Designer Buildings: An Evaluation of the Price Impacts

Designer Buildings: An Evaluation of the Price Impacts

Designer Buildings: An Evaluation of the Price Impacts of Signature Architects Franz Fuerst, Pat McAllister and Claudia Murray University of Reading Henley Business School School of Real Estate and Planning Research Question Do offices designed by architects who have won the AIA Gold Medal and/or the Pritzker Prize command a rental premium? More specifically: Whether, ceteris paribus, commercial offices designed by signature architects achieve rental premiums compared to commercial offices designed by nonsignature architects DEFINITIONS: ARCHITECTURE AND FAME Awards Competitions. Publications. Buildings Focus? Size? Client? The Gold Medal is the highest honor that the American Institute of Architects can bestow on an individual. It is conferred by the national AIA Board of Directors in recognition of a significant body of work of lasting influence on the theory and practice of architecture.

PRITZKER PRIZE AIA GOLD MEDAL DEFINITIONS: AWARDS To honour a living architect whose built work demonstrates a combination of those qualities of talent, vision, and commitment, who has produced consistent and significant contributions to humanity and the built environment through the art of architecture. TRANSMISSION OF PRIZE TO PRICE Prize Cost ,Benefits Image benefits to =occupiers =owners =developers Enhanced prestige, recognizability Higher productivity? Price Better design? (Not considered) Sample: Rents 139 National and/or Chicago designated landmarks in Chicago. Method: Hedonic regression

Findings: Premium only for newer landmark offices , for older buildings the coefficient was negative in all model specifications but insignificant. Vandell and Lane (1989) Sample: Rents and vacancy rates of 55 Class A office buildings in Boston and Cambridge, Massachusetts. Method: Examined the relationship between aesthetic qualities and rents/vacancy rates. Findings: Buildings rated in the highest quintile in terms of aesthetics had rents that were 22% higher than buildings rated in the lowest quintile controlling for differences in location, number of stories etc. Gat (1998) Sample: 50 office buildings in Tel Aviv, rents, physical and location features. Method: Hedonic regression, rating of building and design quality by architects. Findings: Marginal contribution of quality of architecture variable of 5.4% --highest of a non-location variable in the model. PREVIOUS EMPIRICAL RESEARCH Hough and Kratz (1983) CoStar We have identified 490 office buildings designed by signature architects. DATA: In order to identify a control sample against which to compare the sample of buildings designed by signature architects, we have drawn upon the CoStar database (10,500 properties with full set of variables in our sample). METHODOLOGY

RESEARCH METHODS AND DATA This study draws on CoStar's comprehensive national database (1) Hedonic regression modeling enables researchers to isolate the relative contribution of an attribute (or a bundle of attributes) and estimate the price effects. (2) Logistic regression: can be used to identify a suitable peer group based on probability scores. (3) Manual selection from database. Requires close inspection of property details and available comparables. METHODOLOGICAL ISSUES Isolating price premiums Spillover effects Sample selection Skill, brand or expense? Age Quality Tenant quality Location Lease structures Size Specifications Given heterogeneity, key methodological problem is to control for

differences between ODSAs and non-ODSAs. ODSA may lift rental rates in surrounding area . Our definition of a signature architect is unavoidably arbitrary. Comparable buildings may be designed by high profile architects. Iconic buildings need to be considered designed by signature architects of the 1920s and 1930s? Is an identified price difference due to superior design, iconic status or additional expenditure? DISTRIBUTION OF ODSAs IN AMERICA Chicago Boston New York San Francisco Los Angeles Washington % of total Proportion of ODSA in US 20 18 16 14 12 10 8 6

4 2 0 BASIC DESCRIPTIVE STATS ODSA (median) Non-ODSA (median) Rent psf $26.7 $18 Price psf $215 $115 Size 362,771 10,800 Storey 16 2 Age

21 23 Plot Size 1.54 1.16 Occupancy Rate 93.7 79.8 N Variable Variable HEDONIC EQUATION Rental rates ln Ri 0 1 ln Yi 2 ln Oi 3 ln Si 4 ln Li 5 ln Fi ODSAi it Age of building (year of construction or major refurbishment) Size (sq.ft. of rentable space) Height (no. of stories) Size (sq.ft.) Lease type Location (submarket cluster) Building quality

HEDONIC REGRESSION RESULTS - RENTS Coefficient Std. Error t-Stat Prob. C 2.94 0.10 29.40 0.00 ODSA 0.07 0.02 3.39 0.00 Age -0.04

0.00 -17.67 0.00 LOG(RBA) -0.02 0.00 -3.58 0.00 LOG(STORIES) 0.05 0.01` 8.92 0.00 NET LEASE -0.09 0.01 -15.20

0.00 CLASS A 0.22 0.01 19.85 0.00 CLASS B 0.08 0.01 13.30 0.00 OCCUPANCY RATE 0.00 0.00 3.28 0.00 Submarket cluster dummy variables included

Adj. R-sq.: 0.61 F-test: 27.09 Prob. (F) 0.000 REGRESSION RESULTS MID-TOWN RENTS Coefficient Std. Error t-Stat Prob. C 3,43 1.09 3.15 0.00 ODSA 0.09 0.08

1.10 0.27 AGE -0.11 0.06 -1.79 0.08 LOG(RBA) 0.00 0.06 0.05 0.96 LOG(STORIES) 0.02 0.11` 0.17 0.87

OCCUPANCY RATE 0.18 0.20 0.91 0.37 Submarket cluster dummy variables included N=80 Adj. R-sq.: 0.32 F-test: 2.14 Prob. (F) 0.02 HEDONIC REGRESSION RESULTS - PRICES Coefficient Std. Error t-Stat Prob.

C 5.77 0.55 10.51 0.00 ODSA 0.27 0.06 4.72 0.00 LOG(RBA) -0.22 0.01 -16.07 0.00 LOG(STORIES) 0.16

0.02` 9.24 0.00 NET LEASE -0.09 0.01 -15.20 0.00 CLASS A 0.49 0.03 14.78 0.00 CLASS B 0.06 0.02 2.96

0.00 TIME TREND 0.03 0.00 16.07 0.00 AGE (VARIOUS BANDS COEFFICIENTS ALL POSITIVE) Submarket cluster dummy variables included Adj. R-sq.: 0.46 F-test: 8.70 Prob. (F) 0.000 Logit function Probability of ODSA 1.0 0.8 0.6

0.4 0.2 0.0 Logistic regression: Maximum likelihood function, probability (P) constrained between 0 and 1x Results of Logistic Regression Predicted signature Step 1 0 1 Percentage Correct 0 11,015 86 99.2 1 176 201

53.3 Observed signature Overall Percentage a. The cut value is .500 97.7 Results of Logistic Regression Matched peer group characteristics RESULTS USING LOGIT GENERATED SAMPLE ODSA sample Rent Size (sq.ft.) Stories Year built Year renovated Land area % leased

276 479 478 472 109 480 480 Mean 33.67 543,071 22.42 1980 1993 6.58 88.79 Median 26.65

378,538 16.00 1984 1995 1.55 93.70 N Matched sample (p>0.13) Rent Size (sq.ft.) Stories Year built Year renovated Land area % leased N

277 408 408 408 123 408 408 Mean 34.7 502,191 19.76 1982 1996 4.76 88.42 Median 27.94

315,749 16.00 1986 1998 1.81 95.72 CONCLUSIONS/ FURTHER WORK Hedonic analysis suggests a rent premium but peer group analysis based on results of logit model is inconclusive (raw averages vs. probability weighted) Model sales transaction prices Obtaining close comparables by manually selecting suitable properties from the CoStar database for each ODSA (case studies of NYC, LA and Chicago) More in-depth study of specific contribution of signature architects required CONCLUSIONS/ FURTHER WORK More comprehensive assessment of overall added value of signature architecture: Neighbourhood externalities, quality of interior space etc.

Recently Viewed Presentations

  • Transportation Financing in Maryland

    Transportation Financing in Maryland

    Transportation Trust Fund FY 2016 - FY 2021 (Millions of Federal and State $) Sources NOTE: Includes non-budgeted federal assistance. MARYLAND DEPARTMENT OF TRANSPORTATION MARYLAND DEPARTMENT OF TRANSPORTATION With changes from 2011 legislative session, Registration and MVA fees also make...
  • The Earth as a Thermal Engine  Important Radioactive

    The Earth as a Thermal Engine Important Radioactive

    Conventional Wisdom Classification of elements based on geochemical affinity lithophile : affinity for silicates chalcophile: affinity for sulfur siderophile: affinity for iron metal Potassium is strongly lithophile, hence only in the silicate mantle and crust (Bulk Silicate Earth-BSE) No known...
  • CDISC SDTM v 1.5 For New Jersey CDISC

    CDISC SDTM v 1.5 For New Jersey CDISC

    --USCHFL - Unscheduled Flag - Added to support use case in the SEND model (DO NOT USE for Human Trials) --REPNUM - Repetition Number - Indicates the number of times a test was repeated within a given timeframe (i.e. same...
  • Absolute Monarchy - Westerville City School District

    Absolute Monarchy - Westerville City School District

    A system of government headed by a monarch as the only source of power controlling all functions of the state. Absolute Monarchy Artifact Analyze Capital Good A productive resource consisting of human-made materials needed to produce goods and services; ex....
  • Vocabulary Workshop - Augusta County Public Schools

    Vocabulary Workshop - Augusta County Public Schools

    Deadlock (n) n. a standstill resulting from the opposition of two equal forces or factions; (v) to bring to such a standstill Synonym: standoff, stalemate, impasse
  • Strokes and TIAs

    Strokes and TIAs

    Thrombosis, secondary to atherosclerosis. Or . cerebral embolism from AF, valve disease/replacement, post-MI, or DVT embolism with patent foramen ovale (paradoxical embolism) Or . an episode of hypoperfusion. Cerebral haemorrhage (12%) Bleeding into an area of the brain. Intra-cerebral (9%)...
  • Dia 7

    Dia 7

    Digitaal nakijken met Turnitin Werkgroep ECENT 15-5-2013 ... aan dat verwacht wordt / waar naar gestreefd dient te worden Bevordert zelfreflectie / zelfsturing Ervaringen van VU docenten Stimuleert gezamenlijke beoordelingsaanpak Bevordert transparantie / betrouwbaarheid beoordeling Werkt ...
  • Subject code and name Course code and name Name of ... - TAFE NSW

    Subject code and name Course code and name Name of ... - TAFE NSW

    Remuneration. However, after due consultation, it was decided NOT to ban commissions for personal insurances, at this stage. Life companies that do not offer on a direct basis, but via intermediaries such as financial planners, would pay a commission that...