Both the Senate Inquiry report on Census 2016 and the Review of the Events Surrounding the 2016 eCensus (by Alastair MacGibbon, Special Adviser to the Prime Minister on Cyber Security) have just been released - I've embedded both reports below (so they appear in one place).
They are a good read with some useful recommendations for the future.
Reflecting on what has become known as #CensusFail, in comparison to other technical issues experienced by government, the Census 2016 experience probably rates as the most significant public ICT issue experienced by the Australian Government so far this century.
While in the last 15 years the Australian Government has had other incidents, these have been relatively minor, with limited public visibility or impact.
This includes incidents such as the 15 year delay in creating an Apple version of e-Tax (now rectified), launch issues with sites such as MySchool, challenges with access and security within the MyGov system, data breaches from the PCEHR (personally controlled electronic health records) eHealth system and the accidental exposure of private data for asylum seekers.
In contrast, the issues experienced during the 2016 Census have been far more widespread in their public visibility, impact and long-term ramifications for trust in government.
However, to put "the most significant public ICT issue experienced by the Australian Government this so far this century" into perspective - no-one died, no-one was hurt and no-one even lost control of their personal data.
At worst a number of government and IBM staff experienced unhealthy levels of anxiety for several days.
Given the struggles that developing countries have had to get their egovernment ICT working in the first place (with a reported 15% success rate); or the challenges advanced countries like the US have had with national systems (such as ObamaCare); or the experience of states like Queensland, which could not pay some of its Health staff for some time when its new payroll system failed, CensusFail just doesn't rate as an ICT disaster.
The actual operational impact of the 2016 Census problems was merely a short delay for people attempting to fill in the Census online.
Ultimately the ABS still exceeded the desired Census response rate, will still be releasing Census data much faster than ever before, and the agency still saved over $70 million dollars by moving more of the Census online.
However despite not actually rating as a ICT disaster, there was still a real cost to CensusFail - the perceptual and reputational damage from the ABS publicly failing to deliver on its Census Night promise, exacerbated by poor crisis engagement.
As a net result the real impact of CensusFail is on long-term governance in Australian, due to a reduction in trust in public institutions to 'do the job right the first time'.
I'm aware of other agencies now being regularly questioned by their Ministerial offices on whether they have any systems or projects which pose a similar reputational risk to the Australian Government. I've watched as the term 'CensusFail' has become the 'go to' term raised whenever a new government ICT issue is reported.
As a result the trust in government agencies to deliver complex technical solutions has been diminished, and it will take years to recover.
I hope that the recommendations in this Senate report, the lessons from Census 2016, will be top-of-mind for every public servant and Minister engaged in a significant government ICT project for years to come.
Hopefully the right lessons will be learnt - that managing your communications and public engagement well when the ICT gets wonky is critical.
In fact you can even transform a technical failure into an engagement success, if you get your messaging and timing right - strengthening, rather than weakening, trust in government.
They are a good read with some useful recommendations for the future.
Reflecting on what has become known as #CensusFail, in comparison to other technical issues experienced by government, the Census 2016 experience probably rates as the most significant public ICT issue experienced by the Australian Government so far this century.
While in the last 15 years the Australian Government has had other incidents, these have been relatively minor, with limited public visibility or impact.
This includes incidents such as the 15 year delay in creating an Apple version of e-Tax (now rectified), launch issues with sites such as MySchool, challenges with access and security within the MyGov system, data breaches from the PCEHR (personally controlled electronic health records) eHealth system and the accidental exposure of private data for asylum seekers.
In contrast, the issues experienced during the 2016 Census have been far more widespread in their public visibility, impact and long-term ramifications for trust in government.
However, to put "the most significant public ICT issue experienced by the Australian Government this so far this century" into perspective - no-one died, no-one was hurt and no-one even lost control of their personal data.
At worst a number of government and IBM staff experienced unhealthy levels of anxiety for several days.
Given the struggles that developing countries have had to get their egovernment ICT working in the first place (with a reported 15% success rate); or the challenges advanced countries like the US have had with national systems (such as ObamaCare); or the experience of states like Queensland, which could not pay some of its Health staff for some time when its new payroll system failed, CensusFail just doesn't rate as an ICT disaster.
The actual operational impact of the 2016 Census problems was merely a short delay for people attempting to fill in the Census online.
Ultimately the ABS still exceeded the desired Census response rate, will still be releasing Census data much faster than ever before, and the agency still saved over $70 million dollars by moving more of the Census online.
However despite not actually rating as a ICT disaster, there was still a real cost to CensusFail - the perceptual and reputational damage from the ABS publicly failing to deliver on its Census Night promise, exacerbated by poor crisis engagement.
As a net result the real impact of CensusFail is on long-term governance in Australian, due to a reduction in trust in public institutions to 'do the job right the first time'.
I'm aware of other agencies now being regularly questioned by their Ministerial offices on whether they have any systems or projects which pose a similar reputational risk to the Australian Government. I've watched as the term 'CensusFail' has become the 'go to' term raised whenever a new government ICT issue is reported.
As a result the trust in government agencies to deliver complex technical solutions has been diminished, and it will take years to recover.
Hopefully the right lessons will be learnt - that managing your communications and public engagement well when the ICT gets wonky is critical.
In fact you can even transform a technical failure into an engagement success, if you get your messaging and timing right - strengthening, rather than weakening, trust in government.
Census 2016 Senate Inquiry Report as redistributed by Craig Thomler on Scribd
Review of the 2016 ECensus - Final Report as redistributed by Craig Thomler on Scribd
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.