The Dilemma of Regional Economic Policy

1
The Value of a Green Building Certificate for
Office Buildings in CEE
Michal Gluszak, Gunther Maier, Sabine Sedlacek,
Malgorzata Zieba
2
Content

• Introduction
• The project team
• Conceptual background
• Interviews
• Survey instrument
• First results

3
Introduction

• Green Building movement in Central and Eastern
  Europe
• Green Building Councils in almost all CEE
  countries
• Various strategies, various certification schemes
• Internationally active developers and consultants
  got involved
• What is the value of green building certificates
  in the CEE market?
• Very few certified buildings
• Limited information on transactions
• Need to do a contingent valuation study

4
The Research Team

• Currently Cracow (Poland) and Vienna (Austria)
• Interested in expansion to other countries
• Team Cracow
• Michal Gluszak, University of Economics, Cracow
• Malgorzata Zieba, University of Economics, Cracow
• Team Vienna
• Gunther Maier, WU Vienna
• Sabine Sedlacek, Modul University, Vienna

5
Conceptual Background

• Evidence for a positive effect of Green Building
  certificates on values and rents
• Eichholtz, Kok Quigley (2010) (US LEED and
  Energy Star) effective rent premium 7, sales
  price premium 16 the label by itself has a
  positive value above the implied energy savings.
• Fuerst McAllister (2011) (US LEED and Energy
  Star) rent premium 5 (LEED) and 4 (Energy
  Star) sales price premium 25 (LEED), 26
  (Energy Star)
• Wiley, Benefield Johnson (2010) (US LEED and
  Energy Star) rent premium 7 to 17 higher
  occupancy by 10 to 18 selling premium per sqft
  30 (Energy Star) to 130 (LEED).

6
Conceptual Background

• Positive image of Green Buildings
• Addae-Dapaah, Hiang Shi (2009) (Singapore,
  survey of occupants) No effect of awareness and
  appreciation of green benefits beyond cost
  savings and higher building values. Benefits are
  very uncertain.
• Hypotheses
• Green building certificates have a significant
  positive effect on rents and sales prices.
• In less developed markets (CEE) awareness will be
  low

7
Conceptual Background

• Method of choice
• Hedonic price estimation with certificate as
  explanatory variable
• Problem
• Too few green buildings yet in CEE markets very
  limited information on rents and transactions
• Solution
• Expert interviews
• Contingent valuation survey

8
Interviews

• Interviews with 16 commercial property
  professionals
• Semi-structured in-depth interviews
• Main results
• Certificate recognition weak (LEED the most
  popular), some experts were not familiar with
  different certification schemes
• Green profile is not a key attribute in a
  decision process, only one expert spontaneously
  mentioned it as somewhat important when office
  space decisions are concerned

9
Interviews

• Main results (cont.)
• no single expert expected higher rents in
  certified properties.
• no single expert expected that company would move
  to certified building from not certified premises
  even if they were provided assistance.
• some experts suspected green washing
• Barriers supply current economic conditions
• Differences by size (bigger companies are more
  interested in standard and wellbeing of
  employees) and nationality (US and UK companies
  are used to green standards)
• Development is driven by international investors
  (SKANSKA most eminent example)

10
Interviews

• Most important factors in office space choice in
  Poland

No Attribute Frequency
1 rent 10
2 location 9
3 structure and size 7
4 maintenance costs 6
5 parking lots 6
6 tenant-mix 4
7 Facilities 3
8 architectural design 1
9 green building certificate 1
11
Survey instrument

• Survey of companies who have moved to new office
  space within the last 2 years
• Goal identify the WTP (implicit price) for green
  building certificate
• Strategy contingent valuation
• Compare current office space with a similar
  hypothetical alternative which one would you
  have chosen?
• Analysis by use of a conditional logit model

12
Survey (start page)
13
Survey (page 1)
14
Survey (page 2)
15
Survey (page 3, repeated 10 times)
16
Survey

• Generating the hypothetical alternatives
• Criteria are sorted in decreasing expected
  attractiveness (new before old, city center
  before periphery)
• For all criteria except price, operating costs
  and certificate For the new alternative, we
  either stay at the criteria value (40) or go one
  step up (30) or down (30). When out of bounds,
  it is set to the boundary value.
• For certificates When certificate 50 same
  certificate, 50 no certificate when no
  certificate 40 no certificate, LEED, BREEAM
  and DGNB with 20 each

17
Survey

• Generating the hypothetical alternatives
• Sum of characteristics gives a rough measure of
  attractiveness
• Randomly generated price deviations by 0, 5,
  10, 15 or 20 up or down
• Result centered around zero and shifted by
  difference in attractiveness
• Correction over the experiment
• When only the original option is chosen, the
  alternative option becomes cheaper
• When only the alternative option is chosen, it
  becomes cheaper

18
First results

• Based on only TWO respondents from Vienna
• Therefore no significant coefficients, limited
  model quality
• Iteration 0 log likelihood -11.46822
• Iteration 1 log likelihood -10.436821
• Iteration 2 log likelihood -10.328076
• Iteration 3 log likelihood -10.327686
• Iteration 4 log likelihood -10.327686
• Conditional (fixed-effects) logistic regression
  Number of obs 40
• 
  LR chi2(3) 7.07
• 
  Prob gt chi2 0.0697
• Log likelihood -10.327686
  Pseudo R2 0.2550
• --------------------------------------------------
  ----------------------------
• choice Coef. Std. Err. z
  Pgtz 95 Conf. Interval
• -------------------------------------------------
  ----------------------------
• const -.9396476 .8159093 -1.15
  0.249 -2.5388 .6595052
• rent -.1215278 .0835022 -1.46
  0.146 -.2851891 .0421335
• cert01 1.074909 1.112476 0.97
  0.334 -1.105504 3.255322

19
First results

• Iteration 0 log likelihood -13.862944
• .....
• Iteration 12 log likelihood -4.938938
• Conditional (fixed-effects) logistic regression
  Number of obs 40
• 
  LR chi2(8) 17.85
• 
  Prob gt chi2 0.0224
• Log likelihood -4.938938
  Pseudo R2 0.6437
• --------------------------------------------------
  ----------------------------
• choice Coef. Std. Err. z
  Pgtz 95 Conf. Interval
• -------------------------------------------------
  ----------------------------
• loc -.6671478 1.751317 -0.38
  0.703 -4.099666 2.765371
• transp 16.89953 3649.109 0.00
  0.996 -7135.223 7169.022
• age -2.088071 3.2887 -0.63
  0.525 -8.533804 4.357663
• type -1.473766 3.592481 -0.41
  0.682 -8.5149 5.567368
• qual -38.38153 7382.253 -0.01
  0.996 -14507.33 14430.57
• const -.5146494 2.223854 -0.23
  0.817 -4.873324 3.844025
• rent -.296666 .306436 -0.97
  0.333 -.8972694 .3039375

20
Conclusions

• Important question
• Survey instrument is tested and ready
• Lot of additional work needs to be done
• Interesting to add additional CEE countries