Demand for Green Buildings: Office Tenants’ Stated Willingness-to-Pay for Green Features

Abstract

This study analyzes the demand for green office building features among US office tenants. An on-line survey of a random sample of office tenants in 17 major US markets is employed, with 708 responses representing a 23% response rate. Tenants provided their perspective on green buildings and their stated willingness to pay for individual green features. Office tenants have the highest willingness to pay for improved indoor air quality and access to natural light. The results show that public firms, along with those in the Energy and Information Technology industries are most likely to pay for green labeled buildings. Regional and demographic preferences are shown in both willingness to pay and attribute ranking. The findings provide implications for policy-makers and real estate property developers in terms of which green building features are considered to be most important for green building practices, and how demand for green features potentially differ across regions.

Full text

0
Demand for Green Buildings: Office Tenants’
Stated Willingness-to-Pay for Green Features
Spenser Robinson, D.B.A (robin6s@cmich.edu)
Central Michigan University
Robert Simons, Ph.D. (r.simons@csuohio.edu, contact author)
Cleveland State University
Eunkyu Lee (e.lee16@csuohio.edu)
Cleveland State University
Andrew Kern (kern1a@cmich.edu)
Central Michigan University
Final version forthcoming in Journal of Real Estate Research

1
Demand for Green Buildings: Office Tenants’
Stated Willingness-to-Pay for Green Features
ABSTRACT
This study analyzes the demand for green office building features among US office tenants.
An on-line survey of a random sample of office tenants in 17 major US markets is employed,
with 708 responses representing a 23% response rate. Tenants provided their perspective on
green buildings and their stated willingness to pay for individual green features. Office tenants
have the highest willingness to pay for improved indoor air quality and access to natural light.
The results show that public firms, along with those in the Energy and Information Technology
industries are most likely to pay for green labeled buildings. Regional and demographic
preferences are shown in both willingness to pay and attribute ranking. The findings provide
implications for policy-makers and real estate property developers in terms of which green
building features are considered to be most important for green building practices, and how
demand for green features potentially differ across regions.
Key words: office buildings, green, tenant , survey, WTP

2
Demand for Green Buildings: Office Tenants’
Stated Willingness-to-Pay for Green Features
INTRODUCTION
The importance of green buildings and the potential for revenue premiums in Leadership
in Energy and Environmental Design (LEED) and Energy Star buildings is well established and
several theories for those premiums have been put forward (Eichholtz, Kok, and Quigley, 2010,
Fuerst and McAllister, 2011, and Das and Wiley, 2013). However, whether tenants pay for
specific building-level amenities, such as access to natural light, efficient lighting systems, and
proximity to public transportation, or simply for the branding effect of LEED or Energy Star is
relatively unexplored.
This study attempts to fill that gap in the literature by “unpacking” the green building
premium into its component parts and examining tenant demand for 18 specific green building
features. The core questions addressed are which specific green building attributes tenants most
value and demonstrate preferences. A random sample of over 3,000 leases in 329 U.S. office
buildings in 17 geographically diverse MSAs
1
is sampled in an online survey. Survey design stems
from seven focus groups in four geographically representative US cities the results of which are
detailed in Simons, Robinson, and Lee, 2014.
Tenants are asked to provide their perspective on green buildings, preferences for green
building features, and their willingness to pay for those features. The use of contingent valuation
to estimate market preferences is well supported (Carson, 2012) and further detailed in the
literature review. While all survey data has inherent limitations, the results of this survey provide
1
The authors gratefully acknowledge support and funding by CBRE, Inc. as part of their Real Green Research
Challenge. All buildings are institutionally managed by CBRE, but held by a diverse group of owners.

3
the most current information with respect to tenant preferences. The prior literature on revealed
preferences uses primarily building average weighted rent which may include leases originated 3-
10 years prior.
This research introduces the initial findings of the survey, providing implications for real
estate building designers and developers. The revealed attribute level preferences may improve
decision-making for practitioners and provide additional research opportunities for academics.
LITERATURE REVIEW
A growing number of studies suggest that green buildings provide economic,
environmental, and social benefits through several mechanisms including lower operating costs,
employees’ improved productivity, tax credits, and buildings’ positive images (Fuerst and
McAllister, 2011). Kok, Miller, and Morris (2012) find that LEED-certified buildings have a 7.1%
rental premium compared to non-certified buildings. Additionally, their study shows that buildings
yield a highyer rental premium with both LEED and Energy Star certifications. Fuerst and
McAllister (2011) suggest a rental premium of 5% for LEED-certified buildings and 4% for
Energy Star buildings; they also find a sale price premium of 25% for LEED buildings and 26%
for Energy Star buildings with higher levels of certification providing a higher premium.
Reichardt, Fuerst, Rottke, and Zietz (2012) track a rental premium for both Energy Star
and LEED-labeled buildings from 2000 to 2010. The study shows that a significant rental premium
for both voluntary green certified buildings has increased steadily from 2006 to 2008, followed by
a slight decline after 2008 due to the “great recession” economic crisis. Empirical studies showing
the differing premiums over time and across different size and regions include Das and Wiley
(2013) and Robinson and McAllister (2015). To account for potential regional economic
differences, this research follows those economic regions established by Crone (2005), in his

4
configuration of economic regions that share similar industrial and social traits. This is selected
over U.S. census regions, and other schemes (e.g., Malizia and Simons 1991 who use the Salomon
Brothers configuration) because it is more recent and presumably a better measure of current
economic relationships.
Many commercial building tenants also consider buildings as a space to publicize their
environmentally-friendly visions, consistent with a green or sustainable corporate policy.
Corporate social responsibility (CSR) policies can incentivize tenants to extract social benefits
from green buildings, beyond the tenants’ direct corporate profit (Eichholtz, Kok, and Quigley,
2009). Recent research has shown that consumer sentiment, tenant preference in the context of
the subject research, can impact real estate pricing (Marcato and Nanda, 2015)
Lease structure can also affect who (owners vs. tenants) benefits from green building utility
expenses (Jain and Robinson, 2015). In a Triple Net (NNN) lease structure, where tenants pay all
utilities on top of base rent, they may be motivated to pay an increased rent to earn a benefit for
certain cost-saving features. On the other hand, building owners, who pay all utility costs in a Full
Service Gross (FSG) lease, benefit more directly from those cost-saving features, questioning
whether tenants would be willing to pay for them in this structure. The topic of lease premiums
has received increase attention in commercial real estate (Liu and Liu, 2013).
In addition to the empirical studies on the green premium, several studies examine the
effect of public policies on the market penetration of green buildings. Simons, Choi, and Simons
(2009) demonstrate that public policies affect the green building market in different ways. For
instance, their study finds that executive orders are a quicker method for encouraging green
buildings, while legislation is more related to politics.

5
As summarized above, interest in green buildings and their specific features, development
of new space, retrofitting existing space, and return on investment in green features is growing
(Aroul and Hansz 2012, p. 27). But how should developers proceed? Sanderford and Robinson
(2015) find little statistical evidence that building attributes are good predictors of green building
certification which leads to the question of what attributes are valued?
This study uses a method of rental premium estimation widely called contingent valuation which
has been generally accepted for over 30 years. As far back as 1982, Brookshire, Thayer, Schulze
and d’Arge (1982) confirmed “the validity of survey methods as a means of determining a public
good,” through their analysis of real estate values. Contingent valuation is a well-known approach
to measure a consumer’s stated-preference for a good, service, or policy (Carson and Hanemann,
2005). CV estimates individual’s stated Willingness to Pay (WTP) for a subject or policy, such as
a change in environmental amenities, using survey questions that elicit information on how much
each sampled individual would be willing to pay to have the subject or policy implemented
(William, Morey and Lodder 1998, p.715). As the name of CV indicates, a survey research
measures the contingent values revealed by respondents upon hypothetical or constructed projects
or programs (Portney 1994, p. 3). A diverse group of economics and real estate journals have
publications using stated WTP and/or contingent valuation (Simons, 2002). Stigka, Paravantis,
and Mihalakakou (2014) use WTP to evaluate sustainable energy pricing. Simons and Winson-
Geideman (2005) also use CV in estimating real estate values. Lu, Peng, Webster, and Zuo (2015)
use survey based CV to estimate willingness to pay for waste disposal mechanisms. Although
important in determining the value of amenities, some incongruity may occur between the stated
WTP and actual WTP (Lindsey and Knaap, 1999), thus caution should be used in interpreting the
subject results.

6
This study examines a key research gap using CV: what specific building-level amenities,
including location-related factors, drive the rental and sales premiums? Do tenants pay simply for
green branding, and/or to achieve a CSR benefit, or are specific attributes desirable to tenants?
What are they willing to pay for various specific green features? Furthermore, if specific green
attributes are desirable, could developers/owners maximize the value of their own holdings by
adding these features without necessarily achieving a LEED certification? This study fills the
research gap using an on-line tenant survey, and reports the results below.
DATA GATHERING PROCEDURES
Survey Process
Data are collected using a web-based survey distributed to tenants occupying space in over
3,000 leases in 329 U.S. office buildings in 17 geographically diverse MSAs. All properties are
managed by CB Richard Ellis (CBRE) during summer of 2014, but are owned by a diverse set of
institutional owners, each with their own management profile and preferences. Other than through
management, the tenants have no relationship with CBRE. The survey instrument is based on
focus groups conducted as part of the Real Green Research Challenge (Simons, Robinson and Lee,
2014). The focus groups explored which specific green features are valued by real estate market
participants, and how to best collect data on preferences. Seven focus groups, with a total of 49
participants, were conducted in four major metropolitan areas (Chicago, Washington D.C.,
Denver, and the San Francisco Bay area). Participants included building managers, tenant
representatives, project managers, researchers, and architects/engineers. The focus groups for each
area are identical in content, but followed one of two formats: in person or remote webinar-based.

7
Prior research has shown that virtual focus groups are qualitatively similar to their in person
counterparts (Reid & Reid, 2005).
The final survey underwent rigorous pre-testing to account for completion time (target of
15-20 minutes), item clarity, functionality, and other issues commonly found in web-based
surveying. Respondents receive the survey invitation via a preliminary email from their CBRE
building manager. The invitation to participate came a day later, and contained a greeting, general
introduction to the research study, including the opportunity to win one of two iPad® Air devices,
and a survey link. The survey itself includes an informed consent page (required by the
universities’ IRB), where respondents are assured that their responses are confidential. The
surveys are collected over a four month period, and respondents are reminded up to four times
about the opportunity to participate. Overall, 3,015 tenants are invited to participate, and 708
provide complete responses, for a response rate of 23%.
Data collected
Each respondent provides general information about their organization and themselves.
Background information is acquired on the tenant company’s total space within the building,
primary industry (e.g. Construction, Food Services, Education, Advertising, etc.), primary function
(e.g. Executive/Administrative, Manufacturing, Sales, etc.), and number of employees at that
location. Individual demographics include position within the company, years with the company,
education level, age, primary mode of transportation to work, self-reported knowledge of green
building characteristics (categorized low, medium, high), and gender.

8
Respondents are asked about the driving forces behind the company’s decision to locate in
its current building by ranking the top three of 13 attributes, including green building features,
prices, proximity to public transportation, LEED designation, and location.
Six items assess respondents’ overall attitudes towards general green initiatives. These
items are endorsed on a 5-point Likert scale from 1 (Strongly Disagree) to 5 (Strongly Agree).
Sample items are: “I feel like green buildings can comfortably accommodate more people in the
same space than a traditional building;” and “An Energy Star building is more valuable than one
without an Energy Star Certification.” Three additional items pertain to the tenant’s sustainability
initiatives. For example, the first item is: “Have sustainability initiatives, other than building-
related, been discussed at a company meeting you have attended in the last six months?” Two
items are asked regarding the company’s LEED and Energy Star ratings. Both items simply ask
if the respondent knows the company’s ranking/score respectively, and if so, to include it.
Willingness to pay questions for specific green office building features includes two ways
of estimating the economic value of these features. The first is an ordinal ranking and prioritization
of each green item among 18 green features. The second is a direct willingness to pay approach, a
revealed preference technique best viewed in the context of contingent valuation (CV) discussed
in the literature review section. Survey data provides the most cutting edge information in what
real estate executives in related interviews call a fast moving set of preferences. Building average
rent, or individual lease rent, may reflect old information.
Based on this approach, the research team developed a series of questions to measure the
office building tenant’s stated WTP for each green building attribute. Respondents are again
presented with only the list of nine attributes (out of 18) they ranked as being most important from
earlier in the survey and asked to place a percentage value on each item, in the context of their

9
actual current rent and lease structure. The root question is: “How much more, in percentage
terms, would you pay for each attribute?” A sample question is:
If you were comparing a building that has the feature listed below to a building that
does not have the feature listed below, how much more, if any, do you feel your
company would pay for that attribute?(percentage of the total rental price)
Less than -1%, -1%, 0%, 1%, 2%, 3%, 4%, or More than 4%
Profile of Respondents
Exhibits 1 and 2 show a breakdown by region and job category of the 708 respondents
respectively. The sample is from throughout the U.S. Closely proportionate to the number of
buildings and tenants, almost half (45%) of the respondents are from the Far West region, followed
by Energy Belt (29%), Great Lakes (14%), Southeast (9%). The Mideast region is under-
represented (3%). No buildings from New England are available. Response rates by region are
generally about 25%, except for the lower rate in the Mideast (11%).
About 15 percent of respondents are Presidents, Vice Presidents, Chief Executive Officer,
Chief Financial Officer, or Chief Operating Officer (i.e., leadership). Office managers comprise
just over half the responses, and dominate the non-leadership group. Numerous statistical and
qualitative measures indicate leaders and office managers have similar opinions about green
features (see Survey Validity section).
[Insert Exhibit 1 here]
[Insert Exhibit 2 here]
The median number of employees per each tenant company is 17, thus the average of 59
employees is inflated by several larger tenants. A total of 58% of respondents have a Bachelor’s
or higher degree, and 60% have worked for the current company more than six years. More

10
specifically, 35% of respondents have more than 10 years of work experience in the current
company. The age of survey respondents is evenly distributed among different age groups, but
female respondents account for a relatively large portion (69%).
A total of 17% of the total participants (118 companies) belong to the finance and insurance
industry sectors and 11% (78 companies) belong to the legal service sector. In terms of lease
structure, a Full Service Gross lease accounts for 53% (370 tenant companies), a Triple Net lease
accounts for 33% (234 companies) with the balance being modified gross leases.
Survey Validity
The survey performed is the largest and most comprehensive survey to date on tenant stated
willingness to pay for commercial real estate building features. It involves a considerable national
sample and is developed through a series of industry focus groups. The survey is clearly
representative of an institutionally managed portfolio, which should presumably contain more
sophisticated than average tenants.
Briefly, this section reviews overall sample validity relative to the population and provides
a comparison of reported results of leadership to non-leadership (primarily office managers), and
then briefly address construct validity.
No statistical differences between respondents and population are found in base rent,
geography (based on tenant data provided in Exhibit 1), and external factors (such as outdoor air
quality). A statistical difference in mean tenant size is found due to the higher number of small
tenants in the sample. However, the smaller tenants’ overall contribution to building rent is
minimal. Among more corporate-oriented tenants with a standard space of 5,000 SF or above, no
statistical difference between sample and response is found. Another potential bias is that tenants

11
with more sustainability concerns could be more likely to answer the survey. Several tests
including comparing the accuracy of answers between LEED, Energy Star and non-certified
buildings suggest this is not a concern. Further, tenants are incented by the iPad to participate.
The results provide confidence that the sample is representative of the population of tenants.
Another validity concern is that a relatively small portion (about 20%) of the respondents
are “leaders” and that the aggregated results based largely on office manager responses may not
reflect decision maker opinions. Foundational focus groups informed the research team that the
primary respondents would be office managers; senior real estate executives uniformly indicated
that office managers are usually involved in the decision-making process and would “have the
pulse of the boots on the ground.” During that the focus group process, participants, including
executives, leasing brokers, property managers, and office managers all indicated that “office
managers could effectively represent tenant decisions makers” (Simons, Robinson, and Lee 2014).
They stated that for the subject sample, largely small to mid-size firms, the office manager are
often both aware of and involved in lease negotiations. Further, the office managers regularly hear
tenant complaints, and may be more in touch with what tenants want and don’t want, than upper
management, who may be more externally focused.
Nevertheless, responses of leaders and office managers are tested on several key questions
yielding almost identical results. Differences of means tests find few differences between
leadership and office managers responses when comparing their knowledge of the presence of
green attributes, stated willingness to pay (WTP), and the overall ranking of the attributes. A
separate leader dummy variable is included in all probit regression runs. In the multivariate
regressions the leadership group is not statistically different than the general survey in any of the
models. Furthermore, in individual tests of each attribute, the leadership group is found

12
insignificant nearly 90% of the time. In the multivariate regressions the leadership group is not
statistically different than the general survey in any of the four models. Furthermore, in individual
tests of each attribute, the leadership group is found insignificant nearly 90% of the time
2
.
To test for construct validity a series of internal consistency checks including removing
potential outliers and comparing rankings versus WTP estimations are conducted all suggesting
internal consistency. Survey design also maximized clarity with variable definitions, where
plausible. In review of the 18 items, most are clear. They include Bike Racks, Electric Car
Charging Stations, Energy Star Designation, Fitness Facility, Green Cleaning Products, LEED
Designation, Energy Star designation, Recycling Provided, and Shower On site. Further, in the
actual survey, the bounds of efficiency for WTP are defined for Efficient HVAC, Efficient
Lighting, and Water Conservation by asking the WTP proposition in the specific context of a 2.0%
savings on rent, couched in the specific tenants’ rent structure and annual costs. Indoor Air
Quality, Lease structure and Public Transit, definitions are provided to respondents, like “better
than the safe, breathable air required by standard building codes”, “Lease structure that financially
rewards tenant conservation of resources” and “within 5 min walk”, respectively.
The remaining three variables may be somewhat open to interpretation, including Access
to natural light, comfortable temperature control system, and walkability. However, in the building
manager survey (in Exhibit 3 shown later), specific definitions, like “floor to ceiling windows”
“temperature zones of 500SF or less” are used. For walkability, many now access websites like
www.Redfin.com and apply their score, which is becoming an industry standard for this term.
2
A number of chi-square and other tests are performed to thoroughly test for survey bias. This issue is a
critical issue for survey validity and reviewers are provided with extensive support. Further information is available
upon request.

13
Numerous tests along a variety of axes suggest that the sample reasonably represents the
population, that responses are consistent with tenant decision makers’ opinions and that
respondents’ definitions of survey terminology are consistent.
DESCRIPTIVE SURVEY OUTCOMES
Tenants’ perceived vs actual knowledge of green attributes
Tenants are asked to identify if an attribute is present in their building with Yes, No, and I
Don’t Know options provided. To corroborate tenants’ perceptions an external verification on the
accuracy of their overall knowledge of the attributes is conducted. A separate data gathering tool
is administered for CBRE managers, who are asked to identify presence or absence of the 18 green
features. Data are collected on the building level and matched to 547 tenant records (77% of the
sample).
Exhibit 3 shows first the percentage of respondents stating “I Don’t Know” for an attribute
and then providing information on whether a “Yes” or “No” answer is correct, compared to the
on-site building manager response.
[Insert Exhibit 3 here]
Interestingly, respondents know, with considerable accuracy, whether almost all the
building level attributes are present at better than 50%, except for green cleaning. However, they
do not know whether the green labels, LEED or Energy Star are present at a nearly 2/3 rate. This
substantiates the assertion that tenants care more about building level attributes than somewhat
opaque and less observable labels. Virtually all tenants, 84%, who answered whether they are in
a LEED building correctly identified their building status. However, only 30% of those who
answered they are in an Energy Star building are correct. This may have something to do with

14
annual recertification in the Energy Star program and the timing of their lease, whereas LEED is
a 5 year or permanent designation depending on certification type.
Diving further into the brand awareness, only 29% of respondents agree or strongly agree
that a LEED Platinum is more valuable than a simple LEED certified status. Although those in
LEED buildings often know they are, fully 84% could not identify their LEED level (Platinum,
Gold, Silver, and Certified). For Energy Star buildings 58% agree or strongly agree that they are
more valuable than non-certified buildings. The results suggest that a majority of tenants believe
an Energy Star certification adds value, and while LEED has value, the different gradations of
LEED have little market clarity.
Also, proximity to transit and walking access to services are somewhat subject to
interpretation, and the actual building measures used may have been stricter than the tenant
interpretation. The overall accuracy rate of 69% is also surprising, which may show that overall
green awareness is not that high.
The relatively high levels of “Don’t Know” responses may be somewhat surprising at first.
But they are corroborated by the opinions of high level executives in large institutional real estate
firms (e.g. REITs, Pension Funds, Opportunity Funds, etc.). In another related effort, the research
team interviewed these executives and specifically asked if they believe tenants “care” about green
building features; executives answered that the majority of tenants do not, with the exception of
large public firms and the federal government.
Finally, OLS models isolating the stated WTP for a green attribute with the “yes” and
“don’t know” responses are examined against a reference category of “not present” (omitted due
to space constraints). The majority of categories show no statistical differences. However,
respondents currently in buildings with high efficiency HVAC or efficient lighting are more likely

15
to pay for maintaining those attributes. Also, as a demonstration of the positive impact of LEED
and Energy Star, respondents already in buildings with those designations are more likely to pay
a premium for them
3
. Those who are unaware of whether they had improved IAQ, a fitness facility
or a shower are all less likely to want to pay for that amenity. Public transit users and those with
high walkability metrics are more likely to pay a premium for continued access to those amenities.
Those with strong natural light are also more likely pay for it in their office space.
Difference between green attributes considered most important and those currently available
As described earlier the survey asked respondents to value attributes in a number of ways.
First, respondents are asked to delineate between the 9 “Most Important” and 9 “Less Important”
attributes. Then, the respondents are asked to rank the relative importance of those nine top-ranked
attributes from 1 to 9
4
. To consider both the selection of an attribute and its ranking, a weighted
score reflecting the relative ranking of the attributes is generated; a ranking of 1 is given the most
weight down to a zero score for unranked by that respondent. An attribute that is top-ranked by
all respondents would receive a score of 100.
By far, the attribute receiving the highest percentage of most important rankings is
improved indoor air quality (IAQ) with 93%. Access to natural light is the next most valued
attribute by respondents. Interestingly, some of the variables most frequently selected by
respondents, such as recycling, show a slightly lower weight. It appears that although many people
want or expect recycling at the building, it is not as important as other features. Exhibit 4 shows
the list of 18 green building features and their relative importance to tenants. Energy Star and
LEED ranking, independent of any underlying features, are in the middle of the pack.
3
Another potential explanation is self-selection bias of those who already prefer the labels.
4
Although a few selected more, only their first nine is used for most of the analyses.

16
[Insert Exhibit 4 here]
On a ranking basis, improved indoor air quality, access to natural light, and on-site
recycling are the highest. However, the information in Exhibit 4 does not show the presence of
each green building feature in the tenants’ building and rented space. To examine this, the survey
asks if the features are currently available in the building. Then, a ratio of desired and current
availability for each green feature is created shown in Exhibit 5.
Exhibit 5 reveals that tenants perceive a greater need for improved indoor air quality, lease
structures that financially reward tenant conservation of resources, efficient electricity and gas use
for heating and cooling, energy efficient lighting, green cleaning product, and localized
temperature control, than are currently supplied by the market. Presumably, installation of these
features should be recognized in willingness-to-pay higher rents. Also, there are several green
building attributes where current availability exceeds stated demand. Those attributes include bike
racks and public transportation nearby. Exhibit 5b shows the data from Exhibit 5 in a bar chart.
[Insert Exhibit 5 here]
[Insert Exhibit 5b here]
In addition, the authors ran a series of regressions comparing the actual presence of an
attribute to its perceived presence and the effect on WTP (results omitted to conserve space). In
nearly all cases perceived presence of an attribute is highly correlated to its actual presence. Some
that are not as highly correlated are less observable such as Energy Star or rare like electric car
charging stations. In virtually all cases WTP is statistically unrelated to the presence of an attribute.
Thus respondents appear to have independently and reliably estimated the economic impact of an
attribute, per the survey design, regardless of its presence.
Stated Willingness to Pay for Green Building Features

17
Building tenants are asked to state their WTP for each green building attribute in terms of
a percentage over their existing lease. Exhibit 6 shows the results. Office tenants show the highest
willingness to pay for access to natural light (additional 1.3% over their current rental rate),
improved indoor air quality, individualized temperature controls, and lease structure that
incentivized energy savings (all over 1.2%). Furthermore, over 30% of those with positive
responses for IAQ and natural light stated a willingness to pay more than a 2% premium for those
features. These findings are consistent with the ranking portion of the survey.
Interestingly, the extent to which respondents are willing to pay more for individual green
building features generally corresponds to the stated demand (Exhibit 5), except for recycling and
green cleaning products. This result reveals that office tenants consider both recycling and green
cleaning products to be important green building attributes, but are not willing to incur additional
costs for having those features. As further validity support, a side-by-side comparison of the WTP
of the overall sample (N=708) with the leadership group (N=120) shows nearly all attributes are
similarly ranked, with only one attribute (Natural light) statistically different. Finally, the
aggregate premium (the sum on Exhibit 6) for these 18 attributes is 9.3%
5
.
[Insert Exhibit 6 here]
To better explore the impact of lease structure on green attribute, three questions (efficient
heating/cooling systems, efficient lighting, and water conservation measures, noted with an * on
Exhibit 6) remind the respondents that either they (NNN) or the building owner (FSG) would
receive the benefit of 2% savings in utility costs. In an efficient market, a NNN lease holder should
accept a 2% rental increase offset by a 2% expense decrease. Although the results do show a clear
5
As another internal consistency check, the average attribute rankings and the average WTP are compared; in other
words the highest ranked attribute should generally be the highest mean WTP, which it is through the fourth ranked
attribute.

18
difference in WTP for the two lease structures for efficient light and gas, the difference (about
0.4%) is smaller than expected. Under a NNN lease, the respondents state that they are willing to
pay additional 1.24% of the current rental rate for efficient electricity and gas use building features,
while they reveal only 0.88% of willingness to pay with a FSG lease. The willingness to pay for
energy efficient lighting also shows the same pattern as with efficient HVAC systems. Thus, the
outcomes reveal that there is a difference between NNN and FSG leases consistent with theory,
but that difference is smaller than expected. One possible explanation is that, despite repeated
efforts to reinforce the idea, the respondents do not fully appreciate the difference between the
lease types. An alternative explanation could be that other factors beyond the tenants’ bottom line
are in play. Exhibit 7 shows these differences by lease structure:
[Insert Exhibit 7 here]
Surprisingly, even with stated savings, respondents are reluctant to pay for water
conservation. Regionally, the results show that tenants from the Great Lakes region had the lowest
average WTP for water conservation feature (0.33%), with the Far West (0.42%) and East regions
(0.47%) noticeably higher. It may simply be that the stated savings of 2.0% on rent are not readily
achievable from water savings.
MULTIVARIATE REGRESSION
To better answer the core questions of who is interested in green buildings and what
attributes significantly affect stated WTP, a series of regressions is developed. The first model
expands on the well-documented existence of green building premiums by exploring what
company, personal and regional variables have a higher willingness to pay for green-labeled
buildings such as LEED and Energy Star. The second set of models individually regresses each
of the 18 green building attributes on whether the respondent is willing to pay a 2.0% or greater

19
premium (high WTP) for the attribute (from Exhibit 6). The third set of models replicates the
second set with each of the 18 attributes, but with a dependent variable based on rank (from Exhibit
4) rather than WTP. Each of the preceding model structures have dummy dependent variables,
and utilize a probit model format. The models are structured thus:
Probit Regression Model 1: (eq.1)
WTPgreenlabel_i = β0i + β1Regioni + β2Indusryi + β3Size + β4Companyi + β5Leasei + β6Demograpicsi+

i
Probit Regression Model 2: (eq.2)
WTPattribute_ij = β0i + β1Regioni + β2Indusryi + β3 Size + β4Companyi + β5Leasei + β6Demograpicsi +

i
Probit Regression Model 3: (eq.3)
Rankij = β0i + β1Regioni + β2Indusryi + β3Size + β4Companyi + β5Leasei + β6Floor-plani +

i
The dependent variables are:
WTPgreenlabel_i = a binary variable with a value of one if a tenant i has a positive willingness to pay
for green buildings (LEED and/or Energy Star) and zero otherwise;
WTPattribute_ij = a binary variable with a value of one if a tenant i has 2% or more of willingness to
pay for attribute j and zero otherwise; this regression is run separately for each of the 18
green building features.
Rankij = a dummy variable with a value of two if tenant i ranked attribute j as one of the their top
three most important attributes, a value of one of it is ranked fourth through ninth, and zero if it is
in the “less important” half; this regression is run separately for each the 18 green building
features.

20
The independent variables are:
Regioni = a vector of regional categories including Far West, Great Lakes, Energy Belt, Southeast
and Mideast;
Industryi = a vector of industrial sectors including the finance, insurance, real estate, legal, IT and
computer, and other sectors;
Officei = a vector of building or lease variables including tenant square footage;
Companyi = a vector of a company’s characteristics including whether the company’s stock is
publicly held;
Leasei = a vector of lease structures including a Full Service Gross and Triple Net lease;
Demographicsi = a vector of a tenant’s personal information including age, gender, education,
and position in the current company;
Floor-plani = considers if the tenant’s company uses traditional, flex and/or hybrid floor plans;
Where β0 is a constant; and

is the error term.
Exhibit 8 presents the descriptive statistics of the variables used in the models. The typical tenant
occupies 17,600 square feet of space with 59 employees and pays $26.31 per SF/year in rent with
a FSG lease structure. 45% are located in the far west region and 16% are in the finance and
insurance sector. Almost half of the tenants are in professional services. Over 2/3 have a traditional
space layout and 44% of the respondents are 50-59 years old. Office managers and females
comprise over half the sample. LEED or Energy star status reflects about 30% each (some with
both designations).
[Insert Exhibit 8 here]
Multivariate model results

21
The results of the models are presented in Exhibits 9, 10, and 11.
Model 1 (Exhibit 9) addresses WTP for a Green building brand. Results show that out of
34 independent variables, only six are statistically significant. Energy and Information Technology
industries, along with publically traded firms are more likely to state a WTP for green building
labels. The industry findings are largely consistent with prior literature that found energy resource
extraction firms amongst the highest users of green buildings; in this case 2.34 times more likely
(e0.84), statistically significant at a level of 95%. This is often supported by their public corporate
social responsibility (CSR) statements. The Computer-IT sector also is significant (1.58 times
more likely, significant at a 95% level) Not surprisingly, public companies, most of whom will
have published CSR policies are more likely to pay for green-labeled buildings (1.26 times more
likely, significant at a 90% level). Those companies demonstrating some other commitment to
sustainability through their supplier choice (1.43 times more likely, significant at >99%) or a
LEED commercial interior (LEED CI) designation on their internal space also show a significantly
higher stated WTP.
Importantly, a dummy variable for leadership is not statistically different from the
remaining respondents, lending validity that survey outcomes are reflective of decision makers.
Somewhat surprisingly, college educated professionals, those achieving a Bachelor’s degree or
higher are less willing to pay (1.25 times) for green-labeled buildings. The authors consider this
a potential sign that specific building attributes may be more valuable to college educated
professionals than the aggregate baskets of attributes offered by LEED or Energy Star as a brand.
The probit model successfully converges, rejects the global null of zero beta at a 95% significance
level and reports an Aikake Information Criteria (AIC) score of 761.
[Insert Exhibit 9 here]

22
Exhibit 10 presents the results from the series probit regressions run as Model(s) 2. 13
attributes with a statistically reliable N of more than 40 positive WTP > 2.0% responses are shown
6
.
Due to space constraints, the authors do not to show detailed results for all probit regressions on
each individual green attribute, but their underlying format is the same as for Exhibit 11. Shown
in the table are only the variables that are statistically significant at the 90% level of confidence or
better. Exhibit 10 is sorted by the Attribute N column which represents the number of respondents
who indicated a stated WTP of 2.0% or greater. Thus, 250 respondents indicate a WTP of 2.0%
or more for improved indoor air quality.
IAQ results show that the Mideast region and publically traded companies are less likely
to indicate a WTP for IAQ. This could be more indicative of generally higher air quality in the
somewhat smaller Mideast cities relative to the Far West (e.g. Los Angeles) and Mideast (e.g. New
York and D.C.) counterparts.
Finance and legal industries and publically traded companies are less likely to pay for more
natural light. This may relate to the traditional outside window offices and indoor cubicle layout
of these more traditional industries. College educated respondents are more likely to value natural
light.
Individualized temperature control is less likely to be favored in the Mideast, while the
Southeast is more likely to value it. Presumably, the relative ease of distributing heat versus air
conditioning may be a factor in that distinction.
For the efficient lighting feature, only the energy and real estate industries appear as
significant in their WTP for efficient lighting, and both are strongly positive in their stated WTP.
6
Those omitted attributes include green roof, electric car charging, green cleaning products, bike racks, and shower
on site. These are shown on Exhibit 6, center left column, with N<41.

23
Energy may be motivated by a focus on energy usage and product sales while real estate operators
may more often benefit from savings.
High efficiency HVAC is favored by the Southeast region (high air conditioning costs),
real estate and government tenants, and by NNN lease holders. This shows an awareness of cost
savings associated with the NNN lease.
Without detailing each of the remaining variables, other notable results are that NNN lease
holders are more likely to value a lease that rewarded them for conservation. Energy belt region
and energy industry value recycling while the leadership position is less likely to pay for it
7
.
Publicly traded companies have a negative WTP for both public transit and walkability. Each of
the displayed models successfully converges, although with mixed model fit measures. Several of
the models fail to reject the null of a global zero beta, suggesting that the nuanced differences
within demographic strata may be small. That said, all models successfully converge and the
significant variables do indicate some propensity towards WTP for the attribute. Future research
could refine and aggregate the variables into factors or like bundles.
[Insert Exhibit 10 here]
Model 3 results are shown in Exhibit 11 representing results from a series of 18 probit
regressions estimating the likelihood that an attribute is ranked as high (2) or ranked (1) relative
to unranked (0) as described in Eq. 3. The higher the coefficient, the more likely that attribute is
to be ranked as first through third. Similar to above, the table is sorted by the attribute with the
most top three rankings. Also as above, space is too restricted to show full results from all 18
probit regressions, so only statistically significant variables are shown. However, all 18
attributes present enough N for statistically reliable results. The results shown in Exhibit 11 run
7
Note that this is the only model where leadership appears as significantly different in WTP from the rest of the
survey in this analysis.

24
parallel to those presented in Exhibit 10, but examine ordinal ranking as the dependent variable
rather than WTP.
Publically traded companies are less likely to rank natural light highly, while college
educated professional are more likely. The Southeast region, similar to the WTP results, and
also the Lakes region are more likely to rank temperature control. Leadership is also more likely
to rank temperature control highly. The results suggest that while the Great Lakes region and
leadership value individualized temperature control, they may not be willing to pay significantly
more for it.
Other notable results include the Energy Belt and Southeast regions’ relatively lower
value of recycling to other attributes. The legal, energy, government and information technology
fields are all more likely to value tenant reward structures in a lease. A NNN lease is less likely
to value an Energy Star certification, which at first seems somewhat counter to their stated WTP
for high efficiency HVAC, but may indicate a lower value of the designation itself to
respondents. Only the real estate, energy and IT industries show a higher propensity to value
LEED. The Energy belt, Southeast and Great Lakes regions are all less likely to value water
conservation.
The overall results of this series of regressions are generally consistent with those
presented in Exhibit 10, show definite regional preferences and indicate that attribute level
preferences are not homogenous across the United States. Any system that attempts to value
green building attributes should incorporate some level of regional preference. As above, each of
the displayed models successfully converged, although with mixed model fit measures. In this
case the majority successfully rejected the global beta null at traditional significance level, but
some attributes such as individualized temperature control and efficient HVAC failed to reject.

25
This suggests some caution should be used in overweighting the results shown. Although
reasonable to use as guidelines for policy, future aggregation of the attributes is called for.
[Insert Exhibit 11 here]
CONCLUSIONS AND POLICY IMPLICATIONS
This paper adds to the body of literature in several key ways. First, this paper reports initial
results from the first major academic study focused on tenant demand for green buildings beyond
LEED or Energy Star certification. The sample of 708 respondents (gleaned from over 300
buildings in 17 major US markets) nicely represents the overall population and potential internal
sample bias (office mangers vs. leadership) is shown to be small. Second, this research begins the
process of unpacking which green building-level attributes are most important to tenants, to whom
certain attributes are important, and how much more rent, if any, tenants say they are willing to
pay.
The aggregate stated WTP for green features show a 9.3% premium. This slightly exceeds
the 4-7% premiums (revealed preference) found by other scholars (Fuerst and McAllister, 2011;
Kok, Miller, and Morris, 2012) who measured LEED and Energy Star status alone. The current
results show broad acceptance of sustainable building attributes in general with preferences
varying across demographic and regional differences. However, when compared against
individual attributes, it shows a comparatively lower perception of value by tenants for the
currently marketed LEED and Energy Star brands. This suggests that more research is required in
unpacking the bundle of optimal attributes. A regionally differentiated green building scoring
matrix may help shed light on the information gap in the sustainable space rental market.
Of the eighteen green building features, improved indoor air quality and access to natural
light are the highest in perceived importance. High efficiency measures are most likely to be
preferred by certain industries, such as energy and real estate tenants, or NNN lease holders that

26
directly benefit from such measures. Public companies are among the most likely to pay for a
LEED certification. These disparities demonstrate a need in the marketplace for third-party
certified building sustainability metrics that companies can use in the context of their corporate
sustainability missions, beyond just the LEED and Energy Star brands.
Regional preferences such as limited interest in water conservation in the Lakes region or
increased value of individualized temperature control in the warm Southeast region are revealed.
Ideally, any scoring matrix or certification system including multiple building level attributes
should incorporate regional distinctions.
Finally, the presence of some building attributes leads to an increased perception of value
for them. As a potential testament to the value of the Energy Star and LEED certifications, tenants
aware that they have a green certification are more likely to pay higher rent to maintain those
certifications. Natural light, efficient HVAC and lights also are building attributes that show a
higher value to tenants already aware that their building or space has the attribute.
While the survey represents the largest and most comprehensive to date, there are some
admitted weaknesses. The low level of actual decision maker responses is less than ideal. However,
both qualitative and quantitative evidence demonstrate that the overall responses (i.e., office
managers) are consistent with the responses of tenant leadership. The knowledge level of the
responses is consistent with practitioner expectations. Thus, stated willingness-to-pay outcomes
should reasonably reflect prevailing market opinions. Despite efforts to create common language
and clear definitions, respondents may have understood the attributes effects or meanings
differently. It should also be noted that stated WTP does not always translate directly to actual
market prices. Additionally, in terms of green features, omitted variable problems are a concern in

27
any large scale analysis. However, the current survey is based on focus group input, pilot tested,
and is designed to include as much relevant information as possible.
Further research already in progress with this database includes integrating the stated
preferences from the survey with the revealed preferences from rent rolls associated with the
buildings. Future research stemming from the results presented here includes further qualitative
and quantitative examinations of green building demand drivers. Although not central to the
findings herein, the initial results do suggest that productivity or people related qualities may be
of additional academic interest. The heterogeneous results between regional, demographic, and
industry preferences for sustainable features clearly demonstrate a need for further study in that
area. More research into lease structure and whether optimal green lease structures exist are also
suggested by the data.

28
REFERENCES
Aroul, R. R. and J. A. Hansz. The Value of "Green": Evidence from the First Mandatory
Residential Green Building Program, Journal of Real Estate Research, 2012, 34:1, 27-49.
Brookshire, D. S., Thayer, M. A., Schulze, W. D., & d'Arge, R. C. Valuing public goods: a
comparison of survey and hedonic approaches. The American Economic Review, 165-177.
Carson, R. T. and W. M. Hanemann. Contingent Valuation. In: K. G. Mäler and J. R. Vincent
(eds.) Handbook of Environmental Economics, Amsterdam: Elsevier, 2005, 821-936.
Crone, T. M. and A. Clayton-Matthews. Consistent economic indexes for the 50 states, Review of
Economics and Statistics, 2005, 87:4, 593-603.
Das, P. and J. A. Wiley. Determinants of Premia for energy-efficient design in the office market,
Journal of Property Research, 2014, 31:1, 64-86.
Dermisi, S. and J. McDonald. Selling Prices/Sq. ft. of Office Buildings in Downtown Chicago –
How Much Is It Worth to Be an Old But Class A Building? Journal of Real Estate Research, 2010,
32:1, 1-21.
Eichholtz, P., N. Kok, and J. M. Quigley. Why Companies Rent Green: CSR and The Role of Real
Estate, Academy of Management Annual Meeting Proceedings. 2009.
———. Doing Well by Doing Good? Green Office Buildings, American Economic Review,
2010, 100:5, 2492-509.
Fuerst, F. and P. McAllister. Green Noise or Green Value? Measuring the Effects of
Environmental Certification on Office Values, Real Estate Economics, 2011, 39:1, 45-69
Jain, P. and Robinson, S. Professional Ownership Premiums in Real Estate, Working Paper, 2015
Kok, N., N. G. Miller, and P. Morris. The Economics of Green Retrofits. Journal of Sustainable
Real Estate, 2012, 4:1, 4-47.
Lindsey, G., and Knaap, G. "Willingness to pay for urban greenway projects." Journal of the
American Planning Association 65.3 (1999): 297-313.
Liu, C. and P. Liu. Is What's Bad for the Goose (Tenant), Bad for the Gander (Landlord)? A Retail
Real Estate Perspective, Journal of Real Estate Research, 2013, 35:3, 1-17.
Lu, W., Peng, Y., Webster, C., & Zuo, J. Stakeholders’ willingness to pay for enhanced
construction waste management: A Hong Kong study. Renewable and Sustainable Energy
Reviews, 47, 233-240.
Malizia, E. and R. Simons. Comparing Regional Classifications for Real Estate Portfolio
Diversification, Journal of Real Estate Research, 1991, 6:1, 53-78.

29
Marcato, G., & Nanda, A. Information Content and Forecasting Ability of Sentiment Indicators:
Case of Real Estate Market, Journal of Real Estate Research, Forthcoming 2015
Portney, P. R. (The Contingent Valuation Debate: Why Economist Should Care, Journal of
Economic Perspectives, 1994, 8:4, 3-17.
Reichardt, A., F. Fuerst, N. Rottke, and J. Zietz. Sustainable Building Certification and the Rent
Premium: A Panel Data Approach, Journal of Real Estate Research, 2012, 34:1, 99-126.
Reid, D. J., and F. J. Reid. Online focus groups: An in-depth comparison of computer-mediated
and conventional focus group discussions. International Journal of Market Research, 2005, 47:2,
131-62.
Robinson, S. and McAllister, P. Heterogeneous Price Premiums in Sustainable Real Estate? An
Investigation of the Relationship between Value and Price Premiums, Journal of Sustainable Real
Estate, 6, 2015
Robinson, S. and Sanderford, A. Green Buildings: Similar to Other Premium Buildings?
Journal of Real Estate Finance and Economics, Forthcoming.
Simons, R. "Estimating Proximate Property Damage from PCBs in a Rural Market: A Multiple
Techniques Approach." The Appraisal Journal (2002).
Simons, R., Robinson, S., & Lee, E. Green Office Buildings: A Qualitative Exploration of Green
Office Building Attributes. Journal of Sustainable Real Estate, 6(2), 211-232.
Simons, R., E. Choi, and D. Simons. The Effect of State and City Green Policies on the Market
Penetration of Green Commercial Buildings. Journal of Sustainable Real Estate, 2009, 1, 139‐66.
Simons, R. and K. Winson-Geideman. Determining Market Perceptions on Contaminated
Residential Property Buyers Using Contingent Valuation Surveys. Journal of Real Estate
Research, 2005, 27:2, 193-220.
Stigka, E. K., Paravantis, J. A., & Mihalakakou, G. K. Social acceptance of renewable energy
sources: A review of contingent valuation applications. Renewable and Sustainable Energy
Reviews, 32, 100-106.
U.S. Department of Energy. Buildings Energy Data Book, 2012. Retrieved from
http://buildingsdatabook.eren.doe.gov/docs/DataBooks/2011_BEDB.pdf.
U.S. Environmental Protection Agency. The Inside Story: A Guide to Indoor Air Quality. Office
of Air and Radiation, EPA 402-K-93-007, 2009.
U.S. Energy Information Administration. Annual Energy Outlook. 2013, Retrieved from
http://www.eia.gov/forecasts/aeo/.
Wiley, J., J. Benefield, and K. Johnson. Green Design and the Market for Commercial Office
Space. The Journal of Real Estate Finance and Economics, 2010, 41:2, 228-43.

30
William, S. B., E. R. Morey, and T. S. Lodder. Using Contingent Valuation to Estimate a
Neighborhood’s Willingness to Pay to Preserve Undeveloped Land. Urban Studies, 1998, 35:4,
715-27.
Wood, D. J. Corporate Social Performance Revisited. Academy of Management Review, 1991, 16,
691-718.
Wooldridge, J. M. Introductory Econometrics: A Modern Approach, fifth edition, Cengage
Learning. 2013.

31
Exhibit 1 | Distribution of Survey Participants by Region
Region
# of
Building
s
Building
% total
# of
Tenant
s
Tenant
% Total
# of
Complete
Responses
Responden
t % of
Total
Response
Rate
Mideast
17
5%
202
7%
22
3%
11%
Southeast
28
9%
229
8%
61
9%
27%
Great Lakes
43
13%
398
13%
98
14%
25%
Energy Belt
79
24%
741
25%
205
29%
28%
Far West
162
49%
1445
48%
322
45%
22%
Total*
329
100%
3015
100%
708
10%
23%
*Difference of means tests show no statistically significant difference between population and
sample.

32
Exhibit 2 | Survey Respondents’ Work Positions (N=708)
Note: Other positions include human resource managers, property managers, partners, accounting
managers, and finance managers.
CEO, 2%
CFO, 2% Director,
3% COO, 3%
Facilities
Manager,
3% President/
Vice
President,
7%
Others,
12%
General
employee
, 14%
Office
manager,
54%

33
Exhibit 3 | Respondent Knowledge of Green Building Attributes in Building (N = 547)
Attribute
Percent who answered
"Don't Know"
Percent Correct when identifying
an attribute as present or not.*
LEED
60%
84%
Energy Star
63%
30%
Natural Light Access
3%
70%
Walking access to services
3%
44%
Fitness Facility On Site
7%
85%
Public Transit Proximity
10%
59%
Recycling
12%
85%
Shower on site
16%
78%
Bike Racks
32%
66%
Electric Car Charging
38%
86%
Efficient Lighting
39%
70%
Improved Indoor Air Quality
42%
76%
Efficient HVAC
48%
58%
Water Conservation
48%
65%
Green Cleaning
76%
75%
Average
33%
69%
* % Correct of those who answered
Note: The responses shown are for the entire database. However the authors also tested for any statistical
difference between the leadership group and the general database. All tests showed no statistically
significant difference between the answers of the two groups.

34
Exhibit 4 | Stated Demand for 18 Green Building Features
Building Feature
N*
% Total (N=708)
Score**
Indoor air quality
659
93%
67.8
Access to natural light
627
89%
67.1
Recycling provided on-site
583
82%
39.2
Energy efficient lighting
542
77%
40.0
Efficient electrical and gas use for heating and cooling
540
76%
43.7
Walking access to services and restaurant
526
74%
37.9
Comfortable temperature control system
518
73%
47.9
Public transportation nearby
425
60%
33.0
Fitness facility on-site
387
55%
25.9
Lease structure
304
43%
18.9
Green cleaning products
277
39%
15.0
Water conservation
274
39%
12.7
Energy Star designation
254
36%
14.2
LEED Designation
211
30%
10.4
Shower on-site
181
26%
9.7
Bike racks at building
111
16%
5.9
Electric car charging station
54
8%
2.4
Green roof
53
7%
2.1
*Respondents are asked to rank their top nine attributes, but some ranked 10 or more, this N
includes those values.
**Score represents possible weighted ranking out 100. A variable that is ranked #1 by all
respondents would be 100. This then measures both if it was ranked and the level of ranking.

35
Exhibit 5 | Difference between “Preferred Attribute” and “Perceived as Present” of Green
Building Features
Green Building Features
(18 items)
Preferred
Attribute
(A)
Perceived
as
Present*
(B)
Difference between
Preferred Attribute
and Perceived Present
** (A-B)
Ratio***
(A/B)
Indoor air quality
659
343
316
1.92
Lease structure
304
62
242
4.90
Efficient electrical and gas use
for heating and cooling
540
320
220
1.69
Energy efficient lighting
542
387
155
1.40
Green cleaning products
277
136
141
2.04
Comfortable temperature control
518
392
126
1.32
Fitness facility on-site
387
290
97
1.33
Recycling provided on-site
583
528
55
1.10
Energy Star designation
254
213
41
1.19
Green roof
53
22
31
2.41
Water conservation
274
244
30
1.12
LEED Designation
211
202
9
1.04
Access to natural light
627
636
-9
0.99
Electric car charging station
54
99
-45
0.55
Walking access to services and
restaurant
526
595
-69
0.88
Shower on-site
181
337
-156
0.54
Public transportation nearby
425
582
-157
0.73
Bike racks at building
111
348
-237
0.32
* On average, about 30 percent of the total respondents answered that they did not know if the
listed green features are currently available for their employees
** This table is sorted by this column.
*** “>1.0 ratio” indicates that there is a net demand for a preferred green building feature while
“<1.0 ratio” indicates that there is a perception that the attribute is in over-abundance.

36
Exhibit 5b | 18 Green Building Features from the Tenant Perspective
659
304
540 542
277
518
387
583
254
53
274
211
627
54
526
181
425
111
343
62
320
387
136
392
290
528
213
22
244 202
636
99
595
337
582
348
0
100
200
300
400
500
600
700
Considered to be Most Important Currently Available

37
Exhibit 6 | Willingness-to-pay for Green Building Features (N=708)
Building Attribute
Overall Sample
Leadership Subsample
(A)
(B)
(A)
(B)
Willingness
to pay (%)
% of
respondents
(≥ 2% of
WTP)
Willingness
to pay (%)
% of
respondents
(≥ 2% of
WTP)
Access to natural light in my work
space
1.33%
243 (34%)
1.64%**
55 (42%)
Indoor air quality
1.29%
250 (35%)
1.30%
43 (33%)
Comfortable temperature control system
1.27%
187 (26%)
1.47%
43 (33%)
Lease structure
1.17%
111 (16%)
1.21%
20 (15%)
Efficient electrical and gas use for
heating/cooling*
1.09%
175 (25%)
1.00%
31 (24%)
Walking access to services and
restaurant
1.06%
158 (22%)
1.15%
31 (24%)
Public transportation nearby
1.06%
140 (20%)
1.08%
30 (23%)
Energy efficient lighting*
1.03%
179 (25%)
1.06%
28 (22%)
Fitness facility on-site
0.98%
107 (15%)
1.08%
21 (16%)
Water conservation*
0.97%
79 (11%)
1.17%
12 (9%)
LEED Designation
0.82%
47 (7%)
0.74%
7 (5%)
Shower on-site
0.78%
40 (6%)
0.94%
7 (5%)
Green roof
0.70%
11 (2%)
0.71%
2 (2%)
Recycling provided on-site
0.65%
110 (16%)
0.51%
14 (11%)
Energy Star designation
0.63%
53 (7%)
0.55%
5 (4%)
Bike racks at building
0.54%
18 (3%)
0.68%
4 (3%)
Green cleaning products used on-site
0.42%
28 (4%)
0.53%
5 (4%)
Electric car charging station
0.41%
10 (1%)
0.83%
4 (3%)
Column (A) indicates the weighted average of stated willingness to pay for each green building
feature.
Column (B) indicates the number of respondents who have greater than or equal to 2% of stated
willingness-to-pay for each feature (% in total respondents, 708)
* The assumption of 2% annual building operation savings is provided for “Efficient electricity
and gas use for heating and cooling”, “energy efficient lighting”, and “water conservation”
questions, based on the tenant’s lease structure.
** Statistically significant difference between all sample and leadership sub-sample group.
8
8
Comparing stated willingness to pay between all sample and leadership sub-sample group, only one attribute is
statistically different. This result supports the internal validity of the current survey data set.

38
Exhibit 7 | Difference in Stated Willingness to Pay by Lease Structure
0.88% 0.81% 0.91%
1.24% 1.16%
0.92%
0.00%
0.20%
0.40%
0.60%
0.80%
1.00%
1.20%
1.40%
Efficient electrical and gas use
for heating and cooling Energy efficient lighting Water conservation
FSG NNN
Ave.= 1.09%
Ave.= 1.03%
Ave.= 0.97%

39
Exhibit 8 | Descriptive Statistics with Definitions (N=708)
Continuous Variables
Variable
Definitions / Unit
Min
Max
Mean
Std.
Deviation
Rent/SF
Rent per Square Foot
$8.00
$87.00
$26.31
$12.27
Total SF
Square Footage of the tenant space
392
350,000
17,633
33,692
Employees
The number of employees in the tenant space
1
1,260
59
133
Dummy Variables
Variable
Definition
Frequency (%)
Lease NNN
Lease Structure: Triple Net, coded as a “1”
234 (33.3%)
Lease Modified
Lease Structure: Modified Gross
104 (14.7%)
Lease FSG*
Lease Structure: Full Service Gross
370 (52.2%)
Region Energy Belt
CO, LA, NM, OK, TX, UT, WY
205 (28.9%)
Region Mideast
DE, MD, NJ, NY, PA
22 (3.1%)
Region Southeast
AL, AR, FL, GA, KY, MS, NC, SC, TN, VA
61 (8.6%)
Region Lakes
ID, IL, MI, MN, OH, WI, WV
98 13.8%)
Region Far West*
AZ, CA, NV, OR, WA
322 (45.4%)
Industry Finance. &
Insurance
Finance and Insurance Industry
113 (15.9%)
Industry Legal
Legal Industry
78 (11.0%)
Industry Energy
Energy-related Industry
19 (2.7%)
Industry Government
Government
20 (2.8%)
Industry Real Estate
Real Estate-related Industry
49 (6.9%)
Industry Comp & IT
Computer and IT Industry
30 (4.2%)
Function Executive
Function of Office Space: Executive
371 (52.3%)
Function Professional
Function of Office Space: Professional Services
342 (48.1%)
Location Decision
Flex
If a tenant selects "flexible floor" as a main reason of the current
location then coded as "1."
231 (32.6%)
Public Stock
If a tenant company's stock is publically held, then coded as "1."
162 (22.8%)
Sustainability initiative
If sustainability initiatives have been discussed at a company
meeting in the last six months, then coded as "1."
195 (27.5%)
Sustainable supplier
If a tenant company prefers to choose suppliers who market
themselves as sustainable over those who do not, then coded as "1."
331 (46.7%)
Layout Hybrid
Open shared, common workspace areas with sunlight in the central
core areas, combined with much smaller than typical private office
or open plan cubicles.
174 (24.5%)
Layout Flex
No permanent office space but have access to work stations or
private offices by reservation
43 (6.1%)
Layout Traditional*
A large variety of private offices line the outside area of the floor
adjacent to the windows.
484 (68.3%)
Position Leadership
President, Vice President, CEO, CFO, and COO
94 (13.3%)
Edu college
Education: Bachelor's Degree
326 46.0%)
Use public trans
Use of Public Transportation (more than once a week)
94 (13.3%)
Age20s
Age group of 20s
58 (8.25%)
Age30s
Age group of 30s
114 (16.1%)
Age40s
Age group of 40s
186 (26.2%)
Age50s
Age group of 50s
312 (44.0%)
LEED Certification
If a tenant's building is designated as LEED, then coded as "1."
202 (28.5%)
Energy Star Rating
If a tenant's building is Energy Star-certified, then coded as "1."
213 (30.0%)
LEED_CI
If LEED Commercial Interior (CI) certified, then coded as "1."
55 (7.8%)
Notes: The “coded as a 1” group is described above. * reflects the reference category.

40
Exhibit 9 | Probit Model 1: Factors Affiliated with WTP for Green Labeled Buildings
Variable
Estimate
Wald Chi-Square
Pr > Chi-Square
Intercept
-1.0558
7.334
0.0068
Total_SF
-2.3E-06
0.3933
0.5306
Rent_SF
-0.00325
0.2896
0.5905
Lease_Modified
0.1085
0.4274
0.5133
Lease_NNN
-0.0173
0.0153
0.9016
Region_Energybelt
0.1475
1.0367
0.3086
Region_Mideast
-0.073
0.0414
0.8387
Region_Southeast
-0.0784
0.1381
0.7102
Region_Lakes
-0.0413
0.0403
0.8409
Employees
0.000604
0.4387
0.5078
Industry_Finan_Insur
-0.1483
0.7549
0.3849
Industry_Legal
0.2033
1.0183
0.3129
Industry_Energy
0.8399
6.4584
0.011
Industry_Government
0.3814
1.4045
0.236
Industry_Realestate
0.1538
0.5034
0.478
Industry_Comp_IT
0.5284
4.1203
0.0424
Function_Executive
0.1202
1.0082
0.3153
Function_Professiona
0.1455
1.4865
0.2228
Location_Decision_Flex
0.0205
0.0298
0.8629
Public_Stock
0.2307
2.8588
0.0909
Sustainability_initi
0.0919
0.5128
0.4739
Sustainable_supplier
0.3592
9.659
0.0019
Layout_Hybrid
0.1535
1.3114
0.2521
Layout_Flex
0.2278
0.9852
0.3209
Position_Leadership
0.177
1.0192
0.3127
edu_college
-0.2164
3.5026
0.0613
Use_public_trans
-0.0838
0.2209
0.6383
Age20s
-0.271
0.805
0.3696
Age30s
-0.2189
0.6695
0.4132
Age40s
-0.2045
0.6538
0.4187
Age50s
-0.3186
1.7125
0.1907
Space_NumberOfPeople
0.0156
0.0802
0.7771
LEEDCertification
-0.1851
0.9147
0.3389
EnergyStarRating
0.2922
2.2807
0.131
LEED_CI
0.4988
6.3735
0.0116
Likelihood of Global Null Beta=0
54.2041
0.0153
AIC
761.142
The binary dependent variable is a positive WTP for LEED and/or Energy Star-certified
buildings.

41
Exhibit 10 | Probit Model 2: Displays only coefficients significant at a 90% or better level from
Probit models on 13 of the 18 attributes with statistically meaningful WTP of 2.0% or more for
the dependent attribute:
Attribute
Attribute
N
Variable
Estimate
Wald
Chi-
Square
Pr >
Chi-
Square
Model
Fit Test
AIC
Internal Air Quality
250
Region_Mideast
-0.7323
4.4783
0.0343
0.439
921
Internal Air Quality
250
Public_Stock
-0.2153
2.9831
0.0841
Natural Light
243
Industry_Finan_Insur
-0.237
2.7094
0.0998
0.005
913
Natural Light
243
Industry_Legal
-0.3076
3.2368
0.072
Natural Light
243
Public_Stock
-0.3634
8.0392
0.0046
Natural Light
243
edu_college
0.2704
6.7784
0.0092
Indiv. Temp Control
187
Region_Mideast
-0.6624
2.9207
0.0875
0.439
819
Indiv. Temp Control
187
Region_Southeast
0.3495
3.5492
0.0596
Efficient Lighting
179
Industry_Energy
0.5806
3.5464
0.0597
0.029
803
Efficient Lighting
179
Industry_Realestate
0.4648
5.4978
0.019
Efficient HVAC
175
Lease_NNN
0.2648
4.5704
0.0325
0.030
794
Efficient HVAC
175
Region_Southeast
0.4924
6.9742
0.0083
Efficient HVAC
175
Industry_Government
0.5231
3.0056
0.083
Efficient HVAC
175
Industry_Realestate
0.4035
4.0384
0.0445
Walkability
158
Region_Lakes
-0.3711
3.9296
0.0474
0.053
754
Walkability
158
Industry_Realestate
0.4332
4.6432
0.0312
Walkability
158
Public_Stock
-0.2769
3.906
0.0481
Public Transit
140
Public_Stock
-0.2722
3.6462
0.0562
0.106
706
Public Transit
140
edu_college
0.2031
3.0876
0.0789
Lease Structure
111
Lease_NNN
0.3116
5.3797
0.0204
0.352
617
Lease Structure
111
Region_Energybelt
0.2417
3.0317
0.0817
Recycling
110
Region_Energybelt
-0.3637
5.8024
0.016
0.097
614
Recycling
110
Industry_Energy
0.7492
5.2257
0.0223
Recycling
110
Position_Leadership
-0.5362
6.4584
0.011
Fitness Facility
107
Region_Energybelt
0.2827
3.9497
0.0469
0.091
603
Fitness Facility
107
Region_Southeast
0.4823
5.6471
0.0175
Fitness Facility
107
Industry_Legal
-0.4365
3.9176
0.0478
Water Conservation
79
Industry_Government
0.6766
4.5831
0.0323
0.219
497
Energy Star
53
edu_college
-0.3028
4.2678
0.0388
0.451
379
Variables are the significant variables only (space constraints) from 13 separate regressions, the column
Attribute shows the dependent variable. Attribute N shows how many times the dependent attribute had a
WTP of >= 2.0%

42
Exhibit 11 | Probit Model 3: Displays only coefficients significant at a 90% or better level from
Probit models on all 18 building attributes with dependent variable set as 2 if ranked first through
third, 1 if ranked fourth through ninth, or 0 if unranked :
Variable
Estimat
e
Wald Chi-
Square
Pr > Chi-
Square
attribute
Rank
1-3
Rank
4-9
Public_Stock
-0.2873
6.8451
0.0089
attRank_natlight
410
207
edu_college
0.1981
4.6123
0.0317
attRank_natlight
410
207
Region_Southeast
0.424
6.8449
0.0089
attRank_tempcont
244
259
Region_Lakes
0.2612
3.4089
0.0648
attRank_tempcont
244
259
Position_Leadership
0.2492
3.6587
0.0558
attRank_tempcont
244
259
Industry_Finan_Insur
-0.2395
3.9994
0.0455
attRank_HVAC
193
337
Industry_Legal
0.284
3.9493
0.0469
attRank_HVAC
193
337
Region_Energybelt
0.2947
8.0638
0.0045
attRank_walkability
156
344
Region_Lakes
-0.3393
5.6138
0.0178
attRank_walkability
156
344
Industry_Legal
0.3139
4.8169
0.0282
attRank_walkability
156
344
Position_Leadership
0.2151
2.7714
0.096
attRank_walkability
156
344
Region_Energybelt
-0.2374
5.0871
0.0241
attRank_pubtrans
151
263
Position_Leadership
-0.2588
3.8243
0.0505
attRank_pubtrans
151
263
Region_Energybelt
-0.1778
2.7327
0.0983
attRank_Recycling
105
464
Region_Southeast
-0.4745
8.0998
0.0044
attRank_Recycling
105
464
Position_Leadership
-0.3863
8.209
0.0042
attRank_Recycling
105
464
Region_Energybelt
0.3576
11.298
0.0008
attRank_fitfac
83
294
Region_Southeast
0.4614
8.1758
0.0042
attRank_fitfac
83
294
Industry_Legal
0.2591
3.0169
0.0824
attRank_tenantreward
59
226
Industry_Energy
0.5674
4.2812
0.0385
attRank_tenantreward
59
226
Industry_Governmen
t
0.4685
3.1107
0.0778
attRank_tenantreward
59
226
Industry_Comp_IT
0.5144
5.5782
0.0182
attRank_tenantreward
59
226
Lease_NNN
-0.255
5.0223
0.025
attRank_EnergyStar
38
199
Public_Stock
0.2011
3.0505
0.0807
attRank_EnergyStar
38
199
edu_college
-0.1772
3.3443
0.0674
attRank_EnergyStar
38
199
Region_Energybelt
-0.2056
3.1385
0.0765
attRank_greenclean
37
221
Region_Southeast
-0.3986
4.5873
0.0322
attRank_greenclean
37
221
Industry_Finan_Insur
0.2245
3.1179
0.0774
attRank_greenclean
37
221
Industry_Legal
-0.4842
7.6074
0.0058
attRank_greenclean
37
221
Industry_Energy
0.7973
8.0358
0.0046
attRank_LEED
31
156
Industry_Realestate
0.5262
8.0756
0.0045
attRank_LEED
31
156
Industry_Comp_IT
0.4324
3.4242
0.0642
attRank_LEED
31
156
Region_Southeast
0.4159
5.5244
0.0188
attRank_Shower
27
136
Region_Lakes
-0.4721
6.4993
0.0108
attRank_Shower
27
136
Industry_Realestate
-0.4648
3.9894
0.0458
attRank_Shower
27
136
Region_Energybelt
-0.3184
7.285
0.007
attRank_watercons
23
211
Region_Southeast
-0.6126
10.0267
0.0015
attRank_watercons
23
211
Region_Lakes
-0.3433
4.4832
0.0342
attRank_watercons
23
211

43
Industry_Legal
-0.4105
5.5556
0.0184
attRank_watercons
23
211
Industry_Governmen
t
0.4511
2.7233
0.0989
attRank_watercons
23
211
Industry_Comp_IT
0.4948
4.8176
0.0282
attRank_watercons
23
211
edu_college
-0.1886
3.662
0.0557
attRank_watercons
23
211
Lease_NNN
0.2893
4.7252
0.0297
attRank_Bikeracks
15
86
Industry_Finan_Insur
-0.4638
5.8159
0.0159
attRank_Bikeracks
15
86
Region_Energybelt
-0.6036
9.0515
0.0026
attRank_elec_carchg
8
41
Region_Lakes
-0.6709
4.4087
0.0358
attRank_elec_carchg
8
41
Industry_Legal
0.4003
3.4717
0.0624
attRank_elec_carchg
8
41
edu_college
0.522
9.0442
0.0026
attRank_GreenRoof
5
41
Variables are the significant variables only (space constraints) from 18 separate regressions, the
column Attribute shows the dependent variable. Rank 1-3 shows how many times the dependent
attribute was ranked in the top 3 by a respondent and 4-9 shows how many times it was ranked 4-9.