Transcript Balance of International Payments Division

The measurement of
international travel services:
Current methodology and alternatives
(including the use of credit card information)
Balance of International Payments Division
November 9th, 2011
Introduction
 Total travel services relatively small, but volatile, portion
of the current account
 Significant impact on trade in services balance
 Travel series used in several System of National
Accounts components
 Current travel estimates combine traveller surveys and
traveller counts
 Alternative sources to validate or replace current sources
• Potential sources
• Advantages and issues
2
Statistics Canada • Statistique Canada
2016-03-26
Some figures on travel
 25 million foreign travellers to Canada in 2010
• 82% from United States (mostly by automobile of which half on
same-day trip)
• 45% of the spending from United States travellers
 54 million Canadian travellers abroad in 2010
• 84% to United States (mostly by automobile of which two-thirds
on same-day trip)
• 60% of the spending made in United States
 Increase in the travel deficit of Canada
• Deficit rose from C$2.1 billion to C$14.3 billion between 2001
and 2010 ... Mostly on larger deficit with United States
3
Statistics Canada • Statistique Canada
2016-03-26
System of National Accounts requirements
 Balance of payments
• Travel spending by purpose of the trip; details on medical, education,
and non-resident workers; passenger fares by mode of transport;
quarterly series with geographical breakdown
 Gross domestic product
• Exports-imports of services; personal expenditures (spending by broad
category); provincial allocation
 Canadian Tourism Satellite Account
• Supply (industry side) and demand (tourist side); domestic and
international travellers; spending by broad category
 Input-Output System
• Household versus business spending; detailed spending by category ...
converted into about 150 commodities; provincial distribution
4
Statistics Canada • Statistique Canada
2016-03-26
Frontier counts
 Volumes of Canadian and foreign travellers are
estimated through administrative data (frontier counts)
 Air travellers
•
•
•
•
•
Customs declaration card
Residency of the traveller
Length of stay
Some information about the destination (for Canadian travellers)
Large sample of the cards are processed for statistics
 Land travellers
• Counts of people crossing the borders
• Split between same-day and overnight trips
5
Statistics Canada • Statistique Canada
2016-03-26
Traveller surveys
 Questionnaires distributed at borders and in airports
 Five versions of the survey
•
•
•
•
Canadians returning from abroad (long questionnaire)
United States travellers (long questionnaire)
Foreigners from countries other than U.S. (long questionnaire)
Canadian returning from United States on same-day trip by automobile
(short questionnaire)
• United States travellers returning from Canada on same-day trip by
automobile (short questionnaire)
 Key questions
• Country of residence; purpose of trip; length of trip (covers commuters)
 Air Exit Survey of Overseas Visitors
• Interviews conducted directly in airports while passengers wait for flights
(target flights to specific destinations)
6
Statistics Canada • Statistique Canada
2016-03-26
Travel expenditures
 Questions on spending
• Broken down by 5 categories of spending
• Share of spending paid by sector (personal-business split)
• Countries visited
 Characteristics of the respondents combined with
characteristics of the “population” (frontier counts)
• Representation of travellers by ports of exit
• Travel spending published quarterly for 6 areas
• Further details on annual basis
7
Statistics Canada • Statistique Canada
2016-03-26
Transportation: Passenger fares
 Questions on fares
•
•
•
•
Fares paid
Name of the carrier
Type of fare
Origin / destination
 Characteristics of the respondents combined with
characteristics of the “population” (frontier counts)
• Passenger fares estimated on a quarterly basis but integrated
under total transport services for publication purposes
• Some details available annually
8
Statistics Canada • Statistique Canada
2016-03-26
Alternatives: Improving current surveys
 Any of the alternatives discussed would need to
retain frontier counts
 Consolidation of the questionnaires from 5 to 2 versions
• Will simplify the distribution
• Reduce printing costs
 Increase the distribution of questionnaires at certain ports
of exit
 Air exit survey for United States travellers
 Review imputation strategies
 Internet version of the questionnaires (distribute cards
with link)... As a means ↓ response burden and ↑
response rates
9
Statistics Canada • Statistique Canada
2016-03-26
Alternatives: Airlines statistics
 Survey of air passenger origin and destination (quarterly)
• Number of passengers, by airlines (all having activities in Canada), by
Canadian and foreign origin-destinations
• Data compiled for air scheduled and air charter services
• No breakdown between Canadian and foreign passenger
 Not seen yet as an alternative to traveller questionnaires
• No country of residence for the travellers
• Does not capture the full journey
 Fare basis survey (quarterly)
• Collects sample of ticket fares from major Canadian air carriers
• Average fares for different type of fare
• Domestic and international journeys
 Could be used to validate questionnaire results
• Although no information on foreign airline fares
10
Statistics Canada • Statistique Canada
2016-03-26
Alternatives: Airlines admin data
 Leverage any available airlines administrative data
• Likely that high frequency airlines administrative data is available for
internal purposes ... Marketing and route planning purposes as demand
evolves
 Possible detail
•
•
•
•
•
Origin and destination of travellers
Nationality
Personal travel or business travel
Domestic and international journeys
Duration of trips
 Likely only domestic carriers and affiliates
11
Statistics Canada • Statistique Canada
2016-03-26
Alternatives: Household surveys ...
And modelling
 Adding questions on international travel through current
Statistics Canada’s surveys
•
•
•
•
•
Survey of Household Spending
Possibly supplementary module on higher frequency surveys
Would only cover Canadian travel abroad
Lack of spending details
Good control total for non-business spending ... acceptable to
estimate average spending
 Modelling data
• Based on counts, exogenous variables and established patterns
• Likely used in combination with other alternatives
• Challenge: The need to generate sufficient detail for the national
accounts programs
12
Statistics Canada • Statistique Canada
2016-03-26
Alternatives: Credit card information
 Partial aggregated monthly statistics available ... Data are
timely ≤ 15 business days after the end of the reference period
 Used as an indicator and data confrontation tool for Household final
consumption expenditures (GDP)
 Class of card holder (personal / commercial) ... With 6
classes of detail:
• Personal credit card purchases made in Canada by Canadian residents
• Commercial credit card purchases made in Canada by Canadian
residents
• Personal credit card purchases made in Canada by non-residents
• Commercial credit card purchases made in Canada by non-residents;
• Personal credit card purchases made outside Canada by Canadian
residents;
• Commercial credit card purchases made outside of Canada by
Canadian residents.
13
Statistics Canada • Statistique Canada
2016-03-26
Credit card information (cont`d)
 Transactions are grouped for about 40 different merchant
classes such as: Accommodation, Airline, Car rental, Full-service
restaurants, Other retail, Travel agencies, etc.
 Challenges:
•
•
•
•
•
Not all credit suppliers included ... Key as market shares evolve
No formal agreement in place to secure data on an ongoing basis
Residency of the card holder limited
Location of the transaction (only in Canada or abroad)
Transaction detail not perfect: More industry than commodity ...
Possibly not standardized across all credit suppliers
• Commercial-personal use determined by type of card
• Difficult to estimate spending per trip (even with # of transactions
and/or # of trips per cardholder)
• Difficulty comparing estimates to survey results
14
Statistics Canada • Statistique Canada
2016-03-26
Credit card information (cont`d)
 Challenges (cont`d)
• Purchases at Canadian or foreign business represent
transactions while travelling, by phone or by internet → more
than travel spending
• Capital purchases (e.g., deposit on timeshares) may be included
• Timing may be an issue, as transactions reflect the date of the
purchase ... Not necessarily the date of travel (e.g., air tickets)
• No way to identify if the purchase was done from home or abroad
• Purchases from Canadian travel agencies shown as purchases
made in Canada, but part may be for travelling outside the
country
• No information about exact country of purchase
• No information about country of origin of foreign travellers
• Credit cards do not cover all travel transactions
15
Statistics Canada • Statistique Canada
2016-03-26
Gaps ... debit card information
 Debit card transactions (withdrawals while travelling)
likely account for a not insignificant share of travel
spending
 Further, debit cards maybe used differently than credit
cards while travelling
• credit cards for accommodation and other expenditures
• debit card withdrawals for a percentage of meals, leisure
activities and shopping
 Further, debit cards maybe used differently than credit
cards while travelling
 Implications for expenditure allocations by commodity
16
Statistics Canada • Statistique Canada
2016-03-26
Credit card information (cont`d)
 Going forward implies a large amount of work and data
development
• Determine if additional country detail could be made available ...
Country of residency for the cardholder and merchant
• Determine if we can obtain the Canadian region of the merchant
and cardholder
• Establish a better understanding of credit suppliers merchant
classifications in relation to standard industry-commodity
classifications
• Determine if there is a flag/indicator identifying (i) internet, (ii)
telephone and (iii) merchant location
• Establish an approach to estimate average expenditure, per
cardholder, per trip
• Extend efforts to include debit cards
17
Statistics Canada • Statistique Canada
2016-03-26
Conclusion ... Questions, comments
 Statistics on international travel are important not only for
balance of payments / trade purposes
 Canada’s estimates are derived from a combination of
travellers surveys and travellers counts
 Alternative sources require be further studied to validate
or replace current sources
 More work on credit-debit cards, in particular
 Questions ... Comments?
18
Statistics Canada • Statistique Canada
2016-03-26