The Blackspot Paradox: Why We Wait For Crashes Before We Act

Abstract

Road safety investment is predominantly guided by historical crash records. While crash-based prioritisation frameworks have delivered significant reductions in fatalities and serious injuries, they remain fundamentally retrospective. Interventions are frequently triggered only after sufficient crash history has accumulated to justify engineering treatment.

This creates a structural limitation within the decision-making process. Locations exhibiting elevated risk but limited crash history may remain difficult to prioritise despite observable unsafe interactions between road users.

This paper examines the relationship between crash occurrence and risk exposure, outlines the limitations of relying solely on crash statistics for safety prioritisation, and discusses the role of surrogate safety measures as a complementary framework for identifying emerging risk.

Introduction

Road safety investment is inherently a prioritisation problem. Road authorities operate within finite funding environments and must determine where intervention will produce the greatest reduction in risk.

Historically, crash records have formed the foundation of this process. Fatalities, serious injuries and reported crashes provide objective and auditable measures of safety performance. Locations demonstrating elevated crash frequencies can be identified, investigated and treated through established engineering processes.

This methodology has delivered substantial benefits and remains a cornerstone of contemporary road safety practice.

However, crash records represent realised outcomes rather than underlying risk. They describe where failures have occurred, but they do not necessarily describe where failures are most likely to occur next.

As a result, many safety programs remain fundamentally retrospective. Interventions are often prioritised only after sufficient evidence of harm has accumulated.

This creates a challenge for agencies seeking to adopt more proactive approaches to safety management.

Crash Frequency as a Lagging Indicator

From an engineering perspective, a crash should be viewed as the outcome of a sequence of interacting operational conditions.

These conditions may include:

• Traffic demand
• Road geometry
• Signal operation
• Visibility constraints
• User behaviour
• Environmental factors
• Network performance

The collision itself is not the source of risk. Rather, it is the manifestation of risk within a specific set of operating conditions.

Consequently, crash frequency is best understood as a lagging indicator of safety performance.

The absence of crashes does not necessarily indicate that a location is operating safely. Similarly, the presence of crashes does not necessarily indicate that risk has recently emerged.

In many cases, the conditions contributing to collisions may have existed for years before a statistically significant crash history developed.

This distinction is particularly important at locations experiencing changing traffic patterns, emerging land use pressures or increasing vulnerable road user activity.

By the time a location appears within a blackspot program, the underlying risk may already be well established.

The Blackspot Paradox

Blackspot programs have been highly effective in directing investment toward locations with demonstrated safety deficiencies.

However, they create an inherent paradox.

To justify intervention, evidence of risk is required.

The most widely accepted evidence remains crash history.

Yet crash history is generated only after road users have already been exposed to that risk.

As a result, locations exhibiting hazardous operational characteristics may remain difficult to prioritise until those characteristics eventually result in recordable crashes.

This is not a criticism of blackspot methodologies. Rather, it reflects the limitations of the data used to support decision making.

The framework is optimised to identify locations where harm has occurred.

It is less effective at identifying locations where harm is likely to occur in the future.

The consequence is that many safety programs remain reactive by design.

Understanding Risk Exposure

A central challenge in modern road safety practice is distinguishing between risk and outcome.

Crash records measure outcomes.

Risk exposure describes the conditions from which those outcomes emerge.

This distinction is well understood in other areas of transport engineering.

Congestion management relies on indicators such as delay, occupancy and queue length rather than waiting for network failure.

Pavement management relies on deterioration indicators rather than waiting for complete asset failure.

Bridge management relies on structural condition assessments rather than waiting for collapse.

In each case, agencies monitor indicators that precede failure.

Road safety is increasingly moving toward a similar model.

Rather than relying exclusively on collision outcomes, engineers are beginning to measure the interactions and behaviours that occur before collisions take place.

This provides a more direct understanding of operational risk.

The Emergence of Surrogate Safety Measures

Over the past two decades, surrogate safety measures have become an increasingly important area of road safety research and practice.

These measures attempt to quantify risk through observation of road user interactions rather than collision outcomes.

Common examples include:

• Time to Collision (TTC)
• Post Encroachment Time (PET)
• Deceleration Rate to Avoid Collision (DRAC)
• Conflict frequency
• Conflict severity
• Gap acceptance behaviour
• Exposure rates

Unlike crash statistics, these measures can be collected continuously and at significantly higher frequencies.

A signalised intersection may experience a small number of reported crashes over several years while generating thousands of measurable interactions between vehicles, pedestrians and cyclists during the same period.

This provides a substantially richer dataset for understanding operational performance.

Importantly, surrogate safety measures should not be viewed as replacements for crash analysis.

Crash outcomes remain the definitive measure of safety performance.

However, conflict analysis provides visibility of the conditions that exist prior to collision occurrence and therefore offers the potential for earlier intervention.

Implications for Safety Prioritisation

The integration of exposure-based safety measures has several implications for transport agencies.

First, it enables identification of locations exhibiting elevated risk before statistically significant crash histories develop.

Second, it allows interventions to be evaluated using larger and more responsive datasets than crash records alone.

Third, it provides a mechanism for assessing risk involving vulnerable road users, particularly in environments where serious incidents remain relatively rare but the consequences are potentially severe.

Applications include:

• School zone assessments
• Pedestrian crossing evaluations
• Active transport corridors
• Signalised intersections
• High-turning-movement environments
• Emerging development precincts

In each case, exposure-based measures provide insight into risk that may not yet be visible through traditional crash analysis.

From Blackspots to Risk Hotspots

As infrastructure becomes increasingly capable of observing and analysing road user behaviour, opportunities exist to broaden the scope of safety assessment.

Historically, road authorities have focused on identifying blackspots — locations where crashes have already occurred.

Emerging technologies now enable the identification of what may be described as risk hotspots — locations where measurable indicators suggest elevated risk despite limited crash history.

This represents a shift from outcome-based prioritisation toward risk-based prioritisation.

Rather than relying solely on historical evidence of harm, agencies can incorporate leading indicators into the decision-making process.

The question evolves from:

Where have crashes occurred?

to:

Where is risk already measurable?

This distinction is fundamental to the development of more proactive safety frameworks.

Conclusion

Crash-based analysis remains an essential component of road safety engineering and will continue to play a central role in safety investment and prioritisation.

However, crash frequency should be recognised as a lagging indicator of network safety performance.

As infrastructure becomes increasingly capable of measuring road user interactions, opportunities exist to complement traditional blackspot methodologies with exposure-based indicators that provide earlier visibility of risk.

The objective is not to replace crash analysis.

The objective is to broaden the evidence base available to engineers.

By integrating surrogate safety measures alongside traditional crash data, transport agencies can move toward a more proactive model of safety management—one that identifies emerging risk, supports earlier intervention and better aligns with the long-term objectives of Vision Zero and Safe System frameworks.