ATM teams under the gun.
ON A WARM AND SUNNY TUESDAY AFTERNOON AT 2:30 pm bank employees Bob and Karen went to a remote kiosk automatic teller machine (ATM) to make the afternoon deposit pickup. As they approached the kiosk, they saw a lone white male eating a hamburger near the curb by the ATM. The man then approached them wearing a maroon, turtleneck sweater partially over his face, with his hand in a McDonald's paper bag. The man showed Bob and Karen what appeared to be the handle of a gun. He then said, "I have a gun in the bag. Give me the money. Hurry up, hurry up, drop it. The safety latch is off." He then told the team to open the door to the ATM. Bob and Karen dropped the empty bag as they went into the kiosk, slammed the door, and hid behind the ATM. The suspect fled on a motorcycle.
In February 1987, ATM service technician William Massengale, 40 years old, was dispatched late at night to repair a remote ATM that was later discovered to have been jammed with pieces of paper. At 11:55 pm Massengale radioed bank security personnel, telling them he was ready to leave the ATM. Moments later, an alarm from the ATM went off at the bank. A second bank employee was dispatched to the ATM and found the company car Massengale had been driving, with the keys and Massengale's jacket in it. Massengale was found several miles away from the ATM the next morning, dead of multiple gunshot wounds. The Atlanta Constitution reported that more than $50,000 was missing from the ATM.
In October 1986, career bank employee Carolyn McComas was about to perform the daily service on a shopping center kiosk ATM when she was met by a man who had gotton out of a nearby van with the words "Burglar Systems" painted on the side. McComas set off a silent alarm shortly after entering the service side of the ATM. When police arrived about 10 minutes later they found McComas gagged, bound hand and foot, and shot twice in the head. She died shortly thereafter. The man suspected of murdering her fled the scene in the "Burglar Systems" van driven by a second male. No cash was lost in this attack, and no suspects have been apprehended to date.
These three actual incidents(1) illustrate the risk of attacks on ATM service teams in America today. Thus far, one ATM servicing expert says, American financial institutions have led charmed lives with the relatively few attacks that have been reported on ATM service teams.
In 1989, more then 70,000 ATMs are in service in the United States. Using a very conservative service rate of three times per ATM per week, more than 11 million ATM service runs will be completed in America this year. Spencer Nilson, in his Nilson Report, estimated that more than $6 million passed through each ATM in America in 1986 in the form of cash dispensed and deposits accepted. At that rate, American ATM service teams will transport more than $420 billion to and from US ATMs in 1989.
Among the criminal element, news of the relatively low risk and relatively high yield of ATM service team attacks is spreading. News media reports of the dollar amounts lost in some ATM service team attacks can only encourage the planning of new attacks. With nearly 7,000 robbers willing to attack well-fortified bank lobbies in America in 1987 for an average take of less than $6,500, it comes as no surprise that most ATM security experts predict a dramatic rise in attacks on the much more vulnerable ATM service teams, which carry tens of thousands of dollars in cash to and from ATM.
Bank Administration Institute (BAI) ATM Crime Task Force Chairperson Brian McGinley predicts that by the year 2000 "approximately 20 employees per year will be killed or injured as a result of criminal attacks on [ATM] service teams. Average dollar loss per attack will approximate $70,000. Expect additional employee litigation against their employers for negligent security practices." This article reviews current data on ATM service team attacks and makes seven ATM service team attack prevention recommendations.
In 1987, BAI released the results of its national survey of ATM-related crimes of violence. This survey included information from 177 of America's largest financial institution ATM operators. This survey led to 15 detailed reports of ATM service team attacks using the form found in Exhibit 1. These 15 reports represent the largest published data base of information about ATM service team attacks in the world. [Tabular data Omitted]
These reports represent ATM service team attacks known to banks since January 1982. The armored service industry in America is aware of many more attacks on ATM service teams. However, in recent years, reportedly due to competitive pressures, the largest American armored service providers have not released information about their ATM service team attacks. Thus, these 15 detailed reports from America's largest financial institution ATM operators represent the best available data about ATM service team attacks.
THE DETAILED "VIOLENCE AGAINST ATM Service Team Report" forms reveal a number of characteristics. (See Exhibit 2.)
Table : Exhibit 2 Year of Reported Attacks
Number of reported Year ATM service team attacks 1982 1 1983 0 1984 2 1985 3 1986 4 1987 5
The steadily increasing number of attacks probably reflects not only an increase in the real number of these attacks but also better record keeping and better report availability in recent years. The following are some findings made from the reports:
Time of month. Interestingly, all 14 of the service team attacks reported with a date occurred in either the third week of the month (seven attacks) or the fourth (seven attacks). The significance of this trend (if any) is not known.
ATM servicers have registered a great deal of concern about late-night, remote ATM attacks that robbers set up by purposely damaging or disabling ATMs. The data from this study show that only 21 percent of the attacks occurred during the hours of darkness. Furthermore, none of the reported attacks on ATM service teams were between midnight and 8:00 am. Nearly 80 percent of the attacks occurred during regular business hours (between 8:00 am and 4:00 pm), with the majority (57 percent) occuring between noon and 4:00 pm. (See Exhibit 3.)
Table : Exhibit 3 Time of Attacks
Number of Percentage Time block incidents of incidents
4:00 am 0 0%
4:00 am to
8:00 am 0 0%
8:00 am to
noon 3 21.4% noon to
4:00 pm 8 57.1%
4:00 pm to
8:00 pm 0 0%
8:00 pm to
midnight 3 21.4% 14 100%
ATM service team activity. At the time of the attack, 64 percent (or nine) of the service teams were doing "routine ATM cash replenishment and deposit pickup." This further confirms the trend of the robbers preferring to ambush ATM service teams on their regular runs. Only 21 percent (three attacks) involved a "special ATM service run (to clear machine jam, answer alarm, etc.)." In the remaining two incidents, service teams were engaged in other ATM service.
Location of the service team. Exhibit 4 shows the data describing where the ATM service teams were when the attacks began. Two thirds of the attacks began when the ATM service teams were beginning their work at an ATM - between the time they left their vehicle and entered the ATM service area. Only two attacks began when the service team was returning from the ATM to the service vehicle. In only two incidents was the vehicle involved with the attack - once when the team was in the vehicle, and once when the team was not.
Table : Exhibit 4
ATM Service Team
ATM service Number of Percentage team was incidents of incidents
in its vehicle
closed 1 7% leaving
an ATM on
foot 4 27% entering or
service area 6 40% returning
on foot or
vehicle 2 13% not in vehicle
attacked 1 7% unknown 1 7%
Type of attack. Nearly 75 percent of the attacks involved ambushing the ATM service team outside their service vehicle. Two of the reported robberies involved an attack on the vehicle, and one attack involved "pretext or deception."
Perpetrators. In the 15 reported attacks, 18 perpetrators were reported; all of them were male. A breakdown of the race and age characteristics of the perpetrators is listed in Exhibit 5.
Table : Exhibit 5 Race and Age of Perpetrators
Race of perpetrators
White 10 Black 2 Hispanic 2 Other 1 Unknown 3
Age of perpetrators
Less than 20 years 1 20-29 years 9 30-39 years 1 40-49 years 1 50 years or older 0 Unknown 6
ATM service team robbers were predominantly white, 20- to 29-year-old males. By contrast, ATM customer robbers were identified in a 1986 BAI survey as being predominantly black, 20- to 29-year-old males.
Level of threat. Several levels of violence were used in the attacks on ATM service teams, sometimes in combination with one another. (See Exhibit 6.)
Table : Exhibit 6 Type of Violence Used in Attacks
Type of Number of Percentage violence incidents of incidents
weapon(s) 11 73%
assault 4 27%
only 3 20% Abduction 1 7% Other 1 7%
Weapons were used in nearly three fourths of the attacks; however, simple assaults (such as pushing down two unarmed tellers) were sufficient to gain compliance in 27 percent of the incidents. Verbal threats alone were used in 20 percent of the 15 ATM service team robberies. The one abduction was of a single ATM service person who was found shot dead.
The most common type of weapon used in these ATM service team attacks was a handgun, which was used in 67 percent (or 10) of the incidents. Cutting or blunt weapons such as a knife or a club were used in 13 percent (or two) of the incidents. Handguns were the obvious favorite of ATM service team robbers, probably for their ease of concealment, lethality, and availability. No shoulder weapons or explosives were used in the 15 reported attacks.
Loot taken. Loot was taken in two thirds of the 15 reported ATM service team robberies - five times money was taken from the ATM machine, twice it was taken from the service vehicle, and in two other incidents noncash "other property" was stolen. The average cash loss in the incidents where cash was taken was in the low five figures.
Personal property of ATM service team members (such as a weapon, wallet, or ATM keys) was stolen in two incidents. ATM deposits and ATM cards were stolen in one incident, and the service team's vehicle was stolen in one incident.
THE FOLLOWING SECTIONS COMPARE characteristics of the 15 victimized ATM service teams with the typical ATM service teams, service vehicles, and ATM locations as described by the 177 respondents using the BAI ATM survey form in Exhibit [Tabular Data Ommitted] 7.
Size of ATM service team. Exhibit 8 illustrates the differences in size of typical ATM service teams compared to the 15 victimized teams. Attacks on single-person service teams (33 percent) occurred much more often than single-person teams were used (14 percent). Also, attacks on three-person service teams (7 percent) occurred much less often than they were used (16 percent). Single-person service teams were victimized relatively more often than either two-person or three-person teams. suggesting that service team size does indeed play a role in ATM service team robbers' selection of targets. [Tabular data Omitted]
Composition of service team. The 15 victimized teams were composed of 26 team members including 14 bank tellers, two bank security guards, three contract ATM service employees, three armored carrier or contract security guards, and four other personnel. Exhibit 9 compares the percentage of different types of personnel on the 15 victimized teams to the percentage used on typical teams as reported by 177 financial institutions.
Bank tellers appeared more often on victimized teams (41 percent) than they did on typical teams (34 percent). On the other hand, armored carrier or contract security guards appeared less often on victimized teams (12 percent) than on typical teams (17 percent). In addition, victimized service teams were slightly less likely to have bank security guards, contract ATM service employees, or police officers on them than were typical service teams. This suggests that ATM service team robbers may prefer attacking service teams with bank tellers on them rather than service teams that are armed.
Use and characteristics of guards. About half (49 percent) of the typical service teams of the 177 financial institutions reported using guards or police officers. Virtually the same percentage (47 percent) of the victimized ATM service teams used some sort of guard or police officer. Exhibit 10 compares the characteristics of guards from both groups.
Despite the small size of the victimized ATM service team sample, some trends may still be observed. Guards on victimized ATM service teams were more likely to have been unarmed and in plain clothes, but also were more likely to have had portable two-way radios and to have received periodic ATM security training. Guards who were unarmed and in plain clothes were victimized more often than their uniformed and armed counterparts. The appearance of authority and firepower appear to be important deterrents to ATM service team robbers.
Injury to service team members. Of 22 victimized service team members reported on, 18 had no physical injury. Two tellers on one team suffered minor injuries when they were pushed during a simple assault robbery. Two single-person service team members were shot and killed in separate incidents. Both murders appears to be execution-style slayings designed to prevent later identification of the perpetrators. These findings clearly support the advantage of multiperson service teams for team member safety.
ATM service team vehicle. The service vehicles used by 10 victimized ATM service teams were compared to those of 144 typical ATM service teams. This comparison showed little difference between the use of unarmored and armored service vehicles (71 percent unarmored on victimized teams, 75 percent unarmored on typical service teams). Victimized teams did show a higher rate of using unmarked vehicles (100 percent) than did typical teams (78 percent). Interestingly, victimized ATM service teams' vehicles were equipped with two-way radios 70 percent of the time, more often than typical service teams (53 percent).
Crime site characteristics. The fact that 87 percent of ATM service team attacks occurred in medium-sized or large cities comes as no surprise because of the BAI study's focus on America's largest ATM operators. However, the fact that 7 percent occurred in suburban communities and 7 percent occurred in small cities or towns is of interest. (See Exhibit 11.)
Table : Exhibit 11 City and Neighborhood of Attacked ATMs
Large city 60% Medium city 27% Suburban community 7% Small city or town 7%
Shopping center or mall 53% Business district 27%
Retail, grocery, or
convenience store 13% Other 7%
Of special significance is the fact that 53 percent of the attacks occurred in shopping centers or malls, with 13 percent occurring in other nonbank locations such as retail, grocery, or convenience stores. Clearly ATM robbers are preferring to attack service teams in off-premises locations, especially in shopping centers or malls. Only one fourth of the ATM service team attacks occurred in business districts. That is an important finding of this research. Further details of the ATMs where service teams were attacked are given in Exhibit 12.
Table : Exhibit 12 Attack Site ATM Characteristics
Off Bank premises 73% On bank premises 27% Walk-up 93% Drive-up 7%
Public access to ATM Outside a building with
no vestibule 64%
Outside a building with
free-access vestibule 14%
Outside a building with
card-access vestibule 14% Inside a building 7%
Service team access Freestanding ATM enclosed
service area 71% Through the wall 21%
Freestanding ATM without
an enclosed service area 7%
The rate of attack at walk-up and drive-up ATMs almost exactly matches the percentage of walk-up and drive-up ATMs in service in America. ATM robbers again showed a strong preference for off-premises ATMs (73 percent) over ATMs on bank premises (27 percent). Robbers chose exterior ATMs in more than 90 percent of the attacks. Finally, 78 percent of the service team attacks occurred near freestanding ATMs, again underscoring the strong preference of ATM robbers to have quick access to the service team and quick access to escape.
Available security devices and practices. In the 15 service team robberies, 60 percent (nine) of the ATM crime sites had neither transaction nor surveillance cameras. A total of 27 percent (four sites) had transaction cameras, while 20 percent (three sites) had surveillance cameras. Two of the 15 sites had both transaction and surveillance cameras. (See Exhibit 13.) [Tabular Data Omitted]
Transaction cameras are generally mounted inside the ATM and are designed to get a head-and-shoulders picture of the person inserting the ATM card. There were fewer transaction cameras used at ATM service team attack sites (27 percent) than generally installed at the typical ATM sites (44 percent). Because most freestanding ATMs are entered from the side or rear, this probably was not a major factor in the ATM robbers' selection of a crime site.
However, there were fewer ATM surveillance cameras (generally mounted close to the ceiling 10 or 15 ft. away from the ATM) used at victimized ATM sites (20 percent) than at typical ATM sites (28 percent). The ATM surveillance camera has a much broader field of view than the transaction camera and in some cases could get a picture of people entering or loitering near the ATM service door.
Service team alarms. Several types of service team-activated alarms were available at the 15 ATM robbery sites. (See Exhibit 14.)
Table : Exhibit 14 Service Team-Activated Security Devices at ATM Crime Sites
Hidden duress button inside
service area or safe 64% Security telephone 57%
Portable duress activator 29%
Portable two-way radios 14%
Audio monitoring by central
security guard 14%
An impressive variety of alarm devices was available to the attacked service teams, including hidden duress buttons and portable duress activators, which have a good opportunity to be used under attack. The actual rate at which these devices were used could not determined from this research.
It was possible to compare victimized service teams with typical ATM service teams on one variable - the use of portable two-way radios. A total of 63 percent of typical ATM service team guards had portable two-way radios, while just 14 percent of the victimized teams had such radios. The lack of communication while on foot outside the service vehicle appears to be an important factor in the selection of victimized service teams.
IF A PROFILE OF THE MOST COMMON characteristics of the 15 victimized ATM service teams were to be created, it would show a single white male between the ages of 20 and 29 using a handgun and attacking a two-person service team, typically bank tellers, on a routine cash replenishment and deposit pickup between noon and 4:00 pm. The robber would hit the team as it left its unmarked, unarmored vehicle or while it was entering the ATM service area and had no radio contact with security personnel. The ATM would be an off-premises, walk-up ATM that was freestanding outside a building. The ATM would be in a shopping mall and would have neither transaction nor surveillance cameras. Most often the robber would get away with some loot.
The problem of ATM service team robberies is here to stay. Brian McGinley of BAI predicts that "attacks on ATM service teams will be a major problem through the 1990s as financial institutions join the ATM bandwagon but fail to allocate the resources to provide adequate security measures for their servicing personnel."
What can be done to minimize the risk of making ATM service team an attack target? The following are a few suggestions:
* The visibility of the ATMs to the surrounding area must be maximized. All barriers that block visual access of the ATM from the nearest sidewalk, street, or pedestrian walkway must be minimized. Such barriers include shrubbery, landscaping, tinted glass, curtains or blinds, and signs.
* Minimum ATM lighting standards must be established. Consideration should be given to adequate lighting for the immediate ATM surroundings and the nearest parking area and pathway to the ATM. Desired lighting levels must be maintained by light meter or specified fixture output and must be regularly monitored at night for burned-out bulbs or other changing conditions around the ATM.
* Strong consideration should be given to the uniform use of wide-angle transaction cameras. Special consideration may be given to continuous-operation transaction cameras or surveillance cameras. Consideration may also be given to installing a camera on the inside of the ATM service area, with remote monitoring and recording of the images inside the service area. This arrangement denies ATM robbers as well as ATM burglars and dishonest employees the privacy of the ATM service area to complete their crimes.
* Both regular-hours and after-hours servicing and replenishment of ATMs (especially those off-site) should be done by a service team of two persons to minimize the risk of crimes of violence to the team and crimes of fraud.
* ATM servicing policies, including robbery policies, should be written down and taught to all ATM service team members.
* ATM service teams should receive preservice and periodic in-service training on ATM security issues and practices. Strong consideration should be given to equipping proprietary ATM service teams with portable two-way radios or other alarm devices. Consideration may also be given to using appropriately trained and equipped third-party servicers. All reports of ATM service team-related crime must be thoroughly documented, perhaps by using the BAI "Violence Against ATM Service Team Report." All ATM crimes should be tracked by a central person or office in each financial institution. International communication and reporting of serious ATM service team incidents and corrective actions should also be directed to corporate legal counsel.
* ATM service team attack information must be shared between branches of financial institutions in the same city or region and between financial institution security officers and local law enforcement agencies.
Only with the collection and distribution of ATM service team attack information can appropriate and effective countermeasures be identified and the risk of an ATM service team's becoming an attack target be minimized.
About the Author . . . F. Barry Schreiber, PhD, is a professor of criminal justice at St. Clound State University in St. Cloud, MN. He is a member of ASIS.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||violence against automatic teller machine service personnel|
|Author:||Schreiber, F. Barry|
|Date:||Sep 1, 1989|
|Previous Article:||To arm or not to arm.|
|Next Article:||Who you gonna call? Fraudbusters!|