THE STATISTICS DIVISION HISTORY

Contributed by Edward Gillin

December 2006, Rome

The members of the Organizing Conference of FAO, in 1945 in Quebec , Canada , recognized the important role which statistics on food and agriculture would play in the work of the newly established U.N. Agency. Among the many recommendations at that time was one that the Director General established a Standing Advisory Committee on Statistics. The Expert Committee was drawn from countries in all parts of the world. It held its first meeting in 1946 and developed guidelines for the statistical work, which was originally located in FAO's Economics Division. In that year, too, a formal agreement had been reached that FAO would take over the work/mandate of the International Institute of Agriculture (IIA). In fact, in every sense the IIA, established in 1908, had been the predecessor of FAO. Statistics was the major concern of the IIA and its courageous staff had endeavoured to maintain statistical series even during the War years. Their work provided a starting point for the collection of current data from cooperating governments. In 1946, three IIA staff members, and the statistical files, were transferred from IIA offices in Rome to the new FAO Washington Headquarters. They worked with newly recruited staff in restoring and enlarging the flow of information. The Monthly Bulletin and the Production and Trade Yearbooks were the first visible evidence of the successful efforts to secure the cooperation of governments.

An important activity of the now closed IIA had been the promotion of worldwide censuses of agriculture. Common concepts had been developed and an emissary had gone to countries to promote the taking of such censuses on a comparable basis. The IIA was especially concerned to persuade governments that had not yet taken such a census, to include it in their programmes. The target year was to be 1940, but the outbreak of the War ended those efforts. A worldwide Census of Agriculture in 1950 became a matter of high priority for FAO. Manuals were developed providing question ‘wording, explanatory material, common definitions, table outlines and guides for the conduct of a census. Emphasis was put on meeting the needs of individual governments. In recognition of the variation in the capability of governments to undertake a census, the recommendations provided for a basic list of items to be included in every census. A supplementary list was developed to be of assistance to those governments that wished to collect additional information. Although it was late in the decade to prepare for Censuses in 1950, there was widespread recognition of the need for the basic data which they could provide. At the same time the Statistical Office of the United Nations was giving much attention to Population Censuses to be taken in 1950.

Governments involved in the War had been confronted with formidable problems of managing the food supplies to meet the nutritional needs of their people. Building on that experience, the First World Food Survey was issued in 1946 and guidelines were developed for tabulations showing each country's supplies and its needs. The tables were to show indigenous production, harvest losses, imports, exports, and quantities used for feed, seed, and industrial purposes. Losses due to predators or to inadequate storage were to be taken into account. The national reports would eventually lead to the ability to prepare worldwide Food Balance Sheets.

One of the casualties of the War had been the weakening of statistical offices through the loss of experienced personnel and the reduction in training of potential recruits. Two lines of action were recommended; (i) the establishment of regional and ad hoc statistics training centres, and (ii) the assignment of competent persons who could assist in the organization and conduct of statistical operations, when requested to do so by governments.

As early as the spring of 1948, 40 statisticians from Egypt , Iran , Lebanon and Syria gathered in Baghdad for a training course given by an instructor sent by FAO. The favourable reception of this effort led to the establishment of a training programme in Mexico . This was a cooperative effort with participation by FAO, the Director General of Statistics from Mexico , the Statistical Office of the United Nations, and the Inter-American Statistical Institute. The course lasted 3 months. One student was the Director General of Statistics in his country; most of the others came from national statistical offices.

Early on in FAO it became clear that the needs of governments could be met effectively through regional offices which could provide technical assistance on a continuing basis. One officer on the regional staff was to be a Regional Statistician. Familiarity with local conditions and with a common language was added advantages. The Middle East and Latin America were the first regions to receive this service.

The early years were marked by many new needs arising in many different places. Recruiting staff devoted to meeting the challenges was quickly completed. While a few mistakes were made, valuable experience was gained. Foundations were laid for a programme that has matured and changed during the sixty years of its existence.

In 1951 when FAO transferred from Washington to Rome , food and agricultural statistics were part of the work of the Economics Division and constituted one of its Branches when P.V. Sukhatme took over from Conrad Taeuber as its Chief in 1951. Dr. A.H. Boerma (Director General FAO 1968-76) was the Director of the Economics Division at that time. With the growth in activities of FAO, the structure of the organization changed and the Branch was upgraded to the status of Division in 1959 and its work was regrouped in four new Branches — one for dealing with the census of agriculture, another for dealing with production statistics, a third dealing with statistics of trade, prices and. index numbers and the fourth dealing with methodology. In addition, a special unit for social statistics was attached to the Director's Office which dealt with food consumption, population and their statistical analysis. The entire Division was located within the Department of Economic and Social Affairs with a responsibility to serve all FAO Divisions and its regional offices. The Forestry and Fisheries Departments had small statistical units dealing with forestry and fisheries statistics respectively, attached to them.

The recently created Statistics Division was re-structured several times over the years to better orientate its resources in face of changing FAO priorities and data needs. For example, in 1971 the Office of Director already included an executive/administrative unit for project operation work under the Executive Officer; the special unit for social statistics was already the Statistical Analysis Service with social statisticians working on food consumption and population statistics and analysis to estimate and project agricultural population and labour force, supply/utilisation accounts and food balance sheets, economic accounts in agriculture, trade and prices; the Census and Methodology Branch was mainly dealing with censuses of agriculture, census and survey methodology and field projects in agricultural statistics, while the Production Branch was mainly dealing with agricultural production, input and land use statistics.

The nature of the work within each Branch, except that of methodology and the unit in the Director's Office, largely continued to be what it was before, viz, collecting statistics from member countries, tabulating the same in a comparable manner in accordance with agreed concepts and definitions and publishing them in Yearbooks of production and of trade. The results of censuses were published in separate volumes and dealt with the structure of agriculture and means of production, including number of farms and their characteristics in the different countries. Statistics of fertilizer input were singled out in view of their importance for production. In addition, in response to the demand from member countries, a Monthly Bulletin highlighting statistics and situations most appropriate to the season was published. The Division publication programme was so large that of the total budget for documentation for the Organization as a whole, a third was allotted to the Statistics Division.

Tabulating data for publication was not merely a problem of sending a questionnaire to the countries to get the needed data. As more and more countries joined FAO, it became a question in the first instance of evolving agreed concepts and definitions and, secondly, a question of the practicability of collecting the needed data from countries through censuses and surveys. These two were the major activities of the Division. The former was handled through statistical conferences annually convened in each region and at special committees at Headquarters. The preparation of documents to serve these meetings was an enormous task for the various Branches. The follow up of the recommendations by the Conferences was largely the work of the regional statisticians in the course of their normal visits to countries. It a painstaking job and called for patience on one hand and simultaneous testing in the field for practicability. It was the latter aspect of the second activity that became the major thrust of work.

First and foremost, attention was paid to preparing manuals on sample surveys, describing the experience in the developing countries in their applications for collecting agricultural statistics. Secondly, International Training Centres on Sampling in Agriculture had to be set up and appropriate field demonstrations organized in various member countries. Wherever feasible, these Centres had to be supplemented by short seminars to discuss difficulties arising in the course of work. The major difficulty was lack of field technical guidance and of funds for collecting countrywide data. This was met by FAO assuming direct responsibility under the technical assistance programme. So large was developing country interest in this effort that annually some 40 experts were assigned to work in developing countries in data collection through agricultural censuses and surveys. (This numbers of experts was later to double).

In promoting this effort, the Statistics Division took into consideration the stage of statistical development in countries. In countries where no system of agricultural statistics existed, the collection of area and yield statistics was made on the basis of eye estimates supplied by knowledgeable persons. In countries with developing systems of agricultural statistics, interview surveys were organized on a sample basis. The results of these surveys were checked using results of area measurements and crop cutting for estimating yield taken on a sampling basis. In countries with developed systems of agricultural statistics, it was possible to organize objective methods almost from the beginning. By its very nature, development of agricultural statistics is a long and continuous operation due to constraints in funds and trained manpower. It therefore took years to evolve workable systems in many countries, particularly those lacking skilled manpower, technical facilities and funds. As the work progressed, FAO made it a condition that countries requesting technical assistance provide counterpart provision for training and make budgetary provision for organizing country-wide surveys and censuses as part of their planning process.

Among the developing countries, India had the largest and longest experience in the use of objective methods of sampling and agricultural statistics. The Division took a lead in preparing a manual which subsequently became a textbook on sampling surveys. The Census and Methodology Branch extended its work through a series of manuals to cover specialized topics in sampling and censuses. Many of the experts appointed in the developing countries were drawn from India and Pakistan , etc. They drew from their past experience and initiate pilot surveys in countries to which they were assigned. This also helped in extending the coverage on a country-wide basis. Most of the early work was concentrated on improving the frame for sampling. Once this was done, it became feasible to integrate current agricultural statistics with census statistics. With the advent of computers, even this distinction between the census and current statistics got largely obliterated in the case of a few countries. The difficulties initially experienced by countries were effectively surmounted through national allocation of funds and resources as part of national development plans. Many countries set up permanent field staff organizations.

Most of the sampling experts were Indian. India also asked for technical assistance in statistics to enable it to improve the application of statistical methods to agricultural research. The Indian Institute for Agricultural Research Statistics in Delhi is partly the result of this assistance provided by FAO experts. The strides made by India in developing high yielding crop varieties and extending the area under them and their production was greatly facilitated by the help from FAO.

The second major thrust was in the area of analysis and interpretation of social statistics relating to population and food consumption. The Second World Food Survey was published in 1952. FAO statisticians felt that popular statements on hunger and malnutrition were exaggerated. In particular, the proposition that two thirds of the world was hungry and malnourished based on the average energy intakes of countries with their average energy requirements appeared to be a gross overstatement. The Third World Survey published in 1963 showed that when the statistical distribution approach was used in assessing the numbers of hungry in the world, the incidence was much less than two thirds. The Third World Food Survey became a benchmark for the Freedom From Hunger Campaign (FFHC) the following year. Then it was on the heels of the Freedom from Hunger Campaign that the Director General requested the preparation of Indicative World Plan for Agricultural Development (IWP) which was to be centered on crafted food balance sheets expanded to area and yield and called supply utilization accounts (SUA).

In early 70's the Division was re-structured to accommodate for the development of the Interlinked Computer Storage and Processing System of Food and Agricultural Commodity (ICS) and the organisation of the work on compilation of the statistics on agricultural production and trade and the preparation of supply/utilisation accounts and food balance sheets on a country rather than subject matter basis. The re-structure had the Office of Director as before; the Statistical Analysis Service (ESSA) including the preparation of the World Food Surveys but excluding trade, supply/utilisation accounts and food balance sheets; the new Statistical Development Service (ESSS) mainly incorporating work from previous Census and Methodology Branch; and, the re-named Basic Data Unit (ESSB) incorporating work from the former Production Branch including trade, input and land use statistics and the construction of supply/utilisation accounts and food balance sheets. So, following, the release of the IWP, major activities covered the development of a statistical database on food and agricultural statistics (ICS) creating a data bank, a word which at that time was not in wide use. It formalized the data which had been built up for the IWP. This was the predecessor to today's internationally acclaimed FAOSTAT, and represented an application of the mainframe computer to FAO's work as well as the drawing together of production, trade and utilization data to estimate food consumption. It became the source for supply utilization accounts and food balance sheets and was the basis for the publication of FAO statistical yearbooks. Work continued on the study on population and food relationships, planning the revision of the FAO index of agriculture production, work on agriculture producer prices and sector accounts, and the preparation of the upcoming decennial World Census of Agriculture Programme.

Priority was given to the work on the ICS to enable the computerization of the Production and Trade Yearbooks and to better meet various data needs of the Organization. One of the most important elements in the ICS is the establishment of supply/utilization accounts for each country by commodity. They provide a useful tool in checking the consistency of estimates as part of the process of revision. Every output programme became and editing programme. Supply/utilization accounts also provide the basis for the preparation of food balance sheets which give estimates of calorie and protein supplies in various countries.

Based in part on these data, a basic study entitled “Population, Food Supply and Agricultural Development” was prepared as one of FAO's major contributions to the 1974 World Population Year and World Food Conference held in Rome in November 1974. The study also marked a significant step in the preparation of the Fourth World Food Survey. The study provided the basis for the generally accepted estimate of more than 400 million undernourished in the world.

FAO world, regional and country indices of agricultural production at that time were calculated using regional average wheat based price relatives as weights. While this procedure avoided international exchange problems, it involved several conceptual and technical difficulties. Priority was given to a detailed study of alternative weighting systems, which resulted in the decision to revise the indexes of agricultural production using national average producer prices as weight for the compilation of an aggregate agriculture production for each country and, after converting into US dollars, summing these aggregates to provide regional and world indices. Later, this methodology was further refined and international dollar prices derived from the Geary equations on agricultural purchasing power parities were used for both countries and regions/world.

Special emphasis was placed on the development of producer price series in cooperation with the countries concerned as a basis for economic accounts for agriculture, the revision of the agriculture production index, and in agriculture planning. An initial publication presented production account data (1961-71) for some 60 countries. Also in cooperation with the UN Statistical Office, a manual was issued formulating draft world guidelines for use in preparation of economic accounts for agriculture.

The 1980 World Census of Agriculture Programme was presented to the 1975 FAO Conference and after approval by the Conference was disseminated to the countries for use in undertaking agriculture census activities. The 1980 Census round emphasized changes needed in questions asked procedures and tabulations in view of changes in agricultural data needs and the shifts to computer processing.

The experience of the about 30 years of countries' participation in three decennial World Censuses of Agriculture sponsored by FAO demonstrated that fully meaningful and comprehensive information on agricultural population and employment could not be obtained through agricultural holdings only as the unit of data collection. In formulating the 1980 Programme for the World Census of Agriculture, we therefore gave up the concept of farm population based on farm response criteria. The data on agricultural population and employment could better be obtained in conjunction with population censuses, household surveys and labour force surveys. This could also ensure greater coordination with the work programmes' of other international agencies, in particular, the United Nations and the International Labour, Organization. The promotional efforts for the 1980 World Census of Agriculture included organization of national demonstration centres to be hosted by Governments aimed at providing practical training, including field demonstrations to national census personnel in selected developing countries. While the main priority was to train the local staff in field operations, participants from a few neighbouring countries were also invited to attend. This was a new development in FAO's promotional efforts for the agricultural census. Arrangements for similar training were also made with established training institutes.

Again during the 1970's, several more additional important activities were undertaken. Of particular note were the preparation of world-wide estimates and projections of the agricultural population and labour force within the framework of an inter-agency work programme on demographic estimates and projections involving the UN, ILO, FAO and UNESCO and the carrying out of pilot activities on the development of long term agricultural statistics programmes in selected countries.

In 1977 the first major demonstrable results of the new computer system were available with the publication of the 1972-74 Average Food Balance Sheets for 162 countries and territories published in 1977. Also, the FAO Production and Trade Yearbooks and the annual Fertilizer Review, which were made as offset reproduction of master tables processed from the computer, published and released earlier than similar publications of other international organizations. Steps were also taken to install computer terminals in the Basic Data Unit (ESSB) to ensure rapid input and ready access to the computer storage.

The Monthly Bulletin of Agricultural Economics and Statistics was converted into the Monthly Bulletin of Statistics only, processed now entirely from the computer. This eliminated the time lag by about two months required for editing, translating and processing textual articles on economic and commodity reviews.

The general questionnaire issued by the Division for its major annual publication, the FAO Production Yearbook, was tailored through the computer for each individual country and thus differed from the previous general questionnaire in two respects. Firstly, instead of setting out a comprehensive list of commodities and items the same for all countries, it contained only those commodities pertaining to the agriculture of each individual country. Secondly, the new questionnaire presented the data for preceding years available in the computer files for updating and revision.

Work was started on preparing a Glossary of Statistical Terms used in food and agricultural statistics. The need for this project was contemplated in discussions at the ACC Sub-Committee on Statistical Activities and the Conference of European Statisticians. The Council for Mutual Economic Assistance (CMEA) too cooperated in this project.

The Statistics Division actively participated with the UN Statistical Office (UNSO) in the formulation and establishment of a National Household Survey Capability Programme (NHSCP). One of the basic technical reasons for this was to link the collection of data on agriculture and food consumption with those on income distribution as well as with general household surveys contemplated by the NHSCP. Experience in the past has shown that extensive food consumption surveys, particularly in the developing countries; have been rather rare. To ensure active support to the NHSCP, a senior post was established in the Division.

A questionnaire on investment and expenditure in agriculture was formulated for the first time in consultation particularly with the then Policy Analysis Division (ESA) and sent out to a number of countries on an experimental basis.

The scope of the Fourth World Food Survey, published by the FAO in 1977 was broadly similar to that of its predecessor. However, because of the impending study by the FAO on Agriculture Toward 2000, it did not tackle the prospects for the future. Besides, it differed from the Third World Food Survey in two technical aspects. Firstly, the critical limit for estimating the numbers undernourished was arrived at entirely on physiological considerations (basic metabolic rate) and independent of any assumed distribution of energy requirements and its correlation with energy intake. Secondly, it presented estimates of the numbers undernourished for a large number of individual countries instead of providing only global estimates. It of course, emphasized that the country figures given were only by way of example to indicate the order of magnitude of undernutrition at the country level.

Work also started on preparing guidelines for a comprehensive list of social and economic indicators which could be used for monitoring the progress of agrarian reform and rural development in countries. This was to follow up the recommendations of the 1979 World Conference on Agrarian Reform and Rural Development (WCARRD).

In the early 1980's the work on agricultural input statistics was transferred from ESSB to ESSA. Following WCARRD, ESSA absorbed the work on socio-economic indicators, resource flows to agriculture as well as methodological studies on certain statistical indicators pertaining to food and agriculture. However in the mid -1980's the developmental work relating to WCARRD indicators was shifted to ESSS. Furthermore, the project operation work in the Director's Office moved to the Agricultural Operations Division (AGO).

Going into the 1980's there was general expansion of statistical coverage and advancement in the data processing cycle for the basic commodity SUAs processed by the ESSB Unit. Data coverage has increased both country and commodity wise, and the processing cycle has advanced to the stage that there was only one year lag in the data; i.e., by the end of 1986 the ICS contained supply utilization accounts for the year n-2. Advances in the data processing cycle enabled earlier release of the Production and Trade Yearbooks. The receipt of more such data on magnetic tapes as opposed to mail out questionnaires, as well as wider use of computer terminals and online data processing, has increased the efficiency of data processing such that a larger volume of data is processed and analysed without a corresponding increase in resources. In addition to the more timely publication of statistical Yearbooks, there was increased dissemination of such data on magnetic tapes and via a new FAO Statistical Pocketbook and new Statistical Processed Series containing time series of important statistics.

The 1980 World Census of Agriculture Programme was introduced to the 1975 FAO Conference and after approval by the Conference was disseminated to the countries for use in undertaking agriculture census activities. The 1980 Census round emphasized changes needed in questions asked procedures and tabulations in view of changes in agricultural data needs and the shifts to computer processing.

The experience of countries' participation in three decennial World Censuses of Agriculture sponsored by FAO demonstrated that fully meaningful and comprehensive information on agricultural population and employment could not be obtained through agricultural holdings only as the unit of data collection. In formulating the 1980 Programme for the World Census of Agriculture, we therefore gave up the concept of farm population based on farm response criteria. The data on agricultural population and employment could better be obtained in conjunction with population censuses, household surveys and labour force surveys. This could also ensure greater coordination with the work programmes' of other international agencies, in particular, the United Nations and the International Labour, Organization. The promotional efforts for the 1980 World Census of Agriculture included organization of national demonstration centres to be hosted by Governments aimed at providing practical training, including field demonstrations to national census personnel in selected developing countries. While the main priority was to train the local staff in field operations, participants from a few neighbouring countries were also invited to attend. This was a new development in FAO's promotional efforts for the agricultural census. Arrangements for similar training were also made with established training institutes.

A new information system orientation to technical assistance work with developing countries in advancing their statistical capability in food and agriculture This development was accelerated with the publication of the first two volumes of “ Statistical Development Series”: (i) Food and Agricultural Statistics in the Context of a National Information System, and (ii) The Programme for the 1990 World Census of Agriculture. Emphasis was on a closer balance between a decennial census of agriculture and a multi-year cycle of sample surveys as complementary aspects of basic data development at the country level, as well as greater balance between field enumeration, data processing and statistical analysis. Above all, a reemphasis of a user orientation for statistical programmes.

In 1985, the preparation of the Fifth World Food Survey was enhanced by a somewhat greater availability of data from household income and expenditure surveys and food consumption surveys from developing countries, expanded coverage of the food supply estimates derived from food balance sheets, methodological improvements, shared responsibility with the Food Policy and Nutrition Division and a resulting greater emphasis on policy concerns and implications. Methodological improvements included new concepts relating to energy requirements and the use of alternative cut-off points for energy requirements reflecting the continuing scientific debate on how to determine minimal energy requirements in the estimation of the prevalence of undernourishment, and greater attention to the multidimensional causal relationships relating to undernutrition and malnutrition.

In the 1980s, with the advances in computer technology, the growth in world communications and the personal computer became globally available, the Statistics Division sought the development of a more user oriented dissemination databank, known as AGROSTAT. This followed improvements in the quality of commodity supply utilization accounts and food balance sheet statistics (available in the ICS) as well as easier access to improved data series from other organizations, and a tendency to increase the array of socio-economic data in the public domain. So in 1984, for dissemination, the 1 million time series for about 200 countries were made available to analysts across FAO via online remote terminals, from AGROSTAT data files. Data was also available on diskette subscription to AGROSTAT.

Efforts continued on building a global dissemination platform and replacing the aging computer processing system. By the end of the 1980s, the Organization decided, as a part of the development of information mandate, to merge and modernize several of its systems to improve the quality of the statistical data collected by FAO. Moving beyond ICS, these systems had all been performing the same tasks of compiling, processing and validating time series data, used to generate other data and as input for advanced economic analysis. ICS continued to function as the processing system till 1991. This new umbrella system was called FAOSTAT , the Corporate Database for Substantive Statistical Data . The system was implemented between 1990 and 1992. Its initial beta release, in 1992, addressed the data processing needs of the Statistics Division as well as the Fisheries and Forestry Departments. Work continued through to 1995 as the system expanded to incorporate other statistical working systems of the Organization and all development concluded by 1995. But following the launch of the database, demands from users of the statistical data grew exponentially and users demanded faster and more frequently updated data. Internal users in ESS required more stability in the running of the system. Staff also experienced problems with the huge quantity of data which had now grown from 1 million to 3 million time series as electronic trade files and other data flowed into the Division.

On the analytical side of the work performed by the Statistical Analysis Service (ESSA) of the Statistics Division work continued. In 1996 the Division published its Sixth World Food Survey, the last in its series before the World Food Summit (WFS) of the same year. The main conclusion of the Survey was that per caput dietary energy supplies continued to increase in the developing countries as a whole, with the result that, during the two decades from 1969-71, the prevalence of food inadequacy declined: 20 percent of the total population had inadequate access to food in 1990-92 compared with 35 percent two decades ago. Even more remarkable was the improvement in absolute terms, i.e. fewer people faced inadequate food access in 1990-92 compared with 20 years ago, notwithstanding the addition of 1.5 billion people to the population of developing countries during this period. The number of people with inadequate access to food declined from 918 million in 1969-71 to 906 million in 1979-81 and further to 841 million in 1990-92. Nevertheless, this number was still very high in 1990-92, as one out of five people in the developing world faced food inadequacy.

After the WFS (1996) the role of the Division increased with annual assessments of undernourishment which were published in the State of Food Insecurity in the World (SOFI) starting in 1999 for monitoring the WFS target of reducing by half the number of undernourished by 2015. This task was performed by the Statistical Analysis Service (ESSA), re-named in 2004, as the Socio-Economic Statistics and Analysis Service).

As data on household surveys is a very important tool for the estimation of the population undernourished, it has been reinforced, at the beginning of 2000, for strengthening national statistical capacity building in the analysis of the collected food consumption data from the national household income and expenditure surveys (NHS). National Demonstration Centres (NDCs) for deriving food security statistics from NHS have been conducted on regional basis for selected countries in Asia (2002), Africa (2003), CIS (2004) and Near East and North Africa (2005) with a coverage more than 60 countries. Country technical assistance, which was sporadic due to the limited available resources, has considerably increased recently to more than 10 countries, for example, under the EC-FAO program “Food Security Information for Action”. Furthermore, since November 2006, the ESSG has launched its one-month tutorial training at the International Demonstration Centre (IDC) with participants from Palestine . The training activities target both participating individuals and institutions involved in collecting food consumption data on the processing and analysis of the most recent NHS food data. The training provides opportunities in report writing and dissemination of food security statistics among the national users and policy makers. This enables participants a better coordination of all institutions involved in providing food data inputs for estimating the food security statistics. The National Statistics Offices, being the responsible institution of all national data are encouraged to lead and perform efficiently a sustainable estimation of food security statistics including the prevalence of food deprivation.

The Division has also, since November 1994, assisted the Director General by conveying to his office all the statistical information which he may require during his activities. To this end, several products have been prepared for example the quarterly reports “Summary on World Food and Agricultural Statistics”, which began as an internal document of around 50 pages is now doubled, and since July 2001, the “Summary of Food and Agricultural Statistics by Regional Groups”; since the end of 1990s, the annual Country Briefs prepared together with other FAO units in the form of Country profiles by the Technical Cooperation Department. All reports are made available as the “Compendium of food and agricultural indicators” in the ESS web site. The Division also has provided support for the clearance of official figures in documents and ad hoc works for the Director General such statistical data presented in speeches, interviews or documents distributed by the Office of the Director General. This work has required the ability of working under pressure in very short notices.

By the turn of the millennium the technology had changed from central processing to desktop personal computer processing and FAOSTAT systems were becoming obsolete and failing. The FAO Programme Committee in 2001 recognized this concern and recommended to The Director General the modernization of FAOSTAT . Work began in 2004 and was completed in 2005 and new FAOSTAT was released to the public in mid-2006. The new system was technically compatible with the latest data transmitting technologies and was able to generate or to help to derive missing data from limited information received from many member countries. The system used the latest artificial intelligence algorithms to validate the incoming data and fill gaps in missing country data. The methodological framework, processing and dissemination systems of FAOSTAT were redeveloped together with its coverage, revised standard international commodity classifications, statistical and meta databases to provide more up-to-date and reliable statistics. At the same time as wider access to data was available, The Statistics Division took the initiative to combine the various FAO statistical yearbooks into a single FAO Statistical Yearbook, Volume 1 being published in 2004 which met the needs of a wider audience. FAOSTAT, managed by the Statistics Division, is the world's largest and most comprehensive statistical database on food and agriculture. It contains over 1 billion data points, 40 million of which are updated annually. The FAOSTAT site receives over a quarter of the visits to the FAO web site and around 20 gigabytes of data are downloaded every day. This is a thirty fold increase in use over just five years earlier.

Finally, in 2007 the Statistics Division was re-structured again into just two Services the Global Statistics Service (ESSG) and the Country Statistics Service (ESSS), as part of an overall FAO re-structure. ESSG combined ESSB and ESSA. Country Statistics Service (ESSS) covers technical assistance to countries, CountrySTAT as well as the Census of Agriculture. ESSG covers a) data collection, compilation, input and data processing locally b) data dissemination (FAOSTAT) and c) data analysis. Inter alia, the latter covers the relevant methodological development and analysis of food insecurity and other related socio-economic indicators performed by the Food Security Indicators Group. A small country technical assistance on food security statistics was provided due to the limited available resources; however, it has considerably increased recently, for example, under the EC-FAO program “Food Security Information for Action”. Furthermore, an “International Demonstration Centre (IDC)” was established in FAO Headquarters and launched in November 2006). The training activities target both participating individuals and institutions involved in collecting food consumption data on the processing and analysis of the most recent NHS food data. The training provides opportunities in report writing and dissemination of food security statistics among the national users and policy makers. This enables participants a better coordination of all institutions involved in providing food data inputs for estimating the food security statistics.

This historical note will be continued to be updated in future and can be a reference point on the latest changes taking place within the FAO Statistics Division. We hope it is informative and encourages more contact amongst users and producers of data and staff.