Concepts and Approaches

An economy is an integral body of producers and consumers. They are inter-dependent. What consumers are willing to buy will largely determine the output level of the producing industries. Vice versa, what industries produce and how they produce will have significant impacts not only on consumers but also among the industries themselves. The impacts are positive for some but negative for the others, at the same time. The tasks of economic impact analysis are to identify the channel of such effects.

Early development of impact analyses was based the foundation of the Input-Output (IO) table, using fixed linear relationships among all industries in the economy, often referred to as the IO multipliers. Limitations of the IO multipliers are a few. The most common one is that the feedback of changes in wages and prices is not captured in the whole process. There are no limits on resources in the economy, which is an unrealistic economic framework. Price elasticity and income elasticity effectively play no role in the mechanism. Under such setting, if the impacts are assessed for a very small region of a country, where supplies of labour and capital are deemed to be abundant, results may be acceptable. However, if the assessment is concerning a country or large states of a country, results will be severely compromised, as the IO multiplier technique renders uniform effects, either positively or negatively, across the entire economy depending on the initial stimulus. Neither of these is realistic. 

Any remedy? Yes. The computable general equilibrium (CGE) modelling technique combines micro-economic theories and the IO database to allow for cost minimisation in the production. Industries can source inputs from the cheapest source of supply. At the same time, household consumption bundle is modelled subject to changes in household’s income level and the relative changes of commodity prices. Resources can be relocated between industries depending on the prevailing demand levels. Thus, an increase in an industry can attract and take away resources from other industries, thereby adversely affecting them. 

Among the most well-known CGE models is the ORANI model (Dixon et al. 1987) for the Australian economy. The history of CGE development can be found from here, an enjoyable reading. Over time, the national framework in ORANI has been expanded to multi-states model (MMRF) and further down to The Enormous Regional Model (TERM) with an innovative data structure that requires less bilateral flows of goods and services among regions and yet results are still comprehensive. 

The tourism CGE (Insights) model we develop here is a TERM-stylised model for all states and territories of the Australian economy. Depending on the requirement, the Insights model can be further developed to include specific tourism destinations, if required, as it has been done for the study of the economic impact analysis of the Gold Coast 2018 Commonwealth Games in the past. Tourism is explicitly recognised by inter-state, intra-state, day-trip and the top ten inbound markets, namely China, United Kingdom, New Zealand, United States, Singapore, Korea, Malaysia, India, Japan, Hong Kong and the Rest of the World. The details of tourism markets enable accurate tourism impact analyses, as each market has its own expenditure pattern, changes in the number of visitors in a market will have different effects on the individual regional economies.

The Insights Model is a powerful tool. Its analyses encompass results for macro-economic and sectoral variables for all regions in the model. Macro-economic results include all typical regional aggregate variables such as gross regional product, total household consumption, total investment, total exports, total imports, inflation, changes in employment, capital stocks and movements of the real exchange rate across all regions. Sectoral results include output, exports, employment (by occupation) by industry. The model has proven its robustness and significance for comprehensive tourism policy research that provides deep understanding in insightful policy implications of strategic development projects.

Early version of the model has been used to assess the impacts of the 2018 Gold Coast Commonwealth Games before the games and validate the assessment after the games with actual data; the COVID-19 effect of inbound tourism on the Australian economy. A similar modelling approach was implemented to assess the COVID-19 impacts of tourism-induced poverty in Indonesia. Full reports of these projects can be found below.  

This section gives you a brief description, with essential concepts, of what the tourism satellite account (TSA) is all about. Tourism is not recognised explicitly in the system of national accounts (SNA). This raises difficulties and challenges for policy planning. As such, United Nation World Tourism Organisation (UNWTO) has developed the TSA framework so as to capture and reflect tourism contribution clearly. TSA is basically a way to represent tourism in parallel with the SNA while conforming to all concepts and principles of the national accounts. For this reason, it has the name satellite account. On that basis, the TSA data are comparable with data of all other industries in the economy. 

Essentially, TSA encompasses all demand and supply of tourism in a set of tables. Typically, TSA begins from the demand side, using tourism expenditure surveys to capture tourists’ expenditure across all goods and services related to their travelling, including before and after the trips. Goods and services are then classified into two groups that have either (i) direct or (ii) indirect physical contact with visitors. This is an important concept in the TSA framework (for technical readers, it reflects the principles of the IO tables that exclude inputs in the calculation of industry contribution to avoid double counting). While expenditure on hotels and restaurants is straight forward to illustrate the direct contact between visitors and tourism service providers, it is more complicated when considering the purchase of fuel by self-driving visitors, for example. The cost of fuel itself is part of the indirect group, as visitors do not purchase fuel directly from the fuel producers. Instead, visitors buy fuel from petrol stations, a middleman of the retailer services. Thus, visitors only have direct contact with the petrol stations, not the fuel producers. In general, most manufacturing goods consumed by visitors are treated in this way.

Expenditure on manufacturing goods is divided into two components. The costs of the goods (fuel) are considered as inputs of the tourism sector, thus not counted as tourism contribution. However, the retail component (e.g. petrol station) that facilitates the purchases of the goods are included together with expenditure such as hotels, restaurant and car hire etc... in the direct group. The direct group is deemed as contribution of the tourism sector. Expenditure of the direct group is further processed to derive gross value added, wages, net taxes on products and production, and employment for the tourism sector. 

For readers with some technical background, the dichotomy between direct and indirect in the TSA framework is very different from the direct and indirect concept in the IO framework. Explanation of such differences is provided by Tourism Research Australia in their earlier publication (see here).   

Fundamentally, the TSA process extracts goods and services used by tourism activities directly. This enables a consistent comparison between tourism data with data of other industries in the same country, or with tourism data in other countries. By nature, TSA serves the reporting purpose to reflect the size of the sector and the expenditure patterns of different types of visitors. Therefore, TSA data are important for marketing exercises, for tourism development strategies including preparation of employment and training policies. 

Using TSA alone cannot inform policy makers of any effects from changes to tourism on other sectors in the economy. Such effects are the nature of impact analyses, which require an economic model. Tourism CGE models such as the Insights model are most suitable for the task. In an impact analysis, TSA is used to report the first-round effects, and a CGE model is used to provide comprehensive impact analyses of tourism.