Australian National Accounts: Tourism Satellite Account methodology

1 This release presents information on tourism activity in Australia through a Tourism Satellite Account (TSA). A TSA is a standard statistical framework and the main tool for the economic measurement of tourism. The Tourism Satellite Account: Recommended Methodological Framework 2008 (TSA: RMF 2008) provides the common conceptual framework for constructing a TSA. It adopts the basic system of concepts, classifications, definitions, tables and aggregates of the System of National Accounts 2008 (SNA 2008), the international standard for a systematic summary of national economic activity, from a functional perspective.

2 Implicitly, tourism is included in the core national accounts. The products purchased by visitors, and produced by suppliers, are all part of the economic activity measured in the national accounts. However, while all the products that are produced and consumed in meeting tourism demand are embedded in the core accounts, they are not readily apparent because 'tourism' is not identified as an industry or product in international statistical standards. In the Australian and New Zealand Standard Industrial Classification (ANZSIC), 2006 edition, underlying the Australian national accounts, industries are defined on the basis of the goods and services which they mainly produce. On the other hand, the tourism industry is defined according to the status of the consumer. That is, it is the characteristics of the consumer that determine whether the production is included within the scope of tourism.

3 A TSA provides a means by which the economic aspects of tourism can be drawn out and analysed separately using the structure of the main accounts. In fact, one of the major features of a TSA is that it is set within the context of the whole economy, so that tourism's contribution to major national accounting aggregates can be determined, and can be compared with other industries.

4 The estimates of tourism consumption, direct tourism output, value added and GDP presented in this release are recorded at current prices, that is the prices prevailing in the period to which the observation relates. Estimates for tourism consumption, tourism value added and GDP are also presented in chain volume terms and in index form. Changes in current price estimates over time are a result of changes in prices as well as changes in the underlying level of economic activity (or volume).

5 Chain volume estimates show changes in pure volume (or quantity) terms with price effects removed. They enable the comparison of inflation adjusted growth between the tourism industry and other industries. For information on the introduction of chain volume estimates for tourism see the information paper Tourism Satellite Account: Introduction of Experimental Chain Volume Estimates, 2012-13 (cat. no. 5249.0.55.004). The estimates are derived using methodology consistent with ABS National Accounts. This enables comparison with chain volume estimates published in Australian System of National Accounts (cat. no. 5204.0).

6 The concepts and methods used in the Australian TSA are based on international standards developed by an Inter-Secretariat Working Group on Tourism Statistics comprised of the Organisation for Economic Co-operation and Development (OECD), the statistical arm of the Commission of the European Communities (Eurostat), the United Nations Statistics Division (UNSD), and the United Nations World Tourism Organisation (UNWTO). These standards are presented in the TSA: RMF 2008. National statistical offices such as the ABS were consulted during the development and updating of the international standards. The standards are based on the SNA 2008, which is the prevailing international standard for national accounts statistics.

7 The OECD also released a publication in 2000, "Measuring the Role of Tourism in OECD Economies". This publication is consistent with the international standards. However, it also includes the modelling of tourism generated employment as an extension to the TSA. The OECD manual also shows practical examples of how to compile tourism gross value added.

8 Key tourism supply measures such as direct tourism output, direct tourism gross value added, direct tourism GDP and direct tourism employment in this release relate to the direct impact of tourism only. A direct impact occurs where there is a direct relationship (physical and economic) between the visitor and producer of a good or services.

9 In the case of goods purchased by visitors, only the retail margin contributes to key tourism supply measures. This is because it is deemed that only the retailer has a direct relationship with the visitor and is therefore part of the tourism industry. The implication of this treatment is that the value added generated in the chain of supply of goods to visitors up to, but not including, the retail level will be treated as an 'indirect effect' of tourism consumption, while only the value added generated from retail trade activities provided to visitors will be considered as a direct effect.

10 Indirect effects of tourism consumption are out of scope for the Australian TSA. They are a broader notion of tourism consumption that includes downstream effects of tourism demand. For example, when a visitor buys a meal, indirect effects are generated for the food manufacturer, the transporter, the electricity company, etc., that provide the necessary inputs required to make the meal. To fully measure the indirect effects, account should also be taken of changes in incomes which may feed through to further changes in tourism demand. A full analysis of indirect effects is best done using economic modelling. Tourism Research Australia (TRA) have undertaken this work and their latest results are reported in the "Tourism's Contribution to the Australian Economy 1997-1998 to 2012-13".

11 The term 'tourism' in the international standards is not restricted to leisure activity. It also includes travel for business or other reasons, such as education, provided the destination is outside the person's usual environment.

12 An important conceptual distinction concerns the difference between travel and tourism, and consequently between a traveller and a visitor. Travel is a broad concept which encompasses the activity of travellers and includes commuting to a place of work, migration and travel for business or leisure. A traveller is defined as:

'..someone who moves between different geographic locations, for any purpose and any duration.' (International Recommendations for Tourism Statistics 2008, para 2.4).

13 The central statistical entity in tourism statistics is the 'visitor'. The scope of tourism in the international standards comprises the activity of visitors.

14 A visitor is defined in the international standards as:

'...a traveller taking a trip to a main destination outside his/her usual environment, for less than a year, for any main purpose (business, leisure or other personal purpose) other than to be employed by a resident entity in the country or place visited.' (International Recommendations for Tourism Statistics 2008, para 2.9).

15 If a person stays in the one place for longer than one year, their centre of economic and social interest is deemed to be in that place, so they no longer qualify as a visitor.

16 The following types of persons are not considered to be visitors:

• persons for whom travel is an intrinsic part of their job, e.g., bus driver, air crew
• persons who travel for the purpose of being admitted to or detained in a residential facility such as a hospital, prison or long stay care
• persons who are travelling as part of a move to a new permanent residence
• persons who are undertaking military duties
• persons who are travelling between two parts of their usual environment.

17 Visitors can be classified into national and international visitors. National or 'domestic' visitors consist of Australian residents who travel outside their usual environment within Australia. They include both overnight visitors (staying one or more nights at a location) and same day visitors. International visitors are those persons who travel to a country other than that in which they have their usual residence.

18 The one year rule for length of stay for an international visitor is consistent with the principle applied in determining residency in the balance of payments, which generally requires the length of stay in an economic territory to be less than one year to qualify as a non-resident. There are some categories of individuals though that are exceptions to this one year rule in the balance of payments, in particular international students, who continue to be resident in the territory in which they were resident prior to studying abroad even if their course of study exceeds a year. The approach adopted in the Australian TSA is to include as visitors all international students undertaking short term courses with an actual length of stay of less than one year. If a student stays longer than one year (ignoring short-term interruptions to their stay, for example at vacation break), their usual environment is deemed to be the school or university, and they do not fit the definition of a visitor. However, if they travel outside their usual environment, they are considered a visitor.

19 For the purposes of measuring direct tourism gross value added and direct tourism GDP in the TSA, the consumption of Australian residents travelling overseas (outbound visitors) is excluded, except to the extent they consume domestically produced products before or after their overseas trip. This is because their consumption overseas does not relate to the value of goods and services produced within the Australian economy.

20 The usual environment is made up of one or more areas in which a person undertakes their regular activities such as their place of residence, place of work, place of study and other places frequently visited. The usual environment criterion has two dimensions; frequency and distance. Places that are visited on a routine basis (at least once a week) are considered part of a person's usual environment, even if the place visited is located a considerable distance from their place of residence. Further, locations up to 40 kilometres from home for overnight trips and up to 50 kilometres from home (round trip) for day trips are included in a visitor's usual environment in the Australian TSA.

21 Although a considerable amount of tourism spending may take place within the usual environment (i.e. purchase of air tickets, tour packages, pre-paid accommodation), the consumption of most tourism services occurs outside of the usual environment. Visitors have a positive economic impact on their destination by generating additional consumption at the destination over and above that generated by the resident consumers.

22 Tourism expenditure covers actual expenditure by the visitor, or on behalf of the visitor, and is defined in the international standards as:

'...the amount paid for the acquisition of consumption goods and services, as well as valuables, for own use or to give away, for and during tourism trips. It includes expenditures by visitors themselves, as well as expenses that are paid for or reimbursed by others.' (International Recommendations for Tourism Statistics 2008, para 4.2).

23 Consistent with the definition of visitors, tourism expenditure includes expenditure by visitors whose primary purpose is business, whether this is totally or partially paid by businesses or government. Expenditure before or after the trip is likewise included provided the expenditures are related to the trip, such as the digital printing of photos after they return home or the purchase of tourism single-purpose consumer durables in anticipation of trips (e.g. luggage or camping equipment). Consumer durables which are purchased on a trip (e.g. motor vehicles) are also included, even though they are not mainly for use on trips. Since there is no reliable data source in Australia for acquisitions of valuables, this item is not included in the national accounts and is also excluded from tourism expenditure in the Australian TSA.

24 Expenditures by Australians travelling abroad on goods or services provided on Australian domestic territory, either before or after the trip, are included in tourism expenditure. In addition, outbound and inbound services provided by Australian international carriers are also included. All other expenditures by Australians while abroad are excluded from tourism expenditure, consumption and value added. These expenditures are classified as imports of goods and services.

25 Tourism consumption, as defined in the Australian TSA and the revised internationals standards, is broader in scope and also includes imputations for the consumption by visitors of certain services for which they do not make a payment. Imputed consumption included in the Australian TSA include:

• Services provided by one household to the visiting members of another household free of charge, including the value of goods such as food and purchased services provided by host family/friends
• Housing services provided by vacation homes on own account (imputed services of holiday homes deemed to be consumed by their visitor owners)
• Imputed values of non-market services provided directly to visitors such as public museums even though these may be provided free or at a price which is not economically significant.

26 In the Australian TSA, tourism consumption includes consumption by both domestic and international visitors. Domestic consumption is further split into that of households and that of business and government. This dissection is based on who incurred the expenditure rather than who actually paid. For example, a visitor may spend $1,000 on a business trip, of which $800 is reimbursed by the business. While the primary purpose of the trip is business, $800 is recorded as business consumption, with the remaining $200 reported as household consumption.

27 Some further information on particular inclusions and exclusions from tourism consumption are discussed in the 1997–98 issue of this release.

28 Direct tourism gross value added and direct tourism GDP are the major economic aggregates derived in the TSA. The concepts are not the same and it is important to outline the differences between them.

29 Direct tourism gross value added shows only the 'value' which a producer adds to the raw material goods and services it purchases in the process of producing its own output. Direct tourism gross value added is measured as the value of the output of tourism products by industries in a direct relationship with visitors less the value of the inputs used in producing these tourism products. Output is measured at 'basic prices', that is before any taxes on tourism products are added (or any subsidies on tourism products are deducted). Taxes on tourism products include the GST, wholesale sales taxes and excise duties on goods supplied to visitors. Direct tourism gross value added is directly comparable with estimates of the gross value added of 'conventional' industries such as mining and manufacturing that are presented in the national accounts.

30 SNA 2008 states that basic price measures are to be used for comparisons between industries and across countries because it is free of the effects of taxes and subsidies on products which can vary between industries (and countries) and over time. The tax and subsidy component of a product's sale price does not represent value added by the industry producing that product.

31 Direct tourism GDP, on the other hand, measures the value added of the tourism industry at purchasers' (market) prices. It therefore includes taxes paid less subsidies associated with the productive activity attributable to tourism. Direct tourism GDP will generally have a higher value than direct tourism value added. Direct tourism GDP is a satellite account construct to enable a direct comparison with the most widely recognised national accounting aggregate, GDP. While direct tourism GDP is useful in this context, the direct tourism gross value added measure should be used when making comparisons with other industries or between countries.

32 Purchases of capital assets are excluded from tourism demand for the purposes of calculating direct tourism gross value added, as there is no direct relationship between the visitor and the acquisition of capital by the tourism industries.

33 The services that capital investment provide are captured to the extent that the price change for products implicitly include a component to cover the cost of capital. Whilst expenditure on gross fixed capital by tourism industries is of significant analytical interest, collection of industry data presents a number of difficulties. One example is assets obtained under finance leasing arrangements need to be recorded in the industry of the lessee rather than the legal owner. Due to these conceptual difficulties, estimates of tourism gross fixed capital formation are not produced as part of the Australian TSA.

34 Not all products and industries in the standard national accounts product and industry classifications are related to tourism. Therefore, the TSA distinguishes between products and industries that are related to tourism, and those which are not. Tourism related products and industries are further classified into tourism characteristic and tourism connected. There are three categories of industry and product in the TSA, as outlined below.

35 Tourism characteristic products are defined as those products which would cease to exist in meaningful quantity, or for which sales would be significantly reduced, in the absence of tourism. Under the international TSA standards, core lists of tourism characteristic products, based on the significance of their link to tourism in the worldwide context, are recommended for implementation to facilitate international comparison. The core list of tourism characteristic products is consistent with the international classification of products, namely the Central Product Classification, Version 2.1 (CPC V2.1).

36 It is also recommended in the international TSA standards that country-specific tourism characteristic products are identified. In the Australian TSA, for a product to be a country-specific tourism characteristic product, at least 25 per cent of the total output of the product must be consumed by visitors.

37 Tourism connected products are those that are consumed by visitors but are not considered as tourism characteristic products. All products in the Supply and Use table not consumed by visitors are classified as 'all other goods and services' in the TSA.

38 See Appendix - tourism product correspondence for a correspondence between tourism related products in the Australian TSA, and products included in the CPC V2.1.

39 Tourism characteristic industries are defined as those industries that would either cease to exist in their present form, or would be significantly affected if tourism were to cease. Under the international TSA standards, core lists of tourism characteristic industries, based on the significance of their link to tourism in the worldwide context, are recommended for implementation to facilitate international comparison. The core list of tourism characteristic industries is consistent with the newly revised international classification of industries, namely the International Standard Industrial Classification, Revision 4 (ISIC Rev. 4), which aligns closely with ANZSIC 2006.

40 In the Australian TSA, for an industry to be a country-specific tourism characteristic industry, at least 25 per cent of its output must be consumed by visitors. Whether or not an industry is classified as characteristic has no effect on total value added resulting from tourism, as the TSA measures the gross value added resulting from the production of products directly consumed by visitors, not the total gross value added generated by tourism related industries.

41 Tourism connected industries are those, other than tourism characteristic industries, for which a tourism related product is directly identifiable (primary) to it, and where the products are consumed by visitors in volumes which are significant for the visitor and/or the producer. All other industries are classified as 'all other industries', though some of their products may be consumed by visitors and are included in the calculation of direct tourism gross value added and direct tourism GDP.

42 The following points are worth noting about the industry classifications for the TSA:

• whether or not an industry is classified as characteristic has no effect on total value added resulting from tourism
• for an industry to be tourism related it must serve the visitors themselves, that is, there must be a direct relationship between the provider of the product and the consumer. Producers of goods from the manufacturing industry that are not in direct contact with visitors will therefore not be included as a tourism related industry.

43 See Appendix - tourism industry correspondence for a correspondence between tourism related industries and industries included in the ANZSIC.

44 The "Supply and Use" tables for the Australian economy provide the framework in which data for visitor expenditure (demand) and industry output (supply) are integrated and made consistent in the TSA benchmark. Moreover, they provide the means of calculating direct tourism gross value added and direct tourism GDP. The input-output table variant provides a tool for further analysis and economic modelling of tourism.

45 The 'supply' table is a matrix showing (in the rows) the basic price values of products produced by each major industry. It also shows the supply of products from imports, and the net taxes on products and trade and transport margins required to derive supply at purchasers' prices. The 'use' table shows the use of each product, both as intermediate consumption by industries and in domestic final demand and exports. The use table also shows the primary inputs (compensation of employees and gross operating surplus) required by each industry.

46 The Supply and Use tables are brought to balance so that the supply of each product equals its use. The Supply and Use tables on which the TSA is based contain 67 industries and 301 products. To derive the TSA, it was necessary to augment the standard Supply and Use tables. As the objective of the TSA is to focus on tourism related products and the industries that produce them, some disaggregation of the products and industries shown in the standard tables was required. For operational convenience in constructing the TSA, the non-tourism products and industries were compressed, but the details still remain in the underlying Supply and Use tables.

47 An important characteristic of tourism products is that they are not uniquely defined by their nature, but by who purchases them. Therefore, the consumption of each product has to be divided into that part consumed by visitors and that part consumed by non-visitors. This information is used to partition industries into their tourism and non-tourism components, enabling the derivation of direct tourism value added and direct tourism GDP.

48 An important part of the compilation process is to check the consistency of data for visitor expenditures on products with the total supply of products. Apparent inconsistencies have to be resolved by further data investigations and adjustment.

49 More details on the Supply and Use approach to constructing a TSA can be found in the OECD's "Measuring the Role of Tourism in OECD Economies". Readers who require more detailed information on Supply and Use tables more generally should consult SNA 2008.

50 The gross value added for an industry is derived as the gross output for that industry less the intermediate consumption required to produce that output. However, as the tourism industry is defined according to who purchases an industry's output rather than according to the nature of the output itself, tourism consumption is required in order to measure the tourism output of industries. For the current TSA time series, benchmark years are 2006–07, 2009-10, 2012-13 and 2016-17.

51 To calculate direct tourism gross value added, a number of steps are required. These can be summarised as:

• identify which products in the economy are purchased by visitors
• derive an estimate of internal tourism consumption for each tourism product
• remove product taxes and subsidies, margins and imports from internal tourism consumption of each product at purchasers' prices to derive internal tourism consumption at basic prices—this represents the domestic output consumed by visitors
• determine what proportion of the domestic output of each product is consumed by visitors by dividing tourism consumption at basic prices into the total supply of each product at basic prices—this is the tourism product ratio
• for each tourism product, identify the industries that have a direct relationship with visitors. In the case of the supply of goods, this will only include the industries that provide retail trade services. In the case of the supply of other margins (wholesale and transport margins), all industries are deemed to not have a direct relationship with visitors
• apply the tourism product ratio to the output of each product for those industries that have a direct relationship with visitors to derive the direct tourism output of each industry
• estimate the intermediate consumption required to produce each industry's output of tourism products using relationships in the Supply and Use tables. The default assumption is that there is a constant ratio of output to intermediate consumption for both tourism and non-tourism products produced by an industry
• calculate direct tourism gross value added at basic prices for each industry as direct tourism output less the intermediate consumption required to produce the direct tourism output, and sum for all industries in the economy. Direct tourism gross value added at basic prices is directly comparable to the value added for all other industries.
   

52 It is not feasible to collect the detailed supply side data required to produce a timely full scale TSA every year. However, the key aggregates can be updated annually using relationships in the benchmark TSA and demand side data that are available on a yearly basis.

53 The following steps are used in the update years:

• derive an estimate of internal tourism consumption for each tourism product
• remove product taxes and subsidies, margins and imports from internal tourism consumption of each product at purchasers' prices to derive internal tourism consumption at basic prices—this represents the domestic output consumed by visitors
• remove supply of goods by producers that do not have a direct relationship with visitors (equal to the cost to retailers of domestic goods sold directly to visitors) using ratios from the benchmark TSA to derive direct tourism output
• allocate the direct tourism output of each tourism product to producing industry using ratios from the benchmark TSA
• sum the products produced by each tourism industry to derive the output for each tourism industry
• split industry output between value added and intermediate consumption using each industry's input-output ratios taken from the benchmark TSA.
• sum direct tourism gross value added for all industries to calculate the tourism industry's gross value added. Direct tourism GDP is derived by adding tourism net taxes on products.

54 Where there is structural change in tourism related industries or the economy more generally in the update years, it is likely that there will be revisions when the next benchmark is compiled. For example where there is significant change in input costs over time (such as fuel), then the ratio of output to intermediate consumption established at the benchmark may no longer be valid. See paragraph 90 for more detail on this issue.

55 Direct Tourism Gross Value Added -Tourism gross value added chain volume estimates are derived from tourism output minus tourism intermediate use in prices of the previous period. A volume index is then compiled and this index is chained taking account of the price relativities from year to year. The index is then converted to chain volume values by using the current price index ratio from the reference year.

56 Direct Tourism output – whilst tourism output is not being published, it needs to be derived in volume terms to calculate tourism gross value added. Tourism output product level deflators are a mix of consumer price indexes (CPI) and producer price indexes (PPI). These have been used to derive the prices in the previous year for each tourism sub-industry and total tourism output to derive tourism industry output volume index(es). This is then chained (linked) together to derive the chain volume index which is applied to the current price index ratio in order to arrive at chain volume estimates. The indexes are re-referenced to equal 100 in the reference year. In a small number of instances where the chain volume estimates of tourism output and intermediate use (for a particular industry) are leading to unexpected gross value added volume movements, adjustments are made as required. These adjustments may be either to the tourism output or intermediate use chain volume estimates to ensure the gross value added estimates are more consistent with real world observations.

57 Direct Tourism intermediate use – chain volume estimates of tourism intermediate use (by tourism sub-industry) are estimated by deflating the corresponding tourism current price estimates. These estimates are deflated using industry level implicit price deflators (IPDs) derived from the ANA’s unpublished Supply and Use tables (for each tourism related industry). Implicit price deflators are used rather than a broader measure such as All groups CPI, as they are more reflective of the costs associated with producing the relevant industry's output. As noted in paragraph 56 above, adjustments are made as required to selected tourism industries' chain volume output or intermediate use estimates to ensure the gross value added estimates are more consistent with real world observations.

58 Direct Tourism Gross Domestic Product (GDP) - to arrive at tourism GDP, tourism GVA in prices of the previous period is added to tourism net taxes on products in prices of the previous period. A volume index is then compiled and the index converted to chain volume values by using the current price ratio from the reference year.

59 Direct Tourism Consumption by Product - tourism consumption is deflated at the tourism product level, for example, accommodation services and taxi fares. The products are deflated using a mix of CPIs (or producer price indexes where relevant CPIs are not available) and the implicit price deflator for dwelling rent. Volume indexes are derived, chained and converted to chain volume values.

60 Tourism Implicit Price Deflator - the tourism implicit price deflator, or tourism price index, is calculated as the ratio of the current price value of tourism GVA (or Tourism GDP) to its corresponding chain volume value, multiplied by 100. The advantage of using an implicit price deflator for measuring price change for the tourism industry over a general measure of price change such as the CPI, is that the implicit price deflator accounts for compositional shifts in the basket of products consumed (or produced) from year to year, unlike the fixed weighted CPI. Users should be aware that implicit price deflators are revised annually with re-referencing and therefore are not recommended for purposes such as indexing in legal contracts.

61 The main data sources used to compile the benchmark and the update TSA are described in this section. The 1997–98 issue of this release provides additional information on the data sources used to compile a benchmark TSA.

62 The data sources for actual expenditures are the same for both the benchmark and update years.

63 Most of the visitor expenditure data used in the compilation of the TSA is sourced from TRA in the National Visitor Survey (NVS) for expenditure by Australian visitors and the International Visitor Survey (IVS) for expenditure by international visitors. These data are supplemented with data from the balance of payments and national accounts, e.g. expenditure on gambling and betting services and airfares.

64 The international visitor consumption total is consistent with the sum of the balance of payments' passenger transportation and travel services items adjusted for conceptual differences between the TSA and balance of payments. The conceptual differences relate to the restricted coverage in the TSA of student visitors studying in Australia; imputations for non-market services provided to overseas visitors; margins on foreign exchange transactions; and the value of products provided to overseas visitors within private households. IVS data provide the detailed product information.

65 Adjustments to the TRA data are required to break down the following broad expenditure categories for both domestic and international visitors: organised tours, package tours, conference fees, and trips with more than 21 stopovers. Given a lack of suitable direct information, these allocations are made on the basis of models or supplementary information from either ABS or TRA surveys.

66 The major imputed component in tourism consumption relates to the imputed value in 'actual and imputed rent on dwellings', where the imputed value of rent refers to the consumption of housing services provided by holiday houses to the owner. The estimate is calculated by multiplying average annual rents by the total number of holiday houses in Australia from the Census of Population and Housing.

67 An imputation for products provided to visitors in private households is derived using indicators from the ABS Household Expenditure Survey and IVS/NVS data which are moved forward using household final consumption expenditure from the National Accounts. These methods are used for both the benchmark and update estimates.

68 The imputation for non-market services provided by government utilises data on visitor numbers from Attendance at Selected Cultural Venues and Events, Australia (cat. no. 4114.0) publications. Initial estimates of foreign exchange margins were derived using international visitors' expenditure data and data for currency buy/sell rates, with estimates for later years extrapolated using movements in the number of short term arrivals.

69 Imputed estimates of tourist information centre consumption and output are sourced from unpublished Supply and Use data. It is assumed that the economy-wide consumption of tourist information centre services (provided by the public sector to consumers on a non-market basis) is mostly attributable to tourists; non-tourism exclusions include activities such as visits by students for school projects.

70 Internal tourism consumption at basic prices is calculated by removing the net taxes, margins and imports from internal tourism consumption at purchasers' prices. This is done using relationships in the Supply and Use tables. Generally a 10% GST applies to all expenditure by domestic households and international visitors, excluding expenditure on education, health, international airfares, and groceries that are GST exempt. In the update years other product taxes, subsidies, imports and margins are calculated using a variety of sources including government budget reports, balance of payments data and retail trade data.

71 Memorandum items for credits (exports) and debits (imports) of tourism related services are included in Balance of Payments and International Investment Position, Australia (cat. no. 5302.0). These are derived by combining total travel services (business, education-related and other personal travel), passenger fares, and the air transport component of other transportation services which relates to agency fees and commission receipts.

72 In this release information on Tourism Consumption by Australian Residents on Outbound Trips is presented. This table records the value of goods and services sourced from non-residents purchased by Australian residents in association with these trips. The difference in the estimates in this release and the service debits memorandum item published in the balance of payments relate to: any transactions between residents and non-residents which do not relate to the value of tourism goods and services purchased within Australian economic territory; and expenditure by student travellers on long term visas (as they do not meet the definition of a visitor).

73 Balance of payments tourism related services credits are closely related to exports of tourism goods and services in the estimates of the value of domestically produced goods and services presented in the Tourism Consumption by Non-residents on Inbound Trips table in this release. The most significant differences occur because the Australian TSA excludes the expenditure of overseas students with a length of stay of greater than one year and non-resident to resident transactions which occur in other countries, i.e. delivery of services by Australian residents in other countries, both of which are included in the balance of payments. The other differences relate to the TSA imputations for non-market services provided to overseas visitors, margins on foreign exchange transactions and the value of products provided to overseas visitors within private households, these imputations are generally not recorded in the balance of payments.

74 While the Supply and Use tables provide 'control totals' for industry output of products and the inputs required to produce those products, in the benchmark TSA years of 2006–07, 2009-10, 2012-13 and 2016-17 the data have to be disaggregated and rearranged to focus on tourism related products identified from tourism demand data. This has mostly been done using information from the ABS annual business surveys.

75 The ABS annual business surveys collect data for business income and expense items for all broad industry groups in the economy. These surveys were expanded in the 2006–07, 2009-10, 2012-13 and 2015-16 (extrapolated for TSA 2016-17 benchmark year) reference years to provide extra data for tourism characteristic industries. The ABS Economic Activity Survey (EAS), supplemented with taxation data, is the major source of data for the following industries: transport, automotive fuel, motor vehicle hire, travel agents, accommodation, cafes and restaurants, pubs and taverns, clubs, gambling industries and casinos.

76 As the TSA supply table is not updated for the extrapolated years, benchmark coefficients are carried through to the update estimates. The current benchmark reference years (with associated benchmark coefficients) used in the TSA are 2016-17, 2012-13, 2009-10, 2006-07, the latter applying to all periods in the series back to 2004-05 (see paragraphs 89-90 below for where exceptions apply). The 2016-17 coefficients are carried through to 2017-18 in this release, and will continue to be used until the derivation of a new benchmark.

77 Persons employed in tourism related industries will generally provide services to both visitors and non-visitors. Tourism employment is derived for each industry by applying the tourism value added industry ratios from each of the benchmark years to employment estimates for each industry in subsequent years. As part of the 2016-17 benchmark, new industry value added ratios were compiled and the ratios for 2006-07, 2009-10, and 2012-13 were reviewed and adjusted. Due to the significance of revisions to the supply side data, the industry value added ratios from 2006-07 have been applied to all periods in the series back to 2004-05. This method of using the tourism value added industry ratios involves an assumption that the employment generated by tourism in each industry is in direct proportion to value added generated by tourism in the benchmark year.

78 Details by industry of employment, status in employment and gender, are collected in the Labour Force, Australia, Detailed, Quarterly (LFS) (cat. no. 6291.0.55.003) in the February, May, August and November months. Estimates of the number of total, full-time and part-time, and male and female tourism employed persons by industry have been calculated as the average of these four months. Total employment is derived by adding employment in the defence forces to the civilian labour force. Users should note that there may be some downward bias in employment estimates for the tourism industries due to the timing of the collection, i.e., the LFS survey months exclude the major Christmas holiday period.

79 Tourism employment and hours worked estimates are based on original, unadjusted Labour Force data. Detailed industry level data from the Labour Force Survey are required to produce corresponding tourism industry estimates. These data are available as an original series only and are therefore subject to more variability than trend and seasonally adjusted series. Improved seasonal analysis factors have been used in all periods to reduce the variability in the tourism industry estimates.

80 Part-time, full-time, male and female employment ratios are derived for each tourism characteristic and connected industry using LFS ratios and applying these ratios to the total number of tourism employed persons by industry. The underlying assumption is that the LFS distribution of part-time, full-time, male and female employment from within a given industry is consistent with the distribution of part-time, full-time, male and female employment within the tourism share of that particular industry.

81 Hours worked by labour is a key input to the measurement of labour productivity in industry analysis. Indexes of hours worked are preferred to employment numbers as a measurement of labour input as hours worked captures changes in overtime, standard weekly hours, leave, and part-time work. The index of tourism hours worked presents a tourism view of labour input conceptually equivalent to the ANZSIC Division level labour input index presented in the national accounts, and, as such, it allows comparison of labour input growth in tourism with traditional industries. However, the method for producing the tourism index differs from that applied in the national accounts, where the index is constructed from industry estimates of aggregate annual hours worked in the reference year. As aggregate annual hours worked is unavailable at the detailed industry level necessary for application to tourism, the tourism index is instead based on annual averaging of quarterly detailed estimates of actual hours worked. This approach has been found to produce estimates of sufficient quality for the analysis of movement in a data series such as that described by an index and also results comparable with the National Accounts hours worked indexes.

82 Labour productivity estimates in the National Accounts are indexes of real GDP per hour worked. For the whole economy, they are derived by dividing the chain volume measure of real (or chain volume) GDP by hours worked. For the "tourism industry", they have been derived by dividing the chain volume measure of tourism GVA by an index of tourism hours worked and are shown in index form. This is consistent with ANZSIC Division level labour productivity estimates in the National Accounts (although the numerator for the ANZSIC industries is an index of aggregate, rather than actual, hours worked for reasons outlined above). Labour productivity indexes reflect not only the contribution of labour to changes in product per labour unit, but are also influenced by the contribution of capital and other factors affecting production.

83 Some of the tourism industries in the TSA have been compressed in the tables relating to employment because the LFS is not designed to produce estimates of sufficient accuracy for some of the fine level industries in the TSA.

84 Domestic visitor numbers are sourced from the NVS.

85 Data on international arrivals and departures (by country of origin or destination) were taken from the ABS release Overseas Arrivals and Departures, Australia (cat. no. 3401.0).

86 While as much care as possible has been taken to ensure the quality of the estimates in the TSA, users should exercise some caution in the use and interpretation of the results. In order to produce estimates at a finer level of product and industry detail than that normally provided in the national accounts, some of the data have had to be stretched up to the limits of their design capabilities. Moreover, major tourism aggregates such as direct tourism gross value added and tourism employment are not directly observable in practice. They have to be modelled in a Supply and Use framework. The assumptions underlying the estimates can have an effect on their quality.

87 The estimates have been prepared from a wide range of statistical sources. Some are closely related to the desired national accounting basis, but others are not. There are differences in coverage, concepts and timing. Many of the tourism industries and products identified in this release are at a more detailed level, or do not directly concord with the industry and product details in the national Supply and Use tables. While every effort has been made to improve the survey coverage of the finer level tourism industries, the accuracy of these estimates are subject to a higher degree of error than that generally pertaining to the broader level estimates published in the national accounts.

88 Tourism expenditure data are generally obtained from large scale visitor surveys and are a key component of the TSA. These surveys are scientifically designed to produce estimates of good quality but, like all sample surveys, are subject to sampling variability. The relevant NVS and IVS publications describe in detail the associated confidence intervals for a given estimate. In order to adapt the visitor survey data to the concepts and classifications required for the TSA, some dissection and rearrangement of the data has been required. While the rearrangement of the basic data can impact on the quality of tourism consumption estimates for individual products, the aggregate level of tourism consumption should not be affected, although the estimate of direct tourism gross value added could be.

89 The most significant assumption in the compilation of a TSA relates to the use of the tourism product ratios and the tourism industry ratios in the calculation of tourism related monetary and employment aggregates. The default assumption is that the input requirements of tourism and non-tourism output are identical for an industry. While this is likely to be a more valid assumption for fine level industries where industry output is relatively homogenous (such as the taxi transport industry), there will be some instances where the assumption may be less valid. This is more likely to be the case where the tourism specialisation ratio of the industry is low, and a diverse range of products are produced. However, errors resulting from the use of assumptions will tend to offset in the calculation of the broad aggregates such as direct tourism value added and direct tourism GDP.

90 In the TSA update years, the allocation of direct tourism gross output to producing industry and the derivation of direct tourism gross value added are based on relationships that are established in the benchmark years. For the years between benchmarks, the coefficients are smoothed between benchmark years.

91 As the benchmark relationships are likely to become less relevant over time, extrapolated estimates only remain tenable where there is a realignment to periodic benchmarks. The underlying assumption of the benchmark approach is that structural change occurs only slowly. The extrapolation techniques used in the TSA updates may not fully capture structural changes in the tourism industry and the Australian economy. To overcome this deficiency, the benchmark coefficients can be improved where there is strong evidence of structural change in tourism related industries. This is however subject to the availability of detailed data and in the case of the Supply and Use tables, the latest Supply and Use data is only available one year after the current reference period.

92 Estimates of the number of persons employed in tourism related industries have been derived from the LFS. As this is a household survey, it has some deficiencies when used to derive detailed industry estimates. In order to mitigate some of these potential quality problems, estimates of employment have been published at a more aggregated industry level than that provided in the first dimension of the TSA.

93 For productivity measurement, while indexes of actual hours worked are considered to be of good quality, they are published as indexes as levels may be subject to reporting bias. That is, there may be a tendency for respondents in the LFS to either overestimate or underestimate their hours worked. Industry levels of hours worked may also be subject to a reporting bias due both to the number of hours reported, as well as self-selecting the industry they work in. Further, hours worked in tourism are derived using benchmark tourism value added ratios, which again are subject to some of the quality issues mentioned above. Indices of actual hours worked are used as it is reasonable to assume that any bias does not change over time and so does not affect the growth rate.

94 Most figures are subject to revision as more complete and accurate information becomes available.