Previous - Next
PART TWO: SOURCES AND METHODS
Chapter 5 Sources of data and methods of estimation
Sources
of Data
Agricultural and livestock censuses
Crop
production surveys
Cost of production surveys/farm management
studies
General surveys and administrative records
Methods
of estimation
Period
of compilation
Output
Valuation of output
Intermediate
consumption
Compensation
of employees
Taxes
on production
Subsidies
Gross fixed capital formation
Changes
in inventories
Consumption of fixed capital
The goods and service account
Organization of the compilation of SEAFA
5.1 The SEAFA comprises value aggregates relating to different economic activities covering the entire sector of food and agriculture. It is usually presented in the form of time series to reveal changes in the economy over time. Hence, it becomes more essential to examine carefully the coverage of the data on each item or transaction. This chapter deals with the estimation of various components using the data available from various sources.
5.2 The main sources of data for the compilation of SEAFA include censuses of agriculture and livestock, crop estimation surveys, farm management and cost of cultivation studies, household surveys and various returns collected by administrative agencies concerned with prices and production relating to agriculture. Data for compilation of SEAFA can also be derived from the population census, statistics relating to industrial production, balance of trade and balance of payment statistics, index numbers of wholesale and retail prices, quantum and value index numbers of imports and exports, government budget documents and reports available from specialized agencies or boards dealing with marketing of specific crops, fertilizer or pesticides, finances for agricultural activities, cooperative (agricultural) societies, etc. The data available from these sources can be divided into two broad categories. The first category consists of data with complete coverage of the agricultural activity in the economy while the second relates to data from surveys with incomplete coverage which need to be adjusted by structural ratios derived as explained in footnote (number 19) to pare 4.12 of Chapter 4.23 A full documentation of the concepts, definitions, statistical units, samples, etc. is needed for each single piece of data. Access to the basic questionnaire is indispensable. Information on the coverage, the rate of response, the homogeneity of the sample, etc. is also essential. In most cases direct contact with the statisticians who have designed and elaborated a survey is the only way to avoid misunderstanding or misinterpretation. It is necessary to collect the information that could be useful for compilation of SEAFA from both sources. The next step is to examine the coverage and concept of the data and decide how to adjust the information to suit the requirement of SEAFA. The point is illustrated with the help of a few examples in subsequent paragraphs.
Agricultural and livestock censuses
5.3 Agricultural censuses are organized by many countries on a periodic basis to collect basic quantitative information on the structure of agriculture which is needed in development planning, socio-economic policy formulation and the establishment of national priorities. Generally, the censuses are repeated every five or ten years, but some countries conduct a census every year to meet their specific requirements. The information is collected mostly by the method of complete enumeration but sample surveys are also sometimes used by some countries for completing part of the coverage (or the entire operation) to match the resources and needs. The statistical unit for data collection in the census is the holding defined in the programme for the 1980 World Census of Agriculture and is as follows:
"a holding, for agricultural purposes, is a techno-economic unit of agricultural production comprising all livestock kept and all land used wholly or partly for agricultural purposes and operated under the management of one person or more, without regard to title, legal form, size or location. The holding, as a techno-economic unit under a single management, generally uses throughout the same means of production, such as labour, farm structures, machinery or draught animals".
The information collected in the census relates to holding, holder, population and employment, land use, crops, livestock, machinery and equipment and selected practices and facilities.
5.4 It can be seen from the above that the concept of holding in the agricultural census is equivalent to the concept of establishment used in economic accounts. Thus the data collected on crops, livestock and machinery and equipment can be used for compilation of economic accounts. A critical review of country practices for data collection brings out the following important points:
· The coverage of the census differs from country to country. For practical reasons, many countries put a limit on the enumeration of holdings. The limit can be based on area, volume or value of output, number of livestock or trees or labour requirements.
Some countries also include establishments producing only forestry products.
· For crops, the data collected may relate to area sown, area under crops, area of compact plantations, area under crop mixtures, etc. The concept of area as well as the items on which data are collected also varies from country to country. Similarly, under livestock, it is generally recommended to collect data on numbers of livestock
(except bees) by age groups, sex and purpose (for important kinds of livestock).
Regarding machinery and equipment, stationary power-producing machinery, agricultural machinery and transport equipment are usually included although the coverage is again different in different countries.
5.5 The data collected in the census thus provide an overview of the important crops and livestock for any country, but this information cannot be used directly unless the areas or items that have not been covered by the basic frame are estimated and incorporated into the information collected. As discussed earlier, it is first necessary to have an idea about the magnitude of the component that has not been covered and the importance of taking it into account in formulating policies and development plans. For example, let us assume that subsistence cultivation, which may contribute about 10 percent of agricultural output, has not been covered in the census. In formulating policies or plans the omission may become significant if, for example, the 10 percent of output is produced by 30 percent of the labour force engaged in agricultural production. Thus, if plans for reduction in poverty are required, it is essential that more information about the subsistence sector is gathered. It would therefore be necessary to estimate this component carefully, so that it is not overlooked, and present it with other data collected by the census or survey before preparing SEAFA. If this activity is not considered, however, significant a notional adjustment (which can be explained in a footnote) can be made for the sake of completeness.
5.6 The next step in the exercise would then be to obtain a factor with which the total activity can be estimated. This requires careful comparison of the definitions, the data, the reference point for which the data have been collected and the coverage of the census with those of different sources, such as population census. The desired adjustment factor may be built up either on the basis of existing information or special sample surveys. This aspect of data analysis is important not only for the compilation of SEAFA but also while the census is being planned. At the planning stage itself alternative uses of the information should be kept in view so that other required studies can be planned simultaneously.
5.7 The census of agriculture usually also collects data on numbers of livestock and agricultural machinery which are useful for the compilation of SEAFA and need similar detailed examination. Some countries conduct independent censuses for livestock covering all the livestock husbandry activity in the country. In these censuses the livestock holding is taken as the statistical unit (which is also defined as landholding in the agricultural census). In addition to covering the number of livestock by type, sex, age and purpose these censuses also include items such as breed of animals; dairy cattle subdivided into animals in milk and dry animals; numbers of eggs put in incubators; chicks hatched as laying hens separately from those hatched as broilers; facilities, such as fencing and buildings, watering and grazing grounds and machinery and equipment for livestock husbandry etc.
5.8 Reliable estimates of the total annual output of food and non-food producers can be obtained either from the final uses side or from the production side. From the final uses side, the estimates of the value of agricultural output obtained generally consist of:
· sales of crops, fruits and vegetables to marketing boards, cooperatives and processing factories;
· imputed values of seeds and feed produced for own intermediate consumption;
· other disposals, such as sales or barter in local markets, payments of wages in kind, and produce retained for own final consumption.
To secure complete coverage of annual production by this approach, it would be necessary to add to the above the value of growth of immature permanent and standing crops. Using this approach the estimates of the different components are generally based on different sources of varying degrees of reliability which can change over time. It is therefore necessary to monitor each component regularly. The total value obtained in this way may also include trade and transport margins and some products that are already processed. The values of the products may have to be adjusted to arrive at producer's value by making the relevant subtractions. The value of any by-products that are not included in the estimates have to be added. Quantities according to place of cultivation are not generally available with this method of information gathering.
5.9 From the production side the estimates of total output of the producers in question can be prepared following one of two approaches, i.e. either through data collected by mail questionnaire from farmers and other producing units or through crop estimation surveys that are organized for different crops grown in different seasons to estimate productivity (or yield) in a given area. One advantage of the latter method is that, apart from providing the estimates of productivity used to calculate output, it can also provide data on intermediate consumption (i.e. quantity, value and variety of seeds, fertilizers and pesticides used, amount of irrigation provided, etc.) which may not be available from any other reliable source.
5.10 For the compilation of SEAFA it is important to know all about the process of data collection. This is because, when using this data, it is necessary to understand what agricultural services, such as harvesting and threshing, have already been performed and what services are still left in order to get a clear idea of the estimate. It is also necessary to know exactly how much of the pre- or post-harvest losses have been taken into account. If crop cutting is made under laboratory type conditions then it is necessary to make an allowance for losses on the basis of information from independent surveys. Another interesting exercise would be to compile data on items such as seed and use of fertilizer or pesticides to get some idea about the inputs for each crop, if such information is not available from other sources.
Cost of production surveys/farm management studies
5.11 Generally production surveys or farm management studies are designed to collect information about costs incurred on material and labour at different stages of production. Thus the results of such surveys provide data on expenditure incurred or quantity used in producing a crop in terms of labour (cost or hours worked), seed, fertilizer input at different stages of crop development, diesel oil consumed for use of pump sets, tractors and other equipment, payment made for electricity and water bills (if any) and taxes paid to government along with the data on output (product and by-products). These surveys are a good source of data on compensation of employees, intermediate consumption, mixed income or operating surplus, rent, interest, etc. However, before using such information it is necessary to ensure that the design of the surveys takes into account the socio-economic conditions under which agriculture is being carried on.
General surveys and administrative records
5.12 For the compilation of SEAFA large amounts of data are required, not only on outputs and inputs but also on prices, by-products, external trade, final and intermediate consumption of agricultural products, details of secondary activities, etc. Regular Censuses and Surveys such as crop estimation or cost of production surveys, do not meet this demand. In addition, therefore, to meet various administrative or technical needs, countries organize sample surveys or type studies to collect information. A lot of information may also be generated as a by-product of other administrative activities. These sources provide much information that can be used for the compilation of SEAFA, but it is necessary to keep in mind that such surveys or studies do not always cover the whole country or region and they may not even cover the entire activity. For example, a consumption survey may cover only households (and sometimes not even all households, very rich or very poor households often being excluded). Moreover, institutional units, such as hospitals, religious institutions, jails, hotels and restaurants, or units without fixed address, such as beggars, may also not be included. Likewise a survey covering manufacturing establishment which may provide data on the use of agricultural goods in intermediate consumption for agro-based industries, may restrict its coverage with respect to employment size or may be limited to a particular area only. Thus when using the results of such surveys or studies the frame and concepts adopted must be examined carefully. These sources are generally used to work out structural or technological ratios rather than global estimates.
5.13 From time to time FAO has issued a number of manuals covering details that require consideration in planning various censuses/surveys/studies on various of the aspects discussed above. The manuals or study series also give details of practices followed in some of the representative countries which can be helpful when planning new studies. In the discussion given below no effort has been made to list the requisites for planning and organizing the collection of basic data. Users interested in these aspects should consult the relevant FAO publications listed in Annex 5.1.
5.14 SEAFA contains five types of accounts of which the first two i.e. the production and the generation of income accounts, should be prepared for "institutional units" as well as for "establishment-type" units. In most countries, data for compilation of SEAFA are available either for establishments, whose production is meant to be as homogeneous as possible, or for institutional units. In the case of the former, data are collected through crop estimation type surveys while in the case of the latter data are collected through regular household surveys, (mail) questionnaires, tax enquiries, etc. Thus, generally it will be feasible for countries to prepare one or other set more conveniently, depending on the system data collection. However, given one set of accounts, i.e. either by institutional type of or establishment type of unit, it is recommended that countries should attempt to prepare the other set in order to obtain a complete picture of agricultural activity in the country. This should be feasible if studies are conducted to estimate the main secondary activities carried out by agricultural households. Labour force data relating to secondary activities are generally collected in population censuses along with data on principal activities. Using such data the first step should be to estimate the quantum and nature of secondary activities. The two accounts should then be generated by imputing the remaining components utilising the input structure from data where these (i.e. secondary) activities are principal activities. This may require additional data to be collected as case-studies to provide benchmark estimates. In most developing countries lack of data makes it difficult to separate the principal and secondary activity of the farmer. In such cases household inquiries are required to estimate the size and complexity of the situation and make adjustments accordingly. Questions often arise concerning the inclusion of food processing activities undertaken in households for their own use. These include activities such as the milling of maize, wheat and barley, the husking and polishing of rice, the drying of plantain and root crops, the extraction of vegetable oils and the preparation of drinks such as beer and arrack. It may be recalled (see Chapter 3) that the 1993 SNA made it clear that all activities producing goods that can be disposed of on the market should be included in the production boundary, but the production of personal or domestic services for own consumption by members of the households (including the cooking of food for immediate consumption) should not. A working rule can be prepared by looking at the quantum of the activity. Notes on methodology or footnotes should generally be given to make the coverage clear so that the users of the data can draw appropriate conclusions. In the following discussion, while giving the methods of estimation, the distinction between type of unit has not been highlighted unless necessary.
5.15 The period for compilation of SEAFA depends on the availability of various sets of data which are required to compile comprehensive national accounts so that SEAFA can be linked to the latter. The period can be different from the agricultural year in which case adjustments are required to convert the entire data from the agricultural to the accounting year. This is not much of a problem for livestock activity, but crop production activity requires special care. The 1993 SNA has already laid down some of the guiding principles (see Chapter 3). Generally, for most temporary crops, crop-production time is four to six months depending on the variety of the seed and type of crop. Thus, one of the following four situations may be encountered:
· TYPE 1: Crop is sown in the previous year (t-1) and harvested in the current year(t);
· TYPE 2: Crop is harvested in the same year in which it was sown;
· TYPE 3: Crop is harvested at the end of the year in which it was sown so that it does not reach the final users that year;
· TYPE 4: Crop is sown in the current year but harvested in the next year (t+1).
5.16 Type 1 and Type 4 situations are similar. In such cases, the part of the production carried out in the previous or current year should be recorded as work-in-progress and estimated on the basis of the allocation of actual and imputed costs incurred between two years. Thus, the output corresponding to the first two months for Type 1 situation does not form part of the output of the current year, whereas work done in the first two months for Type 4 situation is included. The two entries may be expected to cancel each other (except under conditions of high inflation), unless the pattern of production changes between the two crop years. However, if significant changes in productivity occur the net change in work-in-progress could be estimated as the portion of the cropping period x the area of the crop x the change in productivity. This can easily be derived mathematically. Thus, depending on the availability of data, countries should record work in progress for Type 1 and Type 4 situations. In Type 2 situation there is no need to measure work-in-progress whereas in Type 3 situation the total output should be recorded as an increase in inventories of stocks of finished goods.
5.17 In the case of permanent crops including plantations (where the trees take much longer than a year to grow to maturity and provide repeat products over a number of years), the gross fixed capital formation by the agriculture sector. This capital formation includes agricultural services such as clearing and preparing the land for planting the crop. It may be estimated on the basis of the total costs incurred including actual as well as imputed costs of labour and items of intermediate consumption used.
(a)
Crop output
(b)
Production of animals
(c) Animals sold for slaughter and animal
products
(d)
Agricultural services
5.18 Agriculture covers the production of goods and services meant for sale, payment of wages and salaries in kind, own account final use, addition to stock and for barter. For the compilation of economic accounts, it would be necessary to make estimates of output or production for all activities for which accounts are required to be prepared. The discussion which is of prime interest to the formulation of SEAFA has been divided into two parts -concepts and measurement.
5.19 The 1993 SNA does not recognize the concept of total farm output. In SNA output of goods and services, which are covered in one or other of the transactions recognized by the system, is the outcome of the production process that is sold or bartered, that enters the producer's inventories or is supplied to another establishment belonging to the same enterprise for use as intermediate input or that is retained for own use or given as a gift or charity.
5.20 However, agricultural activity is rather different from other production processes. Many agricultural production processes last less than a year, which is the normal accounting period for SEAFA. The same crop (and this is true also for chicks) is produced more than once and some of the output of the first cycle is taken as input into the second cycle. The preparation of production plans for improvement of production technology (such as the drafting of a seed replacement plan) therefore requires complete information on inputs. In addition, agricultural data collection systems, especially in developing countries, do not always permit purchased inputs to be distinguished from those provided by own account production. The preparation of an inter-industry transaction table also requires this information. Another important characteristic of agricultural production is the wastage that occurs at various stages of production related operations that form an integral part of the production process. Examples are cereal grain threshing and winnowing, on-farm drying and moving and unloading of goods between fields, threshing area, drying area and farm granary or stockhouse. A number of factors affect the final output and are determined by technical constraints such as the type of product and the production process, the size of packing and storage installations, the conditions of the pick-up and transport system and the geographic proximity to processing units. Therefore, although SNA is only concerned with the output that is ready to leave the farm, for a good agricultural plan it is also necessary to monitor the wastage (as wastage reduction means increase in output) and intra-farm intermediate consumption. SEAFA needs to take these factors into account.
5.21 In the light of the foregoing discussion, four possible alternative measures of agricultural output may be distinguished:
Harvested Output
| less | losses (wastage) on producing farm between harvest and the use of or storage of the output. In such cases output covers both the principal product and by-products |
| equals | Total Output |
| less | intermediate uses of harvested output, principal product or by-products within the producing farm (establishment) itself |
| equals | Output (SNA) |
| Less | intermediate uses by a farm (establishment) other than the producing farm equals Output Available for Use Outside Agriculture |
5.22 The harvested output represents quantities actually harvested in fields, orchards, vineyards and other plantations and animal products such as milk, eggs and honey. It does not include quantities lost during or after harvest. Total output (which in the past has sometimes been described as "gross-gross" output) consists of the total quantity of agricultural goods and services ready for storage by the producer, for intra-farm use or for removal from the farm for other uses or sale. If the harvested output is known total output can be calculated by deducting all losses (wastage) during "auxiliary operations" incurred between the time of harvest and the moment of delivery, use or storage on the farm. From total output, SNA output can be calculated by deducting the value of agricultural products used for intermediate consumption during the accounting period in question on the same farm that produces the commodities; for example, seeds, plants, plain fodder, milk consumed by unweaned animals, manure and eggs for hatching. Finally, output available for use outside agriculture is obtained from SNA output by deducting the value of the agricultural products used for productive purposes during the accounting period on a farm other than the farm that produced them. This concept of output relates to the so-called "national farm".
5.23 The 1993 SNA, naturally recommends the use of the SNA concept of output in the production accounts of agricultural establishments. However, for SEAFA the concept of total output keeps the following facts in view:
· The good and/or service as an output of an industry is a product that is available for final or intermediate uses outside of the industry concerned. Thus the auxiliary operations listed as "agricultural and animal husbandry service activities, except for veterinary activities" are an integral part of agricultural activities. In fact, it is not always easy to measure the contribution of these "auxiliary operations" separately from the production of crops or animal husbandry products when these come from the same household establishment. Thus, when the operations are carried out by the farms themselves, they must be classified as auxiliary (i.e. secondary) activities and be included in the output of the establishment. However, when such services are purchased by the producer household, the output of such services is recorded separately under ISIC group 014 (agricultural and animal husbandry service activities except for veterinary activities).
· Losses (wastage) have to be deducted from harvested output. Of course, these losses would not be visible in the account unless clear provision is made to present them. As mentioned above, it is desirable to collect data on losses so that valid inter- or intra- country comparisons can be made with the aim of minimizing such losses. The data can be presented in the form of supporting statements, with the losses being valued at basic prices.
· From the above description it is also clear that the harvested output is only a provisional measure of the output of agricultural activity as it does not take into account post-harvest operations and related losses.
· It is not recommended to use the SNA concept of output for several reasons:
(i) it does not provide the information needed to prepare a seed replacement plan;
(ii) it is difficult to gather the requisite data, especially in developing countries where estimates of some components of intermediate consumption (i.e. seed, feed, etc.) are based on the cost of cultivation, production studies or norms derived from special studies -- the sources of procuring inputs are not recorded in such studies;
(iii) at the time the production takes place it may often be undecided whether or in what proportions the goods produced are destined for the market. In subsistence agriculture own produce is very often used as seed, but during times of shortage this produce is sometimes consumed instead.
· The estimate of output provided by crop-cutting experiments relates to harvested output and not SNA output.
5.24 In view of the arguments given above and the needs of planners it is recommended that SEAFA should use the concept of total output. To maintain closer links with SNA it would be possible to show the SNA concept of output in the accounts and also to show at the same time the intra-farm uses of own production under both "resources" and "uses" in the Production Account and also in the Goods and Services Accounts. In practice, however, this is almost impossible because of lack of data. It would mean amassing data on all the products consumed within the agriculture sector, i.e. seed, plants, manure, plain fodder, straw and pasture grasses grazed directly by animals, etc. In practice, this is not feasible. At best, it may only be possible to evaluate the most important products consumed within the sector.
5.25 For measurement of output, the agriculture sector may be divided into four sub-groups viz., crop production, production of live animals, animal products and production of agricultural services when these have not been covered under crop production. In the following discussion measurement of the quantities of each component is discussed first and their valuation taken up subsequently.
5.26 Crop production for analytical reasons can be decomposed into two components, viz. area and productivity. As mentioned earlier, after careful examination of the available data total crop production can be divided into three separate groups of crops depending on the reliability of the data. While processing the data a checklist may be prepared indicating:
· a list of all the crops grown in the country or region, divided into three groups, viz. crops for which data on both area and productivity are available; crops for which only data on area are available (i.e. productivity data are not available); and crops for which only production data are available, either directly through Marketing Boards etc. or indirectly in the form of processed or finished end products;
· the ancillary data, which can be used to prepare estimates of output for the crops, for which data on productivity are not available;
· whether the subsistence production has been covered;
· the stage of production to which the data for various crops relate (for example, whether the production of paddy is available or paddy rice) and whether these stages are covered under ISIC group 01;
· whether the production of by-products (according to stage of production) has been included. In the processing of crops by indigenous methods several by-products are also produced. However, for the estimation of production only those by-products that are readily identifiable and have some definite economic values are considered. The quantum and value of such by-products are generally available from farm management studies or cost of production surveys;
· whether, when certain non-agricultural activities has been included, they can be separated on the basis of structural ratios derived from data available from other sources, such as large-scale surveys of manufacturing establishment and case-studies.
5.27 The main sources of data on agricultural output have already been listed. However, for a few crops more reliable data are sometimes available from sources that are not meant to provide data for agricultural statistics. For example, for crops such as tea where a number of pickings are made during the year, it becomes costly to organize proper crop estimation surveys but information on the production of processed tea may sometimes be easily available from institutions such as marketing boards or cooperative societies that are involved in bulk trading in such commodities. It may then be more convenient to obtain an estimate of the production of the processed final product (such as processed tea) during the year and to adjust this figure using a structural ratio obtained from data collected in mail inquiries or from reliable producers. In certain other cases, such as the production of fodder crops and grasses where generally no data are being collected in developing countries, the imputed value of the output could be prepared directly on the basis of a special survey listing the area under crop and the estimated value thus arriving at a benchmark estimate that can be extrapolated on the basis of either total area under crop or number of livestock. In some rare cases, estimates of total output for a few crops, such as fruits and vegetables, can be made from consumption data. In such cases, it is necessary to examine the consumption data carefully in order to establish how representative it is and also to convert quantities of processed products consumed into quantities of unprocessed ones produced. Adjustments would also be needed when the product has non-consumption uses and may be exported or imported.
5.28 The worksheet for preparing estimates of total output should carefully record losses that occur during and after harvesting (the latter being needed when preparing goods and services accounts) but during the same accounting period. The intra-farm uses of the goods (such as feedstuffs for animals, fuel for on-farm processing of the goods or seeds used in the same accounting period) produced and any losses during the harvesting operation should be subtracted from harvested production to arrive at SNA output. The post-harvest losses which occur during transportation of the crop produce by traders or manufacturers and losses during storage become part of intermediate consumption of the processing industry or the trade activity. Catastrophic losses caused by natural disasters, etc. should, however, not be treated as intermediate consumption. It may be possible to make approximate estimates of normal, recurrent losses on the basis of small-scale studies.
5.29 Generally (i.e. unless it is necessary to make adjustments for changes in inventories because of the switch from crop to reference year) it should be assumed that there are no change in inventories of finished products in the agricultural sector. This assumption excludes the activities of storage and trading (hoarding for trading) from the production of the agriculture sector.
5.30 The production of live animals depends on changes in the numbers of animals in economic use. These changes are calculated after dividing the total population into different categories according to their economic uses. Classification is needed in order to obtain the production value attributable to the total cost of inputs incurred (water and feed, veterinary care, heating of stables, etc.), over the whole production period, real or imputed payments for factors of production (labour of the holder and employees) and other costs (wear tear and obsolescence of buildings and equipment used by animal husbandry farmers) during the year under consideration. For example, the production of bullocks for slaughter is spread over a number of years as it takes a long time for them to reach their optimum size and weight. Therefore, the value of the slaughtered animal does not reflect only the production costs incurred during the year in which it is slaughtered, but also the accumulated cost of all inputs and factors of production over all the years from its birth. The same is true for other animals reared for slaughter, work, breeding or for producing other products such as milk, wool, fur and eggs. However, only the costs incurred during the year are recorded in the production account and thus only the corresponding changes in the composition of live animals become the output of the current year. To make a correct assessment of this component changes in the number of animals by different categories are evaluated.
5.31 Comprehensive data on number, type and age of animals are collected in Livestock or Agricultural Censuses which may be organized after every five to ten years or so. However, in some countries annual data on the more important types of live animals are also collected. If annual data on changes in the number of live animals classified by type of animals are not available, they can be estimated from periodic benchmark data collected in livestock censuses. Let us define PA as the total volume of output of live animals in any year (i.e. produced between the beginning and the end of the accounting year). By definition:
PA = SI + D C + E - I
where SI is the number of animals slaughtered inside the country for economic purposes (i.e. production of meat, leather, fur, etc.) or dying from natural causes during the year, AC is the annual change in the number of live animals obtained by interpolation or extrapolation of livestock census data, E is the export and I is the import of live animals by the country. In this case the term D C refers to the actual change in the total number of different categories of animals (i.e. classified according to sex, race, age, weight, etc., which are quantifiable characteristics for economic purposes) derived from the Livestock Census. Thus, to estimate the output of live animals it is required to estimate (PA - SI) from data available from the census and foreign trade statistics. However, it may be mentioned here that the data on number and end use of animals that have died from natural causes are generally not available in many countries where animals are used to obtain by-products such as hides and skins. The SNA 1993 recommended that "Incidental losses of animals due to occasional deaths from natural causes form part of consumption of fixed capital" (paragraph 10.87). A more accurate treatment of this component depends on the availability of data and on country practices.
5.32 At this stage it is important to understand that although the total change (+/-) in the livestock population, excluding changes resulting from the slaughter of animals and deaths from natural causes, is part of agricultural production it is also capital formation. The true change in the livestock population after being adjusted for export or import is classified into two groups. The first group consists of changes in numbers of those animals that are treated as fixed assets, namely adult dairy animals, animals raised for their wool or breeding or as draught animals (used for work) that are more than one year old (i.e. when they become economically active). Changes in the population of these animals therefore constitute gross fixed capital formation. The remaining animals, including young stock, are not fixed assets and changes in their numbers are recorded under changes in inventories. This group consists of animals that are being reared for their future use (as fixed assets or for slaughter). Changes in their numbers are recorded under changes in work-in-progress or changes in inventories in finished products (ready for slaughter), as the case may be. Thus the term /`C can be broken down into terms:
D C = D CF + D CS
where F and S represent fixed assets and inventories (stocks) respectively. Similarly the term I, representing import of livestock, can be decomposed as:
I = IA + IF + IS
where F and S have the same meaning and A denotes animals slaughtered immediately after entry into the country under consideration. Thus the above relation for PA can be written as:
PA = (A - IA) + E - (D CF - lF) + (D CS - IS)
5.33 Logically, the output from live animals in a country (PA) should not include animals imported for immediate slaughter on entry into the country since the output to be measured refers to domestic agricultural activity. For the same reason, the value of output does not cover the ax-customs value of imports IF, i.e. animals imported for use as fixed assets. But generally data collected in the Livestock Census do not distinguish between animals produced in the country and imported animals. Thus, while projecting (or interpolating) data available on numbers of livestock from sources such as censuses this aspect needs to be carefully examined. However, generally the necessary details are not available in the census data and in practice even an estimate that does not include such adjustments may give a good approximation to total production.
5.34 Another aspect requiring consideration is that exceptional losses of animals as a result of major outbreaks of disease, contamination, drought, famine or other disaster are recorded in the other changes in volume of assets account and not as disposals (SNA paragraph 10.87). In other words such losses (i.e. deaths without any economic use) should not be treated as output but as catastrophic losses that do not affect the production account of SEAFA or the 1993 SNA. However, such instances, which are rare in nature, could be shown as a memorandum to the gross capital formation figures.
5.35 An alternative estimate of the change in the number of each kind of live animal during a given year attributable to domestic production can also be estimated as the number of live births each year per head (the reproduction rate), times the total number of livestock at the end of the previous year, less the number of live animals exported and the number of domestically produced animals slaughtered or dead as a result of natural causes during the current year. The normal reproduction rate for each type of livestock can sometimes be obtained from offices such as the national Veterinary Department.
5.36 Data on animals slaughtered are generally obtained from slaughterhouses, marketing boards and veterinary inspectors and need to be adjusted to take account of the numbers of livestock slaughtered for own consumption using data obtained through special livestock surveys or other studies covering such activities. Where no information is available, the number of livestock slaughtered both for sale and for own consumption is sometimes assumed to be equal to the number of hides and skins produced which can be estimated from export statistics or from surveys dealing with the manufacture of products from hides and skins. Preferably, the use of such indirect methods should be minimized to reduce errors in the estimates. However, if direct data are not available, estimates of slaughtered animals should be cross-checked with data obtained from consumption surveys, although consumption surveys generally do not differentiate between types of meat. Thus, to make a cross-check, it may be necessary to prepare a cattle equivalence table for output of meat per animal. Estimates for different animals can be combined taking into account the numbers of different types of animal recorded in livestock census that are used for meat purposes.
(c) Animals sold for slaughter and animal products
5.37 A third important category of agricultural output consists of animals sold for slaughter and animal products. The sources of data on animal products such as milk, eggs and wool clips are cost of production surveys, production or marketing surveys and reports of marketing boards. The production of meat, hides, skins, etc. is classified as manufacturing activity. Generally a structural ratio is derived either from current data available from the same sources or from information available from the veterinary department. The ratio is then applied to the total number of animals. For example, an estimate of the total output of cow milk can be made using the total number of cows, the proportion of cows in milk to total cows and the average yield of milk per cow (in milk). Similarly estimates of eggs, wool, fur, etc. can be prepared using the relevant structural ratios and the animal population. However, whenever possible a cross-check of such estimates should be made with data available from consumption surveys duly adjusted for other uses of the product by agro-based industries, non-household (hotels, military, etc.) consumption, wastage and intra-farm own account consumption. Similarly, the number of animals sold for slaughter can be obtained from marketing boards, slaughterhouses or through surveys of manufacturing establishments covering such activities.
