A. General description of enterprise and other surveys
4.2. Use of surveys. Customs records are the main and usually preferred data source of trade information (see IMTS 2010, para. 8.2). Non-customs sources, such as surveys, are recommended to be used as substitutes for available customs records only if they provide a cost-effective means of improving the quality of trade statistics (see IMTS 2010, para. 8.9). Therefore, in most countries, enterprise surveys are not extensively applied for the compilation of merchandise trade statistics. Member states of customs unions can be an exception, as customs records might not exist for trade among the members of the union (see annex III.A). For example, in the case of the member States of the European Union, enterprise surveys are a main source of information for international merchandise trade statistics. Data compilation in the case of a customs union is discussed separately in chapter X.
4.3. Surveys designed to collect missing data.Enterprise or other surveys, like border surveys or household surveys, can be useful in obtaining information on transactions that are not processed through customs. Specifically, the following items recommended in IMTS 2010 for inclusion as part of the international merchandise trade statistics may require surveys as the principal data source: (a) “goods acquired by all categories of travellers, including non-resident workers, to a significant scale as defined by national law” (IMTS 2010, para. 1.16), also referred to as “shuttle trade”, (b) trade in electricity, gas, oil and water (ibid., para. 1.24), (c) goods dispatched or received through postal or courier services (ibid 2010, para. 1.34); (d) fish catch, minerals from the seabed and salvage (ibid., 2010, para. 1.31) and (e) bunkers, stores, ballast and dunnage, acquired or landed by national aircraft and vessels outside the economic territory of the compiling economy (IMTS 2010, para. 1.32).
4.4. Requirements and integrated approach in data compilation. Carrying out enterprise surveys entails the commitment of additional resources on the part of the national statistical authorities.Enterprise surveys require more time in planning, execution and follow-up than that required to obtain information from administrative sources. Moreover, a business register (or other sources of survey frame) should be in place to draw a representative sample of enterprises for the purpose of the surveys. The samples should be adequate for the concerned economic sectors and stratified by enterprise size and geographical areas, as necessary and feasible. Considering the survey requirements and the high costs of conducting surveys, countries are advised to follow an integrated approach to economic statistics in order to make full use of existing statistical information and infrastructure for data compilation. Some information on the standardization of surveys is given in section B below. For further details on the integrated approach to economic statistics, see chapter XI.
4.5. Merging of data from enterprise and other surveys with data from other sources. One major issue for data compilers is how to incorporate data from enterprise surveys with data obtained from other sources (predominantly from customs records) as the levels of detail, timing, etc., might differ significantly. Chapter VII discusses this issue in more detail and provides further guidance.