1. INTRODUCTION

Censuses taken at periodic intervals are an essential part of an integrated system of national industrial statistics. Benchmark surveys of the census type provide information on the structure and level of activity of a country's industrial sector. Appropriately constructed, such inquiries also contribute to the international comparability of industrial data. The role played by the periodic census and its relationship to annual and more frequent inquiries are fully discussed in Chapter I-10.

The items for which data are collected are only one segment of the total census picture. Basic as they are to the census, the items of data are framed in a setting which describes the scope and coverage of the census, its reference period, statistical unit and classification structures for collecting and tabulating the information. In this case study, all of the elements involved in the taking of a census are treated as concepts.

The success of a benchmark census is measured by the care with which the concepts are selected and the skill employed in defining them and in translating them into questionnaire design, processing procedures, and table outlines. The first step a country such as Providencia must take in embarking on a periodic program of industrial censuses is to decide upon and define clearly all the concepts to be used. In this regard, effective guidance is available from the United Nations Statistical Office which has issued a series of international recommendations for industrial statistics. Providencia used those guidelines as the frame of reference in formulating its 1975 Industrial Census Program.

integral part of a national system of industrial inquiries. The enumeration of units was done in a manner that contributed to the creation of an industrial directory of establishments for other current and periodic surveys. Establishment counts were built into the table outlines to facilitate the later selection of samples and determination of budgets for other industrial surveys. The product and materials inquiries, the definition of the statistical unit, and the system of classifying data provided for the compilation of input-output studies (at broad aggregative levels), physical production indexes, and weighting factors for price indexes--all are factors which affect the content and design of the

The

Various international organizations have developed concepts, principles, and recommendations for the taking of industrial censuses. These bodies have issued comprehensive lists of topics to be investigated in industrial censuses together with their definitions. principal source of such guidelines is the United Nations Statistical Office, but significant contributions have been made also by the Inter-American Statistical Institute (IASI) and the Organization for Economic Cooperation and Development (OECD), with specialized efforts by other groups.

3.1 United Nations (UN) guidelines

Providencia made extensive use of the UN principles and recommendations in the planning of its industrial census. The treatment of concepts in this case study conforms mainly with the presentation made in United Nations publications. Benchmark inquiries of the industrial census type are specified in a 1968

United Nations publication. That document covers infrequent comprehensive inquiries, annual inquiries and more-frequent-than-annual inquiries on industrial activity (mining, manufacturing, electricity, gas and water). The system of industrial statistics discussed is limited to data appropriate to the establishment or establishment-type units. Enterprise statistics are not considered, nor are they included as part of Providencia's Industrial Census.

2

The 1968 guidelines were updated and modified by the UN in a three-part publication developed for the 1973 World Programme of Industrial Statistics. In these publications, recommendations are presented not only on the objectives and content of the program but also on the organization and conduct of the census and the individual commodity and material detail to be collected. Regional working groups of the United Nations made important contributions to the drafting of those recommendations.

The 1968 UN Recommendations for Industrial Statistics makes a distinction between items of data to be collected and tabulated for all establishments and those to be collected and tabulated for large establishments only. Within these categories, each item is indicated as being of first or second priority. The 1973 UN Recommendations further subdivided the data items between those suggested for countries with developed industrial statistics and those applicable to countries beginning to develop their industrial statistics programs. Providencia,

1United Nations. Statistical Office. International Recommendations for Industrial Statistics (New York, 1968), Statistical Papers, Series M, No. 48. Rec

2United Nations. Statistical Office. ommendations for the 1973 World Programme of Industrial Statistics (New York, 1973), Statistical Papers, Series M, No. 54.

There

in its industrial census, generally adhered to those data items proposed as first priority for countries with developing programs. were a minimum number of departures from the UN recommendations. Likewise, Providencia generally conformed to the UN placement of items on the long (large establishments) or short (small establishments) forms. It is recognized that these UN guidelines allow for modifications reflecting the special characteristics of particular regions or countries.

The UN statistical programs are designed to encourage the orderly development of national inquiries into the structure and activity of the industrial sector. At its forty-fourth session, the UN Economic and Social Council recognized "the need for data on the structure and activities of industry on a world-wide basis for the purposes of economic and social development" and noted that "the Statistical Commission has made recommendations for a system of industrial statistics for use by Member States."

Where

The

The COINS Subcommittee's recommendations are in basic agreement with the UN guidelines with respect to the topics that should be investigated but they differ in some cases on the extent of classification detail. differences do occur between UN and IASI in concepts or data items to be reported, such variations are noted in this case study. IASI Subcommittee members placed special emphasis on the need for improvements in the quality of data and for achieving prompt and efficient publication of the statistics. Furthermore, they stressed the importance of continuing programs to train specialized personnel in industrial statistics.

Although basically the UN recommendations were followed, it was necessary for the planning staff in the National Statistical Office to evaluate each concept in relation to Providencia's economic and social development programs. Also, the concepts used in the previous Industrial Census of Providencia (1963) were retained wherever possible so as to provide historical comparability.

4.1 Scope of the census

The 1975 Industrial Census of Providencia covered all establishments within its territorial boundaries which were engaged primarily in mining, manufacturing, and the production and distribution of electricity and gas. These activities correspond to major divisions 2, 3, and 4, respectively, of the International Standard Industrial Classification of All Economic Activities (ISIC)" except for the exclusion of water which is a part of ISIC major division 4. In Providencia, as in many other countries, especially those with developing economies, the maintenance of a water system and its distribution to users is a governmental rather than a private function. Information on the activities associated with the supply, storage, and distribution of water (or waterworks), is provided periodically in reports by the government agency administering the facility.

Off-shore activities, such as the operation of petroleum and natural gas wells, are included to the extent that the activities are located inside the territorial waters of

Providencia. Also included as part of the census is the processing of fish on vessels that are engaged solely in the fish processing activity, provided that such activity has its land base within the territorial boundaries

of the country. For Providencia, the territorial waters include all ocean waters up to approximately five kilometers off the coast plus the outer continental shelf to a depth of approximately 100 meters.

When a significant amount of industrial activity in a particular country is performed in non-industrial establishments, efforts are made to isolate and measure that output, although usually in an abbreviated manner. Such output may be found in combination with agricultural activities, with wholesale and retail trade, or with services. As the COINS Subcommittee document states, "the statistical unit should be classified within the sector corresponding to the largest part of its economic production and, to the extent that records permit, separate data should be collected on secondary activities."5

In Providencia, cottage industries (units where the activity is carried out on household premises) are significant in their output of certain industrial goods and services. It is estimated that there are 180,000 of these units in Providencia. Such activities will be covered by a sample of areas, using a separate household industries questionnaire. The case of cottage industries illustrates the interacting of a decision on the scope of the census with the choice of questionnaire, use of sampling, and tabulation plans. Coverage of cottage industries is optional for a country, depending on their contribution to the output of manufactured goods.

"United Nations. Statistical Office. national Standard Industrial Classification of All Economic Activities (New York, 1968), Statistical Papers, Series M, No. 4.

Inter

5Ibid, footnote 3.