Registry Detail

1546

Information

28 May 2002

ISIC 4 questionnaire - Summary of responses A2

A. Conceptual and structural issues 2. There is a relationship between economic activities and products. The former lead to the production of the latter. The definition of products in existing or revised product classifications, such as the Harmonized System (HS) and CPC, can be used to define the boundaries of activity categories. The relationship between the two types of classifications can range from complete independence to strict linkage on a one-to-one basis. To what extent should the ISIC revision take into account relevant product classifications?

The responses to question A2 are grouped into three categories for review. The categories are: (1) product and industry classifications should be independent; (2) product and industry classifications should be related but not strictly related; and (3) product and industry classifications should have a strict relationship. Summary of assignments to the groupings (1) Independent - 7 (2) Related but not strictly - 11 (3) Strict relationship - 18 In order to further evaluate the answers, responses to A2 were cross-tabulated with responses to question A1. Of the seven independent responses, four desired a production concept, one said that the concept should vary by subject area, and two said that the concept should be output based. Of the eleven requesting a loose relationship, one desired a production concept, one desired a unified concept but provided no direction on what that concept should be, three desired a concept based on output, and six said that the aggregation concept should vary by area. Of the eighteen desiring a strict relationship, ten wanted an aggregation concept that varied by area, two desired an output aggregation criteria, two desired a unified concept with no further details regarding the unified concept that should be applied. Four countries that responded to question A2 did not respond to question A1. Of those that expressed a desire for independence of structure or for strict relationship, the intent of the answers was reasonably clear. Those in the middle group, related but not strictly, did not provide clear guidance. The Technical subgroup noted that the answers indicate that a broader education effort may be necessary to get more consistent answers. At a recent conference, the UN asked about the difference between products and industries and it was not widely understood. Many do not understand the relationship or the implications of the tables that we have. In addition, the Technical Subgroup noted that there appears to be a position that while the current treatment may not be the most correct, changes would cause considerable disruption to data. The Technical Subgroup also noted that for practical reasons, it is often easier to get output data from industry groups rather survey all units that may produce a particular service. Engineering services was provided as an example. It is easier to go to engineering firms to request output data than to cover all businesses that may hire engineers or provide engineering services to clients. Many respondents indirectly indicated that this might be acceptable. The Technical Subgroup postulated that the problem was the lack of development in collection systems. Alternatively, the problem might be the lack of a well-defined infrastructure (product list) for services. Collection methods could be acceptable but detailed collection categories are not available. There was no clear agreement on the Technical Subgroup on this point. The Technical Subgroup also noted that there are coding problems separate from linkage problems in industry and product classification systems. The Technical Subgroup also discussed the meaning of the question. Should a linkage between products and industries be an empirical observation (such as a make table) or a structural design consideration. The empirical approach states, “here are the products and these are the industries that produce those products”. On the other hand, if moving toward structural approach for linking products and activities, we would have to define classes in each set so that the observed empirical facts are built into the product classification. For example, there would always be a list of products that are produced by an industry. The Technical Subgroup also considered the synergistic possibilities of empirically related rather than structurally related classifications. It was mentioned that the empirical relationship can change over time and those changes can be evaluated without structural change to either the product or industry classification. Overall, there was not a unified response to question A2. There was a strong sentiment for a loose or strict relationship between product and industry classifications but there were a number of issues discussed by the group that could impact the interpretation of this result. In particular, the issues of a structural relationship vs. an empirical relationship and the general feeling that more education and outreach on the difference between industries and products are important factors when undertaking future work on ISIC and the CPC.