LOCATION GREEN A SPATIAL ANALYSIS OF GREEN BUILDINGS

LOCATION, GREEN. A SPATIAL ANALYSIS OF GREEN BUILDINGS IN EUROPE? Gunther Maier Research Institute for spatial and Real Estate Economics Vienna University of Economics and Business (WU) Costin Ciora The Bucharest University of Economic Studies (ASE) Department of Financial Analysis and Valuation (AEEF) Ion Anghel The Bucharest University of Economic Studies (ASE) Department of Financial Analysis and Valuation (AEEF)

Agenda 1. Introduction / Motivation 2. Research questions 3. Literature 4. Methodology / Results 5. Conclusions / Future improvements 2

Introduction • Work in progress • Organizing a summer school and ERSA 2016 • “Europe” has shrunk to “Germany” • Substantial room for improvement • Improvement of data basis • Improvement of analytical steps • Improved methods 3

Motivation • The high spread of green building certification across Europe has become more visible with the new projects that have been built in the last five years. • About 35% of the EU's buildings are over 50 years old, so there will be a wide spread of green buildings in the future 4

Motivation Source: DGNB 5

Research questions • Where are the green buildings located? • Where in the country? • Where in the respective city? • Is there a function towards distribution of green buildings? 6

Hypotheses: • Green projects are more common in the SW of Germany (DGNB office in Stuttgart) • There is a common distance of 2 -4 km from the city center, in which green buildings are being developed. The “green belt” - As the distance from the CBD increase, thus the green project are more common (Green Building District) • In larger cities (larger CBD) the green belt is located further away from the center 7

Literature Determinants of green building development Eichholtz, Kok & Quigley (2010) - US market study on 10. 000 commercial buildings with LEED and/or Energy Star label, divided into 900 clusters, based on their location, showed an increase in selling price of 16 percent. Fuerst & Mc. Allister (2009) – calculated a sale price premium of 31% for energy Star certified buildings and 35% for LEED certified. Miller, Spivey & Florance (2008) calculated a value premium of 9. 9% for LEED certified buildings and 5. 3% for Energy Star. Location of green buildings Braun, Cajias & Bienert (2014) – ERES 2014, Bucharest Argued that green buildings (US market) are located predominantly and disproportionally in prime locations. 8

Methodology / Results • Green buildings in Germany with specific coordinates • 343 buildings with DGNB certification and with coordinates (projects without address excluded) • Point pattern analysis in order to test for density differences of green buildings in Germany and for spatial clusters of project in the country. 9

Methodology / Results • Green Buildings are more frequent in the South-West of Germany Trend formula: ~polynom(x, y, 2) Estimate S. E. Ztest (Intercept) -177. 44985272 46. 43045860 na [x] -0. 43375161 0. 78792202 [x^2] -0. 02690892 0. 01275400 * [y] 7. 24675651 1. 82216133 *** [y^2] -0. 07377680 0. 01797647 *** [x. y] 0. 01762523 0. 01505135 10

Methodology / Results • Green Buildings are spatially clustered • L: transformation of Ripley’s K • Based on nearest neighbor distance 11

Methodology / Results • Location in the city / What is the respective city? • “Labour Market Regions” in Germany (Eckey, Kosfeld, Türck, 2006) • Identify LMRs with green buildings • Identify the largest city in each LMR – in few cases, also close to the green building • Define the center of this city as center of the region • For every green building, calculate the distance to the nearest center – “distance from CBD” 12

Methodology / Results • Location in the city / distance from the center • More central in larger cities (ANOVA: p < 0. 0005) n median mean stdev All 337 4. 90 8. 47 9. 41 <50, 000 24 10. 88 14. 48 13. 30 50 -100, 000 10 10. 78 17. 33 18. 25 100 -500, 000 89 6. 10 9. 81 9. 64 500 -1 Mio 109 4. 93 8. 05 8. 96 1 Mio-2 Mio 73 4. 26 5. 82 5. 80 >2 Mio 32 3. 28 4. 91 4. 66 13

Methodology / Results • Distance from the center declines with the size of the city – but, a very weak relationship (R 2 = 0. 06) • Effect levels off with larger cities • But still negative at max population Estimate Std. Error t value Pr(>|t|) signif Intercept 12. 11 0. 93 12. 99 < 2 e-16 *** Pop -6. 8 e-06 1. 78 e-06 -3. 83 1. 54 e-4 *** pop^2 1. 4 e-12 5. 0 e-13 2. 77 0. 006 ** 14

Methodology / Results • • Define 1 km distance bands around the center Count the number of green buildings in each band Expected from increasing area: linear increase (2 rp) Expected from hypothesis: increase – maximum – decrease with maximum in 2 -4 km. • Hypotheses (shape) are clearly NOT confirmed for whole sample 15

Methodology / Results • Subdivide by city categories: • < 50, 000 • 50, 000 – 100, 000 • 100, 000 – 500, 000 • 500, 000 – 1, 000 • 1, 000 – 2, 000 • > 2, 000 16

Methodology / Results • Subdivide by city categories: • < 50, 000 • 50, 000 – 100, 000 • 100, 000 – 500, 000 • 500, 000 – 1, 000 • 1, 000 – 2, 000 • > 2, 000 17

Methodology / Results • Subdivide by city categories: • < 50, 000 • 50, 000 – 100, 000 • 100, 000 – 500, 000 • 500, 000 – 1, 000 • 1, 000 – 2, 000 • > 2, 000 18

Methodology / Results • Numbers of green buildings clearly decline with distance from the city center • Contrary to naïve expectation (linear increase) • Strong concentration tendency • whether this is stronger than the concentration of office buildings / new office buildings needs to be answered • In 4 of 6 city classes is the modus in the first kilometer; same for total sample • Hypothesis of a green building band around the CBD is rejected – for absolute numbers • Could still show up in relative terms (relative to (new) office buildings) 19

Conclusions • We see a clear pattern of spatial distribution of green buildings in Germany, but not necessarily as expected: • GBs are more frequent in SW of Germany (location of DGNB office? ) – more outside the larger cities • GBs are clearly spatially concentrated in clusters • Largest number in the first kilometer around the city center – despite the lack of land historic building stock • Hypothesis of a green band around the city was not confirmed 20

Future improvements • Point pattern analysis: • Use German borders as window • Currently rectangular window • Get to know the method better, use more adequate / sophisticated methods • All analytical steps: • Compare with a reference distribution – stock of office buildings, new office buildings • Currently, the reference is an equal distribution • Spatial framework • Expand beyond Germany • Other countries, other certificates 21

Thank you! Gunther Maier Research Institute for spatial and Real Estate Economics Vienna University of Economics and Business (WU) Costin Ciora The Bucharest University of Economic Studies (ASE) Department of Financial Analysis and Valuation (AEEF) Ion Anghel The Bucharest University of Economic Studies (ASE) Department of Financial Analysis and Valuation (AEEF) Corresponding author: Gunther Maier gunther. maier@wu. ac. at 22