Desk ownership in the workplace: The effect of non-territorial working on employee workplace satisfaction, perceived productivity and health

Full text

1
Desk ownership in the workplace: The effect of non-territorial working on
employee workplace satisfaction, perceived productivity and health
Jungsoo Kima,*, Christhina Candidoa, Leena Thomasb, Richard de Deara
a Faculty of Architecture, Design and Planning, Wilkinson Building G04, The University of Sydney,
NSW 2006, Australia
b Faculty of Design Architecture and Building, University of Technology, Sydney, NSW 2007,
Australia
* Corresponding author. E-mail address: jungsoo.kim@sydney.edu.au, Phone: +61 2 8627 0929
(Jungsoo Kim)
Abstract
The concept of non-territorial workplace has been adopted by a growing number of
organisations. It is clear that the main driver for desk sharing practices is the tangible
economic benefits guaranteed by reducing the amount of office space per person. However,
the question of whether or not occupant comfort or productivity are compromised in the
pursuit of space efficiency has never been investigated. This paper draws on a database from
Australian building occupant survey to investigate how desk arrangements (whether or not
one has a pre-allocated desk) can affect occupant satisfaction, self-reported productivity or
health at workplaces. Our statistical model indicates a fall in occupant self-assessed
productivity as spatial factors (such as the office layout allowing easiness of interaction with
colleagues, the ability to adjust/personalise workspace, and the amount of storage space
provided) perform below occupant expectations. Analysis of the results also show that the
association of spatial factors with occupants’ self-assessed productivity (quantified by odds
ratios) was more pronounced among those in non-territorial workplaces, compared to those
who are assigned with a pre-allocated desk. With respect to self-assessed health, the comfort
of furnishing was identified as the strongest predictor for shared-desk users. Our findings
suggest that these spatial factors, rather than the desk ownership itself, play a more significant
role in the non-territorial work arrangement, affecting occupant attitude towards their
building.

2
Keywords
Non-territorial office, Flexi desk, Desk sharing, Activity-based workspace, Indoor
environmental quality (IEQ), Occupant survey
1. Introduction
The cost associated with office accommodation (e.g. rent, heating or cooling, lighting,
interior fitting, furniture, service, etc) can be substantial for many organisations. Typically,
these property-associated costs are considered to be the second highest after employee
salaries for most companies [1]. For example, office occupancy cost per workstation per
annum was estimated at over US$20,000 in London and Hong Kong in 2013 [2]. Over the
decades, the transformation of office work environments from cellular to open-plan offices
has been driven by economic pressure in the management of property costs [3-5]. On
average, office space (m2) allocated per workstation has declined nearly 50% over the last
two decades [6]. Tangible economic benefits have made open-plan layout the dominant office
type across the commercial property sector, although negative effects on comfort and work
performance due to close proximity between occupants have been extensively debated in the
literature across various research disciplines [7-13].
In more recent years, the pursuit of further office space efficiency has broken the link
between workstation and employee through flexi-desking, which refers to workstations that
are shared by more than one individual and typically claimed/booked on a daily/temporary
basis. While an open-plan layout supports more people in the same amount of space by
reducing square metre per workstation, the idea of flexi-desk (also termed hot desking, desk
sharing, or non-territorial working) increases efficiency by increasing the average number of
employees per workstation.
In the traditional model of an office-based organisation, a fixed workstation is
allocated to each employee (i.e. people to workstation ratio of 1:1) based on the assumption
that employees occupy their “own” desks throughout a working day. Under this scenario, it is
observed that a significant proportion of workstations remain unused and therefore office
space utilisation falls, simply because occupants are often absent from their workstations (e.g.
annual leave, sick leave, attending training courses, and meetings outside the office) [14-16].
This problem is exacerbated where large areas are often dedicated to senior staff who are
away from their office more frequently or as a consequence of organisations employing
modern, non-standard working patterns (e.g. full time, part time, or job sharing) and location

3
(e.g. home, client’s place, or other external locations) [15,17]. The concept of non-territorial
working has been widely adopted in corporate office settings as a solution to improve
occupational space efficiency, through flexi desking or desk sharing.
In addition to direct cost savings from more intensive use of office space, as has been
pursued in the public sector [18,19], flexi-desking arrangement can bring indirect benefits to
an organisation by allowing the workplace to be more responsive to the rapid organisational
change, such as expansion, downsizing, or change in team structures [20]. As workstations
are depersonalised in the desk sharing environment, it becomes easier to re-locate staff
members, in contrast to the conventional workplaces [15]. Flexi-desk work environment has
the potential to improve teamwork. When properly designed, the layout generally affords
more space for interaction and collaboration between co-workers and contributes to cross-
departmental collaboration, as staff members are no longer confined to a designated location
and are given ample opportunity to interact with colleagues [21]. It is also argued that the
ability to choose work location generates a sense of autonomy and control over the work
environment, which might result in greater work satisfaction [22]. Furthermore, non-
territorial workplaces are believed by many to improve productivity as workers can choose
the most appropriate setting (e.g., formal/informal meeting area, quiet study room,
project/team room, cafeteria, break-out area, soft seating, etc.) for them to complete specific
tasks. Such an office configuration has been often referred to as Activity Based Workspace
(ABW) in corporate real estate for the last several years.
Notwithstanding the tangible economic advantages from maximising space efficiency,
there are obstacles and issues of concern when implementing the concept of non-territorial
working. First, significant day-to-day and long-term variations in the total number of
employees attending at work [23] can make it difficult to predict the number of desks needed.
This could lead to a loss of productivity if demand for workstations exceeds the number
available. Second, under non-territorial working conditions in which all the work areas are
shared and interchangeable, occupants tend to lose the ability to display their own identities
and define the boundaries of their surroundings [21]. Limited ability to personalise one’s
workspace contributes to a low level of perceived privacy, which in turn can lead to
employees’ emotional exhaustion [24]. Third, a clear desk policy where shared-desk users are
obligated to clear up a desk after each day of work, pack up and store their belongs in
personal lockers is not easily enforced [19]. Fourth, there is potential productivity decrement
as each employee loses time towards the daily process of finding and setting-up a
workstation, and packing-up at the end of the day. Fifth, personal hygiene is also an area of

4
concern because of the shared furniture, keyboard, and phones among multiple individuals.
Finally, there could be cultural resistance to accept the new working style. Breaking the
conventional “me and my desk” culture and withdrawing the ownership of a designated desk
from employees is acknowledged as a barrier to the introduction of non-territorial workplace
[25].
Notwithstanding the rapidly growing popularity of these non-territorial workspaces
[21,26], there have been very few empirical studies examining the potential impact of non-
territorial working environment on office users. Data currently available with regard to ‘flexi-
desking’ are mostly found in reports or booklets prepared by facilities management
consultants, highlighting the projected cost savings when this concept is applied to an office
building. Although one of the primary reasons why a growing number of organisations
implement flexi-desking policy is to maximise space efficiency, it is important to remember
that the underlying assumption is to do so without compromising individual’s comfort and
productivity [19]. However, it is indeed rare to find peer-reviewed research articles that
investigated how flexi-desking policy driven by organisations are actually perceived or
evaluated by the actual building users. In the commercial building sector, ensuring occupant
comfort and providing excellent Indoor Environmental Quality (IEQ) are widely regarded as
key performance targets of facilities management practice. The underlying logic is that IEQ
has a significant impact on the occupants’ comfort and productivity, while staff salaries
account for the largest proportion of total expenses in the life cycle of a building [27,28].
Therefore it is important to investigate how workplace IEQ is perceived by the occupants in
the office buildings in which the non-territorial working policy has been implemented, and in
turn, whether or not their comfort and productivity are compromised by the pursuit of space
efficiency.
This paper attempts to better understand how desk-sharing environments can affect
the office workforce, in relation to their satisfaction with various IEQ factors. In addition, the
paper addresses side effects of the non-territorial working policy, and examines the influence
of this type of workplace on occupants’ perceived productivity and health. Based on
empirical data derived from both quantitative and qualitative analyses performed on an office
building Post-Occupancy Evaluation (POE) survey database, the paper discusses occupants’
attitudes and reactions towards various workplace issues, depending on whether or not they
hold ownership of a pre-allocated workstation.

5
2. Methods
2.1 Occupant survey
The empirical analysis in this paper is based on an occupant survey database from
BOSSA – Building Occupant Survey System Australia. BOSSA is an officially accredited
POE system within the IEQ section of Australia’s building sustainability rating schemes
including National Australian Built Environment Rating System (NABERS) [29] and Green
Star Performance [30]. BOSSA’s online survey tool assesses the office building occupants’
satisfaction levels for key IEQ related workplace issues such as spatial comfort, indoor air
quality, thermal comfort, acoustics, visual comfort, and perceived productivity and health
[31].
BOSSA’s online survey is initiated upon the request from a client (e.g. building
owners, tenants, or property management companies) and building occupants are invited to
participate in the survey by following the web-link embedded in the recruitment email
circulated to all-staff through the participating organisation. The survey typically takes 6-7
minutes for a participant to complete. Survey responses are digitally recorded and time-
stamped. Background information about occupants and their workspaces, such as
participants’ demographics, type of work they’re engaged, time spent at workspace,
workspace layout, and workstation arrangement, is collected at the beginning of the online
survey. During this process, respondents were asked to indicate whether or not they were
assigned with a pre-allocated workstation (i.e. fixed OR no-fixed). Simultaneously, basic
information about the surveyed building such as location, size, HVAC (heating, ventilating
and air-conditioning) systems, materials, design features, sustainability ratings is collected
from the building owner or the facilities management team. Depending on the availability,
floor plans of the surveyed building are collected in order to better understand the building’s
spatial characteristics and space allocations. Table 1 summarises the IEQ survey
questionnaire items used for the current analysis. The respondents rate their satisfaction level
with each questionnaire item on the seven-point bipolar rating scale (e.g. dissatisfied –
satisfied, disagree – agree, uncomfortable – comfortable), which is coded with numerical
values ranging from 1 to 7.

6
Table 1. List of BOSSA questionnaire items employed for the analysis
Category
IEQ items
Survey questions
Rating scale (7-point)
Spatial comfort
Space for breaks
This building provides pleasant spaces (e.g. indoor or
outdoor green space, break-out areas) for breaks and
relaxation.
1= Disagree ~
7= Agree
Interaction with
colleagues
How do you rate your normal work area's layout in terms
of allowing you to interact with your colleagues?
1= Dissatisfied ~
7= Satisfied
Personalisation of
work area
My normal work area can be adjusted (or personalised) to
meet my preferences.
1= Disagree ~
7= Agree
Space to
collaborate
The building provides adequate formal and informal
spaces to collaborate with others.
1= Disagree ~
7= Agree
Comfort of
furnishing
Please rate how comfortable your work area's furnishings
are (including chairs, desk, equipment, etc).
1= Uncomfortable ~
7= Comfortable
Amount of
workspace
Please rate your satisfaction with the amount of space
available to you at your normal work area.
1= Dissatisfied ~
7= Satisfied
Storage space
Please rate your satisfaction with the amount of personal
storage space available to you.
1= Dissatisfied ~
7= Satisfied
Air quality and
thermal comfort
Air quality
Please rate your satisfaction with the overall air quality in
your work area.
1= Dissatisfied ~
7= Satisfied
Temperature in
winter
Please rate your satisfaction with the temperature
conditions of your normal work area in winter.
1= Dissatisfied ~
7= Satisfied
Temperature in
summer
Please rate your satisfaction with the temperature
conditions of your normal work area in summer.
1= Dissatisfied ~
7= Satisfied
Noise distraction
Unwanted
interruption
The work area's layout enables me to work without
distraction or unwanted interruptions.
1= Disagree ~
7= Agree
Overall noise
Please rate your satisfaction with the overall noise in your
normal work area.
1= Dissatisfied ~
7= Satisfied
Visual comfort
Lighting
Please rate your satisfaction with the lighting comfort of
your normal work area (e.g. amount of light, glare,
reflections, contrast)?
1= Dissatisfied ~
7= Satisfied
Access to daylight
Please rate your satisfaction with the access to daylight
from your normal work area
1= Dissatisfied ~
7= Satisfied
General
Degree of freedom
to adapt
All things considered, how satisfied are you with the
degree of freedom to adapt your normal work area (air-
conditioning, opening the window, lighting, etc.) to meet
your own preferences?
1= Dissatisfied ~
7= Satisfied
Overall building
How satisfied are you with this building overall?
1= Dissatisfied ~
7= Satisfied
Productivity
How does your work area influence your productivity?
1= Negatively ~
7= Positively
Health
How does your work area influence your health?
1= Negatively ~
7= Positively
Open-ended
comments
If you would like to add any additional comments, please
type your comments here (this item is only applied to
surveys for Buildings Q and R in Table 3)
N/A
2.2 Sample buildings and participants!
The current analysis was performed on a BOSSA dataset collected in 20 office
buildings located in capital cities in Australia between 2012 and 2014. In general the dataset
comprises contemporary office buildings, with most of the surveyed buildings falling into the
“Premium” grade of the Property Council of Australia’s office building quality guidelines
[32]. The organisations occupying the sample buildings were engaged in different businesses
in diverse industries such as building and construction, education, engineering, financial
services, manufacturing, professional services, public administration, and real estate. The size
of buildings, estimated by Net Lettable Area (NLA), varied from 2,000 to 66,000 square

7
metres. All the buildings in the current dataset were serviced by centralised HVAC systems.
Filtering out missing values and irrational or internally inconsistent responses to assure the
quality of the dataset, a total of 3,974 individual responses were obtained with the average
response rate (i.e. number of completed questionnaires divided by number of email
invitations sent) of 45.2%. Survey respondents’ personal characteristics (gender, age), work
characteristics (type of work, time spend at workspace), and workspace characteristics (office
layout, workstation arrangement) are described in Table 2. The majority of the participants
were open-plan office occupants (96%) with an age range between 31 to 50 years (56.2%),
engaged in professional (36.1%) or administrative role (23.1%), spending full-time
equivalent hours (71.4%) at the workplace (over 30 hours/week). The responses from those
who had a pre-assigned desk were slightly higher (55.8%) than those who did not have a
fixed location of desk (44%).
Table 2. Characteristics of the respondent samples (total n = 3,794)
Occupant profile
and workspace
characteristics
Description
Sample size
(n)
Percenta
ge (%)
Gender
Female
1,989
50.1
Male
1,985
49.9
Age
30 years old or under
1,030
25.9
31 to 50 years old
2,234
56.2
Over 50 years old
710
17.9
Type of work
Administrative
919
23.1
Technical
627
15.8
Professional
1,435
36.1
Managerial
619
15.6
Other
374
9.4
Time spent in
building
Less than 6 months
469
11.8
7 to 12 months
888
22.3
1 to 2 years
906
22.8
2 to 5 years
899
22.6
More than 5 years
812
20.4
Time (hours per
week) spent at
workspace
10 hours or less
255
6.4
11 to 30 hours
881
22.2
More than 30 hours
2,838
71.4
Office layout
Private office
71
1.8
Private office shared with others
44
1.1
Open plan office with high (>1.5m) partitions
479
12.1
Open plan office with low (<1.5m) partitions
2,409
60.6
Open plan office without partitions
926
23.3
Other
38
1.0
Missing data
7
0.2
Workstation
arrangement
Fixed location (includes exclusive and shared
use of the same workstation)
2,219
55.8
No fixed location (e.g. flexi-desk, activity
based workspace)
1,748
44.0
Missing data
7
0.2

8
Table 3 summarises the building-by-building breakdown of responses by workstation
arrangement. For some sample buildings, responses were logged exclusively or
predominantly from occupants assigned with a fixed-desk (Building A, B, F, H, I, J, K, L, M,
N, O, P, S, T), whereas some other buildings registered responses primarily from flexi-desk
users (Building G, Q, R). Building C, D and E recorded comparable numbers of responses
from both workstation types. Table 3 presents the space allocation characteristics for each
building. The amount of space allocated per desk, with respect to Net-Lettable Area (NLA),
work area (spaces primarily used for individual offices or workstations), meeting area
(inclusive of conference rooms, meeting rooms and project tables), and break-out area (such
as lounge, cafeteria, soft seating, and kitchen), was calculated for typical floors of a building,
based on the floor plans provided. Only 11 of the 20 sample buildings are provided with
space allocation data in Table 3 due to floor plan unavailability in the remainder.
Table 3. Workstation arrangement (fixed or no-fixed) distribution of sample buildings
Building
code
Space allocation (m2/desk)
Workstation arrangement
(number of responses)
Response
rate (%)
Total n
NLA/d
esk
Work area /
desk
Meeting
area / desk
Break area /
desk
Fixed
No-fixed
(Flexi)
A
-
-
-
-
142
0
56
142
B
-
-
-
-
61
0
53
61
C
12.6
8.2
1.2
0.7
189
123
31
312
D
10.3
5.5
0.7
0.5
212
171
52
383
E
-
-
-
-
189
179
74
368
F
13.9
6.4
0.9
0.3
197
20
31
217
G
-
-
-
-
13
122
-
135
H
14.4
11.1
0.4
0.6
28
0
65
28
I
12.9
9.4
0.3
0.9
130
1
54
131
J
10.8
6.8
0.9
1.1
150
0
53
150
K
11.4
5.3
0.9
0.3
109
3
20
112
L
-
-
-
-
83
1
28
84
M
-
-
-
-
95
2
22
97
N
-
-
-
-
74
2
38
76
O
-
-
-
-
221
4
38
225
P
-
-
-
-
156
3
25
159
Q
15.5
4.9
1.4
1.4
46
757
-
803
R
6.9
3.1
0.5
0.3
40
355
-
395
S
11.4
8.0
0.9
0.9
54
3
76
57
T
21.3
10.5
3.6
0.7
30
2
53
32
Average
12.9
7.2
1.1
0.7
-
-
45
-
Total
-
-
-
-
2,219
1,748
3,967
The BOSSA’s online platform offers the ability to customise or add questionnaire
items to investigate specific issues of interest to researchers or clients. Survey questionnaires
administered to Building Q and Building R in Table 3 included an additional open-ended

9
question at the end of the survey, asking the respondents to register any comments on a
voluntary basis. The open-ended question encourages occupants to express their views on
workplace related issues and provide information that can be used to diagnose urgent
workplace problems or to identify salient issues in a building. A total of 371 individual
comments were recorded by occupants in Building Q and R, providing a substantial volume
of qualitative data for analysis. This is particularly relevant as Building Q and R contain the
largest number of responses from non-territorial workspaces (Table 3). These two buildings
are exclusively occupied by the same organisation with a people-to-workstation ratio of
approximately 1.3 (i.e. 100 workstations shared by 130 people) on their flexi-desking policy.
They include a range of workspaces such as formal and informal meeting spaces,
collaborative tables, casual seats, quiet rooms, as well as individual workstations in an open-
plan office layout, as well as dining and vending machine areas, storage, concierge, etc. Upon
their arrival at the building, occupants of Buildings Q and R can choose where to work but
are required to clear up and store their belongings, including personal laptops, in personal
lockers when they finish using the workspace.
3. Results and Discussion
3.1 Occupant satisfaction with different aspects of IEQ
The occupant satisfaction levels for a range of IEQ issues were compared between the
two occupant groups across the POE dataset: those who have pre-allocated workstations
(n=2,219) and those who don’t (n=1,748). Fig. 1 illustrates the mean rating scores for
individual questionnaire items (shown in Table 1), expressed on the seven-point scale within
the bounds of 1 (the lowest score) to 7 (the highest score) by the respondents. The ‘flexi-desk’
group outscored ‘fixed-desk’ group across most of the IEQ items addressed in the
questionnaire, giving more satisfied rating scores on 16 out of the 18 items assessed, except
for ‘amount of workspace’ and ‘storage space’. The differences between the two sample
groups reached statistically significant thresholds on all 16 items (p<0.05 for ‘unwanted
interruption’ and ‘access to daylight’, p<0.001 for the remaining 14 items). The biggest mean
difference was observed for ‘space for breaks’ (1.4 unit difference on 7-point scale), followed
by ‘space to collaborate’ (0.8 unit) and ‘air quality’ (0.8 unit). Both groups reported the
identical mean satisfaction score (5.4) on the amount of workspace available. The only topic
that the flexi-desk group reported lower satisfaction compared to the fixed-desk group related
to the amount of personal storage space provided.

10
Fig. 1. Comparison of mean rating scores (within the range of 1 = the lowest to 7 =
the highest) for individual questionnaire items between desk-assigned (fixed desk) group and
flexi-desk (no-fixed desk) group (Error bars: 95% confidence interval).
Fig. 1 indicated that flexi-desk users are generally more satisfied with most of the IEQ
issues in question, compared to fixed-desk users. Furthermore, flexi-desk users also
registered higher average rating scores on the summative questionnaire items including
overall building satisfaction, perceived productivity and health. Despite most of the IEQ
factors in question being rated more positively by the flexi-desk users, simple generalisations
such as ‘flexi-desking’ outperformed ‘fixed-desking’ arrangement should be resisted because
potential confounding factors cannot be controlled in the comparison analysis across a
number of workplaces. There is a considerable variability in relation to interior fit-outs,
design features, office layouts, ambient environmental conditions, quality of building
maintenance services etc. between sample buildings as well as between floors within a
building. These variations make it difficult to investigate the isolated effect of the desk
arrangement on occupants’ responses. However, some rational explanations in terms of
‘spatial comfort’ topics can be derived by the examination of space allocation characteristics
within the sample buildings reported in Table 3. Based on the space allocation data in Table 3,
Table 4 compares the average space allocation (m2/desk) of the two sample groups: 1) those
buildings operating predominantly on a fixed-desk policy; and 2) those buildings operating
predominantly on a flexi-desk arrangement. The average values in Table 4 were weighted by
the number of survey responses collected from each sample building, in order to be
compatible with the rating scores presented in Fig. 1. While the internal space (NLA) per
desk was almost identical between ‘fixed-desk’ samples (12.8m2) and ‘flexi-desk’ (12.7m2)
Mean response
7
6
5
4
3
2
1
4.5
4.8
4.7
3.4
4.9
5.1
3.9
3.7
4.8
4.6
5.0
4.3
5.4
4.9
5.1
4.3
5.2
4.7
4.1
4.4
4.2
3.1
4.7
4.9
3.7
3.6
4.1
4.1
4.2
5.0
5.4
4.5
4.3
3.9
4.8
3.3
Health
Productivity
Building overall
Degree of freedom to adapt
Access to daylight
Lighting
Noise
Unwanted interruption
Temperature in summer
Temperature in winter
Air quality
Storage space
Amount of workspace
Comfort of furnishing
Space to collaborate
Personalisation of work area
Interaction with colleagues
Space for breaks
Flexi deskFixed desk
Workspace arrangement

11
samples, the amount of space allocated for individual work area, meeting area and break area
were different between the two groups. On average, less space for individual work was
allocated in the flexi-desk workplaces (4.3m2) compared to the fixed-desk workplaces
(7.3m2). However, the reduction of individual work area seems to be compensated by a
substantial increase in the amount of space for break-out areas (67% more than the fixed-desk
group) and meeting areas (22% more) in flexi-desk workplaces. Therefore, the noticeable
differences of occupant satisfaction level on ‘space for breaks’, ‘interaction with colleagues’
and ‘space to collaborate’ in Fig. 1 can reasonably be explained by the space use
characteristics of the two sample groups. Interestingly, both group expressed the same level
of satisfaction with ‘amount of workspace’, despite a much smaller amount of individual
workspace in ‘flexi desk’ group. This is probably because the nature of the flexi-desking
arrangement allows occupants to extend their work domain from the conventional individual
workstation to diverse activity-based work settings such as formal/informal meeting spaces,
project tables, casual seats, cafeteria, etc. On the other hand, despite the higher satisfaction
level for most of the spatial issues addressed in the questionnaire, flexi-desk occupants rated
personal storage space lower than the desk-assigned group. Given there is no quantitative
data available with respect to the personal storage facilities provided, we speculate that the
clear-desk policy (not being able to leave work-related materials and items on the top of desk
when the use of desk terminates) and effectiveness of centralised locker facilities (instead of
personal drawers and cabinets for those assigned with fixed-desk) might have negatively
influenced flexi-desk respondents’ evaluation of personal storage space.
Table 4. Sample buildings’ average space allocation (m2) per desk, weighted by the number
of survey responses from each building
Predominant desk arrangement
NLA (m2) per
desk
Individual work
area (m2) per desk
Meeting area
(m2) per desk
Break area
(m2) per
desk
Average (weighted by N) of Building
F, H, I J, K, S and T (predominantly
fixed desk arrangements)
12.8
7.3
0.9
0.6
Average (weighted by N) of Building
Q and R (predominantly flexi desk
arrangements)
12.7
4.3
1.1
1.0
Another intriguing difference between the two sample groups appeared with respect
to indoor air quality and thermal comfort questionnaire items where the scores for flexi desk
respondents was higher than the fixed desk respondents (Fig. 1). Occupants across the

12
sample buildings (all of which had centralised HVAC systems) could be expected to
experience broadly similar indoor ambient conditions particularly with respect to temperature
given that in Australia “Premium Grade” offices are almost always controlled to operate
within a very narrow band about the set point throughout a year - typically the range across
the entire sample buildings as specified in the lease agreements was 21.5~23°C.
Consequently, it could be conjectured that the differences observed between the two groups
potentially result from HVAC systems in the ‘fixed-desk’ sample buildings somehow
underperforming compared to those in the ‘flexi-desk’ sample buildings. However, in order
to look into this issue more closely, t-test was conducted on a subset of the database, where
responses from Building C, D and E, which contain a comparable number of ‘flexi-desk’ and
‘fixed-desk’ occupant groups (Table 3), were analysed. The underlying rationale of this
analysis is that both the occupant groups in each of these buildings are exposed to the
equivalent ambient conditions because the same HVAC system of their base building served
both types of workstation. In other words, the key factors affecting perceived air quality (e.g.
building services system that filters/conditions/supplies the air) and thermal comfort (building
envelop systems and HVAC systems) are technically identical for both groups.
The results of the independent samples t-test are summarised in Table 5, indicating
significant group differences. Therefore, the question arises: what makes these occupants
respond differently on air quality and temperature issues when they experience identical
ambient conditions provided by the same building services system of the same building? One
possible explanation for the higher scores from flexi desk occupants can be their ability to
choose the location of their work area. Fig. 2 illustrates the distribution of participants’
responses on a 7-point rating scale (1=disagree ~ 7=agree), when they were asked if their seat
selection was influenced by the overall indoor environmental quality of the work area. (This
particular branching question is directed at respondents who indicate flexi-desk as their
primary workspace arrangement.) Over 80% of the respondents agreed (those rating 5, 6 and
7 on the 7 point scale) that IEQ affects their decision of seat selection. The results
demonstrate occupants in the non-territorial workplace can use their choice of seat location as
a means of adjusting IEQ to suit their preferences and exercise a certain degree of personal
control over the surrounding conditions, which is well known to be beneficial to employee
satisfaction [33]. For example, occupants can avoid locations that have a known source of
discomfort such as excessive solar radiation near the external glazing and cold draft blown
from air diffusers. In effect, the ‘flexi-desk’ group responses indicate that they have a higher
degree of adaptive opportunity to meet personal preferences than ‘fixed-desk’ group (Fig. 1).

13
While not wishing to overgeneralise, it does appear that the ‘Locus of control’ [34,35], which
suggests that an occupant’s attitude towards the building can be affected by perceived control
over the indoor environment [36], might have played a role in the respondents’ perceived air
quality and thermal comfort.
Table 5. The difference of mean rating score (rated on a 7-point scale ranging from 1 =
dissatisfied to 7 = satisfied) between ‘fixed-desk’ group (n=590) and ‘flexi-desk’ group
(n=473) in Building C, D and E, on indoor air quality and thermal comfort questions
Questionnaire item
Mean rating score
t-test
Fixed desk
Flexi desk
Mean difference
t
sig.
Air quality
3.8
4.2
-0.4
-3.1
p<0.01
Temperature in winter
3.7
3.9
-0.2
-2.2
p<0.05
Temperature in summer
3.8
4.2
-0.4
-4.1
p<0.001
Unwanted interruption
3.7
3.6
0.1
1.5
NS
Noise
3.6
3.6
0.0
-0.6
NS
Lighting
4.6
4.7
-0.1
-0.7
NS
Access to daylight
4.5
4.3
0.2
1.8
NS
Fig. 2. Histogram of survey responses from flexi-desk users on “The overall indoor
environmental quality of my work area influences my seat/location selection”, rated on a 7-
point scale ranging from 1 = disagree to 7 = agree.
It could be argued that other indoor environmental quality aspects such as access to
daylight, managing unwanted interruptions from noise may also influence occupants’
seat/location selection and responses in Fig 2. However in our case study group, it should be
noted that we found no significant differences between fixed-desk and flexi-desk groups for
the noise and lighting variables as shown in Table 5. These results indicate that flexi-desk
arrangement in itself does not negate the manner in which noise or lighting is rated in these
buildings.
Number of responses
400
300
200
100
0
28.2%
33.1%
18.8%
10.0%
3.7%
2.8%
3.4%
7=Agree
6
5
4
3
2
1=Disagree

14
3.2 Open-ended comments
This section presents qualitative data drawn from the open-ended comments
registered by the occupants of two sample buildings, Building Q and R. The collated set of
371 individual comments was read through several times to acquire a sense of the whole
picture, and to identify substantive topics embedded in the occupants’ feedback on their
buildings. Table 6 presents the major themes identified by the respondents’ open-ended
comments. Seven categories were established based on a total of 371 individual comments
recorded (311 from Building Q and 60 from Building R): 1) flexi desking, 2) spatial comfort,
3) air quality and thermal comfort, 4) noise distraction, 5) visual comfort, 6) building
maintenance and cleanliness, and 7) common service facilities and office equipment. The
qualitative data was also examined in terms of the respondents’ attitudes toward the topics
that they have raised. Each open-ended comment was grouped into either ‘positive’ or
‘negative’ categories (Table 6). Those comments taking a rather neutral position were
excluded in ‘attitude’ column in Table 6. Many respondents declared that their productivity
or health was somehow affected as a result of certain workspace aspect. Respondent
comments relating to productivity (such as “reduce/decrease productivity”, “waste of
effective working time”, “decrease collaboration”, etc.) and health symptoms (such as
“headache”, “neck/back pain”, “sore eyes”, “cold/flu”, etc.) were categorised, and the
frequency of such comments is cross tabulated in Table 6.
Table 6. Frequencies of occupant open-ended comments falling into each category of the
workplace topics, cross-tabulated by the respondent’s attitude, and impact on perceived
productivity and health
Topic
Total number of
comments
Respondent attitude
Productivity
affected
Health
affected
Positive
Negative
Flexi-desking
115 (30.6%)
10
97
33
14
Spatial comfort
67 (17.8%)
6
50
7
5
Air quality and thermal
comfort
49 (13.0%)
2
43
1
3
Noise distraction
39 (10.4%)
0
36
11
0
Visual comfort
35 (9.3%)
1
27
1
3
Building maintenance
and cleanliness
28 (7.4%)
3
25
1
0
Common service
facilities and office
equipment
23 (6.1%)
3
20
0
0
Other
20 (5.3%)
3
7
0
2
Total
376 (100%)
28
305
54
27
Of the 376 total open-ended comments, negative feedback was dominant (305
occurrences) across all categories identified in Table 6. It seems that occupants tended to use

15
the opportunity of open-ended comments to lodge complaints or express dislikes rather than
to praise good aspects (305 negative vs. 28 positive comments). In particular, spatial
workplace arrangements and protocols enforced by the flexi-desking policy were the subjects
of most complaints in the open-ended comments.
The most prominent topic across the entire collection of open-ended comments was
identified as ‘flexi-desking’ which accounted for a third of the total comments received.
Only those comments directly citing ‘flexi-desking’ were included in this category.
Interestingly, most comments on this topic were overwhelmingly critical (97 negative
responses out of 115 comments on this topic). Additionally, of the 115 comments on ‘flexi-
desking’, 29% and 12% of the respondents evaluated that their productivity and health were
respectively affected by the flexi-desking work arrangement. Examples of typical complaints
associated with flexi-desk environments were: non-available desk upon arrival at workplace,
difficulties in locating team members, waste of time during workstation set-up and pack-up
process, and limitation in adjusting/personalising one’s work area or furniture. Although
comments on the flexi-desking arrangement somewhat overlap with those comments
categorised as ‘Spatial comfort’ in Table 6, ‘flexi-desking’ was categorised separately in the
present analysis as it was the single most cited theme across all open-ended comments
recorded. A detailed discussion on ‘flexi-desking’ is undertaken in a later section of this
paper.
The second most recurrent topic was ‘spatial comfort’. This category encompassed
issues such as personalisation of workspace, design and ergonomics of furniture, storage
space, availability of meeting rooms, and the colours and textures of interior finishes. Indoor
ambient conditions also rated a mention. Comments using descriptors such as ‘odours’, ‘too
hot/cold’, ‘stuffy’, and ‘insufficient ventilation’ were categorised into ‘Air quality and
thermal comfort’ while ‘Noise distraction’ was used to categorise complaints regarding
interruption/distraction from background noise and the inability to concentrate on task.
Eleven of the 39 respondents in ‘Noise distraction’ category declared that their productivity
was impaired due to unwanted background noise. Mentions of insufficient daylight, glare or
reflections on monitor screens were categorised into ‘visual comfort’. More general
comments such as the quality of cleaning services provided, maintenance of common areas,
the building management team’s responsiveness to complaints, office equipment support, and
design and functionality of service facilities (e.g. lifts and escalators), were appropriately
grouped into ‘Building maintenance and cleanliness’ or ‘Common service facilities and office
equipment’.

16
3.3 Building occupant complaints about flexi-desking arrangements
Table 7. Open-ended comments on the flexi-desking arrangement: major problems, example
comments and frequency (Percentages refer to the total number of negative comments on
flexi-desking not the number of respondents making them).
Problems
Frequency
Examples
Not enough desks
26.8%
“The only issue that I could raise is that it's often incredibly hard to find a desk and
many of us end up in the kitchen all day.”
“The workspace is not being used as effectively as it could be because people park
themselves at a desk even if they won't be there for hours and there isn't enough
adequate seating for the number of occupants.”
“As a rule, if I arrive at work after 8:30 I need to set up at tables in the kitchen area.”
“The main issue with the building is the lack of available desks on busy days. This
can cause wasted time in trying to find a desk or workspace and has led me to work
less from the office”
Difficult to locate
team members:
limits immediate
collaboration and
impedes team
communication
21.6%
“Difficulty with flexi-desking is that it is difficult to locate people when you need to
speak to them as they are always in a different place”
“Finding that a team member has to sit on another floor (because they couldn't get a
seat nearby) does not support collaborative working; and limits ad hoc conversations
to quickly resolve issues”
“I find it counterproductive that my team can’t sit together. Co-operation and
collaboration have suffered as a result.”
“People want to work together in team environments but are forced to rely on email
and phone calls (both poor collaboration tools when compared with face to face)
because the team is scattered all over the place.”
Waste of time
finding, setting-up
and packing-up a
workstation
12.4%
“We are supposed to sit at a different desk every day. This means you always have to
waste time setting up your workplace.”
“Flexi-desk rationalises usage of spaces, but has a negative impact on productivity
because of time to select, set up work station each day, and slow start up time from
lap tops”
Limited ability to
adjust/personalise
workstations
(desk, chair,
screens and other
equipment) to
meet individual
ergonomic needs
11.3%
“I constantly experience neck and upper shoulder pain due to working off a laptop in
a non adjustable chair/desk.”
“I am required to sit at the short term desks between meetings, which is how it's
meant to be, however after a long period of time of looking down at laptop my neck
and back do start to ache.”
“I am concerned that the flexible working arrangements are causing an increase in
sore necks and backs etc. I have had 3 direct reports with complaints.”
“I am so much more productive with 2 large screens. In this progressive flexi desk
environment I frequently have zero large screens, having to get by with a small
laptop screen. Progressive? Yes!, Productive? Noooooooo!!!!”
Insufficient
storage support
5.2%
“We understand the clear desk policy however, we cannot unplug printers etc each
night nor do we have sufficient storage to lock up our stationary each night”
“Not enough lockers for flexi-desking, I had to wait 4 months when I moved and
now moving again, takes a long time.”
Personal hygiene
risk due to desk
sharing
4.1%
“I suspect there has been an increase in illness due to flexi desking (contaminated
surfaces)”
“No matter where I sit I always find a hair on the desk I choose.”
Other
18.6%
“Flexi desk environments are designed to save organisations from floor space costs,
but at the expense of employee comfort, privacy and productivity”
“The implementation of this office defies the psychological need for a sense of
place. Also, the zoning means that flexi-desking is a fiction. Altogether
environmental stress is high”
“In a flexidesk area people should not be able to reserve desks for friends or leave
their laptops on desks overnight to continually secure the best areas to work”
“I originally liked hot desking, but I'm over it now. I start work early so that I can sit
in my preferred location”
This section provides more detailed data on how flexi-desking work environments are
evaluated by building occupants through a sub analysis of the 97 negative comments on the

17
flexi-desking arrangement recorded in Table 6. Table 7 summaries the most common
problems with typical examples of actual comments as well as the frequency of such
comments. The most frequently cited (26.8%) problem of ‘flexi-desking’ in the sample
buildings was the insufficient number of available desks on busy days (as previously noted
the employee to desk ratio in these buildings was 1.3). The respondents complained that the
lack of available desk upon arrival resulted in them have to work at inadequate settings such
as kitchen benches, cafeteria, short-term desks, or couches. The second most cited (21.6%)
downside of the flexi-desking workplace related to difficulties in finding colleagues and co-
locating with team members. The respondents indicated that the current flexi-desk setting
limited immediate collaboration and communication with work teams, particularly when
there is a problem that needs to resolve promptly. The respondents pointed out that the
electronic engagement methods available to them (e.g. emails, messengers and phones)
weren’t as effective as face-to-fact interactions. Some respondents (12.4%) were also very
sceptical about the concept of non-territorial workspace because their effective working time
was wasted during the repeated process of finding, setting-up and packing-up a workstation.
Eleven percent (11.3%) of the respondents indicated that their specific needs to work
efficiently and individual ergonomic needs were not fully supported in non-territorial
workspace environments. Limited ability to adjust or personalise one’s workstation (e.g.
desk, chair, partition screens, etc.) due to the standardisation of flexi-desking settings was
widely criticised in the open-ended comments. Insufficient storage support and personal
hygiene risks due to desk sharing were also mentioned by the occupants as issues of concern.
The analysis of text data as shown in Tables 6 and 7 points to two salient complaints
regarding non-territorial workplace: 1) productivity decreases because of insufficient
workspace (desk, storage), difficulties to locate colleagues, waste of time setting-up/packing-
up, and inability to adjust/personalise workstations; 2) health affected due to inability to
adjust/personalise workstations and uncleanliness of shared desks. Fig. 3 illustrates these
patterns in the respondents’ open-ended comments. Therefore we can hypothesise that these
spatial issues in question may negatively influence occupants’ perceived productivity and
health when combined with flexi-desking policy.
The analysis presented here throws light on potential and real concerns that must be
taken into account in the design of flexi-desk work arrangements. However, it is
acknowledged that the open-ended comments come from a self-selecting sample and do not
necessarily represent the view of the entire sample population. As previously described, the
open-ended question was voluntary and only administered in sample buildings Q and R.

18
Furthermore, of the small fraction of survey respondents who registered open-ended
comments majority were more inclined to complain about their workspaces. As the above
results are based on feedback that is likely to be more negatively biased, further investigation
would be necessary before they can be generalised to draw more robust conclusions. The next
section presents the results of logistic regression analysis performed on the entire BOSSA
database to test the hypothesis defined in Fig 3.
Fig. 3. Schematic diagram of the salient patterns observed in, and impacts implied by open-
ended comments on the flexi-desking arrangement.
3.4 Non-territorial working and occupant perceived productivity and health
In order to investigate whether the link between the ‘flexi-desking’ associated issues
and workplace productivity/health (defined in Fig. 3) can be generalised, further statistical
analysis was carried out on the entire BOSSA database (n=3,974). As most of the issues
arising in the open-ended comments can be directly mapped to BOSSA’s ‘spatial comfort’
questionnaire items listed in Table 1 (e.g. space for work/storage/break, staff
interaction/collaboration, adjustability/comfort of furniture), our analysis here focuses on
these items. A set of logistic analyses were performed with the type of desk arrangement (i.e.
fixed vs. flexi) and ‘spatial comfort’ questionnaire items (satisfaction score rated on the 7-
point scale) as the independent variables (predictor), and the occurrence of negative votes on
‘productivity’ and ‘health’ questions (the bottom 3 categories on the 7-point rating scale) as
the dependent variables.
The results are summarised in Table 8. The results of the logistic regression model
are reported as odds ratios (e.g. An odd ratio greater than 1 indicates that the outcome is more
likely to occur). In the current analysis, the values of odds ratios (OR) for each individual
‘spatial comfort’ questionnaire item (predictor) represents the likelihood of the respondents in
a given desk configuration declaring that their productivity or health were negatively
Flexi-
desking
Not enough workspace (available desk, storage
space)
Difficult to locate team members (limits
immediate collaboration and deteriorate team
communication)
Waste of time finding a desk, setting-up and
packing-up
Limited ability to adjust/personalise workstations
(desk, chair, screens and other equipment) to
meet one’s own needs and comfort standard
Personal hygiene issue due to sharing a desk
(contaminated surface)
Impact
Productivity
Impact
Health

19
influenced, as their rating scores for that ‘spatial comfort’ item decreases by 1 unit on the 7
point scale.
Table 8. Results of logistic regression model for perceived productivity. Odds ratios (OR)
represent the likelihood of the occupants declaring that workspace environments had a
negative impact on their productivity.
Predictors
b
Wald
Sig.
Odds Ratio
95% CI
Desk arrangement (fixed or flexi)
NS
Space for breaks × Fixed desk
0.23
33.43
p<0.001
1.26
1.17-1.36
Space for breaks × Flexi desk
NS
Interaction with colleagues × Fixed desk
0.12
9.70
p<0.01
1.13
1.05-1.22
Interaction with colleagues × Flexi desk
0.26
27.34
p<0.001
1.29
1.17-1.42
Personalisation of work area × Fixed desk
NS
Personalisation of work area × Flexi desk
0.16
13.39
p<0.001
1.17
1.08-1.27
Space to collaborate × Fixed desk
0.17
17.58
p<0.001
1.19
1.10-1.29
Space to collaborate × Flexi desk
0.12
6.04
p<0.05
1.12
1.02-1.23
Comfort of furnishing × Fixed desk
0.22
31.35
p<0.001
1.25
1.16-1.35
Comfort of furnishing × Flexi desk
0.23
23.13
p<0.001
1.26
1.15-1.39
Amount of space × Fixed desk
0.19
17.25
p<0.001
1.21
1.11-1.33
Amount of space × Flexi desk
0.18
14.07
p<0.001
1.20
1.09-1.32
Storage space × Fixed desk
NS
Storage space × Flexi desk
0.16
15.57
p<0.001
1.18
1.09-1.28
R2=0.22(Cox&Snell), 0.32(Nagelkerke). Model χ2(15)=957.63, p<0.001
As seen in Table 8, the main effect of desk arrangement (fixed or flexi) was
insignificant. That is, the occupants’ self-assessed productivity is apparently unaffected by
whether or not occupants have a pre-assigned desk. On the other hand, the regression model
identified the interaction of the desk arrangement type and ‘spatial comfort’ predictors,
significantly increasing (i.e. OR>1) the likelihood of negative evaluations on workplace
productivity as discussed below.
For those have a fixed desk, the most outstanding predictor for ‘productivity
decreased’ was ‘space for breaks’. In this case, the odds ratio (1.26) indicates that as their
satisfaction rating for ‘space for breaks’ falls, occupants are 26% more likely to complain that
their productivity was being negatively influenced. On the other hand, as seen in Table 8,
there was no significant result to associate ‘space for breaks’ with negative ratings for
productivity for flexi-desk users. This is probably because break-out areas allocated in the
‘flexi-desk’ workplaces are already deemed to be satisfactory by their occupants, resulting in
no negative associations. As previously discussed, not only is space for break-out areas
higher in flexi-desk workplaces (Table 4), but more importantly occupants here registered a
high satisfaction score for ‘space for breaks’ (Fig 1) and had no complaints regarding these
spaces (Table 7). The results emphasise while the generous provision of break-out spaces in
flexi-desk workplaces can translate into positive responses, the insufficient provision of

20
break-out areas in conventional fixed-desk environments is clearly associated with lower
ratings for perceived productivity.
Table 8 also shows occupant satisfaction with layout in terms of enabling ‘interaction
with colleagues’ was a significant predictor of self-assessed productivity for both types of
work arrangements. However, it exerted greater influence (OR=1.29) over the ‘flexi-desk’
group compared to the ‘fixed-desk’ group (OR=1.13). The regression model result for the
flexi-desk group is in line with those open-ended comments that pointed out difficulties in
interacting with co-workers as a negative side effect of the flexi-desking arrangement (Table
7).
The differences between ‘fixed-desk’ and ‘flexi-desk’ groups are evident for
‘personalisation of work area’. While not significant to those having a fixed-desk, the ability
to adjust/personalise the workstation to meet personal preferences mattered significantly to
those in non-territorial workplaces. The model predicts 17% increase in the likelihood of
‘productivity decreased’ votes (OR=1.17) when occupants give lower scores for
‘personalisation of work area’ in the flexi-desk environments. The interaction of ‘storage
space’ with ‘flexi-desk’ was also identified as significant, implying potential drop in
productivity when occupants are dissatisfied with storage options provided to non-territorial
workplaces. In contrast, ‘storage space’ was not associated to productivity for the ‘fixed-
desk’ group, presumably because the amount of storage space they enjoy is already deemed
good enough.
Other aspects such as ‘space to collaborate’, ‘comfort of furnishing’ and ‘amount of
space’, were found to interact with both desk arrangements, with lower satisfaction ratings
for these variables increasing the likelihood of negative votes on the productivity scale (i.e.
OR>1).
Table 9 reports the results of logistic regression analysis of the association between
the desk arrangement and the perceived impact of workplace on health. Whether an occupant
is assigned to a fixed desk or not was found to be insignificant with respect to the occurrence
of negative votes on the health rating scale. Conversely, the interaction of ‘comfort of
furnishing’ with the flexi-desk arrangement reported the highest OR of 1.49 in the regression
model. What this means is that, under flexi-desk settings, occupants were 49% more likely to
report their health as being negatively influenced by their workspace environments if their
workstation furnishings such as chairs and desk were deemed to be uncomfortable. These
findings are consistent with complaints regarding furniture and prevalence of health
symptoms arising from postural problems (e.g. neck, back or shoulder pain) registered by

21
occupants of non-territorial workspaces (Table 7). Similar findings were observed for the
fixed-desk group but the likelihood was lower (1.30) than that of the flexi-desk group (1.49).
Table 9. Results of logistic regression model for perceived health. Odds ratios (OR) represent
the likelihood of the occupants declaring that workspace environments had a negative impact
on their health.
Predictors
b
Wald
Sig.
Odds Ratio
95% CI
Desk arrangement (fixed or flexi)
NS
Space for breaks × Fixed desk
0.28
60.10
p<0.001
1.32
1.23-1.42
Space for breaks × Flexi desk
0.18
21.44
p<0.001
1.20
1.11-1.30
Interaction with colleagues × Fixed desk
0.12
11.12
p<0.01
1.13
1.05-1.21
Interaction with colleagues × Flexi desk
NS
Personalisation of work area × Fixed desk
NS
Personalisation of work area × Flexi desk
0.11
7.65
p<0.01
1.12
1.03-1.21
Space to collaborate × Fixed desk
0.10
6.57
p<0.05
1.10
1.02-1.19
Space to collaborate × Flexi desk
0.10
5.01
p<0.05
1.10
1.01-1.20
Comfort of furnishing × Fixed desk
0.26
48.71
p<0.001
1.30
1.21-1.40
Comfort of furnishing × Flexi desk
0.40
76.99
p<0.001
1.49
1.36-1.63
Storage space × Fixed desk
0.10
9.97
p<0.01
1.10
1.04-1.17
Storage space × Flexi desk
0.11
9.75
p<0.01
1.12
1.04-1.20
R2=0.21(Cox&Snell), 0.30(Nagelkerke). Model χ2(13)=945.09, p<0.001
The regression model also indicated that a decline in occupant satisfaction with ‘space
for breaks’ had a significant negative impact on self-reported health for both sample groups.
Nonetheless fixed-desk users were more likely to register negative ratings on ‘health’
(OR=1.32), compared to those in flexible workspaces (OR=1.20).
To summarise, ‘comfort of furnishing’ and ‘space for breaks’ were the two strongest
predictors of negative health reports at workspaces for both fixed- and flexi-desk groups.
‘Comfort of furnishing’ mattered more to those in flexi-desking environments, whereas
‘space for breaks’ was more influential for those working at a fixed-desk. Other variables
also indicated significant relationships with negative health influence but the strength of
impact was modest (OR=1.10~1.13) in comparison to ‘comfort of furnishing’ and ‘space for
breaks’.
The findings corroborate the problems of non-territorial workplace affecting occupant
productivity and health as identified in open-ended comments registered from a sample of
building occupants. Several of those issues raised earlier in the open-ended comments, such
as waste of time during setting up a workstation and unhygienic desk surface, weren’t able to
be dealt with in the regression analyses due to the absence of relevant questionnaire items in
the standard BOSSA survey. However, in general the quantitative analyses reported in Table
8 and 9 validate the major themes and patterns derived from flexi-desk users’ open-ended
comments (see Fig. 3).

22
4. Lessons learnt and suggestions for improvement
Theoretically, providing less space per person can translate into fewer buildings to
support the same number of people. Improvement of office space efficiency by adopting the
concept of non-territorial workplace concept, therefore will contribute to the reduction of
resources consumption associated with construction, maintenance and operation of office
buildings. The empirical evidence presented in this paper suggest that reducing space per
person enforced by non-territorial work policy doesn’t necessarily decrease occupant
workplace satisfaction (Fig.1), perceived productivity and health (Table 8 and 9). Our
findings suggest that other factors beyond the desk ownership play a more significant role in
the non-territorial work arrangement, affecting occupant attitude towards their building –
some positively and some others negatively.
As seen in Fig. 1, occupant satisfaction was higher with many IEQ issues in non-
territorial workspaces than in conventional fixed-desk layouts. The design of non-territorial
workspaces which allocate more space for meeting or break-out areas than the conventional
fixed-desk layouts seem to be contributing to occupants’ spatial comfort. The ability to
choose the workspace location in the non-territorial working arrangement seems to provide
occupants with a higher degree of opportunity to adapt to the local ambient conditions,
resulting in more relaxed attitude and responses of occupants towards the building’s indoor
air quality or thermal conditions.
While our emphasis in the paper is on the question of desk ownership, and the
difference between fixed address and flexi-desk users, shortage of available desks on busy
days in flexi desk offices was identified in our qualitative data as the most salient problem of
non-territorial workspaces. Notwithstanding the absence of clear data as to the ratio of people
to desks for all sample buildings in this research project, it is important to note that typically,
the ratio of fixed desk to employees is one desk per full time equivalent employee, ie, every
employee has a designated workstation when they are at work. In the case of flexi-desk, the
ratio of employees to desks is always greater than one. In addition to the fact that the flexi-
desk workspace is designed on the premise that 100% of the employees are not in a building
at all times, ancillary spaces such as break out areas libraries and meeting spaces are also
intended to serve as activity based workspaces. Anticipating the daily workstation demand
can be very difficult but as seen in our case study, demand in excess of available
accommodation can result in occupant dissatisfaction and loss of productivity. As employee
salaries account for the greatest expenses in an office building, the penalty cost resulting from
displaced employees is very likely to be more expensive than providing extra workstations.

23
Clearly, a more rigorous and realistic approach is needed to determine the optimum capacity
of workstations when employing the non-territorial workspace arrangement.
The non-territorial model also generated negative feedback regarding effectiveness of
collaboration when team members are scattered across the building making it difficult to
locate them. In our case study of non-territorial workplaces, occupants complained that more
distanced communication methods (electronic engagement) were not as effective as face-to-
face transactions and noted that immediate collaboration and ad-hoc conversations were
compromised as a result. Space allocations and spatial layouts of non-territorial workplaces
need to be carefully designed to reflect the range of organisational work processes – solo
work, team work, short-term project work, or inter-department work. Although it might
conflict with privacy, the use of tracking systems (perhaps based on smart phone
technologies) to locate colleagues in a large office building may be worth considering as a
means of mitigating the identified difficulties with interaction.
Providing easy-to-adjust chairs/desks and more thorough cleaning protocol, as a
consideration to those who will have to use different workstations every day, can be a
solution to decrease prevalence of occupants’ health complaints in non-territorial workplaces.
Our results indicated that the limited ability to personalise one’s work area can negatively
affect occupant perceived productivity and health in non-territorial workplaces.
Personalisation is a type of territorial behaviour, expressing ownership or defining boundaries
of one’s workspace typically by displaying items that indicate personal identities [37,38]. It is
known that personalisation can help to mitigate the adverse effect of low levels of privacy on
emotional exhaustion at workplace [24]. Although it is beyond the scope of the current study,
this conflicting matter – personalisation as a territorial behaviour in such depersonalised
settings of non-territorial workplace – deserves further investigation in future.
The open-ended comments registered from flexi-desk occupants indicated various
unwelcome side effects associated with the non-territorial working policy. On the other hand,
the perceived downside of requiring occupants to walk around in flexi-desk environments
could have unintended health benefits. In the conventional desk-based offices, occupants
often remain seated in the same spot for most of their working day, expending very little
energy [39], and past studies have demonstrated that excessive time spent sedentary in
workplaces can be linked to negative health outcomes including cardiovascular disease,
diabetes, obesity and mortality [40-42]. Consequently, public health initiatives [43] have
focused on reducing sitting time in office workers and the use of standing desk has typically
been a prescription to achieve this (e.g. [41,42]). The nature of non-territorial workplaces

24
requires their occupants to be more active and expend more energy inside their building: e.g.
walk around workspaces to find a desk, travel back and forth between personal locker and a
desk whenever starting/finishing work, move over to a project table or a quiet room to get a
specific task done, and walk up/down stairs to find and interact with a colleague. This implies
potential advantages of non-territorial workspace settings with respect to occupant health,
which haven’t been fully explored yet. Further observational research is necessary to better
understand occupant movement patterns and to quantify the amount of time spent non-
sedentary in non-territorial workplaces.
5. Conclusion
This paper examined the effects of non-territorial workplace environments on
occupant satisfaction, self-assessed productivity and health. Our findings suggest that
whether occupants have a pre-allocated desk or not doesn’t necessarily have adverse
influence on occupant evaluation of workspaces. The occupant survey results demonstrate
that it is possible to provide quality indoor workspace environments at high occupancy
densities through flexi-desk arrangements and that the ability to select their workstation
enables flexi-desk occupants to exercise a certain degree of personal control over indoor
environmental conditions. On the other hand, the open-ended comments from flexi-desk
users indicated several downsides of the non-territorial workplace policy. Issues such as
insufficient supply of desks, difficulties in co-locating colleagues, waste of time during
repeated process of finding, setting-up and clearing-up a workstation, and limited ability of
personalising a workstation were identified as the common complaints in our case studies.
The findings of the paper emphasise the importance of user responsive design for
non-territorial workspaces. Our analysis highlights that occupants in flexi-desk arrangement
are more likely to negatively evaluate their workplace productivity when satisfaction
decreases for aspects such as office layout, enabling interaction with colleagues, comfort of
office furniture, ability to adjust/personalise the workspace, and amount of space available for
work or storage. Likewise occupant perceived health will be adversely affected as
satisfaction with comfort of furniture or break-out spaces decreases particularly in non-
territorial workplaces.
Acknowledgement

25
The study was supported by Australian Research Council’s Linkage Project – BOSSA
(LP1102000328). Ethical approval for this study was granted by The University of Sydney
(2012/1786) and the University of Technology Sydney (2012-372R). Paula Strapasson, a research
intern at the IEQ Lab, The University of Sydney, is acknowledged for her effort in analysing spatial
characteristics of the sample buildings.
References
[1] Steiner, J. (2006). The art of space management: Planning flexible workspaces for people.
Journal of Facilities Management, 4(1), 6–22.
[2] Kuljanin, M. (2014). Occupier perspective global occupancy costs - offices 2014, DTZ.
[3] Daniels, S. (1994). The Hot Desk Shuffle. Work Study, 43(7), 7.
[4] Duffy, F. (1992). The changing workplace. (P. Hannay, Ed.). London: Phaidon.
[5] Hedge, A. (1982). The open-plan office: A systematic investigation of employee reactions
to their work environment. Environment and Behavior, 14(5), 519–542.
[6] BCO. (2013). Occupier density study, British Council for Offices.
[7] Brand, J. L., & Smith, T. J. (2005). Effects of reducing enclosure on perceptions of
occupancy quality, job satisfaction, andjob performance in open-plan offices. In
Human Factors and Ergonomics Society 49th Annual Meeting (pp. 818–822).
[8] Brennan, A., Chugh, J. S., & Kline, T. (2002). Traditional versus open office design: A
longitudinal field study. Environment and Behavior, 34(3), 279–299.
[9] Bodin Danielsson, C., & Bodin, L. (2009). Difference in satisfaction with office
environment among employees in different office types. Journal of Architectural and
Planning Research, 26(3), 241–257.
[10] De Croon, E. M., Sluiter, J. K., Kuijer, P. P. F. M., & Frings-Dresen, M. H. W. (2005).
The effect of office concepts on worker health and performance: A systematic review
of the literature. Ergonomics, 48(2), 119–134.
[11] Kaarlela-Tuomaala, A., Helenius, R., Keskinen, E., & Hongisto, V. (2009). Effects of
acoustic environment on work in private office rooms and open-plan offices -
Longitudinal study during relocation. Ergonomics, 52(11), 1423–1444.
[12] Kim, J., & de Dear, R. (2013). Workspace satisfaction: The privacy-communication
trade-off in open-plan offices. Journal of Environmental Psychology, 36(0), 18–26.
[13] Sundstrom, E., Herbert, R. K., & Brown, D. W. (1982). Privacy and communication in
an open-plan office: a case study. Environment and Behavior, 14(3), 379–392.
[14] Laing, A. (1990). Desk sharing; the politics of space. Facilities, 8(7), 12–19.

26
[15] BCO. (2010). Making flexible working work, British Council for Offices.
[16] Johnston, J. (2009). An investigation into the energy consumption of future office
buildings around the world. University of Strathclyde.
[17] Fawcett, W. (2009). Optimum capacity of shared accommodation: yield management
analysis. Facilities, 27(9/10), 339–356.
[18] NAO. (2006). Getting the best from public sector office accommodation, National Audit
Office, London.
[19] OGC. (2007). Efficiency standards for office space, Office of Government Commerce,
London.
[20] Gibson, V. (2003). Flexible working needs flexible space? Journal of Property
Investment & Finance, 21(1), 12–22.
[21] Elsbach, K. D. (2003). Relating Physical Environment to Self-Categorizations: Identity
Threat and Affirmation in a Non-Territorial Office Space. Administrative Science
Quarterly, 48(4), 622–654.
[22] Van der Voordt, T. J. M. (2004). Costs and benefits of flexible workspaces: work in
progress in The Netherlands. Facilities, 22(9/10), 240–246.
[23] Fawcett, W., & Rigby, D. (2009). The interaction of activity, space and cost variables in
office workstation sharing. Journal of Corporate Real Estate, 11(1), 38–51,65–66.
[24] Laurence, G. A., Fried, Y., & Slowik, L. H. (2013). “My space”: A moderated mediation
model of the effect of architectural and experienced privacy and workspace
personalization on emotional exhaustion at work. Journal of Environmental
Psychology, 36, 144–152.
[25] DEGW. (2008). Working beyond walls: the government workplace as an agent of
change. Office of Government Commerce, London.
[26] SHRM. (2013). SHRM Workplace forecast, The Society for Human Resources
Management, Alexandria, VA, USA.
[27] Brill, M., Margulis, S. T., Konar, E., & BOSTI. (1985). Using office design to increase
productivity. Buffalo, New York: Workplace Design and Productivity, Inc.
[28] Kats, G., Alevantis, L., Berman, A., Mills, E., & Perlman, J. (2003). The Costs and
Financial Benefits of Green Buildings: A report to California’s sustainable building
task force.
[29] NABERS. (2015). NABERS Indoor Environment for offices: rules for collecting and
using data V0.9, Office of Environment and Heritage, Sydney. Retrieved from
www.nabers.gov.au

27
[30] GBCA. (2015). Green Star Performance: summary of categories and credits, Green
Building Council of Australia, Sydney. Retrieved from www.gbca.org.au
[31] Candido, C., Kim, J., de Dear, R., & Thomas, L. (2015). BOSSA: a multidimensional
post-occupancy evaluation tool. Building Research & Information, 44(2), 214-228.
[32] PCA. (2012). A guide to office building quality, Property Council of Australia.
[33] Lee, S. Y., & Brand, J. L. (2005). Effects of control over office workspace on
perceptions of the work environment and work outcomes. Journal of Environmental
Psychology, 25(3), 323–333.
[34] Lefcourt, H. M. (1976). Locus of control!: current trends in theory and research.
Hillsdale, New Jersey: L. Erlbaum Associates.
[35] Rotter, J. B. (1966). Generalized expectancies for internal versus external control of
reinforcement. Psychological Monographs, 80(1), 1–28.
[36] Kim, J., & de Dear, R. (2012). Impact of different building ventilation modes on
occupant expectations of the main IEQ factors. Building and Environment, 57, 184–
193.
[37] Altman, I. (1975). The Environment and Social Behavior: Privacy, Personal Space,
Territory, and Crowding. Monterey, USA: Brooks/Cole Publishing.
[38] Brown, G., Lawrence, T. B., & Robinson, S. L. (2005). Territoriality in Organizations.
Academy of Management Review , 30 (3), 577–594.
[39] Matthews, C. E., Chen, K. Y., Freedson, P. S., Buchowski, M. S., Beech, B. M., Pate, R.
R., & Troiano, R. P. (2008). Amount of time spent in sedentary behaviors in the
United States, 2003-2004. American Journal of Epidemiology, 167(7), 875–881.
[40] Alkhajah, T. A., Reeves, M. M., Eakin, E. G., Winkler, E. A. H., Owen, N., & Healy, G.
N. (2012). Sit-stand workstations: a pilot intervention to reduce office sitting time.
American Journal of Preventive Medicine, 43(3), 298–303.
[41] Dunstan, D. W., Wiesner, G., Eakin, E. G., Neuhaus, M., Owen, N., LaMontagne, A. D.,
Moodie, M., Winkler, E., Fjeldsoe, B. S., Lawler, S., & Healy, G. N. (2013).
Reducing office workers’ sitting time: rationale and study design for the Stand Up
Victoria cluster randomized trial. BMC Public Health, 13(1), 1057.
[42] Gilson, N. D., Suppini, A., Ryde, G. C., Brown, H. E., & Brown, W. J. (2012). Does the
use of standing “hot” desks change sedentary work time in an open plan office?
Preventive Medicine, 54(1), 65–7.
[43] Neuhaus, M., Healy, G. N., Fjeldsoe, B. S., Lawler, S., Owen, N., Dunstan, D. W., …
Eakin, E. G. (2014). Iterative development of Stand Up Australia: A multi-component

28
intervention to reduce workplace sitting. International Journal of Behavioral
Nutrition and Physical Activity, 11(1).