Go Around! Why Pilots Press On When They Should Fly Away

AVIATION SAFETY | THOUGHT LEADERSHIP

A go-around is one of the most safety-critical manoeuvres in commercial aviation — and one of the most psychologically resisted. Understanding why pilots continue unstabilised approaches, and what the industry is doing about it, is central to reducing one of aviation’s most persistent accident categories.

Continuing an unstabilised approach is one of the leading precursors to approach-and-landing accidents worldwide. Despite clear stabilised approach criteria, mandatory go-around callouts, and decades of regulatory reinforcement, pilots continue to press on when the correct action is to discontinue. This article examines the human factors, commercial pressures, and organisational dynamics that drive this behaviour — and what a robust safety culture must do to change it.

The Approach and Landing Accident Problem

Approach and landing accidents — collectively referred to as ALAs — represent a disproportionate share of fatal aviation events. Despite accounting for only a small fraction of total flight time, the approach and landing phase consistently generates the majority of hull losses and fatalities in commercial aviation. The Flight Safety Foundation, which has tracked this category for decades, estimates that approach and landing accidents account for more than half of all fatal accidents in commercial aviation when measured across a sustained period.

At the heart of many of these accidents sits a single, identifiable decision point: the moment a crew recognises that the approach is unstabilised — that speed, configuration, descent rate, or flight path do not meet the criteria defined in their Operations Manual — and chooses to continue regardless. This is the continued unstabilised approach, and it is one of the most studied and persistently mismanaged risk events in the industry.

The go-around exists precisely to resolve this situation. It is not an emergency procedure, not an admission of failure, and not an event that requires explanation or justification. It is a normal flight manoeuvre, available to every crew on every approach, that resets the situation and gives the aircraft another opportunity to land safely. And yet, evidence consistently shows that crews resist using it — often right up to the point where the option no longer exists.

What the Data Shows

The go-around rate across commercial aviation is remarkably low — typically cited at between 1 and 4 per 1,000 approaches, depending on operator, aircraft type, and operating environment. On the surface this sounds reassuring. In context, it is alarming. Studies of flight data monitoring programmes consistently show that the rate of unstabilised approaches is significantly higher than the go-around rate — meaning that in the overwhelming majority of cases where a go-around should have been executed, it was not.

Research published by Airbus, drawing on flight data from multiple operators, found that approximately 97 percent of unstabilised approaches resulted in a landing rather than a go-around. Eurocontrol and the Flight Safety Foundation have produced comparable findings. The gap between the frequency of the trigger condition and the frequency of the correct response represents one of the most significant latent risks in commercial flight operations.

Key Statistic

Studies consistently show that the vast majority of unstabilised approaches result in a landing, not a go-around. The gap between the rate at which stabilised approach criteria are breached and the rate at which go-arounds are executed is one of the most significant latent risk indicators in commercial aviation.

The consequences are visible in the accident record. The runway excursion category — aircraft departing the end or side of a runway — is closely linked to continued unstabilised approaches. So is the hard landing category, and a significant proportion of tail-strike events. In many cases, the crew identified the instability. The decision to continue was made consciously. The outcome was not.

Enquire About This Service

Speak to one of our specialists about how AACS can support your organisation.

Standards We Work To

The Human Factors Behind the Decision

Explaining why pilots continue unstabilised approaches requires engaging with some of the most fundamental dynamics in human performance and decision-making under pressure. These are not failures of skill or knowledge. In almost every case, the crew involved understood what a stabilised approach required, and understood that their approach did not meet those criteria. The failure is a failure of decision-making in a specific operational context — and that context is critically important.

Plan continuation bias

Perhaps the most well-documented factor is plan continuation bias — the cognitive tendency to continue with an established plan even when new information suggests that plan should be abandoned. Once a crew is established on the approach, configured, and descending toward a runway they can see, the psychological momentum toward landing is powerful. The plan is to land. The workload of a go-around, the disruption to fuel planning, the need to re-sequence with ATC, and the perceived social cost of explaining the decision all pull in the same direction: continue.

This bias is not a character flaw. It is a documented feature of human cognition under time pressure and task saturation. The approach phase is inherently high-workload, with multiple competing demands on crew attention. At the moment when the go-around decision is most needed, cognitive resources are most constrained. The conditions that create the need for the decision simultaneously impair the capacity to make it.

Commercial pressure and schedule culture

Commercial pressure is a factor that the industry has been reluctant to discuss openly, but which appears consistently in the literature and in the accounts of pilots willing to speak candidly. The pressure to arrive on time, to avoid delays, to protect slot allocations, and to minimise the cost implications of a go-around and subsequent re-approach creates an environment where continuing becomes the path of least resistance — even when the technical case for discontinuing is clear.

This pressure is rarely explicit. Few airlines instruct their crews to continue unstabilised approaches. But the organisational climate— the unspoken norms about what is valued, what is tolerated, and what draws scrutiny — communicates priorities that formal policy does not. A crew that executes a go-around at a slot-constrained airport may face a debrief about the delay. A crew that continues a marginal approach and lands without incident faces no scrutiny at all. The asymmetry of consequences shapes behaviour over time.

The normalisation of deviance

Many pilots who continue unstabilised approaches have done so before, without incident. Each successful landing reinforces the belief that the approach — although technically outside criteria — was manageable. Over time, the criteria themselves come to feel unnecessarily conservative, and the gap between the stabilised approach standard and actual practice widens. This is the normalisation of deviance: the gradual drift away from defined safety standards as repeated non-compliance without consequence makes the deviation feel normal.

The danger is that this normalisation occurs silently, below the threshold of formal reporting, and is invisible to safety management systems unless flight data monitoring programmes are specifically designed and resourced to detect it. An operator can have an impeccable paper-based safety record while a systemic pattern of approach instability accumulates undetected in the flight data.

When Crew Resource Management Fails at the Critical Moment

The go-around decision is rarely a solo act. In multi-crew operations, both pilots are present at the moment the decision is needed. Both have access to the same flight instruments. Both know the stabilised approach criteria. And yet, in accident and incident reports involving continued unstabilised approaches, a consistent pattern emerges: one crew member recognises the problem, and the other — typically the pilot monitoring — does not assert it effectively enough to change the outcome.

This is a Crew Resource Management failure, but one that manifests not in the absence of communication, but in its inadequacy. A first officer who says “we’re a bit fast” when the correct callout is “unstabilised, go around” has technically communicated, but has not challenged effectively. The gradient between captain authority and first officer assertiveness — which decades of CRM training have worked to flatten, with genuine success — still exerts influence at exactly the moments when the stakes are highest.

The Authority Gradient

In several well-documented approach-and-landing accidents, first officers identified the instability and made subdued references to it without triggering a go-around. The capacity to deliver a clear, direct, and unambiguous challenge to a continuing approach — particularly when the captain is flying — remains one of the most difficult CRM skills to train and sustain. It requires not just the knowledge of what to say, but the psychological safety to say it.

Effective go-around culture requires that the pilot monitoring has both the authority and the expectation to call a go-around if the pilot flying does not, and that this callout is treated as a professional obligation rather than an interpersonal challenge. Airlines that have succeeded in shifting this norm report it as a multi-year cultural change programme, not a training module update.

What Effective Organisations Do Differently

Operators that have successfully improved their go-around rates and reduced their unstabilised approach rates share a set of characteristics that go beyond policy documentation. The difference is not in what the Operations Manual says. It is in what the organisation actually does with the data, the debrief, and the culture.

Flight data monitoring as a safety feedback loop

Effective operators use their Flight Data Monitoring programme not merely to detect exceedances after the fact, but to analyse approach stability trends proactively. They establish thresholds not just for individual events, but for rates — tracking what percentage of approaches on a given route, aircraft type, or by a given crew pairing are meeting stabilised approach criteria. When trends emerge, they trigger intervention before an accident forces the issue.

Destigmatising the go-around

The most practically effective cultural intervention is straightforward in principle and demanding in execution: make go-arounds unremarkable. Some operators have introduced explicit recognition programmes for crews who execute go-arounds — not awards, but the simple acknowledgement in a debrief or safety communication that the correct decision was made. Others have restructured their post-flight reporting so that a go-around triggers a positive debrief — “what did you see, what did you do, what was the outcome” — rather than a process that feels investigative.

The goal is to make the go-around the expected response to an unstabilised approach, rather than an exceptional one. This requires consistent and visible leadership behaviour: chief pilots who talk about their own go-arounds, safety managers who share anonymised FDM data showing go-around rates as a positive safety metric, and operations directors who visibly protect crews from commercial pressure when the correct safety decision creates a delay.

Standardised callouts and decision gates

Several operators have reinforced their stabilised approach criteria with explicit decision gate callouts — mandatory verbal confirmations at defined altitudes that the approach is stabilised and the crew is committed to land, or that conditions are not met and a go-around is being initiated. When these callouts are trained, assessed, and enforced consistently, they serve as an external prompt that interrupts plan continuation bias at a structured point in the approach.

The most robust implementations make the callout mandatory in both directions: the pilot flying calls the gate, the pilot monitoring confirms or challenges. This symmetric accountability distributes the go-around decision across the crew rather than concentrating it with the handling pilot, and creates a documented decision point that FDM programmes can verify.

The Regulatory and Safety Management Dimension

Regulators have progressively strengthened their expectations around stabilised approach compliance. ICAO, EASA, and national aviation authorities including the UK CAA have all issued guidance reinforcing the go-around as a normal and expected procedure, and calling on operators to monitor and report go-around rates as a safety performance indicator. The Flight Safety Foundation’s Approach and Landing Accident Reduction toolkit, widely adopted across the industry, provides a structured framework for operator self-assessment.

From a Safety Management System perspective, the continued unstabilised approach presents a particular challenge: it is a risk event that is rarely reported voluntarily, that leaves no mandatory reporting trail unless an exceedance threshold is breached, and that requires proactive FDM analysis to detect at a population level. Operators whose SMS does not include a structured approach to flight data analysis — with dedicated resource, defined thresholds, and a clear feedback loop from data to crew to management — are operating with a significant blind spot in their safety assurance function.

The gap between an SMS that satisfies regulatory documentation requirements and one that actually detects and corrects approach instability trends is precisely the gap that independent audit is designed to identify. An SMS that looks complete on paper but does not surface these patterns in practice is not providing the safety assurance its documentation claims — and regulators, increasingly, know how to look for the difference.

Conclusion

The go-around is not a failure. It is a system working as designed — a crew identifying a condition that does not meet the threshold for a safe landing, and taking the action that aviation’s entire approach safety architecture is built around. The fact that it is so rarely executed in response to the conditions that demand it is not a reflection of poor airmanship. It is a reflection of the human, organisational, and commercial environment in which airmanship is exercised.

Addressing continued unstabilised approaches requires more than policy reinforcement. It requires operators to examine their FDM data honestly, to build cultures in which the go-around is normalised and celebrated rather than quietly avoided, to train CRM in a way that gives first officers the psychological safety to challenge effectively, and to ensure that the commercial pressures of a competitive industry do not silently override the safety standards that the industry publicly commits to.

The runway at the end of an unstabilised approach will still be there on the next circuit. The consequences of pressing on when the picture is not right will not always be recoverable. Go around.