Sustainable Building Certification and the Rent Premium: A Panel Data Approach
Reichardt, Alexander, Fuerst, Franz, Rottke, Nico B., Zietz, Joachim, The Journal of Real Estate Research
This paper investigates whether obtaining sustainable building certification entails a rental premium for commercial office buildings and tracks its development over time. To this aim, both a difference-in-differences and a fixed-effects model approach are applied to a large panel dataset of office buildings in the United States in the 2000-2010 period. The results indicate a significant rental premium for both ENERGY STAR and LEED certified buildings. Controlling for confounding factors, this premium is shown to have increased steadily from 2006 to 2008, followed by a moderate decline in the subsequent periods. The results also show a significant positive relationship between ENERGY STAR labeling and building occupancy rates.
(ProQuest: ... denotes formulae omitted.)
The emergence and rapid growth of voluntary certification systems such as ENERGY STAR and LEED in the United States are reflective of a paradigm shift towards increased environmental awareness in the commercial real estate industry. The main objective of these certifications is to impart information on a building's degree of energy efficiency and sustainability to both occupiers and investors. Although environmental certification has only recently emerged from a niche market to becoming a mainstream phenomenon, a number of prominent pricing studies of green buildings have been conducted recently. Apart from case studies of individual properties, several cross-sectional and pooled studies, which are reviewed below, have demonstrated that certified buildings command higher rental rates compared to non-certified buildings. However, a potential shortcoming of these studies is that pricing dynamics cannot be studied in a cross-sectional framework as it only provides a snapshot of environmental labeling and certification at a certain point in time. More importantly, it is difficult to rule out in a cross-sectional study that any observed price premia were genuinely caused by eco-certification and not by unobserved pre-existing characteristics that subsequently cause both certification and higher prices.
This study takes the analysis of the effects of ENERGY STAR labeling and LEED certification on a property's rental rates and occupancy rates one step further by applying panel data regressions, specifically difference-in-differences (DID) and fixed-effects models. These models allow controlling for unobserved effects, thereby mitigating a potential omitted variable bias present in many cross-sectional studies. Fixed-effects models also provide an estimate of the dynamic behavior of the rent premium over time. A relatively long time series of nearly ten years of quarterly observations is used to estimate a 'green' rental premium index for a large sample of labeled buildings. A key expectation is that the rent premium for labeled and certified buildings has been growing over time fueled by rising concerns for the environment, higher energy prices, and heightened interest in more sustainable properties. However, the sharp decline in the economy in 2007 and the following quarters may have had a dampening effect on rent premiums.
The sample includes 7,140 buildings, 1,768 of which are certified and 5,372 are non-certified control buildings. The buildings are located in the 10 largest metropolitan markets across the U.S. The DID models show a significant rent premium for ENERGY STAR rated buildings from 2004 to 2007. The fixed-effects models suggest an average rent premium of 2.5% for an ENERGY STAR rating and 2.9% for LEED certification over the observation period. Rent premiums for buildings with an ENERGY STAR rating emerge in 2006. They continuously increase until the second quarter of 2008, when the average rent premium reaches 7%, but then decrease in the wake of the economic crisis. ENERGY STAR labels also have a significant and positive effect on occupancy rates.
The remainder of the paper is organized as follows. …