Project Management | Public Affairs Management
This is a case study of the Australian Bureau of Statistics’ 2016 online census project as a portrait of complete management failure by government, the ABS, and its principal contractor, IBM. The case study is split over several pages because of its length. The table of contents below offers quick navigation to the various sections. Scroll down to read the contents of this section.
So little evidence exists to suggest a formal project management method in designing and implementing the online census system that it seems likely the ABS outsourced this function to IBM, along with the enterprise architecture, but that IBM took the narrowest view of them both to present only a solution architecture with an implementation phase.
If the ABS senior management calculated it could outsource its risk in funding the project along with all other project management functions, it miscalculated.
Project management as a discipline is long established, with methods drawn from engineering paradigms, but consolidated into mainstream management practices through methodology sets like the project management book of knowledge, PRINCE2, the project Institute of Management programs, and associated certifications. All these methods include a focus on governance, viability of business case, stage planning, and the incorporation of risk, quality, and exception (emergency or disaster) planning.
Examined from a PRINCE2 project management perspective, there are so many apparent shortcomings in the visible project management discipline that from all three stakeholder perspectives, there must but concerns about the professionalism, competence, and governance on display from the ABS and IBM.
Yet the census has been a regular project for the bureau for decades. There is even a census function listed in the organisational chart. This implies that there has been a long-term census project management function embedded in the organisation, and it seems like a major strategic mistake to have separated that function from the IBM project.
As it stands, the deficiencies in project management are most obvious in:
- Inadequate direction of the project, evident in the lack of fit for purpose products, but also in the absence of timely exception planning and reporting;
- Uncertain control over project stages, stage boundaries, and project closure, flowing from doubts about the integrity of data collected;
- Missing or incomplete product descriptions, quality control documents, and acceptance criteria, evidenced in unreliably tested infrastructure at all levels;
- Ineffective change, exception, risk, and quality management exposed by project products not fit for purpose, undetected misconfiguration, and confusion about likely causes for service disruptions on census day;
- Failure to tailor the project to suit the environment. This means that rather than relying on a literal interpretation and rigid application of a chosen method (the IBM way), it should have been purposefully adapted to suit the special circumstances of a statutory authority outsourcing a pioneering project; and
- Uncertainty about focus on a continued business justification rather than profit motive, given the lack of ABS supervision and oversight.
Figure 9 illustrates expected elements of PRINCE2 project planning, and shows missing components as shaded areas.
Many of the shortcomings mentioned in the section on enterprise architecture above could have been addressed by demanding ABS oversight of a formal project plan.
This could have been achieved by assigning a senior ABS project manager to ensure that the ABS always had in its possession the latest copy of a plan that included, at a bare minimum:
- Product descriptions for all project artefacts and products, including details of precise function, purpose, and quality tolerances (the degree to which final products could vary from initial specifications). Detailed product descriptions can go a long way towards substituting for an architectural overview;
- A risk management register that took into account the high profile nature of the census as a principal source of intelligence about Australian demographics, and acknowledging that this made it a prime target for foreign intelligence services, criminal cartels, and hackers. Moreover, an honest risk assessment that recognised public debate about privacy issues late in the project lifecycle as heightening the risk of hacking;
- A risk mitigation strategy, including the capacity to spread the financial cost of failure to the principal contractor;
- An exception plan for system redundancy and failover sourced from existing public sector resources (such as re-purposed hardware infrastructure, or resources ‘borrowed’ on standby from other departments for the duration of the census); and
- Stage planning that included the transfer of intellectual property and hands-on technology know-how as a prerequisite to the ABS internalising the capability to conduct future online census projects.
Stakeholders will find it difficult to accept any excuse about ABS project management supervisory responsibilities simply because the project was outsourced.
A professional opinion would likely question whether the bureau’s senior management had treated the undertaking as merely an IT project it was not comfortable to supervise rather than as a major business function using IT tools it should have taken the trouble to understand in greater detail.
Shareholders would not accept any excuse for what looks like incompetence in both the ABS and IBM management ranks.
An observer with no project management experience might even wonder whether there was a swindle going on, with no one accepting responsibility for project mismanagement and failure, but taxpayer money changing hands all the same.
If we accept that the ABS did not deliberately set out to fail, it may nevertheless be assumed from citizen and shareholder perspectives that the current structure, leadership, and culture of the ABS prevents fostering the necessary capabilities to operate effectively in business, operational, and social environments radically changed from those addressed with the bureau’s legislative framework, organisational structure, and management culture.
If this is accepted as the principal problem behind the project failure, addressing it requires complex, sophisticated planning and leadership from government.
From a professional perspective, obstacles include government unwillingness to deal with the politics of public sector labour relations, which represent serious barriers to breaking down dated management structures, careerism, and job security detached from competence and performance.
Public Affairs Management
Managing an organisation’s public image is not an optional extra. Chief executives cannot claim that it is not in their job descriptions, or that they were unfairly targeted for negative publicity. In a world with few communications barriers, the potential for instant and damaging publicity is a given that must be managed through a public affairs function that flows directly from strategic management.
Not managing public affairs is to fail at it. Confusing public affairs management with marketing or promotion is to fail at it. Telling lies is an invitation to bring on disaster. These are all mistakes made by the ABS for which its chief executive must accept ultimate responsibility.
David Kalisch’s media performances were wooden, dogmatically bureaucratic, and so unsympathetic most disinterested observers could easily have formed a negative opinion of the entire ABS on that basis alone.
Worse, from a professional perspective it was evident that Kalisch focused solely on Canberra public service gamesmanship, completely ignoring the social and technological aspects of his audience’s possible responses, such as disdain for bureaucratic stuffiness and legalism, mistrust of dictatorial demands, and the capacity to take to social media for instant feedback. He also seems to have been unaware that by presenting himself as quite so removed from public concerns, he was dramatically increasing risk for the online census project by inviting citizens to provide incorrect details, or even hacking the system as a protest.
Kalisch may well have deliberately targeted only his political superiors with his public appearances and statements, seeking only their approval, but he failed to understand that unexpected public reactions call for issues or crisis management to address negative sentiment, as illustrated in Figure 10. Even if Kalisch felt insulated from public opinion by his organisation’s status as a statutory authority, he should have realised that his superiors , as politicians relying on public sentiment, could not rely on a similar aloofness, and were acutely sensitive to a social media backlash. They would likely have thought Kalisch should have made at least some attempt to manage the fallout.
Stepping back a little, it may be that the ABS lacks the capability to act on the described public affairs triggers, but it should have been clear to the senior management that unilaterally altering an established compact with the Australian public about personal data represented a potential for publicity disaster. Particularly after very recently elected representatives likely to alter the balance of Parliamentary power, publicly declared their intentions to disobey ABS instructions.
ABS miscalculation began in the conception of changing the rules: advertising in the national press for public feedback, and conversations with privacy commissioners, may have met the letter of the law to claim ‘public consultation’ had occurred, but that attitude overlooks a changing media landscape in which press advertisements reach a shrinking number of the population.
This traditional approach to public’ consultation presupposes that ABS actions would not be interpreted publicly as as political rather than objectively neutral, and that the ABS would be immune from politics or public opinion in the era of social media and rapid-onset public relations catastrophes. Dynamics that are not at all aligned with the glacial pace of bureaucratic consultation processes.
Kalisch learnt nothing from his public affairs weaknesses prior to the census. In its aftermath he assured the Australian public that no data was lost or compromised during the outages; how could he have possibly known that before he was able to definitively explain the causes of the outages? His assurances came across as bare-faced lies. Even worse, previous assurances about rock-solid data storage security become questionable in light of the online census system failure.
For a CEO paid around $750,000 a year to be unaware of and unprepared for these factors exposes Kalisch as not yet ready to lead the an organisation of the size and importance of the ABS. If public affairs advice was offered and followed, the advisor’s own competence is also questionable.
From a citizen perspective, even without specialist IT or public affairs expertise, Kalisch appeared like a man making promises about factors completely beyond his control, and unable to muster convincing arguments to deflect his many critics. That is the public image of a man out of his depth, or deliberately trying to mislead his audience.
From a shareholder perspective, Kalisch’s public performances added to doubts about extracting any value at all from a census project so tainted by the possibility of data contamination or inaccuracy that it might have ruined the entire census product, online and offline. In effect this is equivalent to creating a loss of value.
Perhaps worst of all for Kalisch himself, a public affairs professional would recognise quite clearly that Kalisch may have invited his own demise for embarrassing the Prime Minister and Small Business Minister in front of a national audience. This is a direct consequence of both politicians supporting Kalisch’s public statements, and then being left to explain project failures at every level.
It did not help that the Small Business Minister compared trusting the ABS with trusting Facebook. There has been significant adverse publicity about Facebook misuses of private information, and there is no reliable data on the number of Australian Facebook users who have an account linked to valid name and address details. Ironically, if the Minister had been able to cite credible ABS figures on that point, his entire argument could have been more convincing.
Perhaps the most immediate steps the ABS can take to improve its public affairs performance is to address a kind of snow blindness preventing it from recognising its publics beyond its own operational, political, and public service culture.
As the chief provider of statistics used in determining the character of Australian representative democracy and therefore its political economy, many more eyes are on what the bureau does and says than a lay observer might imagine (see Figure 11 for an illustration of possible publics).
Beyond purely national spectators, there are foreign ones with commercial, geo-political, and sometimes criminal interests in acquiring ABS data, corrupting or falsifying it, or using it for nefarious purposes beyond the control of Australian law enforcement and regulation. None of the publics illustrated in Figure 11 are necessarily homogeneous, with some groups likely to have overlapping interests.
Not all demands, concerns, needs, or risks associated with public statements observed by these publics can be addressed, but knowing about their existence and considering their importance in each instance that public statements are made is a better strategy than pretending the world is still too isolated for such strategic considerations to matter. This is not a matter of catastrophising about conspiracies: hacking attacks traced back to Chinese and Russian locations are now routine contingencies for large public and private organisations.
Lessons to be learnt by the ABS from its own performance during the 2016 census include making changes that will help to:
- Create a strong and professional public affairs management capability which is not subservient to external commercial interests, or to internal marketing functions and politicking;
- Develop a public affairs process that includes profiles of all possible publics as well as issues and disaster management plans;
- Implement policy to prevent mixed messages in public, and to provide media performance training to a limited number of public spokespeople;
- Limit public statements to organisational products, expertise, and governance but refrain from commenting on legal matters; and
- Routinely evaluate public performance by monitoring conventional and new/social media.