Bongho Choi

National Statistical Office, Korea

1. Basic Survey Frame Currently Used

In Korea, the results of the ˇ°Annual Census on Basic Characteristics of Establishmentsˇ± have served as the main frame for various business surveys. This census was started in 1994, having the reference date as the end of each year.

All units which have a geographically identified workplace with at least one full-time worker and carry out economic activities regardless of profit or non-profit and legal or illegal are covered. However, sole proprietorship in the sector of agriculture and fisheries, activities of military authorities, foreign diplomatic institutions, sales with no fixed place, private households are not covered in the census.

Items included in the census are as follows: 1) name of establishment and representative, 2) address of location, 3) starting date, 4) legal form, 5) type of establishment classified by single or multi-unit status, 6) the address of headquarters in the case of branch offices and factories, 7) legal entity number issued by courts, 8) code for kind of activity in 5 digit, 9) number of employment, 10) annual turnover, 11) business registration number issued by the tax offices.

2. Development of New Database System

KNSO has made many efforts to develop a ˇ°Integrated Business Survey Frame Systemˇ± since the beginning of 2003. The system is composed of two parts. One is the frame for establishments and the other is the frame for enterprises. The new system has the following capabilities:

1) The new database is updated using the returns of the statistical censuses/surveys, available administrative sources and other useful data sources in the private sector. Specifically, the following sources of data are utilized: Annual Census of Basic Characteristics of Establishments, annual surveys for the sectors on mining, manufacturing, construction, transportation, wholesale & retail trade and services, administrative data kept by the Financial Supervisory Board and the Labour Welfare Corporation. Thus, the new system provides the latest information on the population of establishments/enterprises, and could be used as the central sampling frame for surveys on establishments/enterprises.

2) The new system is based on the concept of data ware-housing(DW) and the on-line analytical processing(OLAP). The new database could be used for various analyses on establishments/enterprises in a more convenient way. For example, statistical tables could be produced from the database from time to time; panel analysis could be performed on enterprises; the status of grouping of enterprises could be studied; etc.

3. Number of Statistical Units in Korea

In Korea, the number of legal units liable to the value-added tax is about 4 million, while the number of establishments surveyed is about 3.1 million as of the end of 2002. The number of legal units liable to the Corporation Tax is about 345 thousand, while the number of incorporated enterprises (companies) profiled by the KNSO is about 186 thousand. Although there is no in-depth study, the differences between the legal units and the statistical units seem to be due to the difference in the coverage.

4. Issues

1) It is recognized that administrative sources is useful and effective for expanding the contents of the database. However, KNSO is not able to access the commercial register kept by the courts and the taxation data (data on VAT , etc.) kept by the tax offices.

2) KNSO has tried to introduce the statistical unit on enterprise in its Integrated Business Survey Frame System. However, it was possible for KNSO to create the statistical unit on enterprise only as for the incorporated companies, using the results of Annual Census on Basic Characteristics of Establishments. This means that the non-incorporated businesses (the majority of them are sole proprietorships) are not considered for the analyses since they are too many to be profiled by the KNSO.

3) Some codes for kind of activities of establishments are revealed not to be accurate. It is observed that 10.3 % of codes are wrongly input in the case of 2002 Annual Census on Basic Characteristics of Establishments. One of the reasons for the quality problem on industrial classification is that the coding is done by interviewers themselves.

4) Demographic data (births and ceases) for establishments/enterprises are not currently produced. Annual turnover data are also revealed to be in poor quality. This is why KNSO is not publishing the data on annual turnover in its annual publication on the basic characteristics of establishments.

5) It is found to be difficult to profile the statistical unit on enterprise groups because of the non-availability and complexity of data on the shareholder-information.

5. Future Plans

1) New items such as the number of branches of an enterprise will be added in the Annual Census on Basic Characteristics of Establishments.

2) Company Organization Survey will be introduced in 2005. The survey will cover some complex enterprises at the initial stage.

3) Dialogue with the tax office is continued in order to share the data.

Source: Internal information from the Industrial Statistics Division of KNSO.