Tom Moultrie
@tom-moultrie.bsky.social
Professor of demography at UCapeTown; CRVS systems; cricket; politics. Analyst of South African Covid testing data and excess deaths (2020-)
Yes. But my realisation was that one can't expect the community to miraculously appear unless one helps build it. Anyway. I'm here. I suspect I'll be able to build off whatever you have managed to do.
November 24, 2024 at 4:52 PM
Yes. But my realisation was that one can't expect the community to miraculously appear unless one helps build it. Anyway. I'm here. I suspect I'll be able to build off whatever you have managed to do.
I have - finally - made the leap. Liking this a whole lot more
November 24, 2024 at 2:25 PM
I have - finally - made the leap. Liking this a whole lot more
correction -- the 170k is the total of matched records in the PES. Still trying to compute the PES sample size.
October 10, 2023 at 4:03 PM
correction -- the 170k is the total of matched records in the PES. Still trying to compute the PES sample size.
More later! Happy #CensusResultsDay!
October 10, 2023 at 8:18 AM
More later! Happy #CensusResultsDay!
In short, this report will give us the essential delineations as to the quality of the count. Thereafter, we will proceed to examine the headline figure; the population by age, sex, population group, and province.
October 10, 2023 at 8:17 AM
In short, this report will give us the essential delineations as to the quality of the count. Thereafter, we will proceed to examine the headline figure; the population by age, sex, population group, and province.
Previous releases of the PES report (2011: statssa.gov.za/census/censu...; 2001: statssa.gov.za/census/censu...) described in detail the extent of the undercount by province, population group, sex, and age, and how the actual count was adjusted.
http://statssa.gov.za/census/census_2011/census_products/Census_2011_PES.pdf…
October 10, 2023 at 8:17 AM
Previous releases of the PES report (2011: statssa.gov.za/census/censu...; 2001: statssa.gov.za/census/censu...) described in detail the extent of the undercount by province, population group, sex, and age, and how the actual count was adjusted.
The PES is a small-scale survey which started in June 2022 and ran through to August 2022, and is designed to estimate the undercount of the census. The undercount in the 2001 census was 17.6%; in 2011 it was 14.6%. What will it be in 2022?
October 10, 2023 at 8:17 AM
The PES is a small-scale survey which started in June 2022 and ran through to August 2022, and is designed to estimate the undercount of the census. The undercount in the 2001 census was 17.6%; in 2011 it was 14.6%. What will it be in 2022?
One of the most important indicators of the quality of the census will be found in the Post-Enumeration Survey (PES) Report, also due to be released at 1530 today. The PES can be thought of as a capture-recapture exercise to ascertain the coverage of the census.
October 10, 2023 at 8:17 AM
One of the most important indicators of the quality of the census will be found in the Post-Enumeration Survey (PES) Report, also due to be released at 1530 today. The PES can be thought of as a capture-recapture exercise to ascertain the coverage of the census.
Further, indications are that the response rate to online completion of the census was not great; leading to the mounting of a large-scale fieldwork operation, and repeated extensions of the fieldwork period.
October 10, 2023 at 8:17 AM
Further, indications are that the response rate to online completion of the census was not great; leading to the mounting of a large-scale fieldwork operation, and repeated extensions of the fieldwork period.
We know difficulties were experienced in getting the population enumerated; with fieldwork and data collection being extended in some provinces for several months (in the Western Cape to end May 2022: statssa.gov.za?p=15377). This was unfortunate.
October 10, 2023 at 8:16 AM
We know difficulties were experienced in getting the population enumerated; with fieldwork and data collection being extended in some provinces for several months (in the Western Cape to end May 2022: statssa.gov.za?p=15377). This was unfortunate.
...which were compounded by the decision (in principle, correct) to move to electronic (web-based) enumeration.
October 10, 2023 at 8:16 AM
...which were compounded by the decision (in principle, correct) to move to electronic (web-based) enumeration.
The (10Oct)2021 census was delayed by Covid and the effect it had on logistics. It may have been preferable to run the 2021 census in October 2022, giving more time to address those logistical issues...
October 10, 2023 at 8:16 AM
The (10Oct)2021 census was delayed by Covid and the effect it had on logistics. It may have been preferable to run the 2021 census in October 2022, giving more time to address those logistical issues...
The census was conducted in February 2022. Which irks: running a census on the same date every round makes demographers' lives easier in understanding changes between censuses.
October 10, 2023 at 8:16 AM
The census was conducted in February 2022. Which irks: running a census on the same date every round makes demographers' lives easier in understanding changes between censuses.
The date of release is significant - but for the wrong reasons. Today is 12 years since the last census was conducted (10Oct11), and 11 since the results of that census were released (10Oct12). So it's an anniversary of sorts. But not a particularly useful one.
October 10, 2023 at 8:16 AM
The date of release is significant - but for the wrong reasons. Today is 12 years since the last census was conducted (10Oct11), and 11 since the results of that census were released (10Oct12). So it's an anniversary of sorts. But not a particularly useful one.
The results will be announced at 1530SAST today. Most commentary will focus on the headline number. The only surprise there will be if it is NOT between 60 and 61 million. Demography is quite predictable that way.
October 10, 2023 at 8:15 AM
The results will be announced at 1530SAST today. Most commentary will focus on the headline number. The only surprise there will be if it is NOT between 60 and 61 million. Demography is quite predictable that way.
The census is _expensive_ and _complex_: A report on the 2000 US census described it as the most complex peacetime undertaking of the federal government. This applies more so to developing countries.
October 10, 2023 at 8:15 AM
The census is _expensive_ and _complex_: A report on the 2000 US census described it as the most complex peacetime undertaking of the federal government. This applies more so to developing countries.
The census _matters_. It is an (the) essential component of the equitable share formula for apportioning revenue to provinces; it gives a snapshot of who we are as a country; and provides a host of socio-economic data points.
October 10, 2023 at 8:14 AM
The census _matters_. It is an (the) essential component of the equitable share formula for apportioning revenue to provinces; it gives a snapshot of who we are as a country; and provides a host of socio-economic data points.
"you must have made a mistake. Your results cannot be". Discuss. (It happened to me)
October 9, 2023 at 5:49 PM
"you must have made a mistake. Your results cannot be". Discuss. (It happened to me)