Data quality, concepts and methodology: International travel statistics

View the most recent version.

Background

The statistical series on travel between Canada and other countries originated in the early 1920s primarily as a requirement for the Canadian Balance of International Payments.

The method of collection relied heavily on the detailed administrative records of Canada Border Services Agency (CBSA) and Citizenship and Immigration Canada (CIC) authorities to obtain the number of border crossings and on the officers from these departments to distribute questionnaires to a sample of travellers in order to collect information on their trip spending.

The accuracy of these methods depended on the completeness of the records of traveller movements and the representativeness of the sample expenditures derived for each category of traveller.

The statistical processes continued to rely on the administrative records of co-operating departments until 1972. Several changes in the methods of documenting visitors and resident travellers, resultant from policy revisions by CBSA and CIC programs, have led in the past to changes in definitions and breaks in statistical continuity. The purpose of the change in 1972 was to standardize definitions, methods and requirements for the collection of traveller counts, in light of the movement toward facilitation of traffic flows, and as a means of more adequately controlling the survey results.

The original purpose of the questionnaire surveys was only to collect expenditure data to be applied to the traveller counts in order to derive aggregate expenditures on visitors to Canada and residents returning from abroad. The first questionnaires were short and contained only three or four questions. The response rate was satisfactory and consistent from year to year.

As travel gained in importance to the Canadian economy, the tourism industry voiced the need for more detailed information on travellers for market research and industry planning. The questionnaires were gradually expanded to include as many as 19 questions on different characteristics, including purpose of trip, duration, locations visited, types of accommodation used, activities and expenditures. The last major changes to the questionnaires were made in 1990. However, minor modifications were introduced in 2001.

Prior to 1990, questionnaires for United States residents visiting Canada were distributed to these travellers when returning to their country by United States Customs officials. The questionnaires were processed by the United States Department of Commerce and the results were sent to Statistics Canada. Since 1990, questionnaires are distributed to residents of the United States upon entering Canada during sampling periods by CBSA officials. Completed questionnaires are mailed to Statistics Canada for processing.

Between 2000 and 2001, several methodological initiatives were developed to improve the quality of the estimates based on the questionnaire surveys. These initiatives include a bias adjustment to the population weights, a procedure for imputing international transportation fares and total trip spending, an air exit survey of overseas visitors, a regional weighting of questionnaires from overseas visitors, a new production schedule to accommodate the inclusion of new questionnaires and a new method for calculating the coefficients of variation.

In 2002, the 2000 and 2001 data from the questionnaire surveys were revised to incorporate the results of these initiatives. Since 2002, these initiatives are part of the new methodology used to produce estimates from the questionnaire surveys. This new methodology is summarized under "Questionnaire surveys" in the "Data quality, concepts and methodology — Description of methods" section. Also, documents describing in a more detailed fashion the new methodology and each of the new initiatives are available upon request.

The introduction of this new methodology resulted in a break in the time series of estimates based on the questionnaire surveys. Consequently, data from previous years are not comparable to revised 2000 and 2001 data, and data published since 2002. However, 2000 and 2001 data were revised to allow users to make comparisons to those years.

The methodology used may be subject to two types of bias: a distribution bias, that is, the questionnaires may not be handed to a random selection of travellers and a non-response bias, that is, returns may not be representative of the travelling public.

Although about 88,600 non-resident travellers entering Canada and 101,500 returning residents were covered by the survey to produce the 2009 estimates, these numbers represent less than 1.0% of the total traffic.

Given the size of the sample, the basic purpose of the questionnaire surveys (the estimation of visitors' spending at the Canada and regional level and spending of resident travellers by region of the world) continues to be met with reasonable levels of reliability providing the assumption of negligible biases is not violated. Estimations of expenditures and other characteristics at lower levels of aggregation, such as certain provincial data and detailed cross-classifications, strain the capacity of the survey; the resultant estimates are less reliable and several of them are not sufficiently reliable to be published, as can be seen in several tables of this publication.

Travel and the balance of payments

Spending by non-residents visiting Canada (receipts) and spending by Canadian residents returning from trips abroad (payments) are incorporated in the travel account which is an integral part of the current account of the Canadian Balance of International Payments. In 2009, travel receipts, included in exports as part of "services", represented about 3.1% of all current account receipts while travel payments, included in imports as part of "services", represented approximately 5.1% of all current account payments.

In the context of the Canadian Balance of Payments, receipts in the travel account are defined to include all expenses incidental to travel in Canada by non-residents. Among these are expenditures in Canada for lodging, food, entertainment, local and intercity transportation and all other purchases of goods and services (including gifts) made by travellers. The series thus includes any purchases of personal goods to be exported by travellers. Also included are medical expenses and education expenses of non-residents in Canada as well as foreign crew members' spending in the country.

Payments in the travel account are correspondingly defined to include all expenses incidental to travel abroad by residents of Canada. Among these are expenditures abroad for lodging, food, entertainment, local and intercity transportation and all other purchases of goods and services (including gifts) made by the travellers. The series thus includes any purchases of goods to be imported for personal use by travellers. Also included are medical expenses and education expenses of Canadian residents outside Canada as well as Canadian crew members' spending in other countries.

The travel account does not conform precisely to the concepts recommended by the International Monetary Fund (IMF). Under those concepts, personal expenditures of foreign migrant workers or Canadian migrant workers, including temporary and seasonal workers and commuters, are part of other balance of payments accounts. However, due to survey limitations, these expenditures are included in the Canadian international travel account.

Furthermore, it should be noted that travel receipts exclude international transportation fares paid by non-resident travellers to Canadian carriers. Also, travel payments do not include international transportation fares paid by Canadian residents to foreign carriers. In both cases, these expenditures are included in another balance of payments account, in accordance to the IMF conceptual framework.

Next