In-depth Analyses

All Data


     Charts / Figures

     Thematic Maps






Web Links







2nd Revision


Arguments - Data Quality
The statistics on meat consumption in China necessary to calculate China's (future) food demand are sufficiently reliable. There are some inconsistencies between (official) meat production statistics and data on per capita meat consumption from household surveys. However, the overall trend is clear. All statistics indicate that the consumption of meat has increased significantly in both urban and rural areas; the only discrepancy lies in the amount of this increase.
Crucial Issues
For assessing the quality of statistics on change in diet, two questions are crucial:
WB00860_.gif (262 bytes) What data sources are available to analyze change in diet in China?
WB00860_.gif (262 bytes) How reliable are the data?
blank_3.gif (810 bytes) blank_3.gif (810 bytes) blank_3.gif (810 bytes)
What data sources are available?
There are three major sources of (statistical) information on diet change in China. First, we can analyze agricultural production statistics by commodity, such as the harvested amount of wheat, rice, vegetables, fruits, or meat. These statistics are readily available from the Chinese Statistical Yearbook, and they can give us a good idea of the trends in China's agricultural sector.
Second, we can analyze data from Chinese household surveys. They include information on the amount of certain food items actually available on average for consumption.
Finally, we can use so-called food balance sheets, which not only provide domestic production statistics for the major primary commodities, but also quantify the (domestic) utilization of these commodities within a food system. For instance, for each commodity, they take into account the amount of feed, waste, and industrial processing, all of which reduce the amount that is available for direct human consumption. These food balance sheets have the great advantage that by definition they balance all sources of supply with all utilization of the commodities.
How reliable are the data?
Production data for various agricultural commodities, such as rice, wheat, or meat, are probably quite reliable in China. However, these statistics do not indicate what amounts of various commodities are actually available for human consumption. They do not include net imports of these commodities, and they do not take into account all uses of the commodities - including their use as seeds, as animal feed, and as products for non-food processing, in addition to their use for direct human consumption. Reliability problems in meat consumption statistics
Table 1
Household surveys should provide the most accurate representation of China's food supply. There are, however, some methodological problems: household consumption statistics require that household members provide a detailed, complete, and accurate account of the foods they have consumed during a certain period, both at home and in restaurants. It is doubtful that all the households interviewed are able to provide this information. With this in mind, it is not surprising that available food consumption statistics based on Chinese household surveys are somewhat inconsistent with official production data.
In theory, food balance sheets should take into account the complete food chain, from the field to supply in the shops. They are an accounting scheme that balances, for each agricultural commodity, production, exports, imports, overall domestic supply, use as feed or seeds, waste, inventory change, and direct use for human consumption. The FAO has developed food balance sheets for all countries in the world. The data have been available from the FAOSTAT system since the early 1960s. The balances are based on official data reported by the countries. The FAO extensively cross-checks these data and makes adjustments where necessary, so the food balance sheets are usually considered to be quite accurate. Ideally, the FAO food balance sheet on China should provide a complete macrodescription of the national food system. However, the balance sheets can be only as good as the primary statistical information. Unfortunately, these detailed statistical data  - especially information on the amount of waste - are not always available for all commodities. Missing data have to be estimated so that the sheet can be balanced. Such estimations inevitably introduce inaccuracies. Moreover, food balance sheets only describe the overall national supply; they do not take into account regional variations, which are probably substantial. Another disadvantage of food balance sheets is that they only provide data for the domestic food supply. They do not indicate how much of this overall supply actually reaches the consumer. They also cannot distinguish the supply of urban and rural, or poor and rich households.
When we cross-check the various sources of statistical information on food consumption in China, some inconsistencies become obvious. For instance, if we multiply per capita meat consumption from the household surveys by the population size, we do not get a total meat consumption that matches the total domestic supply (see Table 1). These inconsistencies might be caused by incorrect responses to the survey, by a sample that is not representative of the whole population, by differences in commodity classification, or by incorrect production statistics. However, despite these inconsistencies, it is obvious that all available data point toward a significant change in diet in China.
Related Arguments

Diet Change:   Trends     Impact    Data Quality    Prediction Error    Intervention Possibilities    Intervention Costs

Revision 2.0 (First revision published in 1999)  - Copyright 2011 by Gerhard K. Heilig. All rights reserved. (First revision: Copyright 1999 by IIASA.)