The following steps provide a summary of the method used to create the experimental canola estimates for 2019-20:
1. The DAS canola crop maps for Australia were used to measure the area of canola within each SA2. A small number of SA2s contained very low estimates of canola crop area from the DAS crop maps and had no associated Grains Levy Payer Register data. These SA2s were excluded from the compilation process. In Tasmania and south-east Victoria, where DAS data was not available, the ABS used a combination of industry and ABS survey information to produce area estimates in these regions.
2. Canola records were extracted from the Grains Levy Payer Register over the period covering the 4th quarter of 2019 through to the 3rd quarter of 2020. Through consultation with industry this period was considered to best represent production from the 2019-20 financial year. Basic editing was undertaken on the levy payer data to correct issues where the tonnes sold was reported with an incorrect unit, this was identified by comparing with the dollar value.
- Production statistics were based on the tonnes sold in the levy payer register
- Local value statistics were based on the dollar value in the levy payer register
- Business number statistics were based on unique levy payers in the levy payer register
3. Canola records in the Grains Levy Payer Register were matched to business address information held by the ABS where possible, this provided additional address detail, for example in cases where the Levy data only had a P.O. box.
4. An SA2 code was initially allocated assigned to each levy record using the ABS held business address. If the assigned SA2 did not fall within a canola producing region, identified by the DAS crop mapping data, the address information from the Levy Payer Register was used.
5. For 2.1% of canola records in the Grains Levy Payer Register the assigned location was not within an SA2 where canola was grown based on the crop mapping. The levy production (tonnes), value (dollars), and business number data from records outside of canola growing areas were summed and redistributed across SA2s where canola was grown. The way this was done depended on the location of the original SA2 assigned using an address:
- If the assigned SA2 was within a canola growing SA4, then these records were redistributed across all canola growing SA2s within the SA4. This impacted 0.9% of levy records and 0.7% of the production total.
- If the assigned SA2s were not within a canola growing SA4, but were within a canola growing State, then they were redistributed across all canola growing SA2s within that State. This impacted 1.1% of levy records and 0.6% of the production total.
- Finally, any assigned SA2s that could not reasonably be allocated within a State were redistributed across all canola growing SA2s within Australia. This impacted 0.1% of levy records and 1.3% of the production total.
The redistribution was based on the area of canola within each SA2. For example, if an SA2 contained 1% of the total canola growing area in the SA4 region, it would receive 1% of the unallocated Levy Payer production of canola. This same process was used to redistribute the production value and business numbers associated with those levy records.
6. A small number of canola records in the Grains Levy Payer Register produced very large volumes of canola that related to an area broader than just the SA2 they were assigned to. These records were identified by comparing tonnes of canola produced with the area of canola growing within the SA2. SA2s with a yield of greater than 4 tonnes per hectare were identified as having canola that was grown across multiple SA2s. The production, value, and business numbers from these records were iteratively redistributed across SA2s within the SA4 region where canola was grown until a target yield was satisfied. The target yield of below 4 tonnes per hectare was determined through industry consultation. This process impacted 1.1% of levy records and 2.2% of the production amount as shown in the table summarising redistribution below.
7. SA2 counts of businesses producing canola were adjusted to align with totals reported in the canola Levy Payer Register once production amount, area and value estimates were finalised. This ensures that business numbers were not inflated through rounding in the redistribution process described in Step 6.
8. SA2 production amount (tonnes), production value (dollars), production area (hectares) and business number estimates were summed to SA4s and validated against SA4 regions reported in the 2019-20 ABS Rural Environment and Agricultural Commodities Survey (published in Agricultural Commodities, Australia, 2019-20) to assist in understanding the accuracy of this experimental method.