In workplace cannabis cases, the Employment Relations Authority sometimes invokes “health and safety” without adequately explaining what the drug test proved, what contractual obligation was breached, or why dismissal followed. When those analytical steps disappear, health and safety becomes less a legal reason than a convenient slogan.

Nobody is arguing that employees should be permitted to operate forklifts, heavy machinery or company vehicles while impaired by drugs.  An employee who is genuinely intoxicated at work may create an obvious and unacceptable danger. Employers are entitled, and often legally obliged, to manage that risk.

But acknowledging the importance of workplace safety does not answer every legal question that follows: 

It does not establish that an employee was impaired.  It does not establish that a drug test was lawfully required.  It does not establish that a workplace policy formed part of the employee’s contractual obligations.  It does not establish serious misconduct.  And it certainly does not establish that dismissal was what a fair and reasonable employer could have done in all the circumstances.

Those matters require evidence and reasoning.  Too often, however, “health and safety” appears to be doing the work that legal analysis should have done.

The Missing Middle

Section 103A of the Employment Relations Act requires the justification for dismissal to be assessed objectively. The question is whether the employer’s actions, and the way it acted, were what a fair and reasonable employer could have done in all the circumstances.

That includes considering whether the allegations were sufficiently investigated, whether the employee was told what the concerns were, whether the employee had a reasonable opportunity to respond, and whether the employer genuinely considered that response. 

The Authority also has an express duty when issuing a written determination. It must state the relevant findings of fact, state and explain its findings on the relevant law, and express its conclusions on the issues that require determination.

That does not require a judgment of High Court length. The Authority was designed to provide speedy, practical and relatively informal justice.  But informality is not permission to skip the reasoning.  In a cannabis dismissal case, there should be an identifiable path between:

1. what the test scientifically established;

2. what the employment agreement or policy prohibited;

3. what misconduct the employer alleged;

4. what evidence supported that allegation; and

5. why dismissal was a proportionate response.

Where a determination moves directly from “positive cannabis test” to “health and safety risk” and then to “justified dismissal”, the most important part of the decision has gone missing.

A Urine Test Is Not a Sobriety Test

This distinction was confronted directly in Hooper v Coca-Cola Amatil (NZ) Ltd.  Mr Hooper returned a positive urine test for cannabis. He acknowledged using cannabis outside work. The result was well above the laboratory reporting threshold.  

That may have established cannabis consumption.  It did not establish that he was impaired at work.

The laboratory evidence recorded that the presence of the cannabis metabolite could not determine the amount consumed, when it was consumed, or whether the employee was impaired. There was also no meaningful evidence that Mr Hooper had been performing badly, behaving unusually or presenting as intoxicated in the workplace.  That evidential limitation matters.

A urine test generally detects metabolites remaining after cannabis has been consumed. Depending on the person and the pattern of consumption, those metabolites may remain detectable long after the intoxicating effects have ended.

The result may prove historical consumption. It does not automatically prove present impairment.  That does not make the result irrelevant. An employer may have a lawful policy that expressly prohibits any detectable drug use, imposes a specified threshold, or requires abstinence for employees performing particular safety-sensitive roles.

But where that is the employer’s case, the decision-maker should say so.  It should not use the language of impairment where impairment has not been proved.

The Test Result Cannot Justify the Test

Hooper also exposes a basic problem of chronology.  Where an employer’s policy permits testing only when there is reasonable cause to suspect impairment, that reasonable cause must exist before the test is ordered.  A positive result obtained afterwards cannot travel backwards in time and create the justification that was missing at the point of testing.

The Employment Court rejected the proposition that an ultimately positive result could retrospectively validate a fundamentally flawed decision to test.  That principle should be elementary.

A police search does not become lawful merely because it uncovers contraband. In the same way, an intrusive employment process does not become justified merely because it later produces information the employer dislikes.  The Authority should therefore identify the alleged basis for testing with precision:

• Was it random testing authorised by a properly incorporated policy?

• Was it post-incident testing?

• Was it reasonable-cause testing based on identified behaviour?

• Or was the employer simply suspicious and hoping the result would supply the missing grounds?

These categories are not interchangeable.

Sim And The Importance Of Following The Policy

The distinction was demonstrated in Sim v Carter Holt Harvey Ltd.  Cannabis plants had been found outside a sawmill. The employer responded by subjecting approximately 190 workers to what it described as reasonable-cause testing.

But its own policy required an employee’s appearance, actions or behaviour to suggest that the individual might be affected by drugs or alcohol.

There was no such individualised evidence against the workers tested.  The Authority concluded that the testing was neither valid random testing nor reasonable-cause testing carried out in accordance with the policy. General safety concerns did not authorise the employer to ignore the conditions it had written into its own procedure.  That is what legal reasoning should look like.

The workplace was safety-sensitive. Cannabis use was a legitimate concern. The employer had statutory responsibilities.

None of that eliminated the need to ask whether the employer had followed the actual policy governing its power to test.

“Health and safety” was relevant, but it was not treated as a magic phrase capable of curing every defect.

The Harder Case: Te Kani

The more difficult authority is Te Kani v Foodstuffs North Island Ltd.  Mr Te Kani was involved in a workplace forklift incident. He admitted recreational cannabis use and expected that he would return a positive result. His test reportedly returned a reading significantly above the applicable threshold.

The Authority made an important distinction. Mr Te Kani was not dismissed because the employer had proved that he was impaired at the time of the accident. He was dismissed for breaching the employer’s drug and alcohol policy by exceeding its specified threshold.  That may be a legitimate basis for dismissal.

An employer is not necessarily required to wait until an impaired worker causes a serious injury. A clearly drafted and properly incorporated policy may impose preventative rules, particularly in genuinely safety-sensitive work.  But that is exactly why the reasoning must be explicit.

If the misconduct is exceeding a laboratory threshold rather than attending work while impaired, the Authority should squarely explain:

• how that threshold became a binding term or lawful workplace requirement;

• what the threshold was intended to measure;

• why exceeding it constituted misconduct;

• why the misconduct was sufficiently serious to justify dismissal; and

• what consideration was given to the employee’s explanation and the available alternatives.

The Authority described the threshold breach as representing a material health and safety risk.  But a threshold is not self-explanatory.  A laboratory cut-off may be designed to distinguish a reportable result from background noise. It does not necessarily correspond to intoxication, diminished performance or a scientifically established level of workplace danger.

The Authority may ultimately accept that exceeding the threshold created an unacceptable risk. But it must explain why.

Otherwise, “health and safety risk” becomes the conclusion and the reasoning at the same time.

The Hidden Policy Judgment

The concern is not necessarily that Authority members are deliberately concealing their views.

The problem is that poorly articulated reasoning can allow unexamined policy assumptions to determine the outcome.

Those assumptions might include:

• that any cannabis user is inherently unsuitable for safety-sensitive work;

• that employers should be permitted to enforce abstinence even where the written policy refers only to impairment;

• that a positive metabolite test is a reliable proxy for unsafe behaviour;

• that deterrence is itself a sufficient reason for dismissal; or

• that employers should receive exceptional deference whenever health and safety is invoked.

Some of those positions may be arguable.  An employer may reasonably conclude that certain roles require strict abstinence. Society may decide that some occupations justify highly intrusive testing. A workplace may legitimately impose standards more demanding than the criminal law.

But those are policy choices.  They should be identified, debated and justified.  They should not enter an employment determination disguised as a scientific conclusion supposedly produced by a urine test.

Testing Is An Intrusion, Not A Routine Administrative Step

The Employment Court has recognised that workplace drug testing significantly intrudes upon an employee’s privacy and bodily autonomy.

In Parker v Silver Fern Farms Ltd, the Court emphasised the invasive character of testing and the corresponding need for testing policies to be strictly interpreted and applied.  That principle is easily forgotten when the employee has tested positive.

Once a positive result appears, there is a temptation to treat every earlier step as an irrelevant technicality.  But the legal authority to demand a bodily sample is not a technicality.  The wording of the policy is not a technicality.  The distinction between random, post-incident and reasonable-cause testing is not a technicality.  Nor is the difference between impairment and historical consumption.

These are the questions that determine whether an employer was managing a genuine workplace risk or simply exercising power it did not possess.

The Authority Is Not A Workplace Safety Regulator

The Authority must take workplace safety seriously.  But its role in a personal grievance is not merely to decide whether the employer had a broadly respectable concern.  Its task is to determine whether the employer’s actual decision, based on the information available at the time and reached through the process followed, satisfied the statutory test of justification.

An employer can have a legitimate safety concern and still conduct an inadequate investigation:

• It can have a legitimate safety concern and still misinterpret its policy;

• It can have a legitimate safety concern and still rely on evidence that does not prove the allegation made; and

• It can have a legitimate safety concern and still reach a disproportionate outcome.

“Health and safety” does not relieve the Authority from examining those failures. It makes careful examination more important.

Reasons Matter

The Authority’s determinations affect more than the parties immediately before it.  Employers rely on them when drafting policies and deciding whether to dismiss.  Employees rely on them when deciding whether testing is lawful and whether challenging a dismissal is worthwhile.  Representatives rely on them when advising clients.

The Employment Court relies on the issues identified when a determination is challenged. A party bringing a non-de-novo challenge must identify the alleged errors of law or fact and specify the questions requiring resolution.  Thin reasoning makes all of that harder.  It also creates the risk of inconsistent justice. One employee may lose a case because a metabolite result is treated as proof of a serious safety breach. Another may succeed because the decision-maker carefully distinguishes between use, presence and impairment.

Without transparent reasoning, the difference can look less like the application of law and more like the personal instinct of the Authority member assigned to the case.  That is corrosive to confidence in the institution.

Safety Requires Better Reasoning, Not Less

Cannabis cases are genuinely difficult.  Cannabis can impair performance. The effects are not always obvious. Employers should not have to wait for an accident before responding to a credible danger.  But none of that justifies evidential shortcuts.

A positive test is not automatically proof of intoxication.  A policy is not automatically contractual merely because an employer calls it a policy.  A threshold is not automatically a measure of impairment.  A testing process is not automatically lawful merely because the result was positive.  And dismissal is not automatically justified merely because the words “health and safety” appear in the determination.

The Authority does not need to decide that cannabis is harmless.  It needs to decide the case that is actually before it, according to the evidence, the employment agreement and the statutory test.

Where broader policy concerns influence the outcome, they should be stated openly and defended honestly.

Anything less is not practical justice.  It is legal reasoning being replaced by an incantation.