Using Amber Alerts and Canada’s Alert Ready System for Missing Children

Recently, a missing child was found deceased after a two-week search in Alberta, Canada.

The child was 11 years old, with autism, and was not abducted. An amber alert exemption was issued (this case would not normally qualify) two days after they were reported missing.

Discussion has begun around whether notification systems should be changed to allow a widespread emergency message to be sent to the public in the future to help with these events.

There are three predominant proposals:

1. Put these events through the amber alert system by expanding its scope to include non-abductions.

2. Integrate this in a different notification within Canada’s Alert Ready system.

3. Create or augment a separate system to these two systems.

There are several arguments for and against each of these, mostly surrounding the benefit of increased messaging to find missing persons against the risk of alert fatigue. Before getting into this, context around these different systems is required.

The Amber Alert System

Amber alerts first came about in response to the abduction of Amber Rene Hagerman in Arlington, Texas on January 13, 1996.  AMBER technically stands for Americas Missing: Broadcast Emergency Response, but Canada and other countries have adopted similar policies, often with the same shortened name.

Here is a rough breakdown as to how amber alerts are issued in Canada.

1. Law enforcement (police) receive a report that meets specific criteria. The criteria being:

- Police have confirmed an abduction has taken place.

- Victim is under the age of 18 or has a proven physical/mental disability.

- There is reason to believe the victim is in danger of serious physical injury.

- There is information available that, if broadcast to the public, could assist in the safe recovery of the victim.

2. Law enforcement sends a request for the alert to be issued (usually a separate agency).

3. Request is verified and content is drafted.

4. Request is sent to NAAD (National Alert Aggregation and Dissemination), which powers the Alert Ready system. 

5. Transmission: The system then formats the data and instantly pushes it to cell phones, television networks, and radio stations. Live broadcasts are instantly interrupted and the message is delivered.

6. Follow-up and eventual cancellation of alert.

On the Alert Ready System

As the amber alert is broadcasted through the Alert Ready system, and the second proposal involved the Alert Ready system directly, this become fairly central to the discussion.

The Alert Ready system is Canada’s emergency management notification system. It has the capability to override most broadcasts and personal communications devices effectively forcing a message through to the public.

It is a system generally reserved to severe threats to public safety where there is an imminent danger to the public.

The NAAD system is operated by Pelmorex Corp (the Weather Network’s parent company) and credentials to issue alerts are highly restricted.

There is a general breakdown of what sorts of events are eligible for being issued, but these basically fall into four types:

1. Any hazard that if left unchecked could result in something contained in the CDD (Canadian Disaster Database). This is natural, technological, human caused, national security, terrorism, etc. Some of those are technically subcategories of the others.

2. Localized imminent threats of significant public danger, an example would be dangerous animals.

3. 9-1-1 or other emergency service outages.

4. Amber Alerts.

The system is capable of restricting the alerts to specific geographic areas, scaling the message more broadly or narrowly as the situation demands.

Other Alerting Systems

In the event of a missing child, other systems currently exist but the situations are considered less severe in comparison to Alert Ready style notifications. 

Two that come to mind immediately are vulnerable persons registries and the Child Search Network.

Vulnerable persons registries are a mitigation and preparedness risk control where family and caregivers record said vulnerable person in a database. In the event they go missing, their information can more quickly be accessed and distributed to law enforcement that will be primarily the response agency trying to help find the person.

The Child Search Network (operated by the Missing Children Society of Canada) helps police distribute details to targeted areas to engage the public with helping with a search.

Police will often also do social media blasts in the event of a missing person, especially a case of a high-risk child.

Analysis

Emergency notification is a massive component of emergency management. This issue cuts to the core of this profession and these suggestions directly involve influencing one of its most powerful tools in Canada.

As mentioned in the intro, the main debate is being centered around alert fatigue versus the benefits of increased public awareness.

The problem is fairly similar to tale ‘the Boy who Cried Wolf’. Most will be familiar with the story, but for those unaware:

A boy pranks the locals repeatedly by yelling “wolf” when one does not appear.

Eventually a wolf does appear and the boy yells “wolf”.

No one believes him and the wolf kills the boy.

In this case of alert fatigue the issue is not lying, the issue is an alert that is not relevant to the recipient. Every time an alert is sent and it does not affect the recipient, it creates a case where the individual may perceive that their ‘time was wasted’ or they were ‘unnecessarily annoyed’ by the alert.

Amber alerts have already come under some controversy in the past with members of the public complaining and police spokespeople responding to these complaints with “We won’t apologize for using all the tools available to us to find a missing child”.

The risk reality remains however, that people may increasingly find a way to override their devices or will just ignore alerts altogether.

Amber Alerts, Alert Ready test notifications, and a Tornado forming above one’s house all have the same audio cue, interruption, and text format. There is no significant distinction between these messages to the average layperson before the specific message is read. The alert ready system overrides silent mode, volume settings, and is designed to be extremely obnoxious to get the recipient to read the message.

It is essentially the ‘nuclear option’ for sending a message to the public.

From the perspective of emergency management professionals concerned with alert fatigue:

The main risk is that someone ignores an imminent shelter in place or evacuation warning because they are sick of hearing amber alerts.

From the perspective of someone who sees amber alerts as a net-benefit:

If an individual ignores alerts and dies because the alerts were annoying, that is their fault and they have effectively won a ‘Darwin award’.

My Perspective

When analyzing this situation, there are two factors that come to mind initially:

1. The size of the issue.

2. Opportunity Cost (broken down into following components)

Solving issue against other issues negatively impacted.

Erosion of effective resource (alert fatigue)

Financial cost

To start, if the issue is too small to reasonably justify a massive federal resource, then in practice the risk control does not make sense, it would be out of scope.

Looking at missing persons data: 

~67,000 – 68,000 people are reported missing every year in Canada. 

~88% are found within 7 days.

There is a cumulative total of over 8,300 long-term open missing persons cases from all previous years combined.

There are 40-100 new cases of recovered and unidentified remains every year.

The exact numbers are not publicly known, but with a bit of approximation math:

Canada reports roughly 34,000 missing children annually.

About 99% of cases are successfully resolved within days or weeks. Overwhelmingly involving family disputes, runaways, or the children returning safely.

This leaves 300-500 missing cases children per year that are unresolved for longer than several months.

When looking at how long-term incidents are categorized when the cases are closed upon the discovery of remains: 23% are categorized as accidents, and 9% are categorized as presumed dead.

These numbers are not separated out by how many cases involved amber alerts and do not.

Assuming all of these cases were not amber alerts, and going with the most extreme number, the upper limit would be approximately 160 children per year.

This number assumes:

• Amber alerts were never issued in these cases 

• All of these cases would not have qualified as amber alerts, currently 23% + 9%

• Every unresolved child case every year was a death.

• The upper 500 children number is used.

• Every case where an amber alert wasn’t issued, if it was the child would have been saved.

More explicitly: Assuming the best-case position for saving children if amber alerts of the alert ready system was used, this would amount to 160 children per year.

To be clear, it is extremely unlikely this number is actually this high, but given how difficult it is to find out this information, at least an upper boundary can be somewhat determined.

In many countries, this number would be far too small to consider within scope, but in the case of Canada, with only ~80,000 approximate avoidable deaths, perhaps this could be a big enough impact to justify change.

Onto the opportunity cost:

Allocating resources towards missing persons earlier in the case does not appear to have an immediate direct negative effect on other risks. There is also the potential benefit that every resolved case also reduces the number of long-term open cases, which should be more efficient for police resources.

Given any update to the system would have credential requirements for Alert Ready, and the police already need to take in the report regardless, any additional administrative work should be fairly simple to navigate.

The alert fatigue issue explored earlier is still a present concern, and colleagues in emergency management are probably correct to caution here. This is a serious concern. Important to note, the data does not exist currently as to how many Canadians are already fatigued and how big of a risk an expansion into there would present. 

Emergency management professions can be fairly risk averse and tend to like applying the precautionary principle, so any solution involving Alert Ready will be problematic from this perspective.

Financial cost will likely not be a significant issue. The infrastructure either already exists or can be mapped onto pre-existing systems in most cases.

The purpose of Alert Ready

The other issue that is not really being talked about directly is the purpose of Alert Ready.

Alert Ready is fundamentally a public safety utility, not a personal safety utility. It is supposed to be used to protect the public at large.

A tornado coming down in a neighbourhood is understandably considered a threat to said neighbourhood. Storm activity over a particular area that risks these things are understandable reasons for alerts.

Amber alerts arguably are informing on a threat to the public as the individual who abducted the child may be erratic or threaten members of the public. It is not necessarily only a private matter between a family the police are working to resolve.

A missing person that is expected to be violent or erratic would also qualify here, not because they are missing, but because they could endanger the pubic.

But a missing vulnerable person that is at risk of accidental death, but not a threat to the public is only practically risking themselves. It seems disproportionate to activate a public safety resource for something that is not a public safety risk.

Perhaps it would be sensible to have tiered alert systems to separate these things, but as it stands if all alerts are equally intrusive, the vulnerable child that is no threat to the public would be competing with a nuclear bomb for the publics attention with Alert Ready and amber alerts.

While I appreciate that this may come across as callous, the reality is that the system responsible for protecting the public from disasters is out of scope for the life of a single individual, at least in its current iteration.

Separate types of audio notifications for different levels of severity could address this. A look at pre-existing systems like the Child Search Network might also be productive. At an initial glance, there should be a reasonable solution that satisfies the risks on all angles for this one.

Learn more and watch the video here.

Vezina is the CEO of Prepared Canada Corp. and is the author of Continuity 101. He can be reached at info@prepared.ca.