Agricultural Production Statistics: June 2022 (Final)

Period-specific information

Survey information

This release contains final results for key livestock, forestry, horticultural and arable crops from the 2022 Agricultural Production Census. These final figures are available by region, Territorial Authority and farm type.

The 2022 Census was part of the current programme of agricultural production statistics that started in 2002. Previous censuses were held in 2002, 2007, 2012 and 2017, with annual sample surveys in 2003–06, 2008–11, 2013–16 and 2018–21.

Responses from farmers and foresters in the 2022 Agricultural Production Census resulted in an estimated eligible population of 47,200 geographic locations.

2022 questionnaire changes (from 2021)

Horticulture – in 2022, we included questions on the area of fruit trees and vines planted and the area of vegetables and other horticulture crops harvested.

Farm Practices – in 2022, we included questions on the irrigated area of the farm and systems used, effluent management systems used, formal farm nutrient planning documents, and cultivation, direct drilling, and pasture renewal practices.

Sheep for milking – in 2022, we included a question on the number of dairy sheep used primarily for milk.

Goats - in 2022, we included two questions on goats to capture the number of goats used primarily for milk and those used primarily for meat. The two categories of goats cover all uses for farm goats, and the sum of these types yields the total number of goats published in 2022. The statistics are comparable to previous records of total goats as a continuous time series.

Forestry - during the Agricultural Production Census 2022 collection Stats NZ received additional data from the National Exotic Forestry Description (NEFD). This data related to additional areas of forestry from business was not previously included in the Agricultural Production Survey/Census population.  The additional data has resulted in increased forestry land use for forestry in 2022.

2022 imputation levels

Imputation levels for the 2022 Agricultural Production Census

Response rates

The estimated proportion of eligible businesses that responded to the 2022 Agricultural Production Census for this final release was 72.9 percent.

General information

About the estimates

Figures in this release are rounded. However, we calculate all percentages in this release using unrounded figures. These agricultural production statistics may differ from statistics produced using other sources, such as the National Exotic Forestry Description Survey produced by the Ministry for Primary Industries, the Stock Number Survey from Beef and Lamb New Zealand Limited, and Dairy Statistics from Livestock Improvement Corporation Limited. These surveys use different survey frames and designs.

When converting livestock numbers into stock units, for the various livestock types and class, stock unit conversion ratios from Beef and Lamb New Zealand were used as a guide. These are also known as the ‘Economic Service Conversions’ and can be found in the definitions section of the Beef and Lamb New Zealand benchmarking tool.

Population frame

The agricultural production surveys include all units identified on Statistics NZ's Business Register as having agricultural activity. The Business Register is a list of businesses in New Zealand, based on their registration for goods and services tax (GST) with Inland Revenue.

The compulsory registration level for GST is $60,000, so there is a partial and unquantifiable coverage of units below this level.

Survey population

The survey population for the agricultural production surveys is all businesses engaged in 'agricultural production activity' (including livestock, cropping, horticulture, and forestry), or which own land intended for agricultural activity. The survey population includes businesses engaged in agriculture or forestry production as a secondary activity.

Industries in scope

The survey population specifically includes businesses classified to the following ANZSIC06 codes:

• A01 Agriculture

• A0301 Forestry (excluding native forestry).

It also includes parts of:

• L671200 Non-residential property operators

• M691000 Scientific research services (agriculture-related research that involves land holding (excluding universities))

• P802300 Combined primary and secondary education (agricultural high school operation)

• P802400 Special school education (special needs education involving agricultural production activity)

• P810200 Higher education (agriculture-related research undertaken by universities that involves land holding)

• R912100 Horse and dog racing administration and track operation

• R912900 Other horse and dog racing activities (racehorse training and racing stables operations)

• R892200 Nature reserves and conservation parks operation.

The survey population specifically excludes:

• A019300 Beekeeping

• part of A019900 Other livestock farming nec (worm farming, pet breeding, dog breeding, cat breeding, bird breeding (except poultry, game birds, ostriches, and emus))

• A017100 Poultry meat (except growers who also produce eggs for human consumption).

Population changes

For the 2002 Agricultural Production Census, we sourced the population from the Business Register and the Inland Revenue Client Register. We checked these sources against industry lists and AgriBase to ensure all large units were included in the population.

After the 2002 Census, we use the Business Register for the census and survey populations.

Survey content changes

Over the years, we have changed the core information we gather in agricultural production collections.

Since 2002, these collections have gathered information on livestock, arable farming, horticulture, and forestry with the following exception:

In 2004, 2006, 2008, 2010, 2013, 2015, 2016, 2018, and 2021 the surveys collected information on livestock, arable farming, and forestry (we did not collect horticulture production information).

Stats NZ no longer collects information or compiles statistics relating to forestry production. These statistics can be found on the MPI website. Please note, Stats NZ does continue to collect information and compile statistics relating to forestry land use.

Reliability of sample survey estimates

This release contains statistics from the 2002–2022 agricultural production collections. All results from agriculture surveys are subject to non-sampling errors. Results from agriculture survey collections in non-census years are also subject sampling error.

Non-sampling error arises from bias in the patterns of response and non-response, inaccuracies in reporting by respondents, and errors in recording and classifying data. Non-sampling error comprises coverage error, measurement error (which arise from collection methods), non-response error, and processing error. We use procedures to detect and minimise these types of error, but they may still occur and are not easy to quantify.

Sampling error occurs because we base inferences about the entire population on information obtained from only a sample of that population. As 2003–06, 2008–11, 2013–16 and 2018-21 collections are sample surveys, not censuses, only the results from these collections are subject to sampling error.

Imputation

We impute values for farmers, growers and foresters who do not return a completed questionnaire. Imputation involves replacing missing items with values based on other information available.

The method of imputation we use is random 'hot deck' imputation.

The Agriculture Production Survey uses random hot deck imputation. Using this method, we replace missing values of one or more variables for a non-respondent with observed values from a respondent with similar characteristics. The imputation class is formed from common characteristics such as regional council area, ANZSIC06 group, and production data from previous years. We randomly assign each non-respondent to a respondent in the same imputation cell, and the farm production data of the respondent is copied across to the non-respondent.

To improve the imputation process, we remove respondents with uncharacteristically high levels of agricultural activity from their respective imputation cells. It is unlikely that these non-respondents would have similar characteristics to the removed respondents.

Nonresponse bias

We conducted a new analysis to assess whether the lower response rates to the Agricultural Production Census 2022 compared to previous years could have resulted in nonresponse bias with respect to the total number of dairy cattle, beef cattle, and sheep. The methods employed include response propensity modelling and response propensity weight adjustment.

Response propensity modelling was used to predict the likelihood of a farm responding to the agricultural production census given various farm characteristics. This analysis showed that patterns of response propensity remained similar to those observed in 2017. In other words, the variables that influenced farms to respond in 2017 were still the most important factors in 2022, despite the lower overall response rate.

Additionally, we used response propensity weight adjustments to understand if the current imputation methods can handle the low response rate. Response propensity weight adjustment is a commonly used method to mitigate nonresponse bias. We compared the results obtained from using the current imputation methods to those obtained using weight adjustment at both regional and national levels. The differences between the two methods were quite small, with less than 10% variation at the regional level and less than 1% variation at the national level.

These two analyses support that low response rates did not lead to significant biases in estimating total number of animals and that the current imputation methods produce reliable estimates in the Agricultural Production Census 2022.

Confidentiality

Data collected and information contained in this release must conform to the provisions of the Data and Statistics Act 2022. Published information must maintain the confidentiality of individual respondents. Prior to 2017, figures in tables affected by these provisions are denoted by ‘C’. Since 2017, the confidentiality of the tables has been implemented using an input perturbation method which involves adding noise to data at an individual farm level, and figures in output tables no longer have to be replaced with ‘C’s.

Quality suppression

Data with high sample errors or imputation levels are suppressed and are indicated by 'S' in the table.

More information

See more information about the Agricultural Production Survey.

Statistics in this release have been produced in accordance with the Official Statistics System principles and protocols for producers of Tier 1 statistics for quality. They conform to the Statistics NZ Methodological Standard for Reporting of Data Quality.

Liability

While all care and diligence has been used in processing, analysing, and extracting data and information in this publication, Statistics NZ gives no warranty it is error-free and will not be liable for any loss or damage suffered by the use directly, or indirectly, of the information in this publication.

Timing

Our information releases are delivered electronically by third parties. Delivery may be delayed by circumstances outside our control. Statistics NZ accepts no responsibility for any such delay.

Crown copyright©

This work is licensed under the Creative Commons Attribution 4.0 International licence. You are free to copy, distribute, and adapt the work, as long as you attribute the work to Statistics NZ and abide by the other licence terms. Please note you may not use any departmental or governmental emblem, logo, or coat of arms in any way that infringes any provision of the Flags, Emblems, and Names Protection Act 1981. Use the wording 'Statistics New Zealand' in your attribution, not the Statistics NZ logo.