Synchronization of Traffic Signals as a Means of Reducing

Red-Light Running

David Shinar, Muki Bourla, and Liat Kaufman, Ben Gurion University of the Negev, Beer-
Sheva, Israel

The goal of this study was to determine if traffic signals that are synchronized along
a route are associated with fewer red-light violations than traffic signals that are not
synchronized. A total 3600 cycles of traffic signals at 12 intersections along 2
major urban arteries were observed. Synchronized intersections were effective in
reducing the likelihood of red-light running (RLR) by (a) providing fewer oppor-
tunities than nonsynchronized intersections for RLR and (b) having a lower rate
of RLR relative to the number of opportunities. After adjustment for the number of
opportunities, the odds of entering the intersections in red in synchronized inter-
sections were nearly 1/7 the odds of RLR in nonsynchronized intersections.
Congestion reduced the effectiveness of synchronized intersections relative to non-
synchronized intersections. Male drivers were slightly more likely to run red lights
than female drivers, and the effects of synchronization were fairly constant across
age, gender, and the presence or absence of passengers. Actual or potential appli-
cations of this research involve signal synchronization to reduce aggressive driving
in general and RLR in particular.

INTRODUCTION ward a goal. In the context of driving, the goal

Surveys of British and American motorists in-
dicate that aggressive driving is one of the driv-
er’s greatest concerns as far as traffic safety is
involved (AAA Foundation for Traffic Safety,
1997; National Highway Traffic Safety Admin-
istration, 1998). Although efforts to identify “ag-
gressive drivers” have been made (e.g., Retting
& Williams, 1996), few efforts have attempted to
understand the phenomenon, measure it empir-
ically, and suggest ways of coping with it (other
than enforcement; for a review see Shinar,1998).
In fact, even the definition of aggressive driving
is controversial (Shinar). However, if aggressive
driving is considered to be another manifesta-
tion of aggression, there is ample psychological
theory and data to explain the phenomenon and
to suggest ways of dealing with it.
For the purpose of the present study, we de-
fined aggressive driving as behavior that results
from encountering an obstacle in the path to-
is typically to get somewhere with a minimum
of delays. The frustrating event or object may
be a slow-moving or stopped vehicle, a traffic
jam, a red light, a lane closure because of con-
struction, and so forth. Viewed from the per-
spective of the frustration-aggression model
(Dollard, Doob, Miller, Mowrer, & Sears, 1939),
the most effective means of reducing on-road
aggression would be to eliminate the sources of
frustration.
Although highway designers and traffic engi-
neers constantly work on improving roadways,
increasing their capacity, and eliminating or at
least reducing gridlocks, it is reasonable to hy-
pothesize that in the process of doing so, their
efforts should also reduce aggressive driving,
given that the primary frustration (of being
stopped in traffic) will be reduced or eliminated.
The purpose of this study was to test whether
one such traffic improvement reduces one form
of aggressive driving.

Address correspondence to David Shinar, Department of Industrial Engineering and Management, Ben Gurion University
of the Negev, Beer Sheva 84105, Israel; shinar@bgumail.bgu.ac.il. HUMAN FACTORS, Vol. 46, No. 2, Summer 2004, pp.
367–372. Copyright © 2004, Human Factors and Ergonomics Society. All rights reserved.
368
The particular engineering solution that was
evaluated was the use of synchronized traffic
signal lights. The synchronization consists of
the coordination of traffic signals along several
blocks of selected urban streets, so that their
green phase is synchronized with traffic mov-
ing on that route at or slightly below the speed
limit. This measure is designed to increase traf-
fic flow, and to reduce speed variability and con-
flicts among vehicles along the synchronized
route. The aggressive driving measure that was
used to evaluate the effectiveness of synchro-
nized intersections was the tendency of drivers
to run red lights (red-light running, or RLR).
Synchronized intersections smooth the flow
of traffic, thereby reducing frustration from
having to stop repeatedly (Hulfine & Adams,
1995; Khalaf, Ergun, Ratrout, Al-Senan, & Al-
Ahmadi, 1995). There is even evidence for a
small decrease in intersection-related accidents,
particularly rear-end collisions (Berg, Kaub, &
Belscamper, 1986; Hulfine & Adams, 1995; Kha-
laf et al., 1995; Schlabbach, 1988). However, the
direct contribution of synchronized intersections
to safer driver behavior is less obvious and has
not been studied. The principal issue is whether
synchronization, by causing vehicles to move
in platoons only, merely reduces the number of
opportunities for conflicts at intersections or if
it changes driver behavior as well. The present
study was designed to test whether or not syn-
chronized signals also reduce drivers’ inclina-
tion for RLR. Because some studies have found
that male drivers, young drivers, and drivers
without passengers are more aggressive than
female drivers, older drivers, and drivers with
passengers, respectively (Hyde, 1984; Retting &
Williams, 1996; Shinar, 1998), we also observed
these aspects of individual differences.
METHOD
Participants
The participants in the study were regular
drivers who drove through the selected signalized
intersections during the times of observation.
Drivers were observed going through 3600 sig-
nal cycles.
Observation Sites and Times
We chose 12 intersections after several con-
sultations with the city traffic engineer of Tel
Summer 2004 – Human Factors
Aviv. The desire was to match the synchronized
and unsynchronized intersections as much as
possible in terms of traffic volume, location, and
general purpose. Thus all intersections were
along two main north-south urban thorough-
fares with two through lanes in each direction
and separate drop-off lanes for right and left
turns. Eight of the intersections were in the cen-
tral sections of major streets with synchronized
signals, and four were with nonsynchronized sig-
nals. Data were recorded for the two through
lanes during the daylight hours of the dry sum-
mer months, when the sky is clear. No data
were collected in extreme or unusual situations,
such as when a police officer was present or
when traffic was stopped in one of the lanes
because of an accident. Observation periods
consisted of rush hours (Sunday–Thursday,
7:00–9:30 and 16:30–19:00; Sunday is a regular
workday in Israel), nonrush hours (weekdays
10:00–15:00), and weekend hours (Saturday
9:00–19:00).
Experimental Design
We observed 100 signal cycles on each of
three observation periods at each of the 12 inter-
sections, yielding a total of 3600 traffic signal
cycles. The main independent measure was the
nature of the intersection: synchronized versus
nonsynchronized. For drivers going through a
red light, as well as for a representative sample
of all drivers going through the intersections,
the following data were observed or estimated:
driver gender, perceived age (younger than 25,
25–50, and 50+ years of age), and presence or
absence of other occupants in the car.
Procedure
Two trained observers conducted all obser-
vations. At each intersection, they observed the
traffic moving in one direction on the two
through lanes. The observers positioned them-
selves in such a way that they were obscured
from the drivers’ forward field of view. For each
cycle, they made separate counts of (a) the num-
ber of vehicles that entered the intersection after
the light had turned yellow but cleared it after the
light had turned red and (b) the number of ve-
hicles that entered the intersection after the
light had turned red. The reference for entering
and clearing the intersection was the car’s front
RED-LIGHT RUNNING AND SYNCHRONIZATION
end crossing the intersection’s start and end, as
defined by the imaginary line that is tangential
to the edges of the curbs. A car was considered
RLR only when both observers agreed. For
each such vehicle the driver’s gender, estimated
age, and presence or absence of passengers
was also noted. In addition, to obtain exposure
data on driver gender, estimated age, and pres-
ence or absence of passengers, these data were
noted for the first 100 cars going through the
intersection at each observation period (before
recording RLR). They are assumed to be a repre-
sentative sample of the drivers traveling though
that intersection at that time.
RESULTS AND DISCUSSION
Synchronization and RLR
The principal hypothesis concerned the im-
pact of synchronization on RLR. In all 3600 cy-
cles of observations, 185 drivers were observed
entering the intersection after the light had
changed to red. Of those, 139 were in the four
nonsynchronized intersections and 46 were in
the eight synchronized intersections.
This difference by itself is insufficient to indi-
cate that RLR is more common in nonsynchro-
nized intersections, given that the opportunity
for RLR is lower in these intersections. An op-
portunity for RLR was defined as a situation in
which one or more cars approached the signal-
ized intersection just as the light turned yellow,
thus giving the driver the opportunity to cross
the intersection on red. When two cars, one be-
hind the other, approach an intersection after the
light has turned yellow or red and the lead driver
stops for the light, there is only one opportunity
for RLR because the second driver is not afford-
ed the opportunity to cross the intersection.
The rationale for fewer opportunities for RLR
in synchronized intersections than in nonsyn-
TABLE 1: Number of Opportunities Per Intersection to Run Red Lights
Rush
Signal Type Hours
Synchronized 78
Nonsynchronized 89
Note. Data based on two lanes and 100 cycles at each time period. An opportunity is defined as
situation in which one or more cars reached the intersection just as or immediately after the light
turned red.
369
chronized intersections is that synchronization
forces traffic to move in “platoons,” with most
cars moving through the synchronized green
phase. Thus, with proper synchronization, traf-
fic waiting at the red light consists mostly of
cars that had just entered the synchronized route
from another route. The results, summarized in
Table 1, show that the number of opportuni-
ties for RLR in the synchronized intersections
was much lower than in the nonsynchronized
intersections.
To adjust for the differences in opportunities
for RLR, we conducted a chi-square analysis on
the frequencies of entering the intersection
after the light had turned red in the synchro-
nized and nonsynchronized intersections, rela-
tive to the number of opportunities observed in
these two types of intersections. The difference
was highly significant, χ2(1) =145.06, p < .0001,
yielding an odds ratio of 6.7. Thus the odds of
RLR were nearly seven times higher in the non-
synchronized intersections than in the synchro-
nized intersections.
There were also 1174 observations of cars
that entered the intersection after the light had
changed to yellow and that cleared the intersec-
tion only after the light had changed to red. For
the remaining analyses, these vehicles were
combined with the 185 vehicles that entered the
intersection during the red phase, yielding a
sample of 1359 RLR observations. This was
done for two reasons. The first, a practical one,
is the relatively small number of cars that en-
tered the intersection after the light had turned
red. The second, a behavioral one, is that all these
drivers had enough time to stop in response to
the yellow signal before entering the intersec-
tion but chose not to do so. Instead, they risked
crossing the intersection against the light: Some
cleared the intersection under a red signal, and
others actually entered it under a red signal. It
Nonrush
Hours Weekends Total
53 44 175
66 58 213
370
is important to note, however, that although all
of these drivers did in fact block the intersection
during the red phase, only those who entered
the intersection after the light had turned red
would be legally considered as RLR. When the
two types of behaviors were combined, chi-
square analysis of the number of RLRs in the
two types of intersections relative to the num-
ber of opportunities yielded a highly significant
effect, χ2 = 111.06, p < .0001, and an odds
ratio of 2.1.
Synchronization and Congestion Effects on
RLR
At all three times of observation, the rates
of RLRs (number of RLRs/signal/observation
period) were significantly lower in the synchro-
nized intersections, as illustrated in Table 2.
These results reflect 40% fewer RLRs in syn-
chronized intersections than in nonsynchro-
nized intersections. Furthermore, although the
absolute number of RLRs was greater in rush
hours than in nonrush hours and weekends, the
relative difference between synchronized and
nonsynchronized intersections increased from
rush hours to nonrush hours and to weekends,
χ2(2) = 49.29, p < .0001, χ2 for trend (1) =
49.02, p < .0001. Thus there were 83% more
RLRs in nonsynchronized intersections than in
synchronized intersections during rush hours,
303% more in nonrush hours, and 386% more
on weekends.
As already stated, because of the nature of
signal synchronization, one possible source for
the effect of synchronization could be simply
that synchronization yields fewer opportunities
for RLR. Table 1 presents the number of oppor-
tunities for RLR in each condition. As can be
seen from Table 1, in each of the three observa-
tion periods the number of opportunities for
RLR was in fact greater in the nonsynchronized
intersections than in the synchronized ones,
TABLE 2: Number of RLRs Per Intersection
Rush
Signal Type Hours
Synchronized 56
Nonsynchronized 103
Note. Data based on two lanes and 100 cycles at each time period.
Summer 2004 – Human Factors
and in both cases the number of opportunities
increased with congestion. However, there was
no differentially significant effect of the period of
observation (rush hours vs. nonrush hours vs.
weekends) on the frequency of RLRs, χ2(2) =
1.92, p = .38.
For a more detailed analysis of the source of
differences among the different observation
periods, we looked at the number of RLRs as a
function of the average number of cars queued
in each lane at the end of the red-light phase
(just before it turned green). This provides a
quantitative, direct, and continuous measure
of congestion (unlike the indirect assessment
through period of observation). The results
with the best-fitting functions are presented in
Figure 1. Best-fitting quadratic equations yield-
ed R2 = .93 for the synchronized intersections
and R2 = .84 for the nonsynchronized intersec-
tions. A linear fit to the two sets of data points
yielded R2 = .83 for both synchronized and
nonsynchronized intersections.
These results are consistent with those of
Table 2, but they show more clearly that as con-
gestion increases the absolute number of RLR
increases in both types of intersections, whereas
the difference in RLR between the synchronized
and nonsynchronized intersections decreases
until it finally disappears completely. Thus as
congestion increases, the effectiveness of syn-
chronization decreases until it becomes totally
irrelevant to driver behavior at the intersections.
Synchronization and Individual
Differences in RLR
Gender. The percentage of men in the RLRs
(82%) was slightly (but significantly, p = .0005)
higher than their percentage in the driving pop-
ulation at the same intersections (72%). The
relationship between gender and synchroniza-
tion (see Table 3) was not statistically signifi-
cant, χ2 = 2.75, p = .10, indicating that the small
Nonrush
Hours Weekends Total
13 6 75
53 34 190
RED-LIGHT RUNNING AND SYNCHRONIZATION 371

Figure 1. The increase in RLR as a function of congestion in synchronized and nonsynchronized intersections.

difference between male and female drivers was
unaffected by synchronization.
Age. The percentage of older drivers (50+
years) in the RLRs (23%) was slightly (but sig-
nificantly, p = .027) higher than their percentage
in the driving population at the same intersec-
tions (17%). In contrast, the percentage of young
drivers (<25 years) in the RLRs (9%) was actu-
ally slightly lower than their percentage in the
driving population at the same intersections
(12%), but this difference was not significant
(p = .15). Overall, it seems that age – at least
as it was estimated – was unrelated to the ten-
dency to run a red light. The relationship be-
tween age and synchronization (see Table 4) was
not statistically significant, χ2(2) = 0.25, p = .88,
indicating that synchronization did not have a
differential effect on younger and older drivers.
Presence of passengers. The percentage of
drivers without passengers in the RLRs (64%)
was almost identical to their percentage in the
driving population at the same intersections
(63%). The relationship between presence or
absence of passengers and synchronization (see
Table 5) was not statistically significant, χ2 = 0.12,
p =.73, indicating that the effect of synchroniza-
tion on drivers without passengers was not dif-
ferent from that on drivers with passengers.

TABLE 3: Number of RLRs Made by Male and Female Drivers Per Intersection

Signal Type Men Women Total

Synchronized 56 19 75
Nonsynchronized 152 38 190
Note. Data based on two lanes and 100 cycles across all observation periods.

TABLE 4: Number of RLRs Per Intersection for Each Estimated Age Group

Young Mature Older

Signal Type (<25) (25–50) (>50) Total

Synchronized 6 54 15 75
Nonsynchronized 16 141 33 190
Note. Data based on two lanes and 100 cycles across all observation periods.
372
TABLE 5: Number of RLRs Per Intersection for Drivers With and Without
Passengers
Without
Signal Type Passengers
Synchronized 57
Nonsynchronized 1310
Note. Data based on two lanes and 100 cycles across all observation periods.
CONCLUSIONS AND
RECOMMENDATIONS
The results of this study demonstrate the ef-
fectiveness of signal synchronization beyond its
assumed engineering benefits. When drivers
move along a route with synchronized signals,
they are much less likely to run the red lights
(either enter the intersection on red or enter it
on yellow and clear it on red) than when the sig-
nals on consecutive intersections are not syn-
chronized. Furthermore, the effectiveness of
synchronization overshadowed the individual
differences of driver gender, driver age, and
presence or absence of passengers. Thus, unlike
some previous studies, this study did not obtain
higher (or practically significant) rates of RLR
among men, younger drivers, or drivers without
passengers, and synchronization seems to have
exerted the same effect on all drivers regard-
less of those individual differences.
The benefits of synchronization are greatest
when the traffic is lightest, and they decrease as
congestion increases, to a point where driver
behavior is the same in synchronized and non-
synchronized intersections. This effect parallels
the increase in frustration over lost time with
increasing congestion and is consistent with the
Dollard et al. (1939) frustration-aggression
model.
These findings demonstrate the importance
of engineering and design in the alleviation of
one type of aggressive driving. When the traffic
system is designed to accommodate traffic more
efficiently, not only does it reduce the number
of opportunities for RLR, it also decreases the
individual drivers’ inclination to commit RLRs
when they do encounter the frustrating red light.
It thus demonstrates a basic ergonomics prin-
ciple: Changing the environment may be more
effective than changing its users.
Summer 2004 – Human Factors
With
Passengers Total
18 75
59 190
REFERENCES
AAA Foundation for Traffic Safety. (1997). Aggressive driving: Three
studies. Washington, DC: Author.
Berg, W. D., Kaub, A. R., & Belscamper, B. W. (1986). Case study
of the safety and operational benefits of traffic signal coordina-
tion. Transportation Research Record, 1057, 58–64.
Dollard, J., Doob, W., Miller, N. P., Mowrer, O. H., & Sears, R. R.
(1939). Frustration and aggression. New Haven, CT: Yale Uni-
versity Press.
Hulfine, C. W., & Adams, D. D. (1995, August). Safety benefits of
signal coordination. Presented at the Institute of Transporta-
tion Engineers 65th Annual Meeting, Denver, CO.
Hyde, J. S. (1984). How large are gender differences in aggression?
A developmental meta-analysis. Developmental Psychology,
20, 722–736.
Khalaf, K. A., Ergun, G., Ratrout, N. T., Al-Senan, S. H., & Al-
Ahmadi, H. M. (1995). The effect of signal coordination on
intersection safety. In Proceedings of the 23rd Planning and
Transport Research and Computation Meeting (pp. 297–309).
Brussels, Belgium: European Transport Forum.
National Highway Traffic Safety Administration. (1998, April).
Aggressive drivers view traffic differently, capital beltway focus
groups find (U.S. Department of Transportation Traffic Tech.
Number 175). Washington, DC: U.S. Department of Trans-
portation.
Retting, R. A., & Williams, A. P. (1996). Characteristics of red light
violators: Results of a field investigation. Journal of Safety
Research, 27, 9–15.
Schlabbach, K. (1988). Pulsating green waves – The Dormstadt
experiment. Traffic Engineering and Control, 29, 293–297.
Shinar, D. (1998). Aggressive driving: The contribution of the drivers
and the situation. Transportation Research Part F, 1, 137–160.
David Shinar is a professor of ergonomics in the
Department of Industrial Engineering and Manage-
ment, Ben Gurion University of the Negev. He re-
ceived his Ph.D. in human performance and human
factors from Ohio State University in 1973.
Muki Bourla is a project manager for telecommuni-
cation infrastructure at LM Ericsson Israel Ltd., Rosh
Ha’ayin, Israel. He received his M.B.A. in 2001 from
Ben Gurion University of the Negev.
Liat Kaufman is a project manager at ROI Ltd., Tel
Aviv, Israel. She received her M.B.A. in 2003 at
Tel Aviv University.
Date received: August 5, 2002
Date accepted: December 1, 2003