Who Answers for Our Billions? In Quebec, Oversight Exists, but Accountability Gets Lost

**SAAQclic, SIFA and the battery industry: recent controversies point less to an absence of oversight than to the difficulty of determining who knew what, who made the decisions and what consequences should follow.** --- Whenever a major public project goes off the rails, the same questions return. Who was responsible? Who knew about the risks? Who authorized the project to continue? And why does the public so often learn about the problems only after costs have risen or services have deteriorated? Calls for greater accountability from ministers and public-sector executives do not necessarily mean that every mistake should lead to a resignation. Governing involves risk, and refusing to take any risk can also come at a cost to society. Accountability instead requires a visible chain of responsibility: a person clearly responsible for the outcome, reliable information that moves upward in time, a record of the decisions made and consequences proportionate to the conduct involved. Quebec is not starting from scratch. It has laws, auditors, parliamentary committees and public dashboards. The more specific problem exposed by several recent cases is that responsibility is often divided among so many actors that it becomes difficult to assign when a critical decision must be made. ## An already extensive oversight system Quebec's [Public Administration Act](https://www.legisquebec.gouv.qc.ca/en/document/cs/A-6.01) provides that deputy ministers and certain heads of public bodies are accountable for their administrative management before the National Assembly. The appropriate parliamentary committee must hear them at least once every four years. The responsible minister may attend if the minister considers it appropriate. The government also publishes a [dashboard of infrastructure projects](https://www.tresor.gouv.qc.ca/infrastructures-publiques/tableau-de-bord) and a [dashboard of information-resource projects](https://www.tableaudebordprojetsri.gouv.qc.ca/). The first reports costs, schedules and changes affecting major projects. The second tracks public technology projects worth more than $500,000 once implementation has begun. The organizations that submit the data remain responsible for its accuracy. Since May 2025, the [Directive on the Management of Major Public Infrastructure Projects](https://cdn-contenu.quebec.ca/cdn-contenu/adm/org/sous-secretariat-infrastructures-publiques/publications/directives/directive_gestion_projets_majeurs.pdf) has also required approval stages, options analyses, risk assessments and monitoring of costs, schedules and scope. The responsible minister, the initiating public body, the project manager and associated entities must inform the Secrétariat du Conseil du trésor of risks that could compromise a project. These are substantial mechanisms. But they do not always answer four simple questions: Which individual was given the mandate to deliver the result? What information did that person receive? What decision did they make? And when? ## SAAQclic: Many actors, fragmented information The Gallant Commission report, released in February 2026, provides the most detailed picture of this problem. The [Commission of Inquiry into the Management of the Modernization of the Société de l'assurance automobile du Québec's Computer Systems](https://cesis.gouv.qc.ca/fileadmin/documents/Rapport/CESIS_rapport_recommandations.pdf) traced a program whose oversight was divided among the Crown corporation, its board, its executives, the responsible department, the Treasury Board and several monitoring bodies. According to the report, the total ten-year cost of the CASA program was established at $661 million following the tendering process. The Commission noted that the SAAQ subsequently produced few, if any, comparable estimates of the program's total cost over the same period. In other words, figures continued to exist, but the baseline needed to clearly measure their evolution became weaker. The Commission's 26 recommendations amount to more than a call for additional reports. They seek to rebuild the chain of decision-making. Among other measures, the Commission proposed mandatory written opinions at several critical stages of a digital project: assessing needs, studying the market, drafting tender documents, starting the project, carrying out a major replanning exercise and preparing for deployment. It also recommended a centralized record containing the information shared with government actors, accessible to the responsible minister in real time and preserving approximately ten years of financial history. That distinction is essential. A dashboard showing the current state of a project informs the public; a history of the approved parameters makes it possible to understand what changed and who accepted the change. ## SIFA: When accountability does not keep pace with restructuring The SIFA project—the financial and procurement information system for Quebec's health and social services network—provides another example of dispersed responsibility. On May 7, 2026, the [Autorité des marchés publics](https://www.amp.quebec/communique/ordonnance-2026-02) issued six orders to Santé Québec and the Ministère de la Santé et des Services sociaux concerning deficiencies in the project's budgetary and contractual management. The AMP found that the department had minimized its responsibilities and had not acted sufficiently to ensure the sound management of public funds, despite its role regarding the asset and the project's budget. It also required incomplete or inaccurate information to be corrected in the government dashboard. The case illustrates a particular risk created by public-sector restructuring: authority, contracts, ownership of an asset and budgetary responsibility may be transferred at different times. Unless one person is publicly designated to integrate the whole, each actor can be responsible for one part without anyone being clearly accountable for the result. ## The battery industry: Risk was permitted, but was it sufficiently documented? The [Auditor General of Quebec](https://www.vgq.qc.ca/Fichiers/Publications/rapport-annuel/219/VGQ_juin2026_ch2.pdf) also examined 29 financial-assistance files connected with the battery industry. The assistance authorized in that sample totalled $2.204 billion, 91% of which had been approved by government order. As of September 30, 2025, approximately $1.9 billion had been disbursed. As of March 31, 2026, approximately $700 million had been recorded as government expenses, primarily provisions for losses and non-repayable financial assistance. The audit did not question the political choice to support the industry. It found instead that, in the files examined, significant risks had not always been sufficiently analyzed or documented and that some estimates had been updated late. The costs of three projects had increased by more than 100%. Four companies were under creditor protection, and 89% of the assistance authorized for them had already been disbursed. A note of caution is necessary: the report presents several findings in aggregate or anonymously. It therefore does not support attributing every weakness to one specific company. Nor does it demonstrate that better analysis would necessarily have led the government to reject the investments. It does, however, show why a written record of risk analysis matters: it allows the public to distinguish between a risky but deliberate decision and one made with incomplete information. ## The real blind spot: Between wrongdoing and bad luck The public debate often pits two overly simple positions against each other. The first holds that a minister should "pay the price" for every failure occurring under their responsibility. The second argues that a minister cannot know every detail managed by thousands of public servants and autonomous agencies. Both statements contain some truth. A minister cannot administer every contract. But the minister should be able to demonstrate that a system existed for reporting critical information, that major risks were presented and that the powers granted by law were exercised. Senior executives, for their part, must answer for the quality of the information, contract oversight and operational management. The appropriate consequence should therefore depend on conduct, not solely on the outcome. - An unforeseeable cost overrun that is quickly disclosed may require corrective action without a personal sanction. - Failure to produce a mandatory report should trigger a public warning and suspend any new authorization until the omission is corrected. - Materially inaccurate information, or information withheld through negligence, may justify administrative discipline. - Deliberate concealment should lead to the investigations and sanctions provided by law. - A political decision to continue despite a negative opinion may remain legitimate, but it should be explicitly assumed and subjected to parliamentary and electoral judgment. This graduated approach protects both taxpayers and decision-makers acting in good faith. It avoids turning every mistake into a scandal while preventing administrative complexity from becoming a permanent shelter from accountability. ## Useful—but imperfect—international models In the United Kingdom, major government projects must have a "senior responsible owner": a named individual accountable for the project's objectives, business case and governance. For certain major projects, that person is also accountable to Parliament. The British framework also allows the senior official responsible for public spending to request a written ministerial direction when a minister wishes to proceed despite concerns about regularity, propriety, value for money or feasibility. These directions are normally made public. ([Government Functional Standard for Project Delivery](https://projectdelivery.gov.uk/library-products/government-functional-standard-govs-002-project-delivery/); [public-spending accountability framework](https://www.gov.uk/government/publications/public-spending-the-accountability-framework/public-spending-the-accountability-framework)) Norway has long used external reviews at the concept-selection and cost-estimation stages of major investments. New Zealand applies independent peer reviews at critical points in the life cycle of high-risk investments. These mechanisms do not guarantee success: official British and New Zealand assessments have themselves identified turnover among project leaders, incomplete data and uneven application. The lesson is therefore not to copy a foreign institution. It is to combine three elements that remain incomplete in Quebec: a named person in charge, a stable record of the project and a written decision whenever a critical threshold is crossed. ## Six concrete changes for Quebec A credible reform could build on recommendations already made by the Gallant Commission, the Auditor General and other oversight bodies. 1. **Publicly appoint an integrating project owner.** Every major project would have an executive personally responsible for overall delivery, with a published mandate, term and set of powers. The minister would retain political responsibility; the executive would carry administrative accountability for the project. 2. **Preserve a historical baseline.** The original approved cost, schedule, scope, expected benefits and risks would remain visible. Every modification would be dated and explained without erasing the previous version. 3. **Require signed opinions at critical stages.** The public does not need every working document or commercial secret. It should, however, be able to know an opinion's conclusion, author and date, along with its major reservations, subject to information that is legitimately protected. 4. **Make reauthorization mandatory.** A major cost increase, substantial delay, reduction in expected benefits or significant change in scope should return the project to the authority that approved it. 5. **Require a signature for a contrary political decision.** If a minister chooses to proceed despite a formally negative opinion, a written direction should state the reasons, the risks being accepted and the mitigation measures. The decision would then be accountable without paralyzing the government's ability to govern. 6. **Connect public data.** A unique identifier should link the project dashboard, calls for tenders, contracts, additional spending and approval decisions. Citizens could follow the entire history instead of reconstructing it from several separate systems. ## Transparency also has legitimate limits Publishing every opinion in real time would be unwise. Quebec's [Act respecting access to documents held by public bodies](https://www.legisquebec.gouv.qc.ca/en/document/cs/A-2.1) protects certain opinions, recommendations and confidential commercial or financial information. Those protections support sound decision-making, the government's negotiating position and the rights of businesses. The solution is to make the decision trail public, rather than every preparatory exchange: the identity of the person responsible, the approved parameters, the existence and general conclusion of the opinions, the decision taken and its rationale. Sensitive appendices can remain confidential, be redacted or be examined by an independent body and authorized parliamentarians. Quebec should also avoid creating a new body without meaningful powers, or adding reports to those already produced without changing how decisions are made. Accountability is not measured by the amount of paperwork. It is the ability to connect authority, information, a decision and a consequence. ## Answering before the crisis Quebec's oversight bodies often intervene effectively after the fact: the Auditor General documents, the AMP issues orders, the Québec Ombudsman describes the effects on users, parliamentary committees ask questions and public inquiries reconstruct the course of events. The next step must come earlier. Before costs become irreversible, the system should require those responsible to state publicly: this is the person accountable for the outcome; this is what changed; this is the opinion we received; and this is why we are proceeding. Only then can the public demand for accountability become more than a recurring call for resignation. It can become a method of government: rigorous enough to protect public money, nuanced enough to allow responsible risk-taking and transparent enough for citizens to judge those who decide on their behalf.