Data Collection

Period-specific information: March 2021 quarter

Period-specific information: March 2021 quarter en-NZ
Period-specific information: March 2021 quarter en-NZ



Reference Period

The information in this release is compiled from data collected in the quarterly Local Authority Survey run in the March 2021 quarter.


  1. Local authority statistics, seasonally adjusted income

  2. Local authority statistics, seasonally adjusted income, percentage changes

  3. Local authority statistics, actual income

  4. Local authority statistics, seasonally adjusted expenditure

  5. Local authority statistics, seasonally adjusted expenditure, percentage changes

  6. Local authority statistics, actual expenditure

Seasonal adjustment

New items of income or expenditure that are included by local authorities can impact seasonal adjustment when the seasonal component of the data is unknown. With the introduction of Auckland’s regional fuel tax (RFT) in the September 2018 quarter, we have seen greater volatility in the seasonally adjusted regulatory income and petrol tax series. We expect this volatility to continue for a number of quarters until we have collected sufficient data to understand any changes in the seasonality of this series resulting from the introduction of the RFT.

In September 2019 a review of all series was carried out. Based on the review, several changes to seasonal adjustment were implemented in the September 2019 quarter: the rates series is now seasonally adjusted, while the interest received and interest paid series are no longer seasonally adjusted.

In June 2020, the seasonal adjustment series were reviewed and additive outlier adjustments were applied to account for the impacts of COVID-19 to the following series:

  • Regulatory income and petrol tax

  • Sales and other operating income

  • Total operating income

  • Purchases and other operating expenditure

QLAS survey update

In the September 2020 quarter, we updated the QLAS form to improve data quality. We refined existing QLAS questions to better align with Local Authority Census (LAC) questions, so that the inclusions and exclusions criteria for each QLAS question matched each corresponding LAC question. This means that each June, when we benchmark QLAS data to annual Local authority financial statistics (LAFS) data, the sizes of annual revisions are anticipated to reduce in comparison to previous periods.

Review of methods

Since March 2014, only the largest 38 local authorities have been covered in the Quarterly Local Authority Survey (QLAS), with the contribution of smaller councils modelled. The estimates from smaller councils have been modelled using the smoothed movements of medium sized councils.

In the March 2021 quarter, we have improved our approach to the modelling of small councils for the dividend income series. Over time, for the dividend series in particular, the quarterly estimates of dividends for small councils had grown significantly higher than the actual figures collected annually in the Local Authority Census (LAC). Our new approach for modelling dividends reconciles the small council estimates in QLAS to annual benchmarks of small council estimates from LAC and prevents such a drift occurring over time. The new modelling has been backdated to the start of modelling in the March 2014 quarter.

Aggregate results (for all councils) from the QLAS are benchmarked to the annual LAC through a process of reconciliation. This process to a large extent has mitigated any deficiencies in the small council modelling for historical quarters. Therefore, the revisions caused by updating the modelling method for dividend income are larger from the September 2019 quarter onwards, as this is beyond the latest LAC benchmark year.

The revisions are also seen in the total investment income, total operating income and operating surplus/deficit series, as the dividend income series contributes to these.

The small council modelling approach for other QLAS series will be reviewed over coming quarters with the likely result an update to their modelling methods.

If you use this specific series and require additional information, please contact <>




View Full History
Revision Date Responsibility Rationale
2 30/11/2021 3:52:42 PM