Australian Government - TRA
Australian Government - TRA
/Images/UserUploadedImages/178/img_regionalexpenditure.jpg

Regional Expenditure

International

Expenditure

Determining international visitor expenditure

There are varying levels in which an international visitor’s total trip expenditure may impact on the tourism destination of Australia and its regions. TRA results present three main types of direct tourism expenditure impacts that are the result of the collection of trip spend details from international visitors as they depart Australia. These three direct expenditure classifications are specifically:

  • Total trip expenditure
  • Spend in Australia
  • Regional expenditure.

 

There are subtle differences between each of these spend classifications.

Total trip expenditure

When an international visitor pays for their trip to Australia they are spending money which impacts on the world economy. Total trip expenditure is all monies spent so that the respondent could undertake their trip to Australia. For example, this may include airfares, package tour expenditure, food and beverages and payment for all accommodation, leisure activities, conventions and schooling while in Australia.

Spend in Australia

While international visitors may spend a lot of money on their entire trip, not all of this is spent in Australia. The international visitor is asked to separate the money they have spent in Australia from their total trip expenditure in the IVS. This is in order to determine the amount of money that is being spent by the visitor in the Australian destination. This may be on items such as food and beverages, accommodation, activities, school books, motor vehicles and shopping.

Regional expenditure

The IVS provides information on travel activity and expenditure by international visitors. Information on expenditure by these visitors is only collected for whole trips; it is not regionally specific. In order to determine the impact that the visitor activity is having on a particular region, TRA uses a model based approach to allocate visitor expenditure to the various tourism regions.

Regional expenditure allocation methodology

A very brief summary of the process by which expenditure by international visitors is allocated to regions is given below. A full explanation of expenditure allocation methods can be found in Carter and Collins (2005).

Two types of expenditure data are collected in the IVS:

  • expenditure for the respondent’s entire journey
  • expenditure at a randomly selected location.

 

A ‘location’ is a more specific spatial unit than a Statistical Area Level 2 (SA2). For example, the location Bondi is in the ‘Bondi - Tamarama - Bronte’ SA2.  A sub-sampling approach (selection of a single location for further study) is adopted as a starting point for regional expenditure estimation. This is because of the need for interviews to be done quickly and because it is unreasonable to expect an interviewee to remember expenditure at every stop. It is entirely feasible for a visitor to visit the same location more than once, but expenditure data are collected only if the randomly selected location has been visited only once.
Four major sub-components of total IVS expenditure are identified:

  • expenditure on domestic airfares
  • expenditure on travel packages
  • expenditure on accommodation, food and beverages
  • all other expenditure.

Domestic airfares

Wherever air travel is indicated, airfare expenditure is allocated equally between the departure region and the arrival region. In cases where there is no major airport in the departure and/or arrival region, the share of airfare expenditure is allocated to the nearest region with a major airport. If air travel is specified for the first leg of the trip, the departure airport is assumed to be the airport where the visitor arrived in Australia.

Package expenditure

A major part of any package for travel within Australia is taken up by airfares and other long distance travel fares. A series of studies by the ABS(1995, 1996), Australian Tourism Export Council (2000), Bureau of Tourism Research (Bonnet et al. 1994 and Skene, 1995) and Office of Economic and Statistical Research (2001) estimated the proportion of travel packages spent on things other than long distance fares to be between 26% and 35%. Based on these results, it has been decided that total package expenditure should be split with 30% being attributed to items other than long distance fares. The remaining 70% is assumed to be spent in the visitor’s home country.

The non-fare component is distributed among the regions using the iterative process.

Expenditure on accommodation, food and beverages, and other expenditure

Expenditure on accommodation, food and beverages (AFB) and other expenditure are obtained by summing expenditure on the relevant items. The total expenditure for each item group is allocated to the regions by the iterative process.

During the iteration procedure, expenditure at the randomly selected location for which there is expenditure information is treated as a known value, and is held constant. The amount actually distributed among the remaining regions in the trip is known as net expenditure, which is equal to total expenditure minus random expenditure. If there is no expenditure at a randomly selected location (either by the interviewer failing to ask, or the interviewee failing to reply) net expenditure is equal to total expenditure.

The iterative procedure

Steps in the iteration process are:

  • An initial regional cost indicator (average expenditure per night) is calculated for each region that has an expenditure sample. Three sources of data are used for this calculation: single region trips, random expenditure for one of the regions of two-region trips, and expenditure at a randomly selected location for multiple region (more than two regions) trips. National average expenditure is used for those regions where there is no expenditure sample.
  • For stops where there is no random expenditure, a preliminary estimate of expenditure at that stop in the trip is calculated by multiplying the cost indicator for the region at the stop by the length of stay at that stop. For stops where there is random expenditure, preliminary estimates of expenditure are left blank.
  • Non-blank preliminary estimates of expenditure are rescaled using the formula:

r = p * (E / R)
where
r = rescaled value for this stop
p = preliminary estimate for this stop
E = reported total expenditure for the trip
R = sum of preliminary estimates for the trip

  • For stops where there is random expenditure (that is, the preliminary estimate is blank), the rescaled value is set equal to the random expenditure. Rescaled values sum to the total reported expenditure for the trip.
  • New estimates of regional cost indicators are calculated by summing rescaled expenditure values for each region, and dividing this value by the total number of nights in each region.
  • If estimates of cost indicators for the current iteration differ from estimates of cost indicators from the previous iteration by less than an agreed amount the process is stopped, otherwise the procedure is repeated from step 2.

 

Rescaled values at the last iteration are the final estimates of expenditure at each stop.
At the end of the allocation process, expenditure on the four major expenditure components is estimated for every stop in each trip. Estimates of expenditure by state/territory and region are obtained by summing expenditure estimates at each stop.
It should be noted that expenditure by visitors who are in transit to another country is included in the expenditure estimates presented here.

Domestic

Regional expenditure allocation methodology

 

A brief summary of the process by which expenditure by domestic visitors is allocated to regions is given below. Further explanation of expenditure allocation methods can be found in Travel expenditure by domestic and international visitors in Australia’s regions, 1999-2010, Tourism Research Australia, Canberra. Expenditure in the NVS is collected for the respondent’s entire journey, not for individual stops. For both overnight and day visitors, information is collected on the following three types of expenditure:

•          expenditure by respondent during the trip
•          expenditure by respondent before or after the trip
•          expenditure paid by employer or other who is not travelling.

For each of these three types of expenditure, information is collected on the amount spent on each of the following expenditure items:

•          Packages
•          Taxis (including to/from airport)
•          Airline fares
•          Organised tours/side trips
•          Car hire costs (rental, leasing)
•          Fuel (petrol, diesel)
•          Vehicle maintenance or repairs
•          Any other long distance transport costs (train, coach, ship etc)
•          Any other local transport costs (bus, train, ferry, etc)
•          Accommodation (not relevant for domestic day trips)
•          Takeaways and restaurant meals
•          Groceries etc for self-catering
•          Alcohol, drinks (not already reported with food above)
•          Shopping, gifts, souvenirs
•          Entertainment, museums, movies, zoos etc
•          Horse racing, gambling, casinos
•          Conference fees
•          Education, course fees
•          Purchase of motor vehicles or any other major equipment
•          Other (phone, postage, medical expenses, repairs, dry cleaning etc).

Expenditure on capital goods (for example, motor vehicles, property and office equipment) is not included in the published NVS estimates, as it is not regarded as tourism expenditure. All the other expenditure items in the above list are included in the expenditure estimates for domestic day and overnight visitors.

Overnight trips expenditure allocation process

 

As discussed above, expenditure information in the NVS is collected for entire trips, not for individual stops. A method has been developed by which this amount can be distributed over all the stops on the trip.

The process of allocating money spent on overnight trips begins with the identification of four major sub-components of total expenditure:

•          expenditure attributable to the respondent’s home region
•          expenditure on airfares and other long distance fares
•          expenditure attributable to the destination region or regions
•          expenditure on long trips.

 The ‘home region’ is the region where the survey respondent lives and home region expenditure, by definition, applies to a single region. In taking a trip a traveller may spend some money in the region where they live; this expenditure is identified as home region expenditure. For example, expenditure on taxi fares, fuel and groceries paid for before or after the trip are assigned wholly to the traveller’s home region. Fuel expenses paid by someone other than the respondent and meals paid for before or after the trip are assigned to the home region in the proportion:

1 / (number of stops +1)   –   with the remainder going to destination regions.

Airfare and other long distance fare expenditure is allocated equally between the region where the journey started and where it ended. In the few cases where there is no major airport in the departure and/or arrival region (for example, the Central Coast of NSW or Melbourne East in Victoria), the share of airfare expenditure is allocated to the nearest region with a major airport. For trips where there is more than one destination, information on the transport mode used to get to each stopover is used to allocate airfares and other long distance transport fares to regions. For example, if a traveller uses air travel on three legs of their trip, one third of the airfare expenditure is allocated to the departure and arrival regions for each leg.

Unpublished work by Tourism Research Australia has shown that, on average, about 60% of package expenditure is taken up by airfares and other long distance fares and the remaining 40% by other items like accommodation, food and car hire. The first step in the allocation of package expenditure is to split the expenditure into these proportions. The airfare and other long distance component are allocated together with other airfare expenditure and the remaining 40% is allocated by the iteration process.

In allocating package expenditure using the iteration method, the ‘regional cost indicator’ is average package expenditure per night. The end result of iteration is an estimate of package expenditure at each stop on each trip.

Destination region expenditure excludes expenditure on capital items and includes 40% of package expenditure. Sixteen expenditure items and package expenditure are grouped into nine major expenditure types, which are:

•          Packages
•          Local transport
•          Entertainment
•          Fuel
•          Food
•          Shopping
•          Accommodation
•          Conference fees
•          Any other expenditure.

Each of these expenditure types is allocated by the iterative process.

For expenditure on long trips (trips with more than 21 stops) a single expenditure figure is collected for these trips, which is equal to total expenditure for the entire trip. This amount is allocated to regions in proportion to the nights spent at each stop.

The iterative process

 

Overnight trips expenditure is allocated by iteration in three steps:

1.         A regional cost indicator is calculated for each region that has an expenditure sample – national average expenditure used for those regions where there is no expenditure sample.

2.         Preliminary expenditure estimates for each stop in the trip are calculated by multiplying the cost indicator for the region at the stop by the length of stay at that stop. Preliminary estimates of expenditure are rescaled using the formula:

r = p * ( E / R )
where

r =        rescaled value for this stop
p =       preliminary estimate for this stop
E =       reported total expenditure for the trip
R =       sum of preliminary estimates for the trip

Rescaled values sum to the total reported expenditure for the trip.

3.         New estimates of regional cost indicators are calculated by summing rescaled expenditure values for each region and dividing this number by the total nights spent in that region. If estimates of cost indicators for the current iteration differ from estimates of cost indicators from the previous iteration by less than an agreed amount the process is stopped, otherwise it is repeated from step 2.

Rescaled values at the last iteration are the final estimates of expenditure at each stop.

Experience has shown:

•          Cost indicators for most regions reach a value close to their final value after four or five iterations
•          Fifteen iterations are usually enough to get successive differences well below $1.00
•          The process is not sensitive to initial estimates of cost indicators
•          The main driver of the process is the nights spent at each stop.

Day trips expenditure allocation process

Expenditure allocation for day trips is similar in principle to expenditure allocation for overnight trips, but less complicated as each trip has just one destination region. All day trip expenditure is apportioned between home and destination regions using predetermined proportions, which are the same as those for overnight trips. Unlike the overnight trip analysis, there is no need to further apportion expenditure to multiple destination regions.

 

 

 

 

 

Image: Byron Bay, NSW. Courtesy of Tourism Australia.