January 09, 2019

BERL Data Newsletter - January 2019

The BERL data newsletter has been designed to provide commentary on data issues and news, including methodological changes, new measures, reviews, delays, and other technical issues and news.

This newsletter covers:

  • BERL Local Authority datasets
  • Further delays to the 2018 Census

BERL Local Authority datasets

The 2018 versions of BERL’s local authority dataset and BERL’s local authority tourism datasets are now operational.

BERL Local Authority Dataset

This dataset contains March annual employment (FTEs), value add (GDP), and business unit data for each of the 66 local authorities, and 14 regional areas within New Zealand. There are two main tables that are usable within the dataset, the first is the summary table which contains employment, GDP and business unit data at the ANZSIC06 1 digit and 3 digit levels.

* Imputed value, included in Total GDP only
Sectors (2018) FTEs % GDP (2018$m) % Business units %
Primary 375 0.2 624 3.3 225 0.8
Manufacturing 4,239 2.8 514 2.7 558 2.0
Construction 7,163 4.7 601 3.2 1,818 6.6
Wholesale and Distribution 8,508 5.6 1,196 6.3 1,500 5.4
Retail Trade and Services 21,849 14.4 1,384 7.3 3,924 14.2
Business Services 51,750 34.2 9,315 48.9 16,806 60.7
Arts and Recreation Services 3,474 2.3 370 1.9 687 2.5
Social Services 53,895 35.6 5,049 26.5 2,148 7.8
Sub-total (excluding O.O.D.) 151,253 100.0 19,054 100.0 27,666 100.0
Owner-Occupied Dwellings (O.O.D.)*     1,654      
Total 151,253   20,708   27,666  

The dataset above is construction using Statistics New Zealand Tourism Satellite Account (TSA) data on the percentage of each ANZSIC industry that is attributable to tourism as a base. For each local authority this Tourism industry percentage is adjusted using guest nights, daily commercial accommodation capacity, hotel capacity, percentage of international visitors, and tourism expenditure variables.

To make the adjustment each variable is first converted to a per FTE figure, which allows us to compare this figure to an average New Zealand figure for each variable. Then the results from the five variables are weighted and combined to create a single number. If this number is higher than 1 then the local authority has a higher relative impact from Tourism than the national average, and if the number is lower than 1 then the local authority has a lower relative impact from Tourism than the national average.

For example Queenstown–Lakes District has the highest relative combined score at 2.5, while Clutha District has one of the lowest relative combined scores at 0.8 with the New Zealand average being 1.

This final relative score is used in the dataset calculations to modify the TSA estimates of March annual employment (FTEs), value add (GDP), and business unit data per ANZSIC06 industry.

Further delays to the 2018 Census

Statistics New Zealand announced late in 2018 that work on the first release of 2018 Census data is taking much longer than expected due to the complex nature of the task. Stats NZ had previously reported that they had full or partial information for around 90 percent of individuals and indicated further analysis would be required that would delay the first release.

To deal with the lower response Stats NZ is developing and extending a range of statistical methods to adjust for limitations in the data collected and to understand the quality of the results produced. This includes using other government data about real people (such as births, tax, health, and education records) to compensate for the information it is missing. However, Stats NZ's own research shows that government data has strengths and weaknesses. These include gaps and inconsistencies, for example, who lives together in a household, and whether people are of Māori descent, and their iwi information. The latest summary of how Stats NZ plans to adjust for missing information can be found here.

An external data quality panel was formed to provide advice and guidance throughout the imputation processes, alongside support from international Statistics agencies who have faced similar challenges. The panel of independent experts brings together a broad range of local and international demography, census model, and statistical methodology skills and experience. The appointed panel members are:

  • Richard Bedford, Emeritus Professor, recently retired Professor of Population Geography, Auckland University of Technology and University of Waikato (Chair and spokesperson)
  • Alison Reid, Local Government Senior Researcher, Auckland Council
  • Barry Milne, Expert Data User, COMPASS Research Centre, University of Auckland
  • Donna Cormack, Senior Lecturer, Faculty of Medical and Health Sciences, University of Auckland; Senior Research Fellow, Department of Public Health, University of Otago, Wellington; and Te Mana Raraunga (Māori Data Sovereignty Network)
  • Ian Cope, International Census Expert, ex-Office of National Statistics (ONS), United Kingdom
  • Len Cook, former New Zealand Government Statistician and former National Statistician of the United Kingdom
  • Tahu Kukutai, Professor of Demography, National Institute of Demographic and Economic Analysis, University of Waikato; and Te Mana Raraunga
  • Thomas Lumley, Professor of Biostatistics, University of Auckland.

Over the coming months, Stats NZ will continue to work on the preparation of population data required for the electoral boundary review, due to be conducted by the Representation Commission ahead of the next general election in 2020. Stats NZ will also be working with iwi and Māori-focused organisations as the impact of the lower response on Māori descent and iwi data becomes clearer.

Stats NZ knows that it didn’t make it easy enough for everyone to take part in the census. This and a number of other factors are being investigated as part of the independent review that is currently under way. This comprehensive independent review of the 2018 Census will be carried out by leading New Zealand Management Consultant Murray Jack and Canadian Census Expert, Connie Graziadei. The reviewers are due to report their findings in July 2019.

The objective of the review is to consider the design, implementation, and operation of the 2018 Census, with a focus on participation in and the coverage of the census. See here for the terms of reference.