ABARES Farmland Price Indicator

Farmland holds immense economic and ecological significance in the Australian context. For most farm businesses, land is the most valuable asset, and trends in farmland prices are a major driver of business performance and growth over time.

Moreover, the value of farmland plays a pivotal role in the financial portfolios and overall well-being of Australian farmers (Chancellor and Zhao 2020). In addition, considering that over half of Australia's land mass is dedicated to agriculture, comprehending the changing value of farmland has profound policy implications for land use.

We employ a stratification method that leverages the CoreLogic land transaction data, which serves as the foundation for our calculations. Details of this method can be found here: Measuring Australian broadacre farmland price: A stratification approach.

Update July 2024: The ABARES Farmland Price Indicator model has been expanded to include experimental farmland price estimates for the following:

• Beef farmland – National and by MLA zone, annual.
• Cropping farmland – National and by GRDC region, annual.
• Dairy farmland – National and by 5 out of 7 dairy regions, annual.
• Forestry farmland – National, annual.
• Hobby farmland – National, annual.
• Orchard farmland – National, annual.
• Sugarcane farmland – National, annual.
• Vineyard farmland – National, annual.

These farm types are defined using data labels attached to each transaction by CoreLogic. Outliers and erroneous data are controlled for in the model. Otherwise, the same stratified median approach is followed as per the main land price estimates. Transactions are spatially assigned to their respective GRDC region (cropping farmland transactions only) and MLA zone (beef farmland transactions only), using the latitude and longitude point linked to the land transaction.

These estimates are to be treated as experimental until a full review of the method and working paper is completed. The upper bound of the outlier control is loosened to account for these farmland types, allow for higher amounts of fixed capital. The unit of measurement is the price per hectare of farmland parcels transacted on market, and there are no adjustments made to exclude houses, buildings, or other infrastructure, where present.

Update February 2024: Changes to the ABARES Farmland Price Index publication were prompted by feedback from users, updates to the data feed, and opportunities for methodological refinement and transparency. These changes enable the publication of more frequent (quarterly) data for the headline estimates, and greater spatial detail — with the release of estimates for most ABARES regions. The period of observation is 1992 to 2023, replacing previous 1992 to 2022 estimates which included a 2023 forecast. In addition, new variables provided by CoreLogic have facilitated data processing, so that we can more readily define broadacre farmland parcel transactions and remove non market and non broadacre parcel transactions.

Users expressed an overwhelming interest in ‘price per hectare’ estimates rather than ‘price indexes’ which were previously used in the ABARES Farmland Price Index publication. Price per hectare estimates are easily understood and can be converted to index numbers, therefore, we agree with the user comments and are now publishing in ‘price per hectare’ format. As the product has evolved and no longer contains index numbers, a practical renaming decision has been made to this product – which is now called the ‘ABARES Farmland Price Indicator’.

The method applied to generate the revised estimates still uses a stratified median approach as in Boult et al. (2023), however, changes to weighting and decile construction (specified below) mean that the revised estimates reflect the actual market conditions for broadacre farmland. By comparison, the previous estimates were weighted using the land area of surveyed farms, and instead represented the stock of market and non-market farmland. The revised estimates reflect actual market transactions for broadacre farmland during the relevant period, and have many advantages over the previous approach in terms of flexibility, transparency, speed of estimation, and customisation.

The revised estimates will now include quarterly publication of the headline broadacre land price per hectare at the national level.

Annual estimates of broadacre land price per hectare will include:

• National level
• State level
• ABARES AAGIS agricultural zones (e.g., wheat-sheep zone)
• ABARES AAGIS agricultural regions where transaction volume is sufficient (e.g., Region 131, NSW Tablelands).

Revisions to method

In the previous method, CoreLogic records were assigned to deciles using weights which were based on overall farm size data, which was drawn from ABARES’ farm survey program. The intention was to develop a farmland price estimate using market transactions which represented the overall farmland stock used in broadacre Agriculture. After careful consideration, we have revised this approach to build the deciles using land size of the actual CoreLogic market transactions — therefore generating estimates which are a true representation broadacre farmland market transactions in a particular period, rather than the potential market or farmland stock.

Land transactions are separated into even stratification deciles according to their parcel size by period (year or quarter) and observation category (national, state, zone, region). A default of ten groups (deciles) is used for each period and category, however, regional estimates and pastoral zone estimates instead use three groups (i.e., separating land size into small, medium, and large). Due to a lower number of transaction records per period a three group system is used for the regional level and pastoral zone estimates.

More information on the changes and revised methodology is available here.

Supporting data - ABARES Experimental Farmland Price Indicator - MS Excel [64 KB]

ABARES Farmland Price Indicator - Measurement Framework