NOTICE: This a PRE-PRINT copy of paper 98-0009
presented at the Transportation Research Board
77th Annual Meeting, January 11-15, 1998
Washington D.C.
Duplication of this preprint for publication or sale is strictly
prohited
without written permission of the Transportation Research Board.
ADULT BICYCLISTS IN THE UNITED STATES -
CHARACTERISTICS AND RIDING EXPERIENCE IN 1996
William E. Moritz, Ph.D.
Professor (Emeritus)
Human Powered Transportation
Box 352500
University of Washington
Seattle WA USA 98195-2500
Revised March 30, 1998
ABSTRACT
In December, 1996, 20% (4712) of the League of American Bicyclist
members were surveyed about their cycling experiences during CY 1996. The 33 questions
included: bicycle type and equipment, distribution of bicycle trips by purpose (e.g. work,
on-road recreation), total distance cycled, commuting habits, accidents, and demographic
data. The survey was designed to update one done by Kaplan in 1975. By the March 31, 1997,
deadline over 2400 (52%) had been returned. Of these 19% were rejected due to
incompleteness or inconsistent responses leaving 1956 valid surveys. The 'average'
respondent was a 48 year-old, married (66%) male (80%) professional (48%) who rode 4670 km
in 1996. Just over 9% reported having had a serious crash (resulting in at least $50 of
property damage or medical expense) in 1996. Based on the experience reported by these
cyclists, the 'average' cyclist in this group could be expected to ride for 11 years
before having such a crash. Falls accounted for 59% of the incidents while running into a
fixed object happened 14% of the time. Moving motor vehicles were involved in 11% of the
crashes and another bicycle in 9%. A RELATIVE DANGER INDEX is calculated which shows that
streets with bike lanes have a significantly lower crash rate then either major or minor
streets without any bicycle facilities (38 and 56% respectively). Multi-use trails have a
crash rate about 40% greater than would be expected based on the miles cycled on them
while cycling on the sidewalk is extremely dangerous.
KEYWORDS: Adult bicycling, Bicycle crash experience
ADULT BICYCLISTS IN THE UNITED STATES -
CHARACTERISTICS AND RIDING EXPERIENCE IN 1996
William E. Moritz, Ph.D.
Professor (Emeritus)
Human Powered Transportation
Box 352500
University of Washington
Seattle WA USA 98195-2500
INTRODUCTION AND BACKGROUND
The decade of the 1990's has seen a revival of interest in bicycling
for recreation, health, and transportation. Adults as well as children continue to enjoy
the fun of cycling. Public health officials regularly extol the virtues of vigorous
aerobic exercise that can be easily obtained on a bike. And in many of our urban areas we
are experiencing ever increasing traffic congestion which is motivating some travelers to
use their bikes for commuting and utilitarian trips.
Over the past several years there have been various public policy
initiatives that have raised the visibility of cycling at all levels of government. One
example is the Intermodal Surface Transportation Efficiency Act of 1991 [ISTEA] (1) which
mandates that cycling must be integrated into required transportation plans. Another
example at the federal level is the Clean Air Act Amendments of 1990 where non-attainment
areas are required to develop specific plans to achieve compliance with the Act.
Substituting bicycling for automobile trips would contribute to reducing air pollution.
Growth management acts which link land use and transportation have been
adopted in many parts of the country with the goal of containing urban sprawl. And commute
trip reduction programs are attempting to reduce vehicle traffic volumes during the
so-called 'rush hours.' While the average one-way commute trip is approximately 16 km in
length, almost half of all trips are 5 km or less. Such distances are well within
bicycling range for most adults in this country. (2)
Lastly, in 1994 the USDOT/FHWA released the National Bicycling and
Walking Study which set a goal "to double the percentage of total trips made by
bicycling and walking in the United States from 7.9 to 15.8 percent." (3)
Yet, in spite of all this interest, we know very little about the types
of adults who cycle on a regular basis. The U.S. Consumer Product Safety Commission made
an attempt to publish such information in 1994. (4) This study reports the results of two
surveys which unfortunately targeted different age groups and used different metrics to
assess 'exposure' - perhaps better characterized as 'use' since respondents were asked to
estimate the number of hours they rode.
In each of the past several years between 800 and 1000 bicyclists have
been killed and hundreds-of-thousands have been injured in the U.S. (5) While fatalities
are well documented, there is very little data concerning the far more frequent but less
serious crashes typically experienced by cyclists. One glaring hole in the data is
exposure information (e.g. crashes per kilometer) and thus the relative safety of various
types of on- and off-road bicycle facilities is virtually unknown.
In 1975 Jerrold Kaplan conducted a national survey of adult cyclists
who were members of the (then) League of American Wheelmen [LAW - now known as the League
of American Bicyclists {LAB}]. (6) Kaplan's report attempted to not only develop a profile
of 'regular' adult cyclists during calendar year 1974, but also to gather mileage
information from which to calculate crash rates per mile - both by activity (e.g.
recreation vs. commuting) and by facility type (e.g. major streets vs. 'bike paths'). His
works stands alone in the literature yet is now quite dated.
Since over twenty years has elapsed since Kaplan's work, it was decided
to update and expand Kaplan's survey. The present study added questions on mountain
biking, for example, and included data on the perceived safety of bicycle facilities, an
assessment of motorist's attitudes towards cyclist, formal bicycle safety training, and
some information on commuting habits. Gathering and analyzing crash data was of particular
interest.
SURVEY GOALS AND POPULATION
The goals of this survey were: 1. Gather detailed data on U.S. adults
who cycle regularly; 2. Develop a demographic profile of these cyclists; 3. Analyze their
crash experience in several ways; and 4. Compare these results with Kaplan's (and with a
similar survey conducted in 1995 by this author of adults in Washington state involving
957 adult cyclists). (7)
Since a national sample was desired, it was decided to again use the
membership of the LAB as was done in 1975. In 1975 the League had 8400+ member-households
and each one received a survey. Kaplan achieved a response rate of about 38% which yielded
a sample of 3270. By 1996 League membership had nearly tripled to about 23,500. Since many
national surveys use a sample of about 1,000 it was clearly unnecessary to include all of
the members.
STUDY METHODOLOGY
Selecting Potential Participants:
Kaplan's survey was distributed and returned along with a ballot for a
Board election thus substantially reducing his mailing costs. No such opportunity existed
presently. In light of the $6,500 available for the present project, a smaller sample had
to be created.
One problem with simply selecting every n'th member was that several
states had disproportionally large fractions of the membership when compared to 1990 U.S.
Census data. (Note a similar problem existed in 1975 but that the geographic distribution
of the membership has also changed markedly since then.)
It was decided to design a selection process that would: a) yield
approximately 2,000 responses (assuming a 40+% response rate) and b) sample from each
state in proportion to that state's share of the U.S. population. As a result, 20% of the
overall membership (4712) were selected. Individual state fractions ranged from as low as
6.5% for MA (which had 7.5% of the membership but represents only 2.4% of the U.S.
population) to a high of 100% for MS (0.2% and 1.0% respectively). Within each state
members were first sorted by ZIP code and then every m'th one was selected where m was
designed to yield the desired number for that state. Table 1 presents the breakdown by
state.
TABLE 1. Breakdown of sample by state.
STATE |
COUNT |
% MBRS |
STATE |
COUNT |
% MBRS |
STATE |
COUNT |
% MBRS |
AK |
10 |
15.4 |
KY |
70 |
39.1 |
NY |
341 |
19.2 |
AL |
77 |
51.7 |
LA |
80 |
46.0 |
OH |
205 |
18.0 |
AR |
45 |
84.9 |
MA |
114 |
6.5 |
OK |
60 |
51.3 |
AZ |
69 |
19.4 |
MD |
91 |
8.1 |
OR |
54 |
19.2 |
CA |
564 |
25.6 |
ME |
23 |
18.4 |
PA |
225 |
15.8 |
CO |
62 |
11.7 |
MI |
176 |
19.9 |
RI |
19 |
20.9 |
CT |
62 |
10.4 |
MN |
83 |
19.8 |
SC |
66 |
41.3 |
DC |
12 |
9.0 |
MO |
96 |
26.4 |
SD |
13 |
48.1 |
DE |
13 |
10.2 |
MS |
49 |
100.0 |
TN |
92 |
27.8 |
FL |
245 |
26.2 |
MT |
15 |
32.6 |
TX |
322 |
41.5 |
GA |
122 |
32.3 |
NC |
126 |
30.1 |
UT |
33 |
30.0 |
HI |
21 |
34.4 |
ND |
12 |
66.7 |
VA |
117 |
12.1 |
IA |
53 |
18.5 |
NE |
30 |
25.6 |
VT |
11 |
10.1 |
ID |
19 |
32.2 |
NH |
21 |
13.1 |
WA |
92 |
15.8 |
IL |
216 |
15.6 |
NJ |
146 |
15.3 |
WI |
93 |
19.2 |
IN |
105 |
23.3 |
NM |
29 |
18.8 |
WV |
34 |
38.2 |
KS |
47 |
18.8 |
NV |
23 |
26.4 |
WY |
9 |
24.3 |
Subtotal |
|
|
Subtotal |
|
|
Subtotal |
|
|
| |
|
|
|
|
GRAND |
TOTAL |
4712 |
|
Mailing and Retrieval:
The surveys were mailed on December 31, 1996, using first class postage
to increase the likelihood of delivery and to be able to track the number that were
undeliverable. A business reply panel was included to encourage responses. A deadline of
March 31, 1997, was indicated. (Kaplan's were mailed in mid March, 1975, and the ballot
and survey had to be returned by April 15 in order to be counted.)
Survey Format:
A printed survey was selected to minimize costs and to replicate the
method used by Kaplan (and this researcher in the 1994 Washington study). Note that the
survey was designed to be anonymous with the ZIP code the only possible means of
identifying the respondent. [A copy of the survey is available from the author at the
above address.]
The questions were designed to capture information similar to that
gathered in the earlier surveys as well as ask additional questions. It was also laid out
to facilitate data entry.
Period Covered:
Respondents were asked to report in their riding and experiences for
calendar year 1996. (Kaplan covered calendar year 1974 while the Washington survey covered
calendar year 1994.)
Qualifications to Participate:
The instructions asked that only current League members over the age of
15 respond. Further, if the household had more than one member, then the one who rode the
most miles should fill it out. Both of these provisions duplicate those used by Kaplan. It
should be noted that a couple of female cyclists sent notes complaining that those
instructions discriminated against them because, while they were active cyclists, their
male housemate cycled more than they did. It had been recognized in the design that these
instructions might bias the sample toward men but this was the same approach used by
Kaplan.
SURVEY DESIGN
Information was requested 4 areas:
A. About riding;
B. About commuting;
C. About safety and accidents; and
D. About the respondent.
A Comment space was provided at the end of the survey.
The most likely answers to multiple choice questions were provided and
assigned numerical codes to facilitate data entry and analysis. YES/NO responses were
entered as 1/0 respectively.
A. About Your Riding
The first 11 questions attempted to characterize the equipment they
used, the amount spent on cycling during 1996, traffic law obedience, whether they rode
mostly on weekends or weekdays and after dark or in the rain, and how long they have been
riding. The final 5 questions dealt with the types of riding they did (trip purpose), the
types of facilities used, total miles and hours ridden, and finally a question about how
these cyclists perceive the motorists they share the road with.
B. Commuting by All Modes
Gathered commute trip information (mode, distance, time) as well as
attempted to determine for those not usually bike commuting the most significant reason
for that decision.
C. Safety/Accidents
One question sought their perception of the safety of various bicycle
facilities. The remaining questions dealt with collisions or falls during 1996. Since many
crashes experienced by cyclists do not result in significant injury or property damage, an
attempt was also made to capture data on the more serious crashes. Note that Kaplan did
not define what constitutes a 'serious' fall leaving it to the reader to decide. In the
present survey (as well as the 1994 Washington one) a $50 threshold was established to
reduce the ambiguity in that definition. The remaining questions attempted to characterize
both serious and non-serious crashed by severity, activity, facility, and mode.
D. About You and Your Household
Five questions sought 'standard' demographic information (for example:
age and sex). A question on bicycle safety training, if any, was also included. This was
followed by a couple of questions about additional cyclists in the home, the number of
bikes and motor vehicles available, and a question about their general health since they
started cycling regularly. Finally, years of LAW/LAB membership and ZIP code were
requested.
Comments
A space was provided for any additional information the respondent
wished to provide.
RESPONSE RATE
The response was nothing short of phenomenal. Within 14 days 35% had
been returned and 2 weeks later we were at 46%. At the deadline just over 2400 (nearly
52%) had responded while 74 had come back undeliverable. Return rates varied from 47% in
New England to 65% in the Northwest.
SCREENING RESPONSES AND DATA ENTRY AND CHECKING
Each survey was checked for completeness and consistency. Approximately
5% (121/2403) were rejected because they were damaged beyond use, were incomplete (often
one entire side was left blank), reported no riding in 1996, or failed to provide a
breakdown of facilities use that totaled between 90 and 110%. This latter parameter was
essential for the facilities crash analysis described below.
One problem with a survey of this type is assessing the validity of the
responses since most cyclists do not keep records of their daily trips. Since there is no
way to verify true accuracy, it was decided to at least require some level of internal
consistency in the responses. For example, the sum of the miles ridden per month by trip
purpose multiplied by the claimed number of months ridden should (ideally) equal to total
miles reported. The range of errors encountered was -97% to +2345%. An acceptance range of
+/-40% was selected which resulted in rejecting an additional 13.5% (326/2403).
As a result, 1956 surveys were included in the final data set and their
responses entered into a Lotus 1-2-3 spreadsheet. Every entry was verified to be within
bounds (e.g. months ridden between 1 and 12) and that all parts of multiple part questions
were filled in (e.g. if a crash was indicated, then a responses must appear in several
related questions dealing with the type of crash and facility upon which it occurred).
Incomplete responses were removed.
ANALYSIS
Several types of analysis were applied to the final data set.
Fractional responses were calculated for questions like bicycle type and equipment. For
questions with numerical responses (e.g. years of regular cycling and age), the highest,
average, median and lowest values were calculated. In some of these cases distributions
were determined as well.
One of the major advantages of this type of study (in which users
report distances ridden and crash experience) is the opportunity to determine crash rates
(the number of crashes divided by the distance ridden in CY 1996). In addition, the crash
rates and characteristics of various sub-populations (e.g. males vs. females) can be
evaluated. Finally, the crashes can be tabulated by activity, facility used, and nature of
the crash.
Another way of looking at crashes vs. facilities is to divide the
fraction of crashes reported for a particular facility by the fraction of kilometers
ridden on that facility type. The resulting number, here called the RELATIVE DANGER INDEX
(RDI), would be 1.0 if crashes occurred in proportion to the distance traveled. An RDI
greater than 1.0 indicates a facility on which crashes occur at a higher rate then would
be expected based simply on distance. RDI is thus a convenient way to compare various
facilities.
RESULTS
In the material that follows, the Washington State survey results will
be referred to as WA 94 and Kaplan's results as LAW 74. Where comparable data exists in WA
94 and the LAW 74 surveys, those results will be shown as [WA 94 / LAW 74] immediately
following the present results. If no such comparable data exists a dash ( - ) will appear.
Space simply does not permit a full presentation of the results for all three studies.
Demographics
The 'average' respondent was a 48 [45/38] year-old, married (66%)
[62%/-] male (80%) [74%/88%] professional (48%) [48%/-] with a college degree (80%)
[81%/-]. More than 53% [46%/-] reported a household income in excess of $60K/year. Just
under 4% [1%/5%] reported not owning a motor vehicle and of those 90% [-/-] said they did
so by choice. The average household had two or more cyclists [2 or more/-], two or more
motor vehicles [2 or more/2 or more], and 4.4 [3.7/-] bikes.
Figure 1, showing the age distribution for the 3 surveys, graphically
demonstrates that the two more recent studies involve substantially older cyclists. Most
notable is the large decrease in the fraction of respondents in the 16-35 age groups in
these surveys.
Figure 1. Distribution of respondent’s ages.
Over 75% [-/-] reported that their general physical health had improved
either greatly or somewhat since they started cycling regularly while less than 1% thought
their health had deteriorated (expect by age alone - 7%).
About Their Cycling
Road (49%) [42%/-] and touring (21%) [25%/-] bikes were used most often
with mountain bikes at 12% [15%/-]. (NOTE: LAW 74's bike type question dealt with the
number of speeds - 1, 3, 5+ - and thus is not comparable.) The majority of this group and
the WA 94 sample were well equipped reporting using or carrying: mirrors, computers,
panniers, pumps, spare tubes, and tools. Helmets were reportedly worn on every ride
by 88%. For LAW 74, 31% reported owning a helmet while for the WA 94 group helmet, ownership
was 99.9%. Neither of these surveys asked about helmet use.
Their cycling was evenly split between weekends (52%) [51%/54%] and
weekdays (48%) [50%/46%]. Nearly half (47%) [35%/35%] never ride after dark while only 15%
[5%/24%] never ride in the rain. The average respondent has been cycling regularly for
14.2 years and claims to have ridden 9 months in 1996. [Both WA 94 and LAW 74 binned
experience into ranges with the following results: <1 year: 4%/3%; 1-4 years: 25%/48%;
5-10 years: 33%/28%; greater than 10 years: 39%/20%.]
On-road recreation was reported by 93% [84%/84%] while 41% [41%/48%]
claimed to have used their bikes for utility purposes and 38% [52%/49%] did at least some
work or school travel by bike in 1996 [1994/1974]. Thirty-five percent (35%) [20%/-]
claimed to have mountain biked as well. Table 2 shows this data as well as the maximum and
average monthly distances claimed for each trip purpose. Note the averages are calculated
over the entire sample population (1956 respondents) not just those reporting a particular
type of trip.
The facility on which the most kilometers were ridden was minor streets
without bike facilities (45%) [44%/58%] while 32% [26%/35%] of the kilometers were on
major streets (again without bike facilities). Bike routes, bike lanes, and multi-use
trails each accounted for 6-7%. (In WA 94 and LAW 74 signed bike routes and bike laned
streets were combined with 12%/3% of the kilometers reported while multi-use trails had
17%/4%.) Off-road/unpaved facilities were 4% [2%/-]. An "Other" category appears
in the present study and nearly all responses indicated this meant sidewalks. They
accounted for just 0.3% of the kilometers ridden.
Total cycling kilometers in 1996 were 9,160,000 [4,185,000/12,171,000]
for an average of 4670 [4370/3760] per cyclist. One hearty soul claimed to have pedaled
38,700 [35,500/>16,000] kilometers while the lowest distance reported was 160 [50/-]
km.
TABLE 2. Trip purposes and monthly distances cycled for each type of trip.
| Trip Purpose |
LAB 96 |
WA 94 |
LAW 74 |
| Work/School |
|
|
|
|
Maximum kilometers |
1667 |
1050 |
N/A |
|
Average kilometers |
78 |
113 |
92 |
|
Fraction of kilometers |
15% |
24% |
22% |
|
% reporting this trip purpose |
38% |
51% |
49% |
| Shopping/Personal Business |
|
|
|
|
Maximum kilometers |
1333 |
667 |
N/A |
|
Average kilometers |
22 |
22 |
27 |
|
Fraction of kilometers |
4% |
5% |
6% |
|
% reporting this trip purpose |
41% |
41% |
48% |
| Road Recreation + Exercise for WA
94/LAW 74 |
|
|
|
|
Maximum kilometers |
2500 |
1500 |
N/A |
|
Average kilometers |
330 |
242 |
277 |
|
Fraction of kilometers |
64% |
51% |
66% |
|
% reporting this trip purpose |
93% |
84% |
84% |
| Mountain Biking Recreation |
|
|
|
|
Maximum kilometers |
667 |
667 |
N/A |
|
Average kilometers |
23 |
12 |
N/A |
|
Fraction of kilometers |
5% |
2% |
N/A |
|
% reporting this trip purpose |
35% |
20% |
N/A |
| Road Racing |
|
|
|
|
Maximum kilometers |
3333 |
1000 |
N/A |
|
Average kilometers |
35 |
20 |
27 |
|
Fraction of kilometers |
7% |
4% |
6% |
|
% reporting this trip purpose |
9% |
7% |
9% |
| Mountain Bike Racing |
|
|
|
|
Maximum kilometers |
347 |
333 |
N/A |
|
Average kilometers |
2 |
1 |
N/A |
|
Fraction of kilometers |
0% |
0% |
N/A |
|
% reporting this trip purpose |
3% |
2% |
N/A |
|
TOTAL FRACTION KILOMETERS |
96% |
86% |
100% |
Figure 2 presents the distribution of annual cycling distance reported in the 3
surveys. Most notable here is that the present cyclists appear to be cycling more each
year than 20 years ago.
Figure 2. Distribution of annual cycling distances ridden.
Women had been cycling regularly an average of 12.4 years compared to
14.6 for men and reported somewhat smaller number of months ridden (8.7 vs. 9.6). Women's
responses to distances traveled and trip purposes (as is depicted for the entire sample in
Table 2) were very similar to men's except for the Road Racing category. Here only 4%
reported engaging in this activity vs. 10% for the men and the fraction of kilometers were
3% vs. 8%. Finally, average cycling distance in 1996 was about 4,300 km for women vs.
5,000 for the men.
Commuting
Only 53% [56%/-] reported commuting to work or school and of those 51%
[41%/-] did so most often by car. Bicycle commuting was the next highest mode at 29%
[40%/-]. Average one-way commute distances and times were 17 [17/-] km and 27 [38/-]
minutes. Needing a car at work, dangerous roads, distance, and lack of facilities at
work/school were the most often cited reasons for not bike commuting for the present
sample. For the WA 94 group, time, weather, and distance were most often cited. LAW 74 did
not ask about commuting.
Crashes - Minor vs. Serious
Twenty-nine percent (29%) [32%/26%] reported having had some type of
'accident' in the study year. Just over 9% [10%/21%] reported having had a serious crash
(resulting in at least $50 of property damage or medical expense) in 1996 [1994/1974].
(Recall that LAW 74 left it to the reader to decide what was a 'serious' crash which might
explain their significantly higher response to that question.)
Those reporting a serious crash had an average of 1.2 [1.2/1.2] such
crashes during the year. Only 28% [27%/-] of such crashes were reported to the police.
Such crashes resulted in average and median medical expenses of: $2,970 [$884/-] and $155
[$150/-] respectively. The maximum total medical expense reported was $250,000 [$17,500/-]
but no details were provided by that respondent. Clearly such an amount will exert a
strong upward bias on the average.
For reported property damage in serious crashes, the average and median
expenses were: $316 [$384/-] and $100 [$130/-]. The maximum total property damage was
$5,000 [$5,000/-].
A slightly higher fraction of women reported having a serious crash
than men - 11% vs. 9% - and their serious crash rates were also slightly higher - 29 vs.
22 per million kilometers. In most other respects the crash results for both sexes were
comparable.
Crashes - Modality
For all crashes falls accounted for 59% [48%/41%] with running into a
fixed object being the next most frequent at 14% [17%/'Other':11%]. Moving motor vehicles
were involved in 11% [11%/18%] and another bicycle in 9% [13%/17%] of all crashes
regardless of severity . See Table 3.
TABLE 3. Collision or fall modality for all (serious and minor)
crashes.
|
LAB 96 |
WA 94 |
LAW 75 |
No other object - simple fall |
59% |
48% |
41% |
Moving motor vehicle |
11% |
11% |
18% |
Stationary motor vehicle |
1% |
1% |
4% |
Bicycle |
9% |
13% |
17% |
Pedestrian |
2% |
2% |
1% |
Animal |
3% |
1% |
8% |
Fixed object |
14% |
17% |
N/A |
Other |
1% |
7% |
11% |
Non-reported |
N/A |
1% |
N/A |
TOTAL |
100% |
101% |
100% |
For serious crashes, falls remain the leading type at 38% with moving
motor vehicles (24%) the next most frequent type while a fixed object and another bicycle
each accounted for 13%.
Crashes - Activity
For all three studies the most frequently reported activity at the time
of the most recent crash was on-road recreation (54% [60%/64%]). This is not surprising
given that most of the kilometers ridden were for recreation. For serious crashes in the
present study that fraction increases to 62%. Interestingly, while off-road mountain
biking is involved in 20% of all crashes, for non-serious crashes it jumps to 27% while
on-road recreation drops to 48% for these less serious crashes.
Crashes - Facility
Table 4 presents the fraction of crashes by facility type along with
crash rates per million km. Note again that in WA 94 and LAW 74 bike routes and bike laned
streets were combined into a single category. Minor streets are by far the most likely
place for a bicycle crash but this should not be surprising since, as noted above, these
facilities are used most heavily by these cyclists. The table also reveals that the more
serious crashes are more likely to happen on major streets without bike facilities and
that off-road/unpaved trails are the scene of about one-quarter of the minor crashes.
(Note that ‘Population’ under LAB 96 refers to the entire data sample and is
analogous to the WA 94 and LAW 74 data.)
Of additional interest is the experience of these cyclists (and indeed
those in the two earlier studies) on streets either signed as a bike route or having bike
lanes. Crash rates on these facilities are significantly lower than all other facility
types.
Table 5 presents the RELATIVE DANGER INDEX (described above) for each
facility type and all three studies. The RDI makes it easy to grasp the likelihood of
experiencing a crash on the various facilities relative to the kilometers cycled on each
facility and for comparing the facilities to one another. {NOTE: A common misconception is
that LAW 74 concluded that streets with bike lanes were less safe than streets without
bike facilities. This is untrue. See Table 13, page 76 of Kaplan (6).}
Table 5 suggests that virtually all facilities have become
"safer" since 1974. Several factors may be responsible. In 1974 multi-use trails
and bike lanes were much less prevalent and often poorly designed. The 1996 group is older
and has significantly more years of cycling experience. Thus they may just be safer
cyclists able to better handle a variety of road conditions. Major roads without bike
facilities may have better shoulders or wider outside lanes today than 20 years ago. It is
interesting that the performance of minor street has remained very consistent over time.
TABLE 4. Crashes by facility type.
|
|
LAB 96 |
WA 94 |
LAW 74 |
|
|
Serious |
Minor |
Population |
|
|
| Fraction of crashes on: |
|
|
|
|
|
|
Major w/o bike facilities |
29% |
17% |
21% |
20% |
35% |
|
Minor w/o bike facilities |
41% |
43% |
42% |
43% |
54% |
|
Signed bike route only (BR) |
6% |
2% |
3% |
N/A |
N/A |
|
On-street bike lanes (BL) |
4% |
2% |
2% |
N/A |
N/A |
|
On-street bike fac (BR or BL) |
N/A |
N/A |
N/A |
7% |
2% |
|
Multiuse trail |
8% |
9% |
9% |
18% |
10% |
|
Off road/unpaved |
8% |
23% |
18% |
13% |
N/A |
|
Other (most often 'sidewalk') |
5% |
4% |
5% |
1% |
N/A |
|
Totals |
100% |
100% |
100% |
100% |
100% |
|
|
|
|
|
|
|
| Crash rates per million kilometers |
|
|
|
|
|
|
Major w/o bike facilities |
|
|
41 |
69 |
71 |
|
Minor w/o bike facilities |
|
|
59 |
82 |
65 |
|
Signed bike route only (BR) |
|
|
32 |
N/A |
N/A |
|
On-street bike lanes (BL) |
|
|
26 |
N/A |
N/A |
|
On-street bike fac (BR or BL) |
|
|
N/A |
38 |
36 |
|
Multiuse trail |
|
|
88 |
91 |
181 |
|
Off road/unpaved |
|
|
282 |
590 |
N/A |
|
Other (most often 'sidewalk') |
|
|
1026 |
N/A |
N/A |
TABLE 5. Facility Relative Danger Index - RDI
| |
LAB 96 |
WA 94 |
LAW 74 |
Major w/o bike facilities |
0.66 |
0.75 |
1.00 |
Minor w/o bike facilities |
0.94 |
0.98 |
0.92 |
Signed bike route only (BR) |
0.51 |
N/A |
N/A |
On-street bike lanes (BL) |
0.41 |
N/A |
N/A |
On-street bike fac (BR or BL) |
N/A |
0.54 |
0.53 |
Multiuse trail |
1.39 |
1.03 |
2.71 |
Off-road/unpaved |
4.49 |
8.58 |
N/A |
Other (most often ‘sidewalk’) |
16.34 |
N/A |
N/A |
DISCUSSION
The respondents to all three of these studies are clearly U.S. adults
who frequently use bicycles for recreation and transportation and who could be called bicyclists
rather than just people who occasionally ride a bike. While these results are not
representative of the general adult population, they do suggest what adults in this
country are capable of. While a small fraction of this group engages in competitive
cycling (road {9%}and off-road racing {3%}), the vast majority are primarily recreational
riders.
While the upward shift in the age distribution since LAW 74 might be
cause for some concern about the future of cycling in the U.S., it is also encouraging
that more than half of this sample are over 45 years of age. The large fraction of
respondents who are professionals and who report relatively high household incomes also
indicate that serious cycling is not just the province of 'poor college students' with few
transportation options. By the way, they reported spending an average of $1,100 dollars in
1996 on cycling which, in many locales also generated some sales tax revenue to the states
and local governments.
These cyclists ride year-round, after dark and in the rain. They wear
helmets at a very high rate compared to the general cycling population.
When the crash experiences are combined with the average distance
cycled, the 'average' cyclist in this group could be expected to ride for 11 [6/14] years
before having a crash.
CONCLUSION
This survey provides a snap-shot of the cycling habits of adults across
the U.S. who cycle on a regular basis. In many ways this study is unique in that it
acquired data from which to derive crash rates based on distance. Given that the sample
was constructed to gather responses from across the country, and did so, its results
should be of interest to a wide range of public and private groups.
Additional analysis will be performed on the data including looking at
the effect of such parameters as age, miles-ridden-per-year, total years of cycling
experience, and those reporting a crash to those who did not.
ACKNOWLEDGEMENTS
This work was partially funded through grants from the Cascade Bicycle
Club/Seattle and the American Society of Civil Engineers (Human Powered Transportation
Committee). The League of American Bicyclists provided a copy of their membership list.
David Messerschmidt assisted in the data entry. This project is being carried out by the
Human Powered Transportation program at the University of Washington and Bike-Ed, a
consulting company specializing in non-motorized transportation and cyclist education.
REFERENCES
1. Intermodal Surface Transportation Efficiency Act, Public Law
102-240, December 1991.
2. Nationwide Personal Transportation Survey 1990, Bureau of
Transportation Statistics, USDOT, Washington, DC.
3. Zeeger, Charlie et al, Final Report, The National Bicycling and
Walking Study. Federal Highway Administration (FHWA-PD-94-023). 1994
4. Rodgers, Gregory et al, Bicycle Use and Hazard Patterns in the
United States. U.S. Consumer Product Safety Commission. June 1994.
5. Traffic Safety Facts - 1994, USDOT/National Highway Traffic Safety
Administration. August 1995.
6. Kaplan, Jerrold, Characteristics of the Regular Adult Bicycle User.
FHWA, 1975. (NTIS Document PB 258-399)
7. Moritz, William E., Regular Adult Bicyclists in Washington State.
ASCE Transpor-
tation Congress, San Diego. 1995.
Adult Bicyclists in the U.S.
