107. Three papers were presented to the Commission under this agenda item. The document APCAS/02/10, "2000 Korean Agricultural Census and Future Plans for Agricultural Statistics" was presented by Mr Kim Tae Choong.
108. The Commission was informed that the Korean National Statistical Office (KNSO) conducted the 2000 Agricultural Census in March 2001 and the final report of the census was published in May 2002. Initially, the agricultural census was undertaken every 10 years. However, in view of the rapid changes in the agricultural sector, a census was undertaken in 1995 followed by another in 2000. Major changes were made in the conduct of the 2000 Agricultural Census compared with earlier censuses. First, the agricultural census and fisheries census were conducted simultaneously; Second, new items such as environment-friendly farming, participation of producer's organizations, e-information use, and housing and living conditions were collected in the 2000 Census. Finally, to reduce non-sampling errors, the questionnaire was revised substantially and supervision of fieldwork was tightened.
109. Mr Kim Tae Choong reported that the agricultural situation in Korea had changed. Earlier the country's main concern was food availability; greater attention was paid today to food quality and good health. To keep up with these changes, the government planned to implement the following measures in the future: (a) reducing the frequency of the agricultural census from ten years to five years; (b) developing and conducting ad-hoc surveys to address specific policy needs; (c) defining clearly the differences between household and industry or establishment sectors so they could be measured more accurately; (d) developing techniques for crop forecasting; and (e) collecting and compiling statistics on quality of crops.
110. In document APCAS/02/11, "The Impact/Consequences of Irregular Censuses: Australian Experience", Mr Allan Nicholls presented the Australian experience with irregular censuses. After more than 100 years of annual agricultural censuses, a decision was made to reduce the frequency to every five year. From 1998 to 2000, large-scale sample surveys were conducted in place of an agricultural census. The agricultural census was again conducted in 2001. The paper described the positive aspects of this change, as well as the challenges it posed.
111. The Commission was informed of the reasons for the decision to reduce the frequency of the agricultural census, as well as the survey design for the large-scale sample surveys in intervening years. The main positive impacts of the change were cost savings and significant reductions in the paperwork load placed on farmers, the latter being an important issue that government agencies in Australia face in dealing with clients. Some of the cost savings were re-directed to data collection in a number of high priority areas that had not been adequately covered before, with the "service industries" being the main beneficiaries. The new situation provided an opportunity to review collection and processing methodologies and institute changes which resulted in more statistical rigour in data collection. The changes also resulted in a more flexible programme as modifications in the content of the questionnaire were quicker and cheaper to make in a sample survey environment. Additional items had been collected subsequently through the sample surveys. Finally, timeliness of outputs had improved.
112. The Commission was informed that the shift to five-yearly agricultural censuses also presented a number of challenges. Limited sample sizes in survey years did not allow the generation of reliable estimates of characteristics, either of less common commodities or at low levels of geographic disaggregation. Problems had also arisen in defining a suitable regional classification acceptable to a broad range of users. Users had found that the sample survey data were not always accurate enough to monitor structural, seasonal and other changes. The reduced accuracy of survey data had also impacted on the quality of indicators derived from production data such as those used to measure agriculture contribution to GDP and nutritional intake by Australians.
113. An annual agricultural census was a cost-effective and efficient method of maintaining the population frame. Without it, it was found that the frame had seriously degraded over the four years. Similarly, the proportion of "dead" units on the frame had increased, causing preliminary estimates to be overstated. Special studies were needed to determine the true death rate and the appropriate weights to apply, and similar studies would be needed each year. The lack of regular updating of frame information also impacted on the quality of information used for stratification and thus sample designs were not as efficient as they could be.
114. The lack of a regular census also had a significant impact on response rates. Despite a public relations campaign that achieved good response at the early stages, acceptable final response rates were only achieved after an intensive and very costly follow-up program.
115. In summary, the change to a five-yearly agricultural census had achieved the major benefits of saving on costs and reduction in reporting load placed on data providers, particularly small farmers. The change, however, had also resulted in a number of problems, largely related to frame maintenance. The extent of these problems was not fully evident until the conduct of the 2001 agricultural census. The ABS had learned important lessons from its experience with the first of the irregular censuses. Applying these lessons in the conduct of coming censuses and surveys should result in better-quality data in the future.
116. The Commission expressed interest on the various problems experienced, as well as some suggestions on how the problems may be addressed.
117. The Commission heard the presentation of paper APCAS/02/12, "Preparations for World Census of Agriculture 2010". It was reported that many countries had undertaken their agricultural censuses since 1996, while others would carry out theirs by 2005 using the FAO Programme for the World Census of Agriculture 2000 (WCA 2000) and its two supplements, i.e. Guidelines on Employment and Guidelines on the Collection of Structural Aquaculture Statistics, as references. A number of issues had emerged from the execution of these country censuses. These were: (a) analysis of agricultural census data; (b) electronic dissemination of census results; (c) identification of food insecure and vulnerable areas; and (d) more emphasis on livestock data.
118. The Commission was informed that many countries had spent resources and efforts to enhance the use and analysis of their agricultural census data. These initiatives were done through the organization of national seminars for the presentation of various analyses and studies based on the census data, involving different types of users and data producers.
119. It was also reported that more and more countries were disseminating their agricultural census data through electronic means, including posting data on a website or storing them on CD-ROMs These methods greatly facilitated further tabulation and analysis of the census results.
120. The Commission was briefed on new trends observed in a number of countries where agricultural censuses were used to undertake studies for the identification of food insecure and vulnerable areas. A major study on this topic was done in China. A few other countries were including items relevant to food insecurity and vulnerability information and mapping systems (FIVIMS) in their agricultural censuses.
121. The Commission was informed that more attention had been given to obtaining livestock data from agricultural censuses in a number of countries in Asia as well as in Africa. In addition to including livestock items in their agricultural censuses, some countries had undertaken livestock censuses during intervening years.
122. The Commission was informed of the processes leading to the release of the new publication of the series of Programmes for the World Census of Agriculture. These included the identification of issues or topics for inclusion in the new Programme during 2002/2003, the preparation of documents on relevant issues, organization of an international expert consultation on agricultural census, and publication of the Programme and its supplements in 2005/2006.
123. The Commission was informed that the new Programme might be composed of the main "Programme for the World Census of Agriculture 2010 (WCA 2010)"; and supplements on: (a) agricultural census data for food security; (b) livestock; and (c) techniques for presentation and dissemination of agricultural census results. A publication on horticulture statistics might also be issued to accompany the Programme for the WCA 2010.
124. The Commission was informed that India included items on infrastructure and on land in its agricultural censuses, which could be useful for food security policy. It agreed that a publication on structural livestock statistics would be useful for countries in their agricultural census planning, even as it pointed out that livestock should always be part of the agricultural census.
125. The Commission agreed that land use statistics from agricultural censuses should be consistent with other related classifications and noted that Geographic Information Systems (GIS) could be more widely used in presenting the agricultural census results.
126. The Commission agreed that guidelines on differentiating farm households from agricultural establishments would be useful. It also noted that time use surveys could provide useful complementary data to agricultural censuses. Finally, the Commission recommended that a list of minimum indicators for FIVIMS could be included in the supplement on agricultural census data for food security.