Dataset ID: WEATHER

Supplier source page

Dimensions

The series key contains two concepts, each separated by a full stop.

1. Mnemonic
2. Region

1. Mnemonic

This dimension represents the weather variable being measured. The following lists all the codes used for the weather variables.

• Total precipitation RAIN
• Average temperature TEMP
• Maximum temperature TMAX
• Minimum temperature TMIN
• Average feels like temperature FLT
• Average humidity HUM
• Average wind speed measured at 2 metres above ground W02
• Average wind speed measured at 10 metres above ground W10
• Average solar radiation SOL
• Average cloud cover CLD
• Average humidity tax HUM_TAX

2. Region

Codelist: CL_ZA_MUNICIPALITY_DISTRICT_PROVINCE

This dimension assigns a code to each municipality and each province. The tables below list each city along with its corresponding municipality name and municipality code, as well as each province and its corresponding code:

Attributes

Heterogeneous attributes

• Label (mnemonic + region, string)
• Unit of Measure (string representation)
  • Precipitation: “Millimetres”
  • Temperature (average, min, max, feels like): “° Celsius”
  • Average wind speed: “m/s”
  • Average solar radiation: “MJ/m²”
  • Average cloud cover: “Percent”
  • Humidity: “Percent”
  • Humidity Tax: “° Celsius”

Example

• Using R, you could access the relevant rainfall data for the Msunduzi municipality using the following:

rainfall <- read_dataset(id = "WEATHER", series_key = "RAIN.KZN225")

The data are sourced from NASA’s Black Marble program, which extracts night-time light intensity from the Day/Night Band (DNB) of the Visible Infrared Imaging Radiometer Suite (VIIRS) sensor onboard the Suomi NPP and NOAA-20 satellites.

Dataset ID: NIGHTLIGHTS

Supplier source page

Dimensions

The time dimension runs monthly from January 2012 to one month ago.

The series key contains two concepts, separated by a full stop.

1. Mnemonic
2. Region

1. Mnemonic

• Radiance R

2. Region

Codelist: CL_ZA_AREA

This dimension assigns a unique code to each region at the national, ward, municipality, and provincial levels. See the region code list below for detail.

Attributes

Example

• Using R, you could access the relevant rainfall data for the Msunduzi municipality using the following:

nightlights_KZN225 <- read_dataset(id = "NIGHLIGHTS", series_key = "R.KZN225")

Dataset ID: STP

The data used for this analysis was primarily built from the IRP5/IT3a tax form, which is a submission (mandatory since 2011) completed by each Pay-as-you-Earn (PAYE) registered employer for all employees earning more than R2000 per annum. The anonymised IRP5/IT3a data available within the National Treasury Secure Data Facility includes details relating to each individual (income, gender, age and industry), and importantly, information relating to the work address (currently a postal code). This field is generally populated for 90% of all certificates. This data is then combined with firm level information from the:

• ITR14 tax return form – which includes details relating to firms such as industry, turnover, firm size (according to SARS definition) etc.,
• VAT submissions – which includes details relating to industry, and
• Customs data – which includes details relating to the import/export status of firms. It is important to note that data used from these other forms are reported at a firm level rather than a branch/establishment level.

Project website: https://spatialtaxdata.org.za/

Reference

Nell, A. Visagie, J. Spatial Tax Panel 2014-2024 [dataset]. Version 5.1. National Treasury – Cities Support Programme and Human Sciences Research Council [producer and distributor], 2025.

Detailed, original metadata document.

Through EconData, Codera supplies the median income and employment variables, for free, aggregated by:

• Tax year (from the 2013/2014 tax year to the latest full tax year with a delay of about six months);
• Sex;
• Age categories;
• Municipality (not disaggregated by sex, age category or industry);
• Province;
• District; and
• First character of the SICv7 industry classification.

Attributes

• Source Dataset (“Spatial Tax Panel (National Treasury)”)
• Unit of Measure People for FTE, or Rands for Median Income
• Price Transformation (“Nominal” for Median Income)
• Frequency (Annual)
• Unit of Multiplication (Units)

Dimensions

1. Mnemonic

Codelist: CL_STP

• Full-time Equivalent Employees FE
  The number of full time equivalent (FTE) employees as calculated based on combinations of the “periods worked from”, “periods worked to”, “total periods worked” and “periods in year of assessment” data recorded in IRP5/IT3a certificates. Where one FTE employee equates to an individual that worked for 100% of the temporal period and 0.5 FTE employees equates to a person that worked for 50% of the corresponding temporal period (six months for a tax year or 15 days for a 30 day month).
• Median Income MI
  The median income weighted by full time equivalent employees based on incomes of all employees within an aggregation. This variable uses nominal values, and this can be adjusted to real values by end-users.
• Mean real income RM
  The mean of midpoints of real income bands (Dec 2024 prices), weighted by full-time equivalent employment.
• Aggregate real income AG
  Midpoints of the real income bands (Dec 2024 prices), multiplied by employment within the respective bands.

2. Region (national, province, or district)

Codelist: CL_ZA_MUNICIPALITY_DISTRICT_PROVINCE

And 227 more municipality codes, from CL_MUNICIPALITY, which follow the Municipal Demarcation Board codes (also the third administrative tier in the UN Second Administrative Level Boundaries (UN-SALB)). The municipality data in the STP on EconData only contains totals for the mnemonics, without disaggregating by gender, industry or age.

3. Gender

Codelist: CL_QLFS_SEX

• Any A
• Female F
• Male M

4. Industry (Major division)

Codelist: CL_SIC7_MAJOR_DIVISION

5. Age categories

Codelist: CL_AGE3

• All A
• 15–24 1524
• 25–34 2534
• 35–119 35119