COVID-19 References

Disclaimer: The data in this report is provisional and should be regarded as an early, indicative estimate of population counts, not official statistics. Any statements made in this are purely opinion based on our observation of trends/figures.

All summarised figures/data produced to create these insights are available in our github repository: https://github.com/dataventuresnz/mobility-index/

  1. Find out more here around how we produce our data:
    1. Population Density Privacy Impact Assessment.pdf
    2. New Stats NZ start up sells location data to government agencies eager to understand population movements
    3. Data from cellphone towers used to help Government make infrastructure decisions
  2. Each classification represented in the six graphs are determined by analysing a sample of suburbs of New Zealand and categorising them according to specific population behaviours. In the instance for each of the classifications, we use a selection of areas based on:
    1. Recreational: where people are typically going for the weekends, usually beach areas, lakes, tramping/biking spots.
    2. Residential: where people are typically staying in the evening, and through to the morning which is where they generally live.
    3. Retail: where people are coming together in times of when a mall/shopping district is normally operating.
    4. Tourism: where international tourists are usually visiting during daylight hours.
    5. Transit: where people are crossing over, either a suburb with a train or major bus station, a motorway/highway intersection or interchange. It is usually suburbs that demonstrate congestion at expected times of the week.
    6. Workplace: where people are generally during the day for 9-5 hours.
    For a technical overview of SA2’s and their types, a spreadsheet can be found inside our repository: https://github.com/dataventuresnz/mobility-index.html
  3. A normal week in 2019 is created from a rolling average using four weeks from the same period in 2019, and averaging the count across each hour.
  4. The methodology for constructing the mobility index is available through our GitHub written in R code at https://github.com/dataventuresnz/mobility-index.html