite the 13.7 million hectares discovered by the survey. China still has a per capita cropland endowment half the world average.

The Minister of Land Resources also emphasized the poor quality of much of the land. He said 5.6 million hectares is in regions vulnerable to floods, 4.3 million hectares are on slopes of 25 degrees or more, and the Ministry of Environmental Protection has found that 3.3 million hectares of land has medium-to-high levels of heavy metal pollution. Some has been damaged by mining and subsidence. These problematic lands add up to 12.2 million hectares, slightly less than the 13 million hectares "discovered" by the survey. So, really, said the Minister, we still have just slightly more than the minimum "red line" cropland area. Pheww, what a relief!

The release of the survey results appears to be setting up a new policy that will take land out of agricultural production in 2014. By adding to the land total now, taking land out of production will be less alarming. The Minister gave a pronouncement that polluted land should not be used for growing crops, and he said the government will spend billions of yuan annually to rehabilitate polluted farmland and groundwater. The Minister raised concerns about serious degradation of grasslands, salinization, and desertification, and he said there is room to retire more erodible cropland using the "grain for green" program. It has been rumored that such ecological programs will be a feature of the 2014 "no. 1 document." In December, after the 10th-straight increase in grain production has been announced, agricultural officials began to warn listeners not to expect such increases to continue in the future. Authorities are probably releasing this larger number now so that the total will still be above 120 million hectares when land is retired by these new programs.

Reviewing the history of cropland statistics prompts more suspicions that cropland data are manipulated. In the 1990s China reported a total of about 95 million hectares of cropland. That total was widely believed to be understated, but it was consistent with historical statistics since 1949 (see chart below). The 130 million hectares reported after the 1996 census was far larger than any cropland total previously reported since 1949 (it took several years to report this number, reportedly because there was much argument over what the number should be). In the 2006 agricultural census, the total was cut to 121.7 million hectares and the 2008 land survey found a suspiciously identical number. Both numbers were slightly above the minimum "red line". Every statistical yearbook from 2008 to 2013 reported the 2008 land number.
"New total" is from "2nd land survey". Other data from Statistical Yearbooks.

Does a larger cropland area mean crop production is actually larger than statistics indicate? Or does it mean that yields are overstated?

The chart above shows a related statistic, the area sown to crops. This can exceed the cropland area if some land is planted with more than one crop in a year. However, one would expect the area sown and cropland area statistics to move up and down in sync, more or less. The chart above shows that area cultivated and area sown have not been consistent. Statistics on area sown were never revised after the 1996 revision of cultivated area statistics. (Lore has it that the statisticians had already factored in their estimate of how much cropland area was understated when they estimated area sown.) At the press conference, the vice director of the statistics bureau glossed over this question, saying the cultivated land and area sown to crops are from different surveys. Statistics on area sown to crops have been rising over the past decade while cultivated area fell--until the new revision.

One of the communist party's rent-an-experts from the Academy of Social Sciences insisted that the cropland revision has no bearing on grain production estimates. NBS's description of its methods suggest that the change in sown area is based on village surveys. But how does NBS estimate the national total? If NBS does not base total sown area on cultivated area, how does NBS estimate total area sown to crops? The pieces of the puzzle suggest that total sown area (and by extension, crop production) is a number fabricated by the statisticians.

No comments: