$4.5 million for training not recognized by the CCQ: how responsible is the CAQ?

Training costing nearly $50,000 per person, paid for by the CNESST for injured workers hoping to rebuild their lives. A certificate that does not provide normal access to construction sites. And after six years of spending, a public agency that does not know what happened to the vast majority of participants. A September 2026 investigation by Radio-Canada's *La facture* raises a legitimate question: how did the Commission des normes, de l'équité, de la santé et de la sécurité du travail pay $4.5 million to a single private school since 2020 without adequately measuring whether the training led to lasting employment? It also raises a political question: is this "the CAQ again"? The short answer requires nuance. **All payments documented in the investigation were made under a CAQ government, and both the Labour Minister and the government are accountable for overseeing the system. However, none of the records reviewed establishes that François Legault, Jean Boulet or Cabinet personally selected the school or authorized each file.** Individual decisions belong to the CNESST, a public agency governed by a joint labour-management board. That distinction does not erase political responsibility. It locates it accurately. ## What the investigation actually found [According to Radio-Canada's report](https://www.youtube.com/watch?v=pzd4MJow-c0), the CNESST paid $4.5 million to the École professionnelle de machinerie lourde in Terrebonne to train 111 injured workers between 2020 and 2026. The full program reportedly costs close to $50,000 per participant and lasts about three months. The workers featured in the report had been found unable to return to their previous jobs. As part of their vocational rehabilitation, the CNESST directed them toward heavy-equipment operation, which it considered "suitable employment" capable of restoring a comparable income. Two workers said they understood the training would allow them to return to the construction industry. After earning their certificates, however, they learned that the Commission de la construction du Québec did not recognize the training as equivalent to the vocational diploma required through the normal route into the trade. The [CCQ states](https://releve.ccq.org/fr-CA/metiers/operateur-pelle-mecanique) that a person seeking to become a shovel operator must normally complete the DEP in heavy-equipment operation, provide a job guarantee of at least 150 hours and obtain an apprentice competency certificate. Exceptional routes exist when labour pools open, but the private certificate does not replace the DEP. Radio-Canada also interviewed four other members of the same class. One had obtained temporary work involving machinery; the others had not found work related to the training at the time of the report. ## No, the investigation does not prove that 101 people are unemployed The most troubling number must be stated precisely. Of the 111 participants funded since 2020, **10 had told the CNESST that they found employment related to their training**. For the other 101, the CNESST had no long-term outcome because it stops following cases after the maximum one-year job-search period. That does not mean 101 people are necessarily unemployed or unable to work. Some may have obtained jobs without notifying the CNESST, worked in another field or left the labour market for unrelated reasons. Missing data is not proof of individual failure. It is, however, a serious management failure. An agency spending millions to retrain vulnerable people should be able to say how many found work, in which sector, at what salary and for how long. Without that information, no one can determine whether the program works or whether the money mainly produces certificates with insufficient employment prospects. The CNESST acknowledged to *La facture* that it could not explain what happened after one year and said it intended to better document future outcomes. ## "Not recognized" does not mean "entirely worthless" Another distinction matters. The private training is not recognized by the CCQ as a DEP providing normal access to regulated construction trades. That does not make it illegal, fictitious or unusable everywhere. The school says it clearly told the CNESST that its course was not equivalent to a DEP and argues that graduates can work outside construction sites governed by the R-20 Act: snow removal, municipalities, mining, forestry, quarries, agriculture, sorting centres or non-regulated private work. [Its website describes the program as private training](https://formationdemachinerielourde.com/) and displays supplier or registration numbers from several bodies, including the CNESST and the Commission des partenaires du marché du travail. Those numbers must not be confused with recognition of a diploma by the Education Ministry or automatic eligibility for a CCQ competency certificate. The central issue is therefore not necessarily that the school offers the course. It is whether the CNESST selected it for the right people, clearly explained its limits, verified real job prospects and ensured that the projected work was compatible with their injuries. ## The CNESST's own criteria are being tested The [CNESST's current policy](https://www.cnesst.gouv.qc.ca/sites/default/files/documents/reinsertion-professionnelle-3-05_0_0.pdf) says suitable employment must, among other things: - make use of the worker's residual capacity and professional qualifications; - offer a reasonable possibility of hiring; - present conditions that do not endanger the worker's health or integrity given the injury; - correspond to a position for which the person remains sufficiently competitive in the labour market. The law expressly allows the CNESST to fund vocational training. Such a program must provide the knowledge and skills required to access suitable employment. The Act also requires the Commission to choose "the most economical appropriate solution" among those capable of achieving the objective. In the two cases detailed by Radio-Canada, several facts weaken the suitable-employment premise. Employers contacted by the workers required experience or a DEP. One worker had shoulder limitations, while industry representatives described work involving vibration, cold and repetitive movement. The other was assigned a potential income of $84,600 even though he could not find an employer willing to hire him with his certificate and limitations. These accounts do not establish that all 111 files were mishandled. They do justify a file-by-file audit or, at minimum, an independent review of the fit between the training, functional limitations, employer requirements and theoretical salaries used to calculate compensation. ## A financial mechanism that may create the wrong incentive When workers cannot return to their former jobs, the CNESST may identify suitable employment and estimate what they could earn. Income-replacement benefits can then be reduced according to the difference between their former income and that potential income. If the suitable job is unavailable, the person may receive an extension for up to one year while looking for work. In the report, lawyer Marc Bellemare and the Union des travailleuses et travailleurs accidentés ou malades argue that this mechanism encourages the CNESST to quickly assign a relatively well-paid theoretical job: the higher the estimated potential income, the less the agency must pay in long-term benefits. This is a criticism and interpretation, not proof that public servants deliberately manipulated cases to save money. The CNESST denies that cost-cutting is the objective of its decisions and says it is fulfilling its rehabilitation mandate. The structural risk nevertheless deserves scrutiny. Training should not turn a theoretical salary into an administrative reality when employers do not recognize the skills, when the individual remains uncompetitive or when functional limitations make the work difficult to sustain. ## Is this "public money"? The $4.5 million was spent by a public agency, but it does not come directly from the general fund financed by personal income taxes. Quebec's workplace health and safety insurance plan is financed primarily through employer premiums based on payroll and risk. The CNESST sets an average annual rate to fund the system. Strict accountability is still required. These funds are administered under public law to compensate and rehabilitate injured workers. A poorly spent dollar can make the plan less effective, increase costs collectively borne by employers and, most importantly, leave injured workers without a genuine vocational solution. ## So what is the CAQ responsible for? Its responsibility operates at three levels. **First, the period.** The documented $4.5 million was paid beginning in 2020, entirely while the CAQ was in power. Jean Boulet remains Labour Minister in September 2026. **Second, oversight.** The CNESST has genuine administrative autonomy. Its joint board sets strategic directions, approves budgets and financial statements, adopts policies and oversees internal controls. The government nevertheless appoints the chair and board members, while the Labour Minister appoints a non-voting observer. The CNESST also appears in the Labour portfolio's organization chart. **Third, legislation.** The "most economical appropriate solution" principle dates to 1985; the CAQ did not invent it. But François Legault's government adopted the 2021 Act to modernize the occupational health and safety regime, sponsored by Jean Boulet. That reform amended section 181 without removing the principle. The CAQ therefore had a legislative opportunity to reconsider the balance between economy, real-world effectiveness and the worker's interests. The most accurate conclusion is this: **the problem was not necessarily created by the CAQ, but it continued under its watch while a CAQ minister was politically responsible for Labour and the government was carrying out a major reform of the system.** The government cannot simply answer that each file belongs to the CNESST. ## The unanswered questions Before declaring the entire program a failure, several facts should be disclosed: 1. How many of the 111 participants completed the program? 2. How many obtained machinery-related work after one, two and three years? 3. In which sectors did they work: CCQ construction, mining, forestry, municipalities, snow removal or elsewhere? 4. What was their actual salary compared with the potential income used by the CNESST? 5. How many challenged the suitable-employment decision or had benefits reduced or terminated? 6. What criteria were used to authorize the school as a supplier and maintain the relationship despite the lack of long-term outcome data? 7. Did the CNESST compare the cost and results of this private program with a recognized DEP or other rehabilitation options? The Labour Minister should request the publication of these outcomes and an independent review of the monitoring system. The CNESST should also pause new referrals to any training whose employment outcomes it cannot demonstrate, unless a specific non-construction job has already been identified and is compatible with the worker's limitations. ## A system that must prove it is actually rehabilitating workers Immediately calling the matter a political scandal would go beyond the available facts. What is documented is sufficient, however, to describe a serious accountability failure. The CNESST spent $4.5 million with a single school. It has a declared positive outcome for only 10 of 111 participants. Workers say they learned too late that their certificate did not open the expected door to the construction industry. And the system can terminate benefits based on potential earnings the worker may never actually obtain. The question is therefore not only whether the training is "recognized." It is whether the system recognizes the reality experienced by injured workers. The CNESST must answer for that. And because the documented problem unfolded under the CAQ, the government and its Labour Minister must demand answers, correct the weaknesses and demonstrate that rehabilitation is primarily about a realistic and lasting return to work—not merely the administrative closing of a file.