Capacity Without Accountability? Institutional Trade-offs Under Cabinet Government and Public-sector Reform

Reforming the public sector typically aims for an increase in capacity in the form of speed, enhanced coordination, improved digital solutions, flexible working of employees and tighter leadership from the executive. Those objectives have the potential to enhance the performance of the government. The associated risk needs to be taken as a topic in March 2026. Capacity can increase, accountability can fade – particularly if there is a tendency toward centralized authority, more outsourcing, high levels of emergency response or data systems are hard to audit. It is a relationship, therefore, which Cabinet Government has to manage, concerning what can be done decisively, and what can be done for traceable responsibility.

There are a number of dimensions of capacity. Government requires to be knowledgeable, to have information, legal powers, resourcing, organisation memory, coordination etc. Often reform is placed in the context of tightening up executive control: tight control will lead to improved delivery. Too much or completely controlling agencies can overwhelm professional guidance, cause delays and make agencies wait until a political sign-off. Part of real capacity is knowing how to change when you notice a mistake.

New Zealand has a Cabinet system of government based on decision making by collective consensus and individual accountability of Ministers. The Cabinet Manual advises on the four elements of authority, confidentiality, conflict and collective responsibility. The direction of reform is a greater coordination through cross-agency but the outcomes should not take away individual accountability. Duties, responsibilities, authorizations and functions of the lead ministries, chief executives, appropriations and statutory tasks should be clarified.

Centralization provides opportunities to eliminate duplication and enhance standards. Analog, digital, property, procurement are potential sources for economies of scale. Centralization also brings a risk of centralisation. An inadequate system can impact all agencies as can local knowledge. Independent testing, staged rollout and agency level escalation for resilience.

There is another function of outsourcing that makes for a trade-off. Limited skilled resources can be obtained from contractors. People can become overly dependent on public institutions, which can make them lose institutional memory and make records inaccessible. Knowledge transfer, Audit rights, data ownership, conflict rules, public accountability must all be included in contracts. Especially important is that all firms perform key constitutional functions must be well protected.

The digital government both speeds up and evenness but also decreases transparency. Data and models that are used for automated forms of eligibility, risk scoring, ID verification and case management are not readily apparent to affected people. The methodology of accountability should include explanations, corrections, independent audits, security and human reviews. Procurement must steer clear of proprietary systems in which any type of meaningful scrutiny is not possible.

But flexibility is possible with workforce reform, which has costs associated with fast restructuring. Veteran employees start to resign, roles are not clearly defined anymore, and employee morale drops. Savings is immediately apparent and capability is not noticed until later. Specifying critical skills, transition risk and continuity of services should be a part of reform plans. A merit-based mechanism to appoint plus professional autonomy are still vitally necessary.

Setting goals can help focus but lead to misbehaviors. Agencies can also reduce cases or choose to focus on more easily quantifiable results, or recalibrate their definition of success. Quality, equity, user experience and long term outcomes should all be covered by balanced scorecards. These dashboards don't count for everything – there is independent audit and qualitative evidence as well.

While the freedom of budget can be an enabler to innovation, generosity in budgets can be an agent of diminished, legislative control. Appropriations should be transparent, so that the Parliament can see how the money is spent. There is a need to report, expire and review contingency funds and emergency procurement. When making claims of efficiency, information must reveal assumptions and distribution issues.

Cross-cutting issues can be addressed through collaborative structures but sometimes the responsibility can become diffused. Boards and ventures require mandates and decision rights, budgets and records, and complaint channels. After failure, a public statement with several agencies involved bearing responsibility, does not help matters. Whoever makes each decision can be identified on the accountability maps.

Cabinet confidentiality ensures freedom of discussion, and should not go beyond its operation. The final decisions, the evidence, implementation plans and performance should be made open to the public as long as it is legal to do so. Initiating the release process helps create legitimacy. There must be a review of information held if there is a reason that it cannot be given out at the time, but its sensitivity has reduced.

Integrity and capacity to provide free and frank advice. Using Ministers' best judgment, officials can make better decisions when they don't have to fear finding a "red flag. When the messages are centralized and pressured through politics, advice can be limited. Honesty of the institution is ensured through record keeping, escalation mechanisms and professional leadership.

Through participation, the public can contribute local knowledge and insights on implementation to enhance capacity. The less time allowed for consultation, and the less flexibility in decision-making, the less strong consultation is. First time correct will save re-design and unnecessary expenses. Community organisations must have resources to be able to be involved, particularly if those resources impact on vulnerable groups when reform is occurring in that group.

Independent institutions provide equalisation of capacities. Decisions can be slowed by auditors, ombuds bodies, courts, information commissioners and parliamentary committees, however; in slowing the decision there is risk in not preventing larger failure. Oversight needs to be in time and at the technical level. There's a sense of accountability that is not an externality of performance, it is part of the performance.

Best reform is one that grows, and makes itself accountable. Government should be able to provide speed, accountability, transparency, to record its actions, to make its outcomes understandable, and to fix any mistakes. Direction without personalizing administration, through cabinet.

The Number of Digital Systems Should Increase Without Opacity. Expertise needs to be added to the contractor's team without the loss of public accountability. Giving capacity without accountability can make a big difference for a short time; but achieving durable state capability requires trusted, lawfully constrained institutions that learn in public.

Resilience of institutions should be considered in addition to efficiency. A system which fulfills its normal demands can break down in disaster, cyber attack, or loss of workforce. Things like redundancy, backup procedures, local discretion, and manual alternatives can seem like inefficiencies, until crisis time. Not all duplicate capacity must be declared waste and the pricing should be resilience focussed.

An independent regulatory body can help to safeguard long-term capacity. The safety agencies, statistics, competition, elections and public finance must have their own professional independence. Although the accountability of ministers is desirable, direct control of the Minister's decisions should not be so as to affect the credibility. Statutes should be clarified on what direction and disclosure is allowed.

An opportunity cost should be a part of the evaluation. Experienced staff and agency managers can get stuck on Central projects. Any claimed savings should include those associated with migration, training, data transfer or service outages. Change of an independent review can test for fulfillment of the projected benefits.

During rapid reform there is a need for protection of whistle-blowers. Often, it is staff who reports on these unsafe systems, procurement problems or juggled metrics initially. Reporting processes must be confidential, require unbiased investigation and shielded from repercussions for report-education careers. Leadership should make available exertions of performance, the results of which are realised by staff and a sense of reporting leading to action can be established.

There is a need to make an explicit investment in records management. It's possible that files are lost, responsibility is obscured and access to information is compromised during reorganization. Ownership, retention, metadata, and making it available to the public should be specified in transfer plans. The validation of the digital migration has to be provided to ensure that historical evidence can be used.

The availability and selection of capacities should be limited by equity. Having a more rapid computer processing does not eradicate the discriminatory effects. A centralized purchasing system may be one that doesn't necessarily involve smaller suppliers. When there are workforce reductions, the language and culture capability can be reduced. Impact Assessment needs to consider who gets faster and who gets slower.

There is a need for sunset and review of reforms in governance. If there are any temporary taskforces or emergency powers or special funds, seek closure or new mandate. It is important to regularly evaluate permanent structures. Institutional accumulation can be another way of achieving unaccountable capacity.

Designing institutions should ensure continuation of challenge functions. Access and independence for internal audit, legal counsel, risk officers, privacy experts and evaluation teams. Concentrating those voices under delivery leadership can soften those scurrine eyes.

Where allowed by law, there should be record of ministerial directions. Having good records also safeguards policymakers and officeholders when interpreting between policy and administrative decisions. As of yet, there is no documentation of the pressure, which reduces the level of control and decreases the force of reviews.

There is a need to prioritise according to public value. Reduction in operating cost can correspond to an increase in social cost due to exclusion, delay or error during implementation of the reform. Breathing measures of broader impact and sensitivity analysis for business cases.

A capability transfer should be part of shared services. Agencies should have sufficient local know-how to be able to go head to head with central agencies, and deal with failure. All is for the government can lead to monopoly in it.

Ideally, post implementation review should be done when the systems have settled down but prior to institutionalized lock in. Need to present and provide baseline data and original assumptions in the reviews. There should be a timeline and owner for management responses.

There should be consolidated reporting done on key reform portfolios to Parliament. The spread out reporting the costs along with risk can be concealed. A portfolio view can be used to compare projects and indicating which project is dependent on the same vendors or data platforms.

Experiments at local level within national limits should be allowed. Evaluation originalises the consistency and pilot authority the innovation.

Leadership Succession Plans are the key to keeping the capability in place when key leadership roles change and limit the needy for dependence on one or two reform champions to drive change.