Appendix I: Appendices

View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Archived

This page has been archived on the Web.

Appendix 2

Appendix 3

Appendix 4

Appendix 5

Appendix 6

Appendix 7

Appendix 8

A new conversion

Data conversion is generally not used to conduct an economic analysis of this type. Furthermore, the database produced from the conventional conversion of the Harmonized System to NAICS may potentially have limitations that are hard to identify. Since no test has ever been done to check the robustness of this conversion, a new conversion will be explore, from the North American Product Classification System (NAPCS) to NAICS, and the results of the two converted data sets will be compare.

Trade data are coded according to a product classification system (Harmonized System or HS, and North American Product Classification System or NAPCS), while manufacturers’ sales are coded according to an industry classification system (NAICS, manufacturing codes 31 to 33). Since the two classification systems are conceptually different, it is not easy to integrate these two data sets. The conventional conversion that was used for the economic analysis in this study is from HS codes to NAICS codes. However, few tests have ever been conducted to check the robustness of the data obtained from this conversion. ¹  The concordance of the HS code to the NAICS system was carried out several years ago and was based on the principle of the “industry most likely to export this product”. To verify the robustness of the results of this work based on the choice of a conversion method, we sought to compare the data produced by this method with those yielded by another method.

Since 2012, international trade data have been published using the NAPCS product classification system. ^2 , 3  For our conversion, we took trade data from NAPCS and converted them into NAICS data. To make a link between the two classifications, we adapted another conversion that matches NAPCS classes with the Annual Survey of Manufactures (ASM) List of Goods of 2004. ⁴  This list of goods classifies both goods purchased and goods produced by Canadian manufacturers, and the first three digits of this classification correspond to the first three digits of the 21 manufacturing industries based on the NAICS code. In this classification, products are classified according to their industry of origin. This concordance therefore gave us a very good basis for generating our own conversion. Accordingly, we looked at the classes ⁵  of the NAPCS code on international trade from 2002 to 2012 and changed them into the three-digit numeric NAICS code, if these classes were of a manufacturing nature according to the ASM List of Goods.

The results of this new conversion generated data that were similar and comparable to those in the conventional conversion, across the 21 manufacturing industries and the 10 leading countries to which Canadian manufactured goods are exported, showing the robustness of both conversions. For example, between 2002 and 2012, the difference of the total between the two conversions was no more than 0.67%. ⁶  In cases where some industries yield results that are not perfectly aligned, the trends are nevertheless quite similar. ⁷  Chart 1 shows the totals of the two conversions side by side from 2002 to 2012.

Having obtained comparable results in the two cases, we believe that these two conversions yield similar results in this specific analytical framework, namely the analysis of domestic exports at the three-digit NAICS level. We cannot guarantee the quality of the results should this conversion be used for re-exports, for imports, for more specific NAICS levels (5-digit NAICS) or for provincial data, since we did not test these levels. We make the assumption that the more detailed the conversion, the less robust the results will be. Although not perfect, these conversions generate data that most closely approach the economic reality of the subject analyzed in this paper. Other studies could use this new conversion to do other tests and improve it. These data could also be compared with, or serve as a complement to, the databases generated by the register of exporters and importers ⁸  when it is updated.