Our census data blindside

Photo by Jamar Cromwell on Unsplash

Exploring the Census 2022 data crisis and what we’re doing in response.

Watching the young census workers walk around our neighbourhood, knocking door-to-door in early 2022, delighted me. But even then, the level of despondency in the workers themselves was real - largely given the number of shut doors they were faced with. The undercount has been pronounced, and the alarm bell has been sounded. Earlier this year, UCT researchers in fact reported that the undercount may have broken world records for census data at a dramatic 31% - and have also meant adjustments may have rendered our census data as questionable in its utility. So for instance, based on demographic reconstruction, the national population of 62 million, after adjustment for the estimated undercount, might have been overestimated by around 1 million people (or under 2%). 

Statistics South Africa have resisted some of the critiques, noting that the figures that have been released can be trusted and are rigorous, but have slowed down the release of certain data types (for instance not yet releasing ward level data). When accounting to Parliament, Statistics South Africa blamed: fleet shortages, inadequate human resources and non-responses (I was particularly saddened by their note that they saw high resignation rates given “primarily due high refusal rate”). Perhaps notably absent from the presentation and minutes was overt mention of National Treasury’s role - but though the Census was conservatively budgeted to need R4.8 billion to be conducted, and more comprehensively estimated to require on normal inflationary data R6.6 billion, only R3.35 billion was requested, and only R 3.15 billion allocated. The political economy of these issues of course make it more complex.

It should not, however, be underestimated what a problem this is. The last completed and fully released census data was in 2011 (with Community Survey data from 2016). To give you a sense of the change, the population then was 55.6 million people. Estimates now suggest we are well over 60 million people - how over being the subject of discussion. And our population is notoriously fluid - one statistic we have been able to ascertain at provincial level in our Census 2022 is around internal migration, which has demonstrated that 1 in 4 Gauteng residents being born elsewhere in the country.

Census data is our most foundational open government data project. This must be understood as a quintessential open government failing. And it feeds into a broader South African context of significant data gaps - when trying to report on progress on our Sustainable Development Goals in 2023, the South African government noted that 33% of the targets did not have sufficient or new data for tracking progress. And yet our ability to design policy for the benefit of South Africa and its residents depends on data. 

This must also be understood as an inequality challenge. Reflecting on specific data types, it has been noted by the United Nations:

“Using the lowest estimate of the percentage ‘missing’ in urban slums and extrapolating to all developing countries, there are an estimated 369 million missing from the sampling frames of standardised household surveys worldwide”.

The least visible tend to be the most marginalised. In other words, the people we can’t see in statistics tend to be the ones we most want to advance the developmental policy. 

So what must be done? Whilst Statistics South Africa is working hard on a way forward for Census 2022, here are the immediate actions OpenUp has taken to help address our open data gap - and we encourage all our partner stakeholders to join us:

  1. Advocate for open data.

Open government data is still such an important priority for the development field in South Africa. In our participation at this years Civic Tech Innovation Network Forum 2024 #UnderTheHood - and as a member of the South African Open Government Partnership’s Interim Steering Committee - we had rewarding discussions on ‘Co-Creating Real Participation’. In it, we discussed the opportunity of open data initiatives for local government participation with French Ambassador David Martinon, Zukiswa Kota of the Public Service Accountability Monitor (PSAM) and Clémence Pène of Agence Française de Développement.

  1. Promote citizen data collection.

In a recent collaboration with the Deutsche Gesellschaft für Internationale Zusammenarbeit GmbH (GIZ), we conducted in-depth research into the opportunity for creating and embedding citizen data into the policymaking processes of South Africa. The research helped to demonstrate that citizen data presents a potential opportunity to enhance policy making across all stages of the policy lifecycle in South Africa. By incorporating the lived experiences, and localised knowledge, of citizens policymakers can develop more responsive, effective, and trusted policies. As one respondent from our research noted:

“Citizen data is very important in terms of planning, in terms of research, in terms of resource allocation as well”.

And we have been implementing these strategies with communities and civil society data collectors for several years across a variety of sites. Localised, ground-up data collection presents an incredible opportunity - and OpenUp have developed rigorous methods for doing it. 

  1. Support statistics agencies.

There are also a variety of ways we have collaborated directly with the public sector to co-create open data platforms, and build each other's skills. In our recent training collaboration with the Global Partnership for Sustainable Data’s Capacity Accelerator Network (CAN), we have helped build the data and digital skills of an African-wide cohort, with representation from several national statistics agencies from across the region.

In sum, we have to preserve the sanctity of census data as a foundational open government data project. But understanding how important open data is for our policy and decision-making environment, means we should also all simultaneously be embracing productive strategies to help address data gaps where they occur. Contact us anytime about your own open data projects.

Share this post:
Email iconTwitter icon

Exploring the Census 2022 data crisis and what we’re doing in response.

Watching the young census workers walk around our neighbourhood, knocking door-to-door in early 2022, delighted me. But even then, the level of despondency in the workers themselves was real - largely given the number of shut doors they were faced with. The undercount has been pronounced, and the alarm bell has been sounded. Earlier this year, UCT researchers in fact reported that the undercount may have broken world records for census data at a dramatic 31% - and have also meant adjustments may have rendered our census data as questionable in its utility. So for instance, based on demographic reconstruction, the national population of 62 million, after adjustment for the estimated undercount, might have been overestimated by around 1 million people (or under 2%). 

Statistics South Africa have resisted some of the critiques, noting that the figures that have been released can be trusted and are rigorous, but have slowed down the release of certain data types (for instance not yet releasing ward level data). When accounting to Parliament, Statistics South Africa blamed: fleet shortages, inadequate human resources and non-responses (I was particularly saddened by their note that they saw high resignation rates given “primarily due high refusal rate”). Perhaps notably absent from the presentation and minutes was overt mention of National Treasury’s role - but though the Census was conservatively budgeted to need R4.8 billion to be conducted, and more comprehensively estimated to require on normal inflationary data R6.6 billion, only R3.35 billion was requested, and only R 3.15 billion allocated. The political economy of these issues of course make it more complex.

It should not, however, be underestimated what a problem this is. The last completed and fully released census data was in 2011 (with Community Survey data from 2016). To give you a sense of the change, the population then was 55.6 million people. Estimates now suggest we are well over 60 million people - how over being the subject of discussion. And our population is notoriously fluid - one statistic we have been able to ascertain at provincial level in our Census 2022 is around internal migration, which has demonstrated that 1 in 4 Gauteng residents being born elsewhere in the country.

Census data is our most foundational open government data project. This must be understood as a quintessential open government failing. And it feeds into a broader South African context of significant data gaps - when trying to report on progress on our Sustainable Development Goals in 2023, the South African government noted that 33% of the targets did not have sufficient or new data for tracking progress. And yet our ability to design policy for the benefit of South Africa and its residents depends on data. 

This must also be understood as an inequality challenge. Reflecting on specific data types, it has been noted by the United Nations:

“Using the lowest estimate of the percentage ‘missing’ in urban slums and extrapolating to all developing countries, there are an estimated 369 million missing from the sampling frames of standardised household surveys worldwide”.

The least visible tend to be the most marginalised. In other words, the people we can’t see in statistics tend to be the ones we most want to advance the developmental policy. 

So what must be done? Whilst Statistics South Africa is working hard on a way forward for Census 2022, here are the immediate actions OpenUp has taken to help address our open data gap - and we encourage all our partner stakeholders to join us:

  1. Advocate for open data.

Open government data is still such an important priority for the development field in South Africa. In our participation at this years Civic Tech Innovation Network Forum 2024 #UnderTheHood - and as a member of the South African Open Government Partnership’s Interim Steering Committee - we had rewarding discussions on ‘Co-Creating Real Participation’. In it, we discussed the opportunity of open data initiatives for local government participation with French Ambassador David Martinon, Zukiswa Kota of the Public Service Accountability Monitor (PSAM) and Clémence Pène of Agence Française de Développement.

  1. Promote citizen data collection.

In a recent collaboration with the Deutsche Gesellschaft für Internationale Zusammenarbeit GmbH (GIZ), we conducted in-depth research into the opportunity for creating and embedding citizen data into the policymaking processes of South Africa. The research helped to demonstrate that citizen data presents a potential opportunity to enhance policy making across all stages of the policy lifecycle in South Africa. By incorporating the lived experiences, and localised knowledge, of citizens policymakers can develop more responsive, effective, and trusted policies. As one respondent from our research noted:

“Citizen data is very important in terms of planning, in terms of research, in terms of resource allocation as well”.

And we have been implementing these strategies with communities and civil society data collectors for several years across a variety of sites. Localised, ground-up data collection presents an incredible opportunity - and OpenUp have developed rigorous methods for doing it. 

  1. Support statistics agencies.

There are also a variety of ways we have collaborated directly with the public sector to co-create open data platforms, and build each other's skills. In our recent training collaboration with the Global Partnership for Sustainable Data’s Capacity Accelerator Network (CAN), we have helped build the data and digital skills of an African-wide cohort, with representation from several national statistics agencies from across the region.

In sum, we have to preserve the sanctity of census data as a foundational open government data project. But understanding how important open data is for our policy and decision-making environment, means we should also all simultaneously be embracing productive strategies to help address data gaps where they occur. Contact us anytime about your own open data projects.