1
RESULTS OF THE 2008- 09
CAMPUS TRAVEL SURVEY
Institute of Transportation Studies
and
Transportation and Parking Services
University of California, Davis
Report prepared by
Kristin Lovejoy, Institute of Transportation Studies
with
Susan Handy, Environmental Science and Policy
Cliff Contreras, Transportation & Parking Services
September 24, 2009
2
TABLE OF CO ( TE ( TS
Table of Contents ............................................................................................................................... ...... 2
List of Tables......................................................................................................................... ................... 3
List of Figures ............................................................................................................................... ........... 4
Executive summary........................................................................................................................ .......... 5
About the Campus Travel Survey ......................................................................................................... 5
Main findings ............................................................................................................................... ........ 5
Mode split ............................................................................................................................... ......... 5
Carpooling and vehicle occupancy ................................................................................................... 7
Distances traveled ............................................................................................................................. 8
VMT and carbon emissions .............................................................................................................. 9
Awareness of TAPS and other transportation services...................................................................... 9
Introduction................................................................................................................... ......................... 10
About the campus travel survey.......................................................................................................... 10
Development of the survey instrument ............................................................................................... 10
Sampling procedure ............................................................................................................................ 10
Survey administration and recruitment of participants ....................................................................... 12
Response rate and attrition.................................................................................................................. 13
Screening respondents for eligibility .................................................................................................. 15
Sociodemographic composition of respondents completing the survey............................................. 15
Weighting responses by role ............................................................................................................... 17
Reference week ............................................................................................................................... ... 18
Findings....................................................................................................................... ........................... 19
Number traveling to campus ............................................................................................................... 19
Mode split ............................................................................................................................... ........... 20
Vehicle occupancy and ridesharing..................................................................................................... 24
Number of vehicles arriving on campus ............................................................................................. 26
Average Vehicle Ridership.................................................................................................................. 26
Comparison of 2008- 09 and 2007- 08 mode splits .............................................................................. 27
Time arriving on campus .................................................................................................................... 28
Vehicle parking, on and off campus.................................................................................................... 29
Number of riders by transit agency provider ...................................................................................... 30
Self- reported travel time ..................................................................................................................... 31
Residential location and distance from campus .................................................................................. 31
Aggregate person- miles and vehicle- miles traveled ........................................................................... 34
Carbon emissions ............................................................................................................................... 35
Vehicle type........................................................................................................................... ............. 39
Bicycle type........................................................................................................................... ............. 40
Crashes ............................................................................................................................... ................ 40
Bicycle theft ............................................................................................................................... ........ 41
Overnight bike parking ....................................................................................................................... 43
Perceptions about existing and potential TAPS programs .................................................................. 45
Acknowledgements............................................................................................................... ................. 46
References ............................................................................................................................... ............... 46
Appendices..................................................................................................................... ........................ 47
Appendix A: Survey instrument, 2008- 09 Campus Travel Survey..................................................... 47
Appendix B: Comparison of mode choice questions in 07- 08 versus 08- 09 surveys......................... 68
3
Appendix C: Text of the recruitment emails ....................................................................................... 69
Appendix D: Calculation of Average Vehicle Ridership ( AVR) ......................................................... 69
Appendix E: Geocoding and network distances ................................................................................. 73
Appendix F: Fuel energy assumptions used for calculation of CO2 emissions .................................. 74
LIST OF TABLES
Table 1. Change in mode split, 2008- 09 versus 2007- 08.......................................................................... 6
Table 2. Average Vehicle Ridership, 07- 08 and 08- 09.............................................................................. 8
Table 3. Miles traveled and carbon emissions .......................................................................................... 9
Table 4: Sampling and response rate........................................................................................................ 11
Table 5. Sociodemographic characteristics of survey respondents......................................................... 16
Table 6. Weight factors, applied by role.................................................................................................. 18
Table 7. Weather and other events occurring during survey reference week.......................................... 18
Table 8. Percent not traveling to campus on an average weekday and reason given.............................. 19
Table 9. Number traveling to campus by day ......................................................................................... 20
Table 10: Primary weekday mode split ( percent using each mode on an average weekday) ................. 21
Table 11. Percent using each mode as a primary mode at least once during the five- day week............. 21
Table 12. Percent using each mode at some point during a “ typical” commute to campus, by role ...... 22
Table 13: Number of modes used during “ typical” trip to campus......................................................... 23
Table 14: Modes typically used by single- mode versus multi- modal commuters .................................. 23
Table 15. Estimated number of bikers, from various survey questions .................................................. 24
Table 16: Percent driving alone versus ridesharing on an average weekday.......................................... 25
Table 17: Average carpool size ............................................................................................................... 25
Table 18. Projected number of vehicles arriving on an average weekday, by role ................................. 26
Table 19: Average Vehicle Ridership ( AVR), 2008- 09 and 2007- 08 ...................................................... 27
Table 20. Primary weekday mode split, 2008- 09 and 2007- 08 .............................................................. 28
Table 21. Percent change in primary weekday mode split, 2007- 08 to 2008- 09, by mode and role ...... 28
Table 22. Arrivals during the peak period, by day.................................................................................. 29
Table 23. Percent arriving during the peak period on an average weekday, by role............................... 29
Table 24: Percent arriving in vehicles that parked on and off campus, on an average weekday............ 30
Table 25: Use of specific transit providers at some point during “ typical commute” ............................ 31
Table 26: Reported number of minutes spent traveling to campus, by role............................................ 32
Table 27: Residential location................................................................................................................. 32
Table 28. Distance from campus, by role................................................................................................ 33
Table 29. Distance from campus, by mode used..................................................................................... 33
Table 30. Mode used, by distance from campus ..................................................................................... 34
Table 31. Aggregate person- miles and vehicle- miles traveled, by mode and role, daily and annually .. 36
Table 32. Assumed carbon emitted per passenger- mile for each mode .................................................. 37
Table 33. Estimated daily carbon emissions by mode and role .............................................................. 38
Table 34. Estimated annual carbon emissions by mode and role............................................................ 39
Table 35. Type of vehicles used for commuting to and from campus .................................................... 39
Table 36. Type of vehicle technology used for commuting to and from campus ................................... 40
Table 37. Source of bikes used for commuting to campus ..................................................................... 40
Table 38. Bike, walk, and auto crashes ................................................................................................... 41
Table 39. Bike crashes within the last year, by role................................................................................ 41
Table 40: Victims of bike theft, by role................................................................................................... 42
Table 41. In total, how many bikes have been stolen from you on the UC Davis campus?................... 42
4
Table 42. Where had you originally acquired the bike( s) that were stolen? ........................................... 42
Table 43. Do you regularly leave a bike on campus overnight? ............................................................. 43
Table 44. Time elapsed since campus bike was last ridden .................................................................... 44
Table 45. What are the main reasons you leave a bike on campus overnight? ....................................... 44
Table 46. Potential that overnight bikes will be abandoned ................................................................... 44
Table 47. Awareness of transportation services ...................................................................................... 45
Table 48. Awareness of transportation services, 2008 versus 2007 ........................................................ 45
Table 49. Interest in car sharing and bike sharing programs................................................................... 46
Table 50. Number of one- way person trips over the five travel days by mode, 2007- 08....................... 71
Table 51. Weight factors by role for 2007- 08 data.................................................................................. 72
Table 52. Geocoding results using automatic address locator ................................................................ 73
Table 53. Fuel energy assumptions used for calculating carbon emissions ............................................ 74
LIST OF FIGURES
Figure 1. Overall mode split 2008- 09 ....................................................................................................... 5
Figure 2. Overall mode split 2007- 08 ....................................................................................................... 5
Figure 3. Bike mode share 08- 09 and 07- 08, by role................................................................................ 6
Figure 4. Private vehicle mode share, 08- 09 and 07- 08, by role .............................................................. 6
Figure 5. Mode split from within Davis, 08- 09 ........................................................................................ 7
Figure 6. Mode split among students from within Davis, 08- 09 .............................................................. 7
Figure 7. Mode split among employees from within Davis, 08- 09 .......................................................... 7
Figure 8. Mode split from outside Davis, 08- 09....................................................................................... 7
Figure 9. Projected number of vehicles arriving on campus on an average weekday, by occupancy ...... 8
Figure 10. Projected number of people traveling from various distances, by role group......................... 9
Figure 11. Percent who have heard of each service, 08- 09 and 07- 08...................................................... 9
Figure 12. Frequency of completed surveys, November 7– 17, 2008 ..................................................... 13
Figure 13. Number of valid responses by question................................................................................. 14
Figure 14. Sample screenshot of a page from the online survey ............................................................ 67
5
EXECUTIVE SUMMARY
About the Campus Travel Survey
The campus travel survey is a joint effort by the Transportation & Parking Services ( TAPS) on campus
and the Sustainable Transportation Center, part of the Institute of Transportation Studies at UC Davis,
meant to be administered annually each fall by a graduate student at the Institute of Transportation
Studies. The main purpose of the survey is to collect annual data on how the UC Davis community
travels to campus, including mode choice, vehicle occupancy, distances traveled, and vehicle/ bicycle
parking. It also offers an opportunity for TAPS to assess awareness of campus transportation services
and perceptions of mobility options. This year’s survey is the third administration of the campus travel
survey, which was first administered in the spring of 2006- 07 as a pilot effort.
The 2008- 09 survey was administered online in October 2008, distributed by email to a stratified
random sample of 14,031 students, faculty, and staff. About 30 percent ( 4,133 individuals) responded to
this year’s survey, with about 26 percent actually completing it. For the statistics we present throughout
this report, we weighted the responses by role group ( freshmen, sophomore, junior, senior, master’s
student, PhD student, faculty, and staff) so that the proportion of respondents in each group reflects
their proportion in the campus population.
Main findings
Mode split
We estimate that about 92 percent of the campus community travels to campus or works from home on
an average weekday. Among these, about 40 percent bike to campus, 32 percent arrive by private
vehicle, 21 percent ride public transit, 6 percent walk or skate, and 1 percent work from home ( Figure
1). These figures represent the percent of travelers using this mode for most of their trip to campus ( the
portion taking up the greatest time or distance) on an average weekday.
Figure 1. Overall mode split 2008- 09
Bike
40.2%
Drive alone
24.3%
Bus
19.9%
Skate
0.3%
Work from
home
1.4%
Train
0.8%
Carpool /
ride
7.5%
Walk
5.6%
Sample n = 3,603
Projected N = 36,298
Figure 2. Overall mode split 2007- 08
Bike
37.5%
Drive alone
28.8%
Bus
19.0%
Walk
4.9%
Carpool /
ride
5.5%
Other /
multimodal
3.7%
Work from
home
0.7%
Sample n = 3,917
6
Table 1. Change in mode split, 2008- 09 versus 2007- 08
Role group Bike Drive alone Carpool Bus Walk
Work from
home
Undergraduates 5.4% - 4.7% - 0.7% 1.5% 1.4% n/ a
Grad students - 4.4% 4.7% 0.2% - 0.6% - 0.8% n/ a
Faculty 0.6% - 1.8% - 0.3% 0.8% 1.0% 0.2%
Staff - 0.5% - 3.3% 4.5% 1.1% - 0.5% 0.9%
Overall 2.7% - 3.0% 0.5% 0.9% 0.7% 0.8%
Statistically significant difference with p < 0.1 in a two- category χ 2 test of the frequency of those using this mode versus
those using any other mode in 2007- 08 versus 2008- 09.
Statistically significant at p < 0.05.
A comparison of findings from the 2008- 09 and
2007- 08 surveys show a statistically significant
increase in the percent biking ( up about 3
percent) and a decrease in the percent driving
alone ( down about 3 percent) between the fall
2007 and fall 2008 surveys, mostly among
undergraduates ( Figure 1, Figure 2, and Table
1). Among staff, there was also a statistically
significant increase in carpooling, up about 5
percent between 2007 and 2008.
An examination of mode choice by role group
shows substantial variation across classes. In
particular, while the overwhelmin majority of
freshmen ( 74 percent) report using bikes, this
percentage drops substantially among
sophomores ( 43 percent), perhaps associated
with their move off campus ( Figure 3). This
trend is similar to that observed in the 2007- 08
survey, though somewhat less pronounced, with
substantially more sophomores still using bikes
in 2008 compared with 2007.
Meanwhile, the percent of undergraduates
relying on private vehicles to get to campus
( either driving alone, carpooling, or getting a
ride by someone who dropped them off)
increases steadily over the four class years
( Figure 4). Employees are most reliant on
private vehicles, with about 71 percent of staff
and 47 percent of faculty driving or getting a
ride on an average weekday.
Figure 3. Bike mode share 08- 09 and 07- 08, by
role
19%
38%
52%
48%
38% 37%
43%
74%
20%
37%
57%
51%
33% 34% 32%
74%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Freshmen
Sophomores
Juniors
Seniors
Masters
PhD
Faculty
Staff
2008- 09
2007- 08
Figure 4. Private vehicle mode share, 08- 09
and 07- 08, by role
71%
47%
31%
35%
24%
17%
11%
4%
69%
49%
26%
30% 30%
24%
17%
5%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Freshmen
Sophomores
Juniors
Seniors
Masters
PhD
Faculty
Staff
2008- 09
2007- 08
7
Figure 5. Mode split from within Davis, 08- 09
Bus
23.7%
Bike
50.4%
Drive alone
12.9%
Walk
7.0%
Carpool /
ride
4.8%
Train
0.1%
Work from
home
0.8%
Skate
0.4%
Sample n = 3,021
Projected N = 30,731
Figure 6. Mode split among students from
within Davis, 08- 09
Bus
6.2%
Skate
0.0%
Work from
home
2.2%
Train
0.0%
Carpool /
ride
7.1%
Walk
4.0%
Drive alone
35.9%
Bike
44.5%
Sample n = 2,474
Projected N = 24,903
Figure 7. Mode split among employees from
within Davis, 08- 09
Bus
27.4%
Bike
51.6%
Drive alone
9.8%
Walk
7.6%
Carpool /
ride
2.5%
Train
0.1%
Work from
home
0.5%
Skate
0.5%
Sample n = 547
Projected N = 5,509
Figure 8. Mode split from outside Davis, 08- 09
Drive alone
66.3%
Bike
2.6%
Bus
5.9%
Walk
0.5%
Carpool /
ride
17.5%
Train
3.4%
Work from
home
3.7%
Skate
0.1%
Sample n = 908
Projected N = 8,831
Focusing just on the mode split among those
traveling from within Davis ( estimated to be
about 76 percent of the campus population or
24,900 people, including 88 percent of students,
66 percent of faculty, and 42 percent of staff),
about half bike to campus ( Figure 5). The
modes chosen by those living within Davis who
do not bike vary substantially by role group
( Figure 6 and Figure 7). Students, especially
undergraduates, are more likely to take the bus
( 27 percent of students versus 6 percent of
employees), while employees are more likely to
arrive by car ( 43 percent among employees,
versus 12 percent among students).
Of those traveling from outside Davis ( about 24
percent of the campus population, or 8,800
people), the overwhelming majority ( 84
percent) arrives by car on the average weekday.
About three- quarters of these drive alone to
campus, while the remaining share carpool with
at least one other person or get a ride.
Carpooling and vehicle occupancy
Among those traveling to campus ( or working
from home) on an average weekday, we
estimate about 32 percent arrive by private
vehicle ( about 11,540 people), either driving
alone or carpooling. Among these, about 76
percent drive alone, 18 percent carpool, and 5
percent get a ride with someone who drops
them off. Among those who carpooled at any
point during the reference week, the average
number of passengers was 2.59 ( including the
driver). Most people who were dropped off on
campus by a driver who continued on
elsewhere were the sole passenger dropped,
with an average of 1.33 passengers ( excluding
the driver) arriving per vehicle.
Based on the numbers of passengers
respondents reported, we can estimate the total
number of vehicles arriving on campus on an
average weekday, counting one vehicle for each
person driving alone, and a partial vehicle
proportionate to the number of occupants in
vehicles with more than one person. Projecting
our findings to the whole population, we
estimate that about 10,313 vehicles arrive on
8
campus each day, or about 3.84 vehicles per
person affiliated with the Davis campus ( Figure
9). Among these, 82 percent ( 8,429 vehicles)
report parking on campus, 9 percent ( 932
vehicles) park off campus, 5 percent ( 494
vehicles) drop passengers off without parking,
and 4 percent ( 459 vehicles) did not report
where they parked ( no response or “ other” in
response to this question after indicating
vehicle use in a prior question).
Average vehicle ridership ( AVR, as calculated
by the South Coast Air Quality Management
District) is roughly a ratio of the number of
person- arrivals to vehicle- arrivals on campus
over a five- day workweek. If everyone drove
by themselves to campus, the campus AVR
would be 1.0; and so higher AVR values
( greater than 1.0) indicate more carpooling
and/ or use of alternative modes of
transportation. Campus- wide AVR for 2008- 09
( including both on and off- campus residents,
students and employees) is estimated to be
3.51, up slightly from 3.20 in 2007- 08 ( Table
2). This is roughly 3.51 person- arrivals for
every vehicle arriving on campus. Excluding
on- campus residents and students, the AVR in
2008- 09 is estimated to be 1.69 in 2008- 09
( compared with 1.67 in 2007- 08).
Distances traveled
Based on the minimum network distance
between respondents’ reported residential
locations and campus, we estimate that the
average distance traveled to campus is about
6.6 miles overall; and 3.6 miles among
undergraduates, 7.8 miles among graduate
students, 12.0 miles among faculty, and 12.1
miles among staff. Distances are much greater,
once beyond the city limits of Davis with an
average of just 1.8 miles traveled from within
Davis and 23.3 miles for those traveling from
outside of Davis. Undergraduates are most
likely to live close by, with 86 percent living
within 3 miles of campus, compared with 70
percent of graduate students, 48 percent of
faculty, and 34 percent of staff ( see Figure 10).
Figure 9. Projected number of vehicles arriving
on campus on an average weekday, by
occupancy
0
1000
2000
3000
4000
5000
6000
7000
Undergrads
Grad students
Faculty
Staff
Outside Davis
Within Davis
Ride
Carpool
Drive alone
2,858
1,542
( AVR=
779
5,134
5,815
4,498
Table 2. Average Vehicle Ridership, 07- 08 and
08- 09
Overall AVR
Role group
2007- 08 2008- 09
Freshmen 26.39 33.40
Sophomores 6.78 10.67
Juniors 4.46 6.56
Seniors 3.77 4.67
Masters 3.49 2.94
PhD 4.20 3.36
Faculty 2.23 2.35
Staff 1.58 1.62
Student 5.04 5.91
Undergraduates 5.04 7.37
Graduate students 3.94 3.21
Employees 1.67 1.71
Outside Davis 1.33 1.33
Within Davis 5.61 6.32
Overall 3.20 3.51
9
Figure 10. Projected number of people traveling from
various distances, by role group
0 2,000 4,000 6,000 8,000 10,000
Freshmen
Sophomores
Juniors
Seniors
Masters
PhD
Faculty
Staff
Less than 3 miles
3 to 20 miles
More than 20 miles
Figure 11. Percent who have heard of each service, 08- 09
and 07- 08
44%
63%
39%
34%
33%
13%
14%
49%
34%
49%
84%
58%
0% 20% 40% 60% 80% 100%
Discounted transit passes
Carpool/ vanpool program
Emergency ride home service
24 free parking days for carpoolers/ transitpoolers
Online ridematching service
Yolo TMA Commuter Club
www. sacregion511. org
TAPS motorist assistance program
Comet in- vehicle parking meters
Bike lock- cutting service
UC Davis Bike Auction
Ten bike tire air stations around campus
2008- 09
2007- 08
VMT and carbon emissions
Based on distance from campus, mode
choice, and vehicle occupancy, we
generate rough estimates of the total
number of vehicle- miles traveled
( VMT) on the way to and from
campus on an average weekday, as
well as the carbon emissions
associated with this travel. We
estimate that the campus community
covers about 425,000 miles per day
roundtrip, generating about 279,000
vehicle- miles of travel (). This
generates approximately 9 pounds
equivalent of carbon per person daily,
ranging from 0.9 pounds per freshman
to 19.1 pounds per staff person, or
30.3 pounds, on average, among those
driving alone to campus.
Awareness of TAPS and other
transportation services
Respondents were presented a list of
services and asked to indicate for
each, “ It’s new to me,” “ I’ve heard of
it, but never used it,” or “ I’ve used it.”
Awareness of services appears
somewhat higher in 2008- 09
compared with 2007- 08, as shown in
Figure 11.
Table 3. Miles traveled and carbon emissions
Generated daily ( on an average weekday)
Role group People Aggregate
person- miles
Aggregate
vehicle- miles
Lbs. equiv.
of CO2
Lbs. CO2 per
person- mile
Lbs. CO2
per person
Undergraduates 22,612 138,162 69,613 107,448 0.78 4.75
Grad. Students 5,562 65,527 38,005 49,534 0.76 8.91
Faculty 2,079 29,642 18,234 23,005 0.78 11.07
Staff 9,309 191,623 153,029 177,451 0.93 19.06
Total 39,562 424,954 278,881 357,438 0.84 9.03
10
I ( TRODUCTIO (
About the campus travel survey
The campus travel survey is a joint effort by the Transportation & Parking Services ( TAPS) on campus
and the Sustainable Transportation Center, part of the Institute of Transportation Studies at UC Davis,
meant to be administered annually each fall by a graduate student at the Institute of Transportation
Studies. The main purpose of the survey is to collect annual data on how the UC Davis community
travels to campus, including mode choice, vehicle occupancy, distances traveled, and vehicle/ bicycle
parking. It also offers an opportunity for TAPS to assess awareness of campus transportation services
and perceptions of mobility options.
This year’s survey is the third administration of the campus travel survey. The survey was first
administered in the spring of 2006- 07 as a pilot effort, with a second survey conducted in the fall of
2007- 08 ( see Congleton 2009).
Development of the survey instrument
The content of the survey was based on the 2007- 08 survey, retaining key questions relating to mode
choice, residential location, and the roles of participants. An ongoing attempt to refine question
wording has meant that some variables are not directly comparable across years. ( See Appendix A for a
copy of the full 08- 09 survey instrument and Appendix B for an overview of differences in the
questions about mode choice between the 08- 09 and 07- 08 surveys.) In addition to core measures
intended to be included annually, this year’s survey included additional questions focusing on the
following topic areas:
• Extent and reasons for leaving bikes on campus overnight ( questions Q0061 through Q0067)
• Where cyclists have acquired their bikes ( questions Q0019, Q0060, and Q0063)
• Measure of crashes across modes ( bike, walk, drive/ ride; questions Q0069 through Q0071)
• Multitasking during the commute trip ( question Q0021)
• Interest in campus- based bike sharing and car sharing ( questions Q0077 through Q0082)
The online survey was prepared using the Lime Survey software ( http:// www. limesurvey. org/), hosted
on a server at the Institute of Transportation Studies administered by Ning Wan. A sample screenshot of
the online appearance of the survey is shown in Appendix A, followed by a full ( text- based) copy of the
survey instrument.
Staff at TAPS, at the Office of Resource Management and Planning, at Student Affairs Research and
Information, as well as faculty, staff, and students affiliated with the Institute of Transportation Studies
provided feedback on survey content, and assisted with pre- testing the online survey.
Sampling procedure
The goal of the sampling procedure was to draw a sufficiently large sample for reliable statistical
estimates within the following groups: freshmen, sophomores, juniors, seniors, master’s/ professional
students, PhD students, faculty, and staff. We based our assumptions about the total number in each of
these groups on the 2007- 2008 three- quarter averages generated by the Office of Resource
Management and Planning ( shown in the first column of Table 4). We used standard statistical
11
techniques to determine the minimum sample size needed for estimates with a +/- 5% margin of error. 1
We then assumed a 20- percent response rate within each strata ( shown in the second column of Table
4). The number recruited to participate in the study comprised about 35 percent of the overall campus
population, including 20 percent of staff, 78 percent of faculty, and 37 percent of students, and totaling
14,031 people overall.
A stratified random sample was drawn from ostensibly complete lists of UC Davis email addresses
maintained at two different departments within the university. The sampling of student email addresses
was conducted by the Student Affairs Research and Information office. Student addresses were
screened based on students’ level and departmental affiliation, including all academic and professional
students except medical students, who are not based on the Davis campus. The sample of employee
( faculty and staff) email addresses was drawn by Data Administration staff using the Campus Data
Warehouse. Employees were screened to exclude those affiliated with the Medical Center or field
stations, those without salary, Emeritus faculty, Extension School faculty, temporary employees, and
employees without email addresses. In each case, the respective offices drew the sample and submitted
to Kristin Lovejoy an Excel spreadsheet containing only those names and email addresses of
individuals selected for inclusion in the sample.
Table 4: Sampling and response rate
2008- 09 Survey
Role group
Campus
population a Invited
Responses
( at Q0001) e
Response rate
( at Q0001) e
Response rate
( at Q0084) e
Response rate
2007- 08 Survey f
Freshmen 4,597 1,774 474 26.7% 22.3% 26.3%
Sophomores 4,498 1,771 413 23.3% 20.6% 21.8%
Juniors 5,796 1,802 440 24.4% 21.5% 21.4%
Seniors b 7,721 1,831 360 19.7% 17.1% 20.2%
Masters students c 1,926 1,657 338 20.4% 18.0% 19.1%
PhD students d 3,636 1,738 704 40.5% 35.3% 28.2%
Faculty 2,079 1,622 558 34.4% 29.6% 37.0%
Staff 9,309 1,846 846 45.8% 39.2% 49.8%
Total 39,562 14,031 4,133 29.5% 25.5% 28.0%
a Population figures are 2007- 2008 three- quarter averages provided by the Office of Resource Management and Planning.
b Includes senior and post- baccalaureate ( teaching credential) students.
c Includes all academic- program masters students, plus professional- program students in Masters of Law, JD, MBA ( full
time and working professional program), Forensic Science, Masters of Advanced Study, and Master of Preventative Vet
Med. Excludes all School of Medicine students.
d Includes all academic- program doctoral ( D1 and D2) students, plus professional- program students in Veterinary
Medicine ( DVM). Exludes all School of Medicine students.
e Includes valid responses through question Q0001 and question Q0084 of the survey, respectively.
f As reported in Congleton ( 2009).
1 For each strata, the minimum sample size, n, was calculated as
5
z S
e
z S
n 2 2
2 / 2
2 2
/ 2
α
α
+
= , where 5 is the total population, S 2
is the population variance, α / 2 z is the ( 1– α/ 2) th percentile of the standard normal distribution for degree of certainty 1– α,
and e is the acceptable margin of error of the estimate ( Lohr 1999, p. 40). This formula assumes a two- sided test and
includes a finite population correction. We assumed S 2 = 0.25 ( since a binary variable assuming a given value with
probability p has maximum S 2 ≈ p( 1– p) when p= 0.5); we assumed acceptable margin of error of +/– 5% ( e = 0.05); and
we aimed for 95% confidence level ( α= 0.05 or α / 2 z ≈ 1.96). Values of 5 used were those shown in Table 6.
12
Survey administration and recruitment of participants
We invited students, faculty, and staff in our sample to participate in the survey via email to their UC
Davis addresses. Members of the sample were sent at most two emails, an initial email and a follow- up
email approximately one week later. All emails were sent from the address
“ travelsurvey@ ucdavis. edu,” which was set to forward to Kristin Lovejoy’s UC Davis email account. 2
The initial email invitation was sent via bulk mail to all recipients at once with the assistance of the UC
Davis Postmaster. After finding that response to the initial launch overwhelmed the web server hosting
the survey, we decided to send the follow- up email by hand ( using the alias travelsurvey@ ucdavis. edu
via Kristin Lovejoy’s account) in staggered batches of approximately 200 at a time.
The initial email was sent to all members of the sample at about 10AM on Thursday, November 6,
2008, officially launching the survey. We then sent members of the sample a second follow- up email at
some point during the day either on Monday ( November 10) or Wednesday ( November 12) if they had
not yet completed the survey as of midnight the night before, as determined by entry of an exact UC
Davis email addresses into one of the last questions of the survey. Matches between these and the
emails in the initial recruitment list were identified using Microsoft Access and used as the basis for
removing respondents from the list, with about 2550 successful matches. In total, about 9880 members
of the sample were sent reminders on November 10, with approximately 200 email reminders sent
every 15 minutes between 7: 45AM and 7: 45PM. At about 7: 45PM, Gmail shut down Kristin’s account,
and the remaining 1,600 reminders were withheld until Wednesday, November 12 ( since Tuesday,
November 11 was the Veteran’s Day holiday). The remaining 1600 reminders sent on Wednesday were
all to students, as all reminders to faculty and staff had been prioritized to be sent prior to 6PM on
Monday.
In both the initial and the follow- up recruitment emails, recipients were notified of the opportunity to
be entered into a drawing to win one 8GB iPod Nano upon completing the survey. Faculty and staff
recipients were addressed “ Dear UC Davis Employee” and students were addressed “ Dear UC Davis
Student.” The follow- up email included language apologizing for technical difficulties in the initial
launch ( described in detail below) and encouraging respondents to try again. Copies of both
recruitment letters are in Appendix C.
TAPS allocated $ 150 for incentives to participate in the survey, down from past years due university-wide
budget concerns. We decided against offering a cash prize because legal issues make it difficult
for the university to do so. We wanted to offer something that would be comparably appealing to
employees and students, as well as something that was neutral with respect to individuals’
transportation choices. We settled on an iPod, which was also one of several prizes offered in the 2007–
2008 survey. We opted for the 8GB iPod Nano, since it fit the $ 150 price point and because Apple’s
iPod products are thought to be desirable. Entry into a drawing for the iPod was mentioned in the initial
and follow- up recruitment emails, as well as on the first welcome page of the online survey, where the
mention of the iPod was hyperlinked to the section of Apple’s website featuring this product.
We conducted the drawing using the email addresses participants gave upon completing the survey.
2 A technical problem prevented mails sent to travelsurvey@ ucdavis. edu from forwarding to Kristin Lovejoy’s email
account the first day of the survey launch, November 6. In particular, anyone attempting to send mail to the address
travelsurvey@ ucdavis. edu during the day on November 6, including anyone attempting to reply to the email invitation,
had their message bounce. This was a result of the migration from Geckomail to DavisMail that coincidentally occurred
at approximately 9: 45AM on November 6 for Kristin Lovejoy’s ( student) email account, interfering with the forwarding
of mails to her account from the survey account travelsurvey@ ucdavis. edu. IT corrected this problem by November 7.
13
Thus, only respondents who completed the survey and entered an email address were included. Anyone
who attempted the survey and found they were ineligible ( based on the initial screening question
Q0003, “ Do you go to the Davis campus regularly, either for work or classes?”) had been forced to skip
to the end of the survey, but was still invited to give an email address to enter into the drawing. In total,
there were 3,779 non- missing entries of email addresses, although 113 of these respondents had opted
out of the drawing. We generated random numbers for the remaining 3,666 address and then selected
the one with the lowest value as the winner, who we notified via email on November 18, 2008 and
issued the prize shortly thereafter.
Response rate and attrition
A total of 4,133 respondents at least commenced the survey ( responding to question Q0001), with
3,577 completing through question Q0084 ( see a copy of the survey instrument in Appendix A).
Minimum sample sizes were achieved within all strata except senior students ( and master’s students by
question Q0084), with response rates ranging from a high of 46 percent ( 39 percent by question
Q0084) among staff to a low of 20 percent ( 17 percent) among seniors ( see Table 4). The overall
response rate was 30 percent ( 26 percent). A histogram showing the numbers of respondents
completing the survey at a given hour over the course of the study period is shown in Figure 12 ( note
that incomplete surveys are not shown in this figure).
Figure 12. Frequency of completed surveys, 5 ovember 7– 17, 2008
0
50
100
150
200
250
300
350
400
Before 11AM
11AM- 12PM
12- 1PM
1- 2PM
2- 3PM
3- 4PM
4- 5PM
5- 6PM
6- 7PM
7- 8PM
8- 9PM
9- 10PM
10- 11PM
11PM- 12AM
before 6AM
6AM- 12PM
12PM- 6PM
6PM- 12AM
before 6AM
6AM- 12PM
12PM- 6PM
6PM- 12AM
before 6AM
6AM- 12PM
12PM- 6PM
6PM- 12AM
before 6AM
6AM- 12PM
12PM- 6PM
6PM- 12AM
before 6AM
6AM- 12PM
12PM- 6PM
6PM- 12AM
before 6AM
6AM- 12PM
12PM- 6PM
6PM- 12AM
before 6AM
6AM- 12PM
12PM- 6PM
6PM- 12AM
before 6AM
6AM- 12PM
12PM- 6PM
6PM- 12AM
before 6AM
6AM- 12PM
12PM- 6PM
6PM- 12AM
before 6AM
6AM- 12PM
12PM- 6PM
6PM- 12AM
6- Nov 7- Nov 8- Nov 9- Nov 10- Nov 11- Nov 12- Nov 13- Nov 14- Nov 15- Nov 16- Nov
It is likely that technical difficulties at least partially depressed overall response. In particular, response
during the first hours after the survey launch on November 6 overwhelmed the web server hosting the
survey, causing slow performance as well as some users’ inability to access the site altogether. By
12: 30PM on November 6, approximately 540 respondents had started the survey but not finished,
perhaps indicating that technical difficulties were encountered. Although this figure dipped to a low of
14
about 430 by the following morning, it seems that at least 400 started the survey but abandoned it
during the first day of the launch, either due to technical difficulties or for other reasons. Responses by
question number are shown in Figure 13. In particular, the numbers shown are 4133 minus the number
of missing responses for a given question. ( In this figure, if a respondent was not shown a question,
then their response is not counted as missing; as a result, some questions shown to relatively few
respondents have particularly high numbers valid responses.) Response was particularly low on the
questions about multitasking activities people do during their trip to campus ( Q0021 series of
questions). Questions Q0085 ( years at UC Davis) and Q0092 ( highest level of education) are
particularly high because derived from role group among students.
Figure 13. 5 umber of valid responses by question
Q0008 n = 3,935
Q0001 n = 4,133
Q0087 n= 3,475
Q0029 n= 3,739
Q0075a n= 3,573
Q0021e n= 3,355
3,200
3,300
3,400
3,500
3,600
3,700
3,800
3,900
4,000
4,100
4,200
2
9
9
10
10
11
11
12
12
16
20
20
20
21
21
21
25
30
34
38
42
46
57
60
61
64
65
66
71
72
72
75
75
75
81
86
86
90
91
Question number
In addition, it is presumed that other respondents may have tried to access the survey but failed to be
able even to start it. For instance, according to web log files, there were approximately 1,500 people
who attempted and failed to access the web server between 10: 30 and 11: 30AM on November 6. At
approximately 12: 30PM, we added a message to the first page of the survey stating, “ Note: If your
survey is loading slowly, please try back again a little later. We apologize for the inconvenience, and
thank you for your patience.” This message was retained as part of the survey for the remainder of the
study.
Because any emails or replies sent to the address “ travelsurvey@ ucdavis. edu” were bouncing
throughout the day on November 6, rather than forwarding to Kristin Lovejoy’s email address as
planned ( see footnote 2), we expect that less feedback reached us than was attempted on the part of
survey respondents. Even without the option of replying directly to the email invitation, approximately
20 survey respondents contacted one or all of Kristin Lovejoy, Susan Handy, and/ or Cliff Contreras ( the
three names listed in the signature of the recruitment email, presumably reached by looking up their
email addresses in the directory) to notify them of difficulties they encountered with the survey on
November 6.
15
Screening respondents for eligibility
While incomplete survey responses were retained in the dataset, cases were excluded from the dataset
based on two criteria. First, because we planned to weight the results by role group ( freshmen,
sophomore, junior, etc.), we excluded cases whose role could not be verified based on responses to
questions Q0001 and Q0002. Of the 226 cases with missing data for question Q0002, 211 could be
recoded based on write- in descriptions in the “ other field.” In particular, we chose to assign fifth- year
students and one “ international” student as seniors; law and MBA students as master’s students;
veterinary students and Ed. D/ CANDEL students as PhD students; and post- docs and visiting scholars
as staff. Five of the remaining 15 cases ( with missing data for both Q0002 and Q0002_ other) had email
addresses entered, which we used to look up the individuals in the UC Davis directory, and assign a
Q0002 value based on the information listed.
Secondly, cases were excluded from the dataset if the respondent indicated that they did not regularly
travel to campus, based on responses to Q0003 and write- in responses to Q0013 (“ What was the main
reason that you did not travel on campus last week?”), revealing they do not travel to campus regularly.
Ideally, we would have liked to include in the sample anyone who usually travels to campus regularly
( including those temporarily elsewhere, such as on sabbatical or maternity leave), but exclude those
whose main work is elsewhere. Unfortunately, question Q0003 left it somewhat ambiguous as to what
was meant by “ regularly” as well as how those away for the quarter ( or for some other temporary leave
period) should response. In total, 214 respondents reported that they do not travel to campus regularly.
We did not ask these respondents more about their circumstances, and they may be some mix of people
who always work elsewhere and people who are only away temporarily.
The sample frame we developed ( that is, the email list from which a random sample was drawn by the
Registrar and Data Warehouse for students and staff, respectively) was intended to only include those
whose affiliation implies a location on the Davis campus. However, it is expected that some stationed
off- campus would be included in the sample erroneously, either due to errors/ micategorization or
because the database has no way of identifying certain individuals’ off- campus locations. Therefore,
despite the fact that the sample frame was intended to exclude these individuals, some of these 214 may
be those who slipped through this screening. As mentioned above, others of the 214 may be undergoing
some temporary situation keeping them away, such as sabbatical, field work, semester abroad, or
family leave. In addition to these 214 answering “ no,” to Q0003, among those who respond “ yes” to
Q0003, 97 reported that they did not travel to campus at all during the reference week, and 13 of these
revealed in Q0013_ other that this was because they do not actually regularly work on campus,
traveling to campus once per month or less. We chose to exclude these from the sample as well.
( However, we retained in the sample those who indicated that they were away temporarily, such as for
sabbatical, bereavement leave, study abroad, or a joint teaching appointment at another UC campus.)
Sociodemographic composition of respondents completing the survey
Table 5 shows sociodemographic characteristics of the unweighted sample. The sample includes more
female than male respondents. This means that males are underrepresented among student respondents,
especially undergraduates, though somewhat overrepresented among employee respondents. In
particular, females comprise about 66 percent of the sample compared with 56 percent of the
population of undergraduates; 57 percent of respondents versus 51 percent of the population of
graduate students; and 52 percent of respondents versus 59 percent of the population of employees. 3
3 Figures for the composition of the campus population by gender are drawn from “ Student Headcount by Gender, Fall
16
For the purposes of analysis, we assume that respondents are roughly similar to the rest of the
population within their role group ( freshmen, sophomore, etc.) with respect to these attributes and
others that may matter for transportation choices. For this reason, we weight the sample by role, as
described below.
Table 5. Sociodemographic characteristics of survey respondents
Role group
Characteristic
Undergraduates Grad students Employees All
Gender: valid n 1460 913 1204 3577
% male 34.0% 43.2% 48.2% 41.1%
Age: valid n 1420 898 1154 3472
% < 20 years old 69.9% 0.1% 0.4% 28.7%
% 20 to 29 years old 29.2% 75.6% 12.5% 35.6%
% 30 to 39 years old 0.6% 20.4% 22.2% 12.9%
% 40 to 49 years old 0.1% 2.6% 26.6% 9.6%
% 50 to 59 years old 0.1% 1.1% 26.2% 9.0%
% 60+ years old 0.2% 0.2% 12.1% 4.2%
Household size: valid n 1424 910 1192 3562
% alone 4.9% 17.5% 13.2% 10.9%
% 2 people 21.7% 46.7% 40.2% 34.4%
% 3 to 5 people 55.5% 34.1% 44.6% 46.3%
% 6 or more people 11.0% 1.8% 1.9% 5.6%
% in a dormitory 6.9% 0.0% 0.1% 2.8%
Highest level of education: valid n 1687 1042 1173 3902
% High school or less 0.0% 0.0% 2.6% 0.8%
% Some college 100.0% 0.0% 7.4% 45.5%
% 2- year degree 0.0% 0.0% 4.6% 1.4%
% Bachelor’s degree 0.0% 0.0% 20.3% 6.1%
% Some grad school 0.0% 32.4% 3.8% 9.8%
% Grad degree 0.0% 67.6% 61.2% 36.4%
Total household income: valid n 0 0 1091 n/ a
$ 0 - $ 19,999 n/ a n/ a 0.7% n/ a
$ 20,000 - $ 39,999 n/ a n/ a 5.2% n/ a
$ 40,000 - $ 59,999 n/ a n/ a 10.9% n/ a
$ 60,000 - $ 79,999 n/ a n/ a 13.1% n/ a
$ 80,000 - $ 99,999 n/ a n/ a 14.6% n/ a
$ 100,000 - $ 119,999 n/ a n/ a 16.2% n/ a
$ 120,000 - $ 139,999 n/ a n/ a 9.9% n/ a
$ 140,000 - $ 159,999 n/ a n/ a 8.6% n/ a
$ 160,000 - $ 179,999 n/ a n/ a 5.4% n/ a
$ 180,000 - $ 199,999 n/ a n/ a 5.5% n/ a
Greater than $ 200,000 n/ a n/ a 9.8% n/ a
Total respondents ( total n) 1687 1042 1404 4133
The statistics shown are unweighted, based on responses to survey questions Q0084, Q0089, Q0091, Q0092, and
Q0093. Questions Q0092 ( education) and Q0093 ( income) were not asked of students. Percentages reported are among
valid ( non- missing) responses to each question.
2008,” “ Employees by Gender and Ethnicity, Fall 2008,” and “ Teaching Faculty by Gender, Fall 2008” available on the
UC Davis Facts website, online at http:// facts. ucdavis. edu/. These population counts include medical ( non- Davis
campus) affiliates who are excluded from the survey sample. In addition, the employee count includes employed
students, who are not included as employees in the survey sample.
17
Weighting responses by role
Respondents were assigned role categories based on their responses to question Q0001, “ What is your
primary role on campus? ( Faculty, Staff, Student)” and question Q0002, “ Are you an undergraduate or
graduate student? ( Freshman, Sophomore, Junior, Senior, Master’s student, PhD student, Other).” We
recoded 72 “ Other” responses into the appropriate category as indicated by what respondents wrote in
by hand, grouping 5 th - year students with seniors; law and business students with Master’s students; and
Ed. D and Vet Med students with PhD students. We also assigned categories to 5 respondents who had
skipped responding to questions Q0001 and/ or Q0002 by using the email addresses they gave to look
up their role in the UC Davis directory. After these recodes, there was only 1 respondent with an
unknown role excluded from the sample.
For all results presented in this report, responses are weighted to be representative of the campus
population by role ( freshmen, sophomores, juniors, seniors, master’s/ profession students, PhD students,
faculty, and staff). That is, we apply a weight factor to each case in a given role group so that the
group’s proportion in the sample is the same as their proportion in the overall population. To
accomplish this, the appropriate weight factor is a ratio of the population share to the sample share for
each role group. That is, with 5 total population, n in the sample, and 5 i in role group i in the
population ( for instance, freshmen), and ni of role group i in the sample, we apply the weight factor Wi
= ( 5 i/ 5 ) / ( ni/ n) to all cases in role group i. Applying the weight factors alters the apparent distribution
of respondents by role, but the overall sample size is unchanged. In instances where we would like to
expand the sample to a projection of the full population, we weight each case by an expansion factor Ei,
equal to ( 5 i / ni). Applying the expansion factors alters both the distribution of respondents by role, and
inflates the sample to the size of the population, or 39,562.
Although the number of valid responses varies from question to question ( that is, n and ni), we found
that which n we use to come up with weight factors does not substantially influence the results for most
variables. ( For instance, generating weights based on the n= 3,935 valid responses to question Q0008
versus weights based on the n= 3,577 valid responses to question Q0075 causes only a small variation
in the results, up to 0.1 percentage points. This is not surprising, to the extent that attrition is
comparable across role groups.) Therefore we use the same set of weight factors for most variables,
based on the distribution of roles among the n= 3,935 valid responses to question Q0008, the first
question among the core series of questions regarding travel to campus during the reference week.
However, for variables relying on geocoding of respondents’ residential location, we generated a
separate set of weight factors, based on the 3,451 cases successfully geocoded ( by zip code and cross
streets in given in questions Q0050 and Q0051 or Q0087 and Q0088, or dorm name given in question
Q0005; see Appendix E) and with non- missing mode data from question Q0011. Both sets of weights
are shown in Table 6.
18
Table 6. Weight factors, applied by role
Based on valid responses to
question Q0008
Based on valid responses to question Q0011
and successful geocoding of home location
Role group
( i)
Population
( 5 ) Valid
responses
( n)
Weight
factor
( 5 i/ 5 )/( ni/ n)
Expansion
factor
( 5 i / ni)
Weighted
sample
size
Valid
responses
( n)
Weight
factor
( 5 i/ 5 )/( ni/ n)
Expansion
factor
( 5 i / ni)
Weighted
sample
size
Freshmen 4,597 444 1.0298 10.3536 457.2 423 0.9480 10.8676 401.0
Sophomores 4,498 403 1.1101 11.1613 447.4 339 1.1574 13.2684 392.4
Juniors 5,796 419 1.3759 13.8329 576.5 353 1.4323 16.4193 505.6
Seniors 7,721 345 2.2260 22.3797 768.0 291 2.3144 26.5326 673.5
Masters 1,926 329 0.5823 5.8541 191.6 287 0.5854 6.7108 168.0
PhD 3,636 673 0.5374 5.4027 361.7 601 0.5277 6.0499 317.2
Faculty 2,079 523 0.3954 3.9751 206.8 464 0.3908 4.4806 181.4
Staff 9,309 799 1.1588 11.6508 925.9 693 1.1718 13.4329 812.0
Overall 39,562 3,935 1.0000 10.0539 3,935.0 3,451 1.0000 11.4639 3,451.0
Reference week
The main statistics we measure are based on questions asking respondents about their activity during
each of the seven days of the week prior to receiving the invitation to complete the survey. The
calendar week serving as the reference week varied depending on when the respondent chose to
complete the survey. Following the initial invitation to complete the survey, distributed on November 6
( Thursday), the reference week specified in the survey instructions was October 27– November 2
( Monday– Sunday). Following the follow- up invitations distributed on November 10 and 12 ( Monday
and Wednesday), the reference week specified was November 3– 9 ( Monday– Sunday). Table 7 notes
events occurring these days that may have affected respondents’ travel and other activities, including
rain on the Friday of the first reference week, the Halloween holiday on the Friday of the first reference
week, and the presidential election held on the Tuesday of the second reference week.
Table 7. Weather and other events occurring during survey reference week
Temperature ranges, precipitation, and notable events a
Day
Week 1: October 27- November 2 Week 2: November 3- 9
Monday 43 – 84 º F 51 – 65 º F
Wind ( 24 mph gusts)
Tuesday 42 – 83 º F 46 – 63 º F
Wind ( 34 mph gusts)
Election day
Wednesday 43 – 82 º F 37 – 59 º F
Some fog
Thursday 52 – 79 º F 44 – 68 º F
Friday 57 – 72 º F
Rain later in the day ( total 0.36”)
Halloween holiday
Davis elementary schools have the day off
43 – 70 º F
Some fog
Saturday 57 – 70 º F
Rain ( 0.44”) & wind ( 25 mph gusts)
45 – 68 º F
Sunday 55 – 66 º F
Rain ( 1.44”) & wind ( 21 mph gusts)
48 – 65 º F
Wind ( 23 mph gusts)
a Weather data are for Sacramento, as reported in the Farmer’s Almanac, available online by city and date at
http:// www. almanac. com/ weatherhistory.
19
FI ( DI ( GS
This section summarizes some of the results from the survey. Throughout this section, data presented
are weighted by role, as described above, and therefore the sample sizes reported are referred to as the
“ weighted sample.” Many statistics are presented by role group as defined above ( freshmen,
sophomores, juniors, seniors, master’s students, PhD students, faculty, and staff). In addition, some are
also broken down by students ( including freshmen through PhD student role- group categories),
undergraduates ( freshmen through senior role- group categories), graduate students ( master’s and PhD
student role- group categories), employees ( faculty and staff role- group categories), within Davis ( those
living on campus or elsewhere in Davis among all role- group categories), and outside Davis ( those
living outside of Davis among all role- group categories).
( umber traveling to campus
About 92 percent of the sample traveled to campus on the average weekday during their reference week
( see Table 8), with a low of about 84 percent traveling to campus on Friday ( Table 9). For any days not
traveled, respondents were asked to indicate the reason ( Table 8).
Table 8. Percent not traveling to campus on an average weekday and reason given
Among those not traveling, percent giving each reason
Role group
Not
traveling Day off
Working
from
home
Traveling
for work Sick Vacation CWW a Other b
Weighted
sample
Projected
pop.
Freshmen 4.7% 43.2% 2.9% 17.3% 11.5% 5.8% 3.8% 15.3% 457 4,594
Sophomores 3.6% 58.8% 5.5% 13.7% 1.4% 1.4% 2.7% 16.4% 447 4,493
Juniors 7.9% 70.9% 1.2% 3.6% 3.0% 1.8% 4.2% 15.0% 577 5,791
Seniors 8.6% 49.3% 3.4% 8.1% 12.2% 2.7% 0.7% 23.6% 768 7,733
Masters 12.6% 62.0% 12.0% 8.2% 1.4% 1.0% 2.4% 13.0% 192 1,929
PhD 10.2% 28.0% 37.0% 9.6% 2.9% 2.9% 2.9% 16.6% 362 3,642
Faculty 19.3% 22.9% 27.9% 20.8% 0.4% 4.5% 10.5% 12.8% 207 2,078
Staff 13.6% 18.6% 15.5% 11.4% 11.8% 10.2% 6.5% 26.2% 925 9,300
Students 7.5% 51.8% 9.9% 8.8% 6.4% 2.6% 2.5% 17.9% 2,802 28,182
Undergraduate 6.6% 56.1% 2.9% 8.6% 8.1% 2.7% 2.4% 19.0% 2,249 22,611
Graduate 11.0% 41.5% 27.1% 9.0% 2.3% 2.1% 2.7% 15.2% 553 5,571
Employees 14.6% 19.7% 18.5% 13.7% 9.1% 8.8% 7.4% 23.0% 1,132 11,378
Outside Davis 18.4% 39.2% 17.2% 7.9% 5.5% 4.9% 4.7% 20.7% 909 9,144
Within Davis 6.9% 36.5% 10.9% 13.4% 9.2% 5.6% 4.7% 19.7% 3,025 30,416
Overall 9.6% 37.7% 13.7% 10.9% 7.6% 5.3% 4.7% 20.1%
Weighted sample 376 142 51 41 28 20 18 76 3,934
Projected pop. 3,782 1,425 518 414 287 201 177 762 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based on responses to
questions Q0008, Q0009, Q0013, and Q0023 through Q0027. For respondents who did not travel any days during the
reference week ( based on Q0008), the reason for not traveling given in Q0013 was assumed to apply to all five
weekdays. For respondents not traveling only on certain days, the reason by day was measured in questions Q0023
through Q0027. We calculate the share of days among the five weekdays that each individual did not travel to campus
for a given reason; the average of these is equivalent to the percent of people not traveling for a given reason on an
average weekday.
a Day off as part of a 3/ 36, 4/ 40, or 9/ 80 compressed work week.
b Respondents selecting “ Other ( e. g. jury duty),” plus those who wrote in a description of an “ other” reason that could not
be otherwise categorized ( in question Q0013 only), plus those not responding to the question.
20
Table 9. 5 umber traveling to campus by day
Day
Number traveling
( weighted sample)
Percent of
sample
Population
projection
Monday 3,604 91.6% 36,241
Tuesday 3,623 92.1% 36,440
Wednesday 3,641 92.6% 36,619
Thursday 3,634 92.4% 36,546
Friday 3,294 83.7% 33,127
Saturday 761 19.3% 7,655
Sunday 755 19.2% 7,593
No days 83 2.1% 836
Total 3,934 100.0% 39,562
Based on responses to Q0009. Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6).
Mode split
For physical trips to campus, mode was determined by asking “ What was your primary means of
transportation from home to your first campus destination each of these days? If you used more than
one means of transportation on a given day, indicate whichever one you used to travel the greatest
amount of distance” ( question Q0011). Thus the modes identified are those used for most of the trip,
and only on the way to campus at the beginning of the day. For each respondent, we calculated percent
of days out of the five- day week that a given mode was used. ( For instance if someone biked one day,
her bike share would be 20 percent.) The overall mode split reported in Table 10 represents the average
shares across all respondents, which is equivalent to the percent of all people using each mode on an
average weekday.
Because we consider working from home ( telecommuting) to be a substitute for physical travel, it is
included as a “ mode” in our mode split percentages. Telecommuting data are based on responses to
questions Q0013 and Q0023– Q0027. If working from home was indicated in Q0013 as the reason for
not traveling, we assumed that the individual did so all five weekdays. We excluded all other reasons
for not traveling from the mode split. Therefore for the mode split reported, the denominator is all
people who physically traveled to campus plus those who worked from home on a given day.
As shown in Table 10, we find that on an average weekday, about 40 percent of people bike to campus
( projected 14,592 people), 32 percent arrive by car ( 11,543 people), and 21 percent ride transit ( 7,513
people). The share biking is highest among freshmen and lowest among staff and those living outside
Davis. Transit ridership ( especially bus) is particularly high among undergraduates.
Table 11 shows the percent of people using each mode as their primary mode on the way to campus at
least one day during the reference week. For instance, while we estimate that about 290 ride the train to
campus on an average weekday ( Table 10), we estimate that 795 ride the train at some point during the
average week ( Table 11). The denominator for the percentages shown in Table 11 is the entire campus
population, including those who did not travel to campus.
21
Table 10: Primary weekday mode split ( percent using each mode on an average weekday)
Among those traveling, percent:
Role group
Percent
traveling a Bike Walk Skate
Drive
alone
Carpool
or get a
ride
Bus Train
Work
from
home
Weighted
sample
Projected
pop.
Freshmen 95.4% 73.7% 12.0% 1.3% 2.1% 1.5% 8.6% 0.7% 0.1% 456 4,597
Sophomores 96.6% 42.7% 2.4% 0.3% 6.6% 4.8% 42.8% 0.3% 0.2% 446 4,498
Juniors 92.2% 38.1% 5.7% 0.6% 12.7% 4.5% 38.2% 0.2% 0.1% 575 5,796
Seniors 91.7% 36.6% 8.8% 0.2% 17.3% 6.4% 29.9% 0.4% 0.3% 768 7,721
Masters 88.9% 48.2% 4.7% 0.0% 30.6% 4.5% 8.3% 2.0% 1.7% 192 1,926
PhD 93.6% 52.4% 5.3% 0.2% 25.4% 5.1% 5.7% 1.9% 4.0% 362 3,636
Faculty 86.1% 37.5% 4.2% 0.0% 39.8% 6.9% 2.5% 2.7% 6.3% 206 2,079
Staff 88.5% 19.3% 1.6% 0.0% 53.9% 16.6% 5.4% 0.8% 2.4% 924 9,309
Students 93.2% 46.9% 6.9% 0.5% 14.0% 4.7% 25.6% 0.7% 0.8% 2,799 28,174
Undergraduate 93.5% 45.9% 7.4% 0.5% 10.8% 4.6% 30.2% 0.4% 0.2% 2,246 22,612
Graduate 92.0% 51.0% 5.1% 0.1% 27.2% 4.9% 6.6% 1.9% 3.3% 553 5,562
Employees 88.1% 22.6% 2.1% 0.0% 51.4% 14.9% 4.9% 1.1% 3.1% 1,130 11,388
Outside Davis 84.8% 2.6% 0.5% 0.1% 66.3% 17.5% 5.9% 3.4% 3.7% 908 8,831
Within Davis 93.8% 50.4% 7.0% 0.4% 12.9% 4.8% 23.7% 0.1% 0.8% 3,021 30,731
Overall 91.7% 40.2% 5.6% 0.3% 24.3% 7.5% 19.9% 0.8% 1.4% 3,929
Projected pop. 36,298 14,585 2,027 123 8,826 2,715 7,218 286 518 36,592
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based on responses to
questions Q0011 ( for physical means of transportation) and questions Q0013 and Q0023 through Q0027 for those
working from home. Percentages are calculated as the percent of five weekdays that an individual respondent used a
particular mode; the average over all respondents represents the percent using this mode on an average weekday.
a Includes all those physically traveling to campus as well as those working from home, which we treat as virtual travel,
but excludes those not traveling to campus for any other reason ( sick, day off, etc.).
Table 11. Percent using each mode as a primary mode at least once during the five- day week
Percent of people using each as their primary mode to campus at least once
Role group
Bike Walk Skate
Drive
alone
Carpool
or get a
ride
Bus Train
Work
from
home
Other
no
travel
Sample
size
Projected
pop.
Freshmen 80.1% 21.7% 2.0% 2.9% 4.3% 14.0% 2.0% 0.2% 9.9% 456 4,597
Sophomores 54.5% 5.2% 0.5% 11.2% 12.7% 57.0% 0.2% 0.2% 9.2% 446 4,498
Juniors 46.9% 11.5% 1.4% 18.9% 9.6% 49.3% 0.2% 0.2% 23.0% 575 5,796
Seniors 44.9% 12.8% 0.6% 26.1% 14.5% 42.6% 1.2% 0.9% 21.7% 768 7,721
Masters 52.6% 7.6% 0.0% 41.9% 9.1% 13.1% 2.7% 4.0% 31.3% 192 1,926
PhD 60.6% 10.0% 0.3% 34.6% 12.0% 10.3% 3.1% 9.8% 17.7% 362 3,636
Faculty 43.3% 6.3% 0.0% 49.6% 12.3% 4.6% 5.0% 15.7% 29.1% 206 2,079
Staff 22.5% 3.1% 0.0% 62.1% 22.1% 7.4% 1.0% 4.8% 27.2% 924 9,309
Students 55.1% 12.0% 0.9% 20.6% 10.8% 35.4% 1.3% 1.9% 18.2% 2,799 28,174
Undergraduate 54.5% 12.7% 1.1% 16.6% 10.8% 41.4% 0.9% 0.5% 17.2% 2,246 22,612
Graduate 57.8% 9.1% 0.2% 37.2% 11.0% 11.2% 3.0% 7.8% 22.4% 553 5,562
Employees 26.3% 3.7% 0.0% 59.8% 20.3% 6.9% 1.7% 6.8% 27.6% 1,130 11,388
Outside Davis 4.4% 0.8% 0.2% 73.7% 21.9% 7.4% 5.3% 7.4% 36.3% 908 8,831
Within Davis 59.6% 12.3% 0.8% 19.4% 11.0% 33.1% 0.3% 2.1% 16.3% 3,021 30,731
Overall 46.8% 9.6% 0.6% 31.9% 13.6% 27.2% 1.4% 3.3% 20.9% 3,929
Weighted sample 1,840 379 25 1,254 533 1,069 57 130 821 3,929
Projected pop. 18,530 3,814 254 12,627 5,365 10,762 571 1,304 8,266 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based on responses to
questions Q0011 ( for physical means of transportation) and questions Q0013 and Q0023 through Q0027 for those
working from home.
22
In addition to questions about respondents’ travel during the particular reference week preceding the
survey, we also asked respondents more detailed questions about their “ typical travel between home
and campus destinations, whether or not it’s what you did last week” ( questions Q0029– Q0048). In this
section, respondents reported if they typically used multiple modes on their way to campus and what
modes they use on each segment of their trip. ( Any walking longer than 7 minutes was counted as a
separate mode.) Table 12 shows the percent of people who report using each mode for at least some
portion of their typical travel to campus. For most modes, the percent of people identifying a given
mode as part of their typical travel ( Table 12) is roughly similar to the percent using each primary mode
on the average weekday during the reference week ( Table 10), with the exception of biking and
walking, which presumably are the most likely modes to be paired with other modes as a part of a
multimodal trip to campus. In total, about 11 percent of the sample reported that they typically use
more than one mode ( Table 13). About 50 percent of multimodal trips involve bikes and 58 percent
involve transit ( Table 14).
Table 12. Percent using each mode at some point during a “ typical” commute to campus, by role
Role group Bike Walk Skate
Drive
alone
Carpool
or ride
Bus or
train
Weighted
sample
Projected
pop.
Freshmen 79.1% 13.9% 1.2% 2.7% 2.4% 9.2% 423 4,597
Sophomores 53.0% 3.9% 0.5% 5.5% 4.5% 43.0% 423 4,498
Juniors 45.7% 8.9% 1.5% 14.9% 4.0% 40.2% 554 5,796
Seniors 44.0% 14.7% 0.3% 18.7% 4.3% 31.5% 728 7,721
Masters 51.8% 10.2% 0.0% 35.5% 3.2% 11.8% 182 1,926
PhD 63.9% 7.9% 0.2% 27.1% 4.5% 9.5% 347 3,636
Faculty 44.0% 5.2% 0.0% 45.2% 9.6% 7.0% 198 2,079
Staff 24.4% 5.2% 0.0% 56.1% 18.7% 6.6% 874 9,309
Students 54.5% 10.5% 0.7% 15.5% 3.9% 27.4% 2,657 28,174
Undergraduate 53.2% 10.9% 0.8% 11.9% 3.9% 31.6% 2,129 22,612
Graduate 59.7% 8.7% 0.1% 30.0% 4.0% 10.3% 529 5,562
Employees 28.0% 5.2% 0.0% 54.1% 17.0% 6.7% 1,072 11,388
Outside Davis 11.8% 7.4% 0.2% 72.2% 19.9% 12.4% 865 8,831
Within Davis 57.5% 9.4% 0.6% 12.8% 4.0% 24.2% 2,865 30,731
Overall 46.9% 8.9% 0.5% 26.6% 7.7% 21.4%
Weighted sample 1,749 333 18 992 286 799 3,730
Projected pop. 18,552 3,535 195 10,518 3,035 8,480 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based on responses to
questions Q0031, Q0032, Q0034, Q0037, Q0040, Q0043, and Q0046.
23
Table 13: 5 umber of modes used during “ typical” trip to campus
Percent typically using…
Role group
1 mode 2 modes 3+ modes
Weighted
sample
Projected
pop.
Freshmen 91.3% 7.8% 1.0% 424 4,597
Sophomores 89.5% 9.4% 1.0% 423 4,498
Juniors 85.6% 12.9% 1.5% 556 5,796
Seniors 86.2% 12.5% 1.2% 728 7,721
Masters 87.5% 10.2% 2.2% 182 1,926
PhD 88.4% 8.5% 3.1% 347 3,636
Faculty 90.1% 7.0% 3.0% 199 2,079
Staff 90.3% 7.4% 2.3% 874 9,309
Students 87.8% 10.7% 1.5% 2,660 28,174
Undergraduate 87.7% 11.1% 1.2% 2,131 22,612
Graduate 88.1% 9.1% 2.8% 529 5,562
Employees 90.3% 7.3% 2.4% 1,073 11,388
Outside Davis 78.7% 15.6% 5.6% 865 8,831
Within Davis 91.5% 7.9% 0.6% 2,868 30,731
Overall 88.5% 9.7% 1.8% 3,732
Projected pop. 35,017 3,844 701 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Data are based on responses to
questions Q0030, Q0036, Q0039, Q0042, Q0045, and Q0048.
Table 14: Modes typically used by single- mode versus multi- modal commuters
Percent using this mode for any part of trip
Mode Among single-mode
users
Among multi-mode
users
Overall
Weighted
sample
Projected
pop.
Bike 46.6% 48.8% 46.9% 1,749 18,552
Walk 4.5% 43.5% 8.9% 333 3,535
Drive alone 24.8% 40.4% 26.6% 18 195
Carpool 6.8% 14.2% 7.7% 992 10,518
Any transit ( bus or train) 16.9% 56.3% 21.4% 799 8,480
Overall 88.6% 11.4% 100.0%
Weighted sample 3,303 426 3,730 3,730
Projected population 35,039 4,523 39,562 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). The number of modes used
( categorization of “ single- mode” versus “ multi- mode” commuter is based on responses to question Q0030. Whether a
mode was used for any portion of the trip is based on responses to questions Q0031, Q0032, Q0034, Q0037, Q0040,
Q0043, and Q0046.
Table 15 compares results on the percent of the population using bikes during the reference week
versus during their reported typical commute, potentially in combination with other modes. Based on
the reported primary mode during the reference week, about 49 percent of respondents used a bike at
least one of the days as their primary mode of transportation, with an average of 37 percent using a bike
on any given day. The percent of respondents who report that they “ typically” use a bike as their sole
means of transportation between home and campus is between these two figures at 41 percent ( a
projected 16,334 individuals), with an additional 6 percent ( a projected 2,207 individuals) reporting
that they typically use a bike in combination with other modes. These reports of typical use would
bring an estimated 18,541 bikes to campus on a daily basis. There may be an additional number of
bikes brought to campus as part of an atypical pattern ( that is, people that only occasionally use a bike,
either alone or in combination with other modes).
24
Table 15. Estimated number of bikers, from various survey questions
Single ( primary) mode reported
during reference week a
All mode( s) reported as part
of typical trip to campus b
Role group
Percent biking on
average weekday
Percent biking at least
once during week
Percent biking
as sole mode
Percent biking
combined with
other mode( s)
Projected
population
Freshmen 70.3% 80.1% 71.8% 7.1% 420
Sophomores 41.2% 54.5% 45.7% 7.3% 395
Juniors 35.1% 46.9% 39.1% 6.5% 506
Seniors 33.6% 44.9% 39.1% 4.9% 674
Masters 42.9% 52.6% 47.3% 4.5% 168
PhD 49.0% 60.6% 56.6% 7.3% 318
Faculty 32.3% 43.3% 38.4% 5.6% 182
Staff 17.1% 22.5% 20.8% 3.6% 812
Students 43.7% 57.7% 48.2% 6.2% 2,479
Undergraduate 43.0% 57.5% 46.9% 6.2% 1,994
Graduate 46.9% 58.0% 53.4% 6.3% 486
Employees 19.9% 30.7% 24.1% 3.9% 994
Outside Davis 2.2% 4.4% 2.1% 9.7% 770
Within Davis 47.3% 62.8% 53.1% 4.3% 2,703
Overall 36.9% 48.6% 41.3% 5.6%
Weighted sample 1,448 1,911 1,541 208 3,473
Projected population 14,585 19,242 16,334 2,207 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6).
Vehicle occupancy and ridesharing
Among those traveling to campus ( or working from home) on an average weekday, we estimate about
32 percent arrive by private vehicle ( either driving alone, carpooling, or getting a ride) ( see Table 10
and Table 16). Among these, about 76 percent drive alone, 18 percent carpool, and 5 percent get a ride
with someone who drops them off ( Table 16). Among vehicle- users, freshmen and sophomores are
least likely to drive alone ( 58 percent, in both classes) relative to the other role groups, whereas
graduate students and faculty are most likely to drive alone ( 85 percent, in both groups). Freshmen and
sophomores are more likely to carpool ( 22 percent and 27 percent, respectively) and to get rides ( 19
percent and 15 percent, respectively) relative to the other role groups.
Both those arriving in carpools ( multiple people in the vehicle arriving on campus together) and those
getting a ride to campus ( where the driver continues on to another destination after the drop- off) were
asked how many other people were in the vehicle. The percent of vehicle users arriving in 2- and 3-
plus- person carpools and of those getting a ride as the sole passenger or multiple passengers dropped
off is shown in Table 16. The average vehicle occupancy for carpools and rides is shown in Table 17.
Among those who carpooled at any point during the reference week, the average number of passengers
was 2.59 ( including the driver), with a high of 3.42 among freshmen and a low of 2.30 among staff.
Most people dropped off on campus were the sole passenger dropped, with an average of 1.33
passengers dropped off per ride to campus ( excluding the driver) ( see Table 17).
25
Table 16: Percent driving alone versus ridesharing on an average weekday
Among those traveling, percent:
Among those in Role group vehicles, percent:
Percent
“ traveling” a
Arriving
by private
vehicle b
Driving
alone
Carpool
of 2
Carpool
of 3+
Ride: 1
dropped
Ride: 2+
dropped
Weighted
sample
Projected
pop.
Freshmen 95.4% 3.6% 58.4% 13.0% 9.1% 9.1% 10.4% 456 4,597
Sophomores 96.6% 11.4% 57.7% 18.5% 9.0% 9.9% 5.0% 446 4,498
Juniors 92.2% 17.2% 74.0% 17.8% 3.3% 4.2% 0.6% 575 5,796
Seniors 91.7% 23.8% 72.9% 13.6% 2.9% 9.0% 1.6% 768 7,721
Masters 88.9% 35.2% 87.2% 7.2% 2.1% 3.5% 0.0% 192 1,926
PhD 93.6% 30.5% 83.4% 9.8% 1.6% 4.8% 0.5% 362 3,636
Faculty 86.1% 46.8% 85.2% 11.3% 1.3% 2.0% 0.2% 206 2,079
Staff 88.5% 70.5% 76.4% 16.9% 3.5% 3.1% 0.1% 924 9,309
Students 93.2% 18.7% 75.1% 13.3% 3.4% 6.6% 1.6% 2,799 28,174
Undergraduate 93.5% 15.4% 70.2% 15.5% 4.3% 7.8% 2.3% 2,246 22,612
Graduate 92.0% 32.1% 84.7% 8.8% 1.8% 4.3% 0.3% 553 5,562
Employees 88.1% 66.2% 77.5% 16.2% 3.2% 3.0% 0.1% 1,130 11,388
Outside Davis 84.8% 83.8% 79.1% 15.3% 3.5% 1.8% 0.3% 908 8,831
Within Davis 93.8% 17.7% 73.1% 14.4% 3.1% 8.1% 1.4% 3,021 30,731
Overall 91.7% 31.8% 76.5% 14.9% 3.3% 4.5% 0.8%
Weighted sample 3,605 1,146 876 171 38 52 9 3,929
Projected pop. 36,298 11,540 8,826 1,722 380 525 88 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Mode of transportation ( drive
alone, carpool, or ride) is based on responses question Q0011. Vehicle occupancy is based on responses to question
Q0016 for carpools and Q0017 for those getting a ride. Responses for Q0016 and Q0017 were assumed to apply to all
days that a respondent carpooled or got a ride during the reference week, as indicated in Q0011. For each respondent, we
calculated the percent of days out of the five- day week that a given mode/ occupancy- level was used; the average shares
across all respondents is equivalent to the percent of people using each mode/ occupancy- level on an average weekday.
a Includes those physically traveling to campus by any means of transportation plus those working from home ( as in Table
10), but excluding those not traveling to campus for any other reason.
b Including driving alone, carpooling, or getting a ride, as indicated in question Q0011.
Table 17: Average carpool size
Average occupancy among those that carpooled / rode at least once Weighted sample
Role group Carpool occupants
( including driver)
Ride passengers
( excluding driver)
Carpool Ride
Freshmen 3.42 1.73 37 66
Sophomores 2.70 1.50 88 73
Juniors 2.46 1.20 105 62
Seniors 2.98 1.23 118 105
Masters 2.35 1.11 25 11
PhD 2.30 1.22 44 35
Faculty 2.78 1.04 25 10
Staff 2.30 1.09 192 51
Students 2.72 1.37 416 352
Undergraduate 2.80 1.40 347 306
Graduate 2.32 1.19 69 46
Employees 2.36 1.08 218 61
Outside Davis 2.57 1.17 205 40
Within Davis 2.61 1.35 428 372
Overall 2.59 1.33 634 413
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Vehicle occupancy is based on
responses to question Q0016 for those carpooling/ vanpooling and to question Q0017 for those who got a ride.
26
( umber of vehicles arriving on campus
The results on the number of people driving alone, carpooling, getting a ride, and the number of people
per vehicle can be combined to estimate the total number of vehicles arriving on campus. In particular,
we estimate the count of private vehicle arrivals as the number of people driving alone, plus fractional
vehicles counted in proportion to vehicle occupancy. That is, if a respondent reports arriving in a three-person
carpool, we count this as 0.33 vehicles arriving on campus on behalf of that respondent. We
weight and expand the sample to project the total number of vehicle arrivals for the entire campus
population, using the weighting factors shown in Table 6. We estimate that 10,313 vehicles come to
campus on an average weekday, or about one vehicle for every 3.52 person traveling to campus ( Table
18). About 970 of these contain carpools and 552 are vehicles just dropping passenger( s) off.
Table 18. Projected number of vehicles arriving on an average weekday, by role
Projected number of vehicle arrivals
on an average weekday
Role group
Drive
alone
Carpool Ride
Total
vehicles
Total
population
( people)
Estimated
percent of people
traveling a on
average weekday
Ratio
total
people/
vehicles
Ratio
traveling a
people/
vehicles
Freshmen 93 15 23 131 4,597 95.4% 35.13 33.52
Sophomores 286 59 60 405 4,498 96.6% 11.10 10.72
Juniors 678 92 41 811 5,796 92.2% 7.15 6.59
Seniors 1,226 128 157 1,511 7,721 91.7% 5.11 4.69
Masters 525 26 21 571 1,926 88.9% 3.37 3.00
PhD 866 56 50 971 3,636 93.6% 3.74 3.50
Faculty 712 50 17 778 2,079 86.1% 2.67 2.30
Staff 4,406 545 183 5,134 9,309 88.5% 1.81 1.61
Students 3,673 375 352 4,401 28,174 93.2% 6.40 5.97
Undergraduate 2,283 294 281 2,858 22,612 93.5% 7.91 7.40
Graduate 1,390 81 71 1,543 5,562 92.0% 3.61 3.32
Employees 5,118 595 200 5,912 11,388 88.1% 1.93 1.70
Outside Davis 5,132 558 125 5,816 8,831 84.8% 1.52 1.29
Within Davis 3,659 412 427 4,497 30,731 93.8% 6.83 6.41
Overall 8,791 970 552 10,313 39,562 91.7% 3.84 3.52
Data are weighted ( and expanded) by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based
on responses to question Q0011 ( for mode used: drive alone, carpool, or getting a ride), questions Q0013 and Q0023
through Q0027 ( for days not traveling and working from home), and questions Q0016 and Q0017 ( for vehicle
occupancy).
a Includes those physically traveling to campus by any means plus those working from home.
Average Vehicle Ridership
Another way of calculating the ratio of person- arrivals to private- vehicle- arrivals is the average vehicle
ridership ( AVR), as calculated by the South Coast Air Quality Management District, computed here for
comparison with figures from other UC campuses ( see Appendix D for details on the calculation of
AVR). The main difference between the ratios in Table 18 and the AVR figures reported in Table 19 is
that in calculating AVR is the numerator ( person- arrivals). Table 18 counts those working from home as
virtual travelers for all role groups, while Table 19 does not include students working from home
( potentially making student AVR lower than it otherwise might); Table 19 additionally counts as
person- arrivals any employees taking days off as a part of a compressed work week schedule
( potentially making employee AVR higher than it otherwise would be). In general, a way to interpret
AVR is that if everyone drove by themselves to campus, the campus AVR would be one, and so higher
values ( greater than 1.0) indicate more carpooling and/ or use of alternative modes of transportation.
27
Among those traveling from off campus, campus- wide AVR is estimated to be 2.99, up slightly from
2.75 in 2007- 08. This means that for every car arriving on campus, there are about three people arriving
on campus ( using some means of transportation or another).
Table 19: Average Vehicle Ridership ( AVR), 2008- 09 and 2007- 08
Off- campus only All ( on and off- campus)
Role group 2007- 08, as in
Congleton ( 2009)
2007- 08, new
calculation
2008- 09
calculation
2007- 08, new
calculation
2008- 09
calculation
Freshmen 4.53 5.32 5.35 26.39 33.40
Sophomores 7.89 6.46 10.24 6.78 10.67
Juniors 5.09 4.05 6.26 4.46 6.56
Seniors 4.39 3.55 4.39 3.77 4.67
Masters 5.31 3.22 2.71 3.49 2.94
PhD 4.33 3.55 2.86 4.20 3.36
Faculty 2.57 2.23 2.34 2.23 2.35
Staff 1.66 1.58 1.60 1.58 1.62
Student 1.67 4.76 5.04 5.91
Undergraduate 5.31 4.24 5.80 5.04 7.37
Graduate 4.66 3.43 2.81 3.94 3.21
Employees 1.82 1.67 1.69 1.67 1.71
Outside Davis 1.33 1.32 1.33 1.33
Within Davis 4.60 5.17 5.61 6.32
Overall 4.17 2.75 2.99 3.20 3.51
For 2008- 09 figures, data are weighted by role based on the 3,935 valid responses to question Q0008 ( see Table 6). See
Appendix D for details on AVR calculation.
Comparison of 2008- 09 and 2007- 08 mode splits
Mode- choice results across the two last survey years are not directly comparable because of differences
in question and answer- choice wording offered to participants in each survey year ( see Appendix B).
One important difference across years is that 08- 09 respondents were asked to indicate a single mode
that they used for most of the trip, whereas 07- 08 participants were given the option to indicate that
they used more than one mode. In addition, 08- 09 participants were not given an “ other” category
option, whereas the 07- 08 participants could indicate “ other.” At the same time, 08- 09 respondents
were offered the additional categories of train/ rail, getting a ride, and skating, whereas 07- 08
respondents were not explicitly offered these.
For the purposes of comparing the two years, we recoded 527 descriptions of multimodal trips from the
07- 08 data into one of the other mode groups ( reducing the number of travel days coded as multimodal
from 1141 to 441, or about 60 respondents per travel day). We combined the remaining “ multimodal”
and “ other” responses from 07- 08 that could not otherwise be re- categorized into a single group. In
addition, we grouped the “ skate” responses in the 08- 09 data in with walking, and getting a ride in with
carpooling or driving alone ( as described above). Finally, in combining the data collected for all five
travel days into a single mode- split percentage for each mode, we apply the same method used for the
2008- 09 data ( described above) to the 2007- 08 data, calculating the average percent of days that
individuals use a given mode over the course of five days, equivalent to the average percent of
individuals using a given mode on an average weekday. ( This differs from the methodology used to
calculate mode split in Congleton ( 2009), which defines primary mode as the one used most often out
of the five days, favoring the “ lower- impact” mode whenever ties occur.) Thusly, roughly comparable
versions of overall mode split across years is shown in Table 20.
28
Table 20. Primary weekday mode split, 2008- 09 and 2007- 08
Year Bike Drive alone Carpool Bus Walk Train
Other/
multimodal Telecommute
Weighted
sample
2008- 09 40.2% 24.3% 7.5% 19.9% 5.6% 0.8% n/ a 1.4% 3,605
2007- 08 37.5% 28.8% 5.5% 19.0% 4.9% n/ a 3.7% 0.7% 3,917
Percentages shown are from among those who traveled to campus during the reference week ( and therefore excluding
6.3% and 8.3% of the 07- 08 and 08- 09 samples, respectively). As in Table 10, percentages represent the average share of
days ( in five- day week) that each mode is used as the primary mode on a given day for each respondent. Data are
weighted by role so that the proportion in each role group in the sample matches the proportion in the campus
population in each year. ( For 2008- 09 data, the results are weighted based on the 3,935 valid responses to question
Q0008, as shown in Table 6. For 2007- 08 data, the results are weighted as shown in Table 51).
Table 21 shows the year- over- year percent change in the number using each mode, including the results
of χ 2 tests for significant differences across years ( indicated by asterisks). 4 Statistically significant
changes include an increase in biking, an increase in working from home, and a decrease in driving
alone. Employees and staff show an increase in carpooling. Changes in other modes are not statistically
significant across survey years.
Table 21. Percent change in primary weekday mode split, 2007- 08 to 2008- 09, by mode and role
Role group Bike Drive alone Carpool Bus Walk
Work from
home
Freshmen - 0.4% - 0.8% - 0.2% 0.7% 2.6% n/ a
Sophomores 10.2% - 5.2% - 0.4% - 1.4% - 0.7% n/ a
Juniors 4.5% - 7.3% 0.0% 5.3% - 0.5% n/ a
Seniors 4.6% - 4.2% - 1.6% 1.5% 3.0% n/ a
Masters - 3.0% 4.6% 1.0% 0.8% - 1.5% n/ a
PhD - 4.9% 4.9% - 0.3% - 1.4% - 0.5% n/ a
Faculty 0.6% - 1.8% - 0.3% 0.8% 1.0% 0.2%
Staff - 0.5% - 3.3% 4.5% 1.1% - 0.5% 0.9%
Students 3.7% - 2.7% - 0.6% 0.7% 0.9% n/ a
Undergraduate 5.4% - 4.7% - 0.7% 1.5% 1.4% n/ a
Graduate - 4.4% 4.7% 0.2% - 0.6% - 0.8% n/ a
Employees - 0.4% - 2.9% 3.5% 1.0% - 0.1% 0.8%
Outside Davis - 0.4% - 2.9% 4.0% - 0.9% 0.1% 2.0%
Within Davis 2.6% - 1.9% - 0.2% 1.0% 0.7% 0.5%
Overall 2.7% - 3.0% 0.5% 0.9% 0.7% 0.8%
Statistically significant difference with p < 0.1 in a two- category χ 2 test of the frequency of those using this mode versus
those using any other mode in 2007- 08 versus 2008- 09.
Statistically significant at p < 0.05.
Time arriving on campus
Table 22 and Table 23 show the percent of respondents traveling to campus on a given day by role and
whether they arrived during the morning peak, considered for the purposes of this survey to be between
the hours of 6am and 10am. Among those traveling to campus on an average weekday, about three-quarters
arrive during the peak period, or a projected 26,855 people.
4 In particular, we conducted χ 2 tests for the independence of the number of respondents falling into each mode group ( as
a binary categorical variable for each mode, for instance, “ bike” versus “ all other”) by survey year ( 2007- 08 versus
2008- 09). p < 0.05 indicates 95% confidence in rejecting the hypothesis of independence across groups.
29
Table 22. Arrivals during the peak period, by day
Arrivals between 6am – 10am, among Total
Day those traveling to campus
Did not travel
to campus
Percent Projected number
Weighted
sample
Projected
population
Monday 8.4% 76.7% 27,791 3,929 39,562
Tuesday 7.9% 73.0% 26,602 3,929 39,562
Wednesday 7.5% 77.3% 28,305 3,929 39,562
Thursday 7.7% 73.3% 26,772 3,930 39,562
Friday 16.4% 75.0% 24,811 3,930 39,562
Saturday 80.8% 27.8% 2,114 3,930 39,562
Sunday 80.9% 22.5% 1,699 3,932 39,562
Data are weighted ( and expanded) by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based on
responses to question Q0012.
Table 23. Percent arriving during the peak period on an average weekday, by role
Total
Role
Did not travel
to campus
Arrivals between 6am –
10am, among those
traveling to campus
Weighted
sample
Projected
population
Freshmen 4.7% 61.7% 456 4,597
Sophomores 3.6% 66.1% 446 4,498
Juniors 7.9% 69.0% 575 5,796
Seniors 8.6% 69.7% 768 7,721
Masters 12.6% 78.2% 192 1,926
PhD 10.2% 77.1% 362 3,636
Faculty 19.3% 86.9% 206 2,079
Staff 13.6% 92.0% 924 9,309
Overall 9.6% 75.1%
Weighted sample 376 2,667 3,929
Projected population 3,782 26,855 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based on responses to
question Q0012.
Vehicle parking, on and off campus
Question Q0018 asked “ Where did you ( or whoever drove you) park?” among any respondents who
indicated having driven, carpooled, or gotten a ride to campus during the five- day reference week
( question Q0011). Among those who answered this question ( 95.4 percent of the total arriving by
vehicle, excluding the “ unknown” column below), about 85 percent arrived in vehicles that parked on
campus, an estimated 9,863 persons on an average weekday. About 9 percent reported parking off
campus, or about 1,049 persons on an average weekday. ( Note that the number of vehicles in each case
may be fewer, to the extent that there is more than one person per vehicle). Students, especially
undergraduates, were more likely to report parking off campus than were employees. Freshmen were
particularly likely to report being dropped off.
30
Table 24: Percent arriving in vehicles that parked on and off campus, on an average weekday
On an average weekday
Among those arriving by vehicle a ,
percent of people parking:
Role group Percent of
people arriving
by vehicle a On campus Off campus Dropped off Unknown b
Weighted
sample
Projected
pop.
Freshmen 3.5% 72.6% 3.9% 20.8% 2.6% 456 4,597
Sophomores 11.0% 73.2% 11.3% 9.5% 5.9% 446 4,498
Juniors 15.8% 79.9% 14.2% 3.3% 2.4% 575 5,796
Seniors 21.8% 73.9% 10.6% 8.2% 7.2% 768 7,721
Masters 31.2% 87.4% 5.4% 3.9% 3.3% 192 1,926
PhD 28.6% 80.4% 11.2% 6.3% 2.0% 362 3,636
Faculty 40.3% 91.0% 4.0% 1.5% 3.1% 206 2,079
Staff 62.4% 83.3% 7.7% 4.1% 4.9% 924 9,309
Students 17.4% 77.9% 10.6% 6.9% 4.4% 2,799 28,174
Undergraduate 14.4% 75.4% 11.4% 7.7% 5.4% 2,246 22,612
Graduate 29.5% 83.0% 9.1% 5.4% 2.5% 553 5,562
Employees 58.3% 84.3% 7.2% 3.7% 4.7% 1,130 11,388
Outside Davis 71.0% 83.3% 9.8% 2.5% 4.3% 908 8,831
Within Davis 16.6% 79.4% 7.2% 8.4% 4.9% 3,021 30,731
Overall 29.2% 81.6% 8.7% 5.1% 4.6% 3,929 39,562
Weighted sample 1,146 935 99 58 52 3,929
Projected pop. 11,540 9,412 1,001 588 528 39,562
Overall percent of
vehicles c
81.7% 9.0% 4.8% 4.4%
Projected number of
vehicles c
10,313 8,429 932 494 459
Data are weighted by role based on the 3,935 valid responses to question Q0008 ( see Table 6). Results are based on
responses to question Q0018. In particular, the parking location indicated in question Q0018 is assumed to be true for all
days that the respondent drove, carpooled, or got a ride in question Q0011. As with mode split, we calculated the share
of the five days that each respondent parked, and then the average of this over all respondents is equivalent to the share
of all respondents parking on an average weekday.
a Including all those who reported driving alone, carpooling, or getting a ride in question Q0011.
b Respondent indicated using a vehicle in Q0011 but did not respond to question Q0018.
c Number of vehicles counted as one vehicle for each person driving alone plus partial vehicles proportionate to the
number of occupants in vehicles with more than one person ( as in Table 18).
( umber of riders by transit agency provider
Table 14 details which transit agencies respondents report using as a part of their typical commute. The
overwhelming majority of transit- users ride Unitrans ( 86 percent), corresponding to a projected 7,309
daily riders. The results indicate that the next most utilized services include Amtrak ( 506 riders),
Yolobus ( 301 riders), and the UCD/ UCDMC shuttle ( 207 riders).
31
Table 25: Use of specific transit providers at some point during “ typical commute”
Among transit riders
Mode used for at least one leg of
the journey to campus
Percent
of total Overall Undergrads
Grad
students
Employees
Weighted
sample
Projected
pop.
Unitrans bus 18.5% 86.4% 93.8% 55.6% 40.4% 691 7,330
Amtrak train 1.3% 6.0% 2.1% 28.4% 25.1% 48 507
YOLOBUS 0.8% 3.6% 2.6% 5.1% 11.3% 28 301
UCD/ UCDMC Shuttle 0.5% 2.4% 0.4% 11.0% 15.1% 20 207
Sacramento Regional Transit 0.2% 1.1% 1.0% 2.0% 1.1% 9 93
BART 0.2% 1.0% 0.5% 4.9% 2.7% 8 87
Fairfield Suisun Transit bus 0.2% 0.7% 0.3% 0.0% 4.8% 6 60
Amtrak motorcoach ( bus) 0.1% 0.6% 0.2% 3.1% 2.7% 5 49
Davis Community Transit 0.0% 0.2% 0.2% 0.0% 0.6% 2 16
UC Berkeley - UC Davis shuttle 0.0% 0.0% 0.0% 0.0% 0.0% 0 0
Other public transportation 0.2% 0.8% 0.5% 2.1% 2.2% 6 65
Overall 21.4% 100.0% 100.0% 100.0% 100.0% 799 8,480
Weighted sample 3,730 799 2,129 529 1,072 3,730
Projected population 39,562 8,480 7,153 571 762 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Data are based on responses to
questions Q0031, Q0032, Q0034, Q0037, Q0040, Q0043, and Q0046 regarding “ typical” travel to campus.
Self- reported travel time
Question Q0029 asked respondents to indicate how many minutes it usually takes them to get from
home to their first campus destination ( in categories of five- minute intervals up to an hour, then 1- 2
hours, or 2 hours or more). Taking the midpoints of each category as the travel time, the average
number of minutes respondents report spending on their trip from home to campus ranges from 12
minutes among freshmen to 25 minutes among faculty ( Table 26). About 10 percent estimate spending
more than a half hour, or about 20 percent of employees.
Residential location and distance from campus
The survey provided two ways of measuring respondents’ residential locations. The first was asking
them whether they lived on campus, elsewhere in Davis, or outside Davis ( question Q0008), as shown
in Table 27. The results suggest that a projected 5,860 live on campus, 24,640 live elsewhere in Davis,
and 8,721— predominantly employees— live outside Davis.
The survey also requested that respondents give their zip codes and an intersection near their homes.
These were geocoded, when possible, and used to estimate respondents’ distance from campus ( see
Appendix E). Table 28 summarizes these estimated distances by role group. We discarded several data
points greater than 175 miles that seemed implausible. Among the remaining cases, the average
distance from campus is estimated to be 6.6 miles overall, though 23.3 miles among those outside of
Davis. About 9 percent of the overall population and 17 percent of employees travel more than 20
miles. A full 70 percent travel fewer than 3 miles ( Table 28). Note that because different methodologies
were used to collect respondents’ residential address and to geocode and analyze network distances in
2008- 09 versus 2007- 08, distance estimates from this year’s data are not comparable to 2007- 08.
32
Table 26: Reported number of minutes spent traveling to campus, by role
Percent reporting…
Role group Average
minutes a
Less than 10
minutes
10- 29
minutes
30- 59
minutes
1 hour
or more
Weighted
sample
Projected
population
Freshmen 12.1 55.6% 39.8% 2.4% 2.2% 424 4,597
Sophomores 16.0 16.5% 77.7% 4.7% 1.0% 423 4,498
Juniors 17.5 21.0% 68.3% 8.7% 2.0% 556 5,796
Seniors 17.8 25.1% 63.9% 8.0% 3.1% 728 7,721
Masters 20.8 16.9% 68.1% 9.6% 5.4% 182 1,926
PhD 20.5 20.6% 64.8% 9.1% 5.4% 347 3,636
Faculty 25.1 15.9% 64.2% 9.7% 10.1% 199 2,079
Staff 24.4 8.6% 67.1% 21.1% 3.2% 874 9,309
Students 17.1 26.6% 63.6% 7.0% 2.8% 2,660 28,174
Undergraduate 16.2 28.4% 63.0% 6.4% 2.2% 2,131 22,612
Graduate 20.6 19.3% 65.9% 9.3% 5.4% 529 5,562
Employees 24.5 10.0% 66.6% 19.0% 4.5% 1,073 11,388
Outside Davis 37.1 1.1% 50.3% 35.6% 13.0% 865 8,831
Within Davis 13.8 28.1% 68.7% 2.8% 0.4% 2,868 30,731
Overall 19.2 21.8% 64.4% 10.4% 3.3% 3,732
Projected population 8,634 25,493 4,124 1,311 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based on responses to
question Q0029.
a Based on responses to question Q0029, which was categorical, by assuming that the travel time for each individual is the
midpoint of the category reported or for the highest category, “ 2 hours or more,” that travel time is 2 hours ( 120
minutes).
Table 27: Residential location
Role group
On
campus
Elsewhere
in Davis
Outside
of Davis
Multiple
locations
Weighted
sample
Projected
population
Freshmen 87.3% 9.7% 3.0% 0.0% 480 4,597
Sophomores 4.6% 91.0% 4.1% 0.2% 456 4,498
Juniors 4.1% 85.3% 9.4% 1.2% 597 5,796
Seniors 6.2% 79.0% 13.2% 1.7% 795 7,721
Masters 9.3% 69.9% 19.0% 1.8% 193 1,926
PhD 15.1% 65.3% 18.6% 1.0% 370 3,636
Faculty 0.2% 66.7% 32.2% 0.9% 216 2,079
Staff 1.2% 42.8% 55.5% 0.5% 944 9,309
Students 20.3% 68.3% 10.4% 1.0% 2,891 28,174
Undergraduate 22.1% 68.7% 8.3% 0.9% 2,328 22612
Graduate 13.1% 66.8% 18.7% 1.3% 563 5562
Employees 1.0% 47.3% 51.1% 0.6% 1,161 11,388
Overall 14.8% 62.3% 22.0% 0.9%
Weighted sample 600 2,524 893 35 4,052
Projected population 5,857 24,640 8,721 344 39,562
Overall 2007- 08 a 14.7% 61.0% 24.3% n/ a 40,601
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based on responses to
question Q0008.
a Based on geocoded residential location in the 2007- 08 survey, as reported in Congleton ( 2009, Table 3- 3, p. 11).
33
Table 28. Distance from campus, by role
Percent
Role group Less than
3 miles
3 to 20
miles
More than
20 miles
Maximum
distance ( miles)
Average
distance ( miles)
Weighted
sample
Freshmen 97.1% 2.3% 0.7% 27.7 1.1 420
Sophomores 90.9% 7.0% 2.1% 89.2 2.8 395
Juniors 83.3% 10.5% 6.2% 75.4 4.4 506
Seniors 78.4% 15.1% 6.5% 89.0 4.8 674
Masters 70.0% 21.6% 8.4% 83.1 7.3 168
PhD 69.3% 21.3% 9.5% 174.7 8.1 318
Faculty 47.5% 36.8% 15.7% 85.4 12.0 182
Staff 33.5% 49.1% 17.5% 72.7 12.1 812
Students 82.8% 11.9% 5.3% 174.7 4.4 2,479
Undergraduate 86.0% 9.6% 4.3% 89.2 3.6 1,994
Graduate 69.5% 21.4% 9.1% 174.7 7.8 486
Employees 36.0% 46.8% 17.1% 85.4 12.0 994
Outside Davis 2.5% 58.4% 39.1% 174.7 23.3 770
Within Davis 88.5% 11.5% 0.0% 9.9 1.8 2,703
Overall 69.4% 21.9% 8.7% 174.7 6.6
Weighted sample 2,411 761 301 3,473
Projected population 27,465 8,670 3,427
Data weighted by role for 3,451 cases successfully geocoded by zip code and cross streets ( given in Q0050 and Q0051
or Q0087 and Q0088, or dorm name given in Q0005) and with non- missing mode choice data in question Q0011 ( see
Table 6). The methodology used for calculating network distance is described in Appendix E.
Table 29 and Table 30 show the correspondence between distance and mode choice. In particular, Table
29 shows the percent who live various distances from campus, among those who use each mode, while
Table 30 shows the percent using each mode, among those who live various distances from campus.
For instance, we see that among those who biked at least once during the reference week, 92 percent
live within three miles of campus; Meanwhile, among those who live three miles from campus, 51
percent bike to campus on average weekday. We also see that the average distance from campus is
greatest among those taking the train ( 41 miles) and working from home ( 21 miles).
Table 29. Distance from campus, by mode used
Among those using this mode at least one weekday
Percent who live
Mode
Less than
3 miles
3 to 20
miles
More than
20 miles
Maximum
distance ( miles)
Average
distance ( miles)
Weighted
sample
Bike 91.7% 8.1% 0.3% 75.8 1.8 1,667
Walk 97.4% 2.1% 0.5% 79.7 1.4 344
Skate 89.8% 0.0% 10.2% 21.0 3.3 23
Drive alone ( SOV) 38.1% 43.3% 18.6% 174.7 12.3 1,192
Carpool/ ride ( HOV) 42.5% 43.8% 13.7% 87.8 11.2 332
Bus 83.3% 14.7% 2.0% 39.2 3.1 941
Train 17.2% 11.2% 71.6% 93.0 40.5 50
Work from home 38.9% 29.9% 31.2% 111.2 20.7 112
Other did not travel 54.0% 27.5% 18.5% 174.7 11.8 689
Overall 69.4% 21.9% 8.7% 174.7 6.6 3,473
Weighted sample 2,411 761 301
Projected population 27,465 8,670 3,427
Data weighted by role for 3,451 cases successfully geocoded by zip code and cross streets ( given in Q0050 and Q0051
or Q0087 and Q0088, or dorm name given in Q0005) and with non- missing mode choice data in question Q0011 ( see
Table 6). The methodology used for calculating network distance is described in Appendix E.
34
Table 30. Mode used, by distance from campus
Percent using each mode on an average weekday
Mode Among those living:
Less than 3 miles 3 to 20 miles More than 20 miles
Overall
Bike 51.4% 11.2% 0.5% 38.1%
Walk 7.5% 0.3% 0.2% 5.2%
Skate 0.4% 0.0% 0.3% 0.3%
Drive alone ( SOV) 10.6% 51.4% 54.0% 23.3%
Carpool/ ride ( HOV) 2.4% 12.6% 10.7% 5.3%
Bus 21.4% 13.0% 4.4% 18.1%
Train 0.1% 0.3% 7.0% 0.8%
Work from home 0.7% 1.6% 4.7% 1.2%
Other did not travel 5.6% 9.7% 18.2% 7.6%
Overall 100.0% 100.0% 100.0% 100.0%
Weighted sample 2,389 761 301 3,451
Projected population 27,465 8,670 3,427 39,562
Data weighted by role for 3,451 cases successfully geocoded by zip code and cross streets ( given in Q0050 and Q0051
or Q0087 and Q0088, or dorm name given in Q0005) and with non- missing mode choice data in question Q0011 ( see
Table 6). The methodology used for calculating network distance is described in Appendix E.
Aggregate person- miles and vehicle- miles traveled
We estimate the number of miles traveled based on the shortest- distance path along the road network
between a respondent’s reported home location and campus— assuming that respondents take this
shortest path to and from campus on the days they report having traveled to campus. This likely
underestimates the total number of miles traveled to and from campus, since it does not take into
account side trips respondents might make on the way to or from campus ( for instance stopping at the
store, to pickup children, or visit friends), or trips away from campus during the middle of the day
( such as to go to lunch or to an off- site meeting).
We estimate the number of miles ( person- miles, versus vehicle- miles, described below) traveled each
day as the doubled network distance between respondents’ geocoded home location and the Silo on
campus ( as described in Appendix E), multiplied times the percent of weekdays a respondent travels to
campus. Thus if a person lives 10 miles from campus and traveled to campus all five days, her average
daily person- miles would be 20 miles; if she traveled to campus only one day, her average daily person-miles
would be 4 miles. We further attribute person- miles to each mode based on the share of weekdays
a respondent used each mode during the reference week. Thus, if a respondent biked one day and drove
four, we would count 20 percent of his miles as bike miles and 80 percent as driving miles. Summed
across all respondents, this represents person- miles traveled by each mode on an average weekday. We
weight and inflate responses by role group to estimate a projection of the total person- miles traveled in
the entire population.
To estimate the number of person- miles traveled annually, we first assume that respondents travel the
same number of days per week and using the same modes as in the reference week for the entire 36
weeks of the academic year. Then to estimate summer travel, we rely on responses to questions Q0056
and Q0057 about the number of weeks and average number of days per week traveled to campus
during the summer, but assuming they used modes used during the survey reference week. For
instance, annual miles biked = ( distance from campus × 2) × ( share of days biked during reference
week) × [( 36 weeks × 5 days/ week) + ( weeks traveled to campus during the summer × days/ week
traveled per summer)].
35
To estimate vehicle- miles traveled ( VMT), we assume that each person- mile contributes a fractional
vehicle- mile equivalent to one divided by vehicle occupancy, for any travel in a private vehicle or
public transit vehicle ( including driving alone, carpooling, getting a ride, riding a bus, and riding a
train). We assume that travel by walking, biking, or skating contributes no VMT. Vehicle occupancy for
carpooling and getting a ride varies for each respondent, as reported in questions Q0016 and Q0017 for
those carpooling/ vanpooling or getting a ride, respectively. If a respondent lives 10 miles from campus
and traveled in a 3- person carpool all five weekdays, her average daily VMT would be ( 10 miles × 2) /
3 = 6.67 miles. Occupancy for those driving alone and for those who got a ride and were the only
person dropped off on campus by the person giving them a ride was assumed to be one.
For bus and train occupancy, we assume average occupancy for all trips on those modes. In particular,
we estimated average bus occupancy based on annual ridership data from Unitrans, since the majority
of bus riders use Unitrans. According to 2007 figures from the National Transit Database, Unitrans
provided 6,560,904 annual passenger miles and 736,797 vehicle revenue miles, suggesting an average
of about 8.90 passengers per mile. 5 Thus, for someone who lives 10 miles from campus and traveled by
bus all five weekdays, average VMT per day is ( 10 miles × 2) / 8.90 = 2.25 vehicle- miles.
We estimated train occupancy based on annual ridership data from the Capitol Corridor, since they
provide the majority of train rides to campus. According to figures in the Capitol Corridor Business
Plan Update, the Capitol Corridor provided 110,036,259 passenger- miles and 1,183,109 train- miles of
service in FY2007- 08, suggesting an average of about 93.0 passengers per mile. 6 So if a respondent
lives 100 miles from campus and traveled by train all five days, her average VMT per day is estimated
to be ( 100 miles × 2) / 93.0 = 2.15 vehicle- miles.
Our estimates for the aggregate number of person- miles traveled and vehicle- miles traveled, by mode
and role, are shown in Table 31.
Carbon emissions
Because our results include estimates of miles traveled by mode, we can use estimates of emissions per
mile traveled on each mode to estimate the overall carbon dioxide ( or other) emissions associated with
campus travel. We estimate the amount of CO2 produced by campus travelers by assuming that each
mode of travel generates a certain quantity of carbon ( pounds equivalent) per mile and multiplying this
times our estimate of person- miles by each mode for an average weekday ( see Table 31). To simplify,
we do not take into account emissions associated with the manufacture of bicycles or vehicles, or of
home energy use for those working from home. Thus we assume biking, walking, skating, working
from home, or otherwise not traveling contributes no emissions. We use estimates for pounds
equivalent of carbon produced per passenger- mile by car, bus, and train generated by
TravelMatters. org, as summarized in Table 32. 7 Estimates of the aggregate amount of carbon produced
by roundtrip travel to and from campus on an average weekday are shown in Table 33; annual estimates
5 U. S. Department of Transportation, Federal Transit Administration, 2007 National Transit Database, Annual Transit
Profile, Unitrans - City of Davis/ ASUCD ( NTD ID 9142), available from
http:// www. ntdprogram. gov/ ntdprogram/ data. htm.
6 Capitol Corridor Intercity Passenger Rail Service Business Plan Update FY 2009- 10 – FY 2010- 11, Appendix C,
available from http:// www. capitolcorridor. org/ about_ ccjpa/ business_ plan. php.
7 From TravelMatters website, Individual Emissions Calculator Methodology, available online at
http:// www. travelmatters. org/ calculator/ individual/ methodologyhttp:// www. travelmatters. org/ calculator/ individual/ meth
odology.
36
are shown in Table 34. As with our estimates of total miles traveled on which these are based, side trips
made on the way to or from campus, and any trips made in the middle of the day are not taken into
account.
Table 31. Aggregate person- miles and vehicle- miles traveled, by mode and role, daily and annually
Aggregate roundtrip
miles traveled
Aggregate roundtrip
vehicle- miles traveled
Projected
populaton a
Daily average Annual total
Vehicle- miles per
passenger- mile
Daily average Annual total
By mode
Bike 14,585 49,535 10,370,030 0 0 0
Walk 2,027 5,002 1,023,682 0 0 0
Skate 123 690 127,164 0 0 0
Drive alone 8,826 242,831 54,144,273 1 242,831 54,144,273
Carpool or ride 2,190 56,212 13,017,935 Varies b 30,712 7,031,163
Bus 7,218 45,218 9,420,881 0.112 c 5,064 1,055,139
Train 286 25,466 5,132,099 0.0108 d 274 55,170
Work from home 518 22,420 4,412,479 0 0 0
Other no travel 3,264 76,364 15,543,233 0 0 0
By role
Freshmen 4,597 10,042 1,833,720 2,811 515,858
Sophomores 4,498 23,761 4,457,387 7,354 1,377,840
Juniors 5,796 43,121 8,310,617 25,076 4,728,812
Seniors 7,721 61,238 12,425,800 34,371 6,912,707
Masters 1,926 20,452 4,014,797 12,684 2,464,498
PhD 3,636 45,075 9,355,770 25,321 5,313,534
Faculty 2,079 29,642 6,402,173 18,234 4,023,826
Staff 9,309 191,623 46,435,800 153,029 36,948,670
Students 28,174 203,689 40,398,091 107,618 21,313,249
Undergraduate 22,612 138,162 27,027,523 69,613 13,535,217
Graduate 5,562 65,527 13,370,568 38,005 7,778,032
Employees 11,388 221,265 52,837,973 171,263 40,972,496
Outside Davis 8,831 320,280 71,170,207 250,298 55,975,700
Within Davis 30,731 104,674 22,065,857 28,584 6,310,045
Overall 39,562 424,954 93,236,064 278,881 62,285,745
Data weighted by role for 3,451 cases successfully geocoded by zip code and cross streets ( given in Q0050 and Q0051
or Q0087 and Q0088, or dorm name given in Q0005) and with non- missing mode choice data in question Q0011 ( see
Table 6). Mode use is based on responses to question Q0011 ( for travel mode) and questions Q0013 and Q0023 through
Q0027 ( for days working from home or not traveling to campus), as reported in Table 10. Distances between
respondents’ homes and campus are calculated network distances based on zip code and cross streets given in questions
Q0050 and Q0051 or Q0087 and Q0088, or dorm name given in Q0005, as described in Appendix E.
a Projected number people in the overall population using this mode on an average weekday or projected number of
people in this role group.
b Based on carpool size or number of passengers dropped off, as reported by respondents in questions Q0016 and Q0017.
c Based on estimated average occupancy of Unitrans buses, as described in the text, ( 1/ 8.90 ≈ 0.112).
d Based on estimated average occupancy of Capitol Corridor trains, as described in the text, ( 1/ 93.0 ≈ 0.0108).
37
Table 32. Assumed carbon emitted per passenger- mile for each mode
Mode Pounds CO2 per mile c
Driving alone 1.1
Carpool or getting a ride a 1.1/ vehicle occupancy
Bus b 0.90
Train 0.46
a Adjusted for each respondent according to the reported number of passengers in the carpool.
b This figure is the basis for the “ high” estimate of bus emissions in Table 33 and Table 34.
c Source: TravelMatters website, Individual Emissions Calculator Methodology, available online at
http:// www. travelmatters. org/ calculator/ individual/ methodology.
5 ote about per- mile carbon emissions for bus and train trips
It should be noted that bus and train passengers don’t actually generate additional carbon emissions by
riding the bus or train on a particular day, given that the bus or train is going to operate anyway,
regardless of his choice to ride that day. If the service will be provided regardless of whether he rides,
then emissions associated with that choice are marginally zero. Switching from driving to riding the
bus in this scenario removes all the carbon that would have been generated by driving a car, and adds
nothing. However, overall transit operations may have implications for the campus’s carbon footprint.
In particular, if Unitrans adds bus service to support campus commuters, then carbon emissions may be
reduced to the extent that the new riders would have otherwise driven cars and that the buses are full.
On the other hand, if Unitrans attracts riders that might have otherwise biked or walked or if buses are
relatively empty, then additional bus ridership may actually increase the campus’s overall carbon
footprint.
A second caveat regarding the per passenger- mile emissions levels estimated by TravelMatters. org is
that they are based on nationwide averages of bus service operations and nationwide averages of
Amtrak operations. The assumed fuel economy for transit vehicles and the typical vehicle occupancy
may or may not be accurate for Unitrans and the Capitol Corridor services, which provide the majority
of bus and train rides for UC Davis commuters. In particular, we might expect per- passenger emissions
on Unitrans to be lower than national averages, because of more reliance on compressed natural gas
( CNG) rather than diesel fuel for Unitrans buses, and because of the relatively high numbers of riders
per bus, on average. In particular, according to the National Transit Database ( for fiscal year 2007),
Unitrans buses consumed 235,300 gallons of CNG and 17,600 gallons of diesel 8 while providing
6,560,904 passenger- miles of service. 9 Assuming 22.14 and 0.89 pounds of carbon per gallon of diesel
and CNG, respectively, 10 then Unitrans operations generated 599,248 pounds of carbon in fiscal year
2007, or just 0.091 pounds per passenger- mile of service. Estimates assuming this lower level of
emissions per passenger- mile are shown in the “ low” columns of Table 33 and Table 34, whereas
estimates assuming the higher estimate of 0.90 pounds per passenger- mile are shown in the “ high”
columns.
Estimates specific to Capitol Corridor operations are harder to estimate due a lack of available data.
Fuel economy per train is unknown ( and difficult to estimate since Amtrak nationwide uses an
unknown mix of diesel and electric power, and no data are published on Capitol Corridor fuel use). We
might expect fuel economy on Capitol Corridor trains to be higher than intercity Amtrak trains, because
8 U. S. Department of Transportation, Federal Transit Administration, 2007 National Transit Database, Annual Data Tables,
RY 2007, Table 17. Available from http:// 204.68.195.57/ ntdprogram/ data. htm.
9 U. S. Department of Transportation, Federal Transit Administration, 2007 National Transit Database, Annual Transit
Profile, Unitrans - City of Davis/ ASUCD ( NTD ID 9142). Available from
http:// www. ntdprogram. gov/ ntdprogram/ data. htm.
10 See Appendix F of this report.
38
they may have fewer baggage and service cars. Another factor is the number of passengers per train,
which appears to be lower on average on Capitol Corridor trains than on Amtrak trains nationwide, but
this may be due to their pulling a fewer number of cars per train. ( Nationwide, Amtrak provided 5,784
million passenger- miles in 2007 and 36,484 thousand train- miles, or an average of about 159
passengers per train. 11 In the same year, the Capitol Corridor provided 110,036,259 passenger- miles of
service via 1,183,109 train- miles, or about 93 passengers per train. 12 ) For lack of a better estimate, we
use the TravelMatters. org estimate of 0.46 pounds per passenger mile in Table 33 and Table 34.
Using these assumptions, we estimate that travel to campus generates a total of 357,438 pounds
equivalent of carbon on an average weekday, or about 9 pounds per person ( Table 33). This sums to
about 35,831 metric tons per year, or 0.91 metric tons per person ( Table 34). Employees, and those
living outside of Davis in general, generate substantially more carbon emissions than students, on
average.
Note that the estimate presented here are not directly comparable to those based on the 2007- 08 data
( as in Congleton 2009), because of differences in the methodology for estimating distances traveled
( see Appendix E) as well as for estimating the associated amount of carbon generated.
Table 33. Estimated daily carbon emissions by mode and role
Pounds equivalent of CO2 generated on an average weekday
Role group Drive
alone
Carpool
or ride
Bus
( high) a
Bus
( low) a
Train Total b
Average per
person b
Projected
population
Freshmen 2,300 644 1,064 108 110 4,226 0.92 4,597
Sophomores 6,064 1,063 6,970 707 330 15,135 3.36 4,498
Juniors 23,346 3,097 8,264 839 355 35,902 6.19 5,796
Seniors 33,533 2,331 14,040 1,425 856 52,185 6.76 7,721
Masters 13,214 612 628 64 1,578 16,095 8.36 1,926
PhD 25,678 1,785 2,181 221 3,573 33,439 9.20 3,636
Faculty 17,823 2,090 593 60 2,438 23,005 11.07 2,079
Staff 145,156 22,160 6,956 706 2,473 177,451 19.06 9,309
Students 104,135 9,533 33,147 3,364 6,803 156,982 5.57 28,174
Undergraduate 65,243 7,136 30,338 3,079 1,652 107,448 4.75 22,612
Graduate 38,891 2,397 2,809 285 5,151 49,534 8.91 5,562
Employees 162,979 24,251 7,550 766 4,911 200,457 17.60 11,388
Outside Davis 243,640 29,338 14,966 1,519 11,696 301,159 34.10 8,831
Within Davis 23,474 4,446 25,731 2,611 18 56,280 1.83 30,731
Overall 267,114 33,783 40,696 4,130 11,714 357,438 9.03 39,562
Average lbs/ person 30.26 15.43 5.64 0.57 40.90 9.03 9.03
Projected population 8,826 2,190 7,218 7,218 286 39,562 39,562
a High estimates assume 0.90 pounds/ passenger- mile ( as estimated by TravelMatters. org). Low estimates assume 0.091
pounds/ passenger- mile, as estimated using Unitrans data on annual fuel use and passenger- miles of service provided ( as
reported for fiscal year 2007 in the National Transit Database). See text for further explanation and references.
b Total and average based on “ high” estimate for bus emissions.
11 U. S. Department of Energy, Center for Transportation Analysis, Transportation Energy Data Book, Edition 28, 2009,
Table 9.10. Available from http:// cta. ornl. gov/ data/ chapter9. shtml.
12 Capitol Corridor Intercity Passenger Rail Service Business Plan Update FY 2009- 10 – FY 2010- 11, Appendix C,
available from http:// www. capitolcorridor. org/ about_ ccjpa/ business_ plan. php.
39
Table 34. Estimated annual carbon emissions by mode and role
Metric tons equivalent of CO2 generated annually
Role group Drive
alone
Carpool
or ride
Bus
( high) a
Bus
( low) a
Train Total b
Average per
person b
Projected
pop.
Freshmen 192 54 88 9 9 351 0.08 4,597
Sophomores 513 92 599 61 27 1,291 0.29 4,498
Juniors 1,985 271 746 76 34 3,112 0.54 5,796
Seniors 3,041 227 1,311 133 70 4,782 0.62 7,721
Masters 1,163 56 59 6 137 1,419 0.74 1,926
PhD 2,447 168 207 21 310 3,152 0.87 3,636
Faculty 1,788 206 62 6 210 2,272 1.09 2,079
Staff 15,887 2,435 774 79 275 19,450 2.09 9,309
Students 9,341 867 3,010 305 586 14,108 0.50 28,174
Undergraduate 5,731 643 2,744 278 140 9,537 0.42 22,612
Graduate 3,610 223 266 27 446 4,572 0.82 5,562
Employees 17,675 2,642 836 85 485 21,722 1.91 11,388
Outside Davis 24,642 3,056 1,485 151 1,069 30,403 3.44 8,831
Within Davis 2,373 452 2,361 240 2 5,428 0.18 30,731
Overall 27,015 3,508 3,846 390 1,071 35,831 0.91 39,562
Average tons/ person 3.06 1.60 0.53 0.05 3.74 0.91 0.91
Projected pop. 8,826 2,190 7,218 7,218 286 39,562 39,562
a High estimates assume 0.90 pounds/ passenger- mile ( as estimated by TravelMatters. org). Low estimates assume 0.091
pounds/ passenger- mile, as estimated using Unitrans data on annual fuel use and passenger- miles of service provided ( as
reported for fiscal year 2007 in the National Transit Database). See text for further explanation and references.
b Total and average based on “ high” estimate for bus emissions.
Vehicle type
Anyone who reported traveling to campus by driving, carpooling, or getting a ride during the reference
week ( based on question Q0010) was asked to indicate the vehicle type and technology ( questions
Q0014 and Q0015). Table 35 and Table 36 summarize the results. About 6 percent of respondents
reported using a hybrid, electric, or alternative- fuel vehicles, while about 20 percent used a truck or
SUV.
Table 35. Type of vehicles used for commuting to and from campus
Among those using a vehicle a , percent of people using:
Percent
using a
vehicle a
Regular
car
SUV Truck
Van /
stationwagon
Motorcycle
/ scooter
Other / no
response
Weighted
sample
Projected
pop.
Overall 54.5% 66.6% 13.2% 7.1% 7.2% 0.6% 5.3%
Weighted sample 2,143 1,427 284 152 153 13 115 3,933
Projected pop. 21,560 14,351 2,853 1,529 1,544 129 1,153 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based on responses to
questions Q0014.
a All those who reported driving alone, carpooling, or getting a ride for any travel between home and campus during the
reference week, as reported in question Q0010. This is a larger percent than those who used these modes as their
primary mode to campus on an average weekday ( Table 10) or on at least one of the days during the reference week
( Table 11).
40
Table 36. Type of vehicle technology used for commuting to and from campus
Among those using a vehicle a , percent of people using:
Percent
using
vehicle
Regular gasoline
or diesel
Hybrid Electric Alt. fuel
Weighted
sample
Projected
pop.
Overall 54.5% 90.3% 4.0% 0.2% 0.2%
Weighted sample 2,143 1,936 86 4 5 3,933
Projected pop. 21,560 19,471 870 44 48 39,562
Data are weighted by role based on the 3,935 valid responses to Q0008 ( see Table 6). Results are based on responses to
questions Q0014.
a All those who reported driving alone, carpooling, or getting a ride for any travel between home and campus during the
reference week, as reported in question Q0010. This is a larger percent than those who used these modes as t