|
Article Excerpt
Understanding the determinants of home-efficiency improvements is significant to a range of energy policy issues, including the reduction of fossil fuel use and environmental protection. This paper analyzes retrofit choices by assembling a unique data set merging a nationwide household survey from Germany with regional data on wages and construction costs. To explore the influence of both heterogeneous preferences and correlation among the utility of alternatives, we estimate conditional-, random parameters-, and error components logit models that parameterize the influence of costs, energy savings, and household-level socioeconomic attributes on the likelihood of undertaking one of 16 renovation options. We use the model coefficients to derive household-specific marginal Willingness to Pay estimates, and with these assess the extent to which free-ridership may undermine the effectiveness of recently implemented programs that subsidize the costs of retrofits.
1. INTRODUCTION
Home renovation is generally asserted to be a highly effective means for households to lower expenditures on energy through increased efficiency. From a public policy perspective, energy efficiency in the residential sector confers the additional benefit of reducing reliance on fossil fuels, thereby contributing to both energy security and environmental stewardship. In Germany, as in other industrialized countries, the residential sector accounts for upwards of 30% of energy end use, the overwhelming share of which is consumed for space heating and hot water preparation. Consequently, the improvement of home insulation and heating equipment in the existing building stock, which directly impacts the energy required for heating services, is seen to afford considerable scope for reducing the country's energy consumption.
Over the past decade, the German government has implemented several financial support programs to encourage such retrofitting activities. Homeowners have received access to low-interest loans and--in a recently launched program initiated in 2007--can alternatively apply for grants. An important question in gauging the policy merits of such measures concerns the homeowner's Willingness to Pay (WTP) for the energy savings that accrue through renovations. Given this information, one can analyze whether and to what extent the subsidization program suffers from free rider effects. Free ridership occurs if the subsidized household would have undertaken the energy-conserving activity even in the absence of the subsidy, that is, if the household's WTP exceeds cost (Train 1994). Despite its relevance to the assessment of publicly-financed programs, WTP estimates for energy-savings and the associated implications for free-ridership have received scant scrutiny to date.
The purpose of the present study is twofold. First, we estimate the determinants of home retrofits and derive therefrom estimates of the marginal WTP for energy savings. Second, we assess the extent to which free rider effects threaten to undermine the social benefits of the subsidization program. These objectives are pursued using a unique data set of 2530 owners of German single-family homes, which combines real investment cost for 16 retrofit measures, engineering estimates of the respective energy savings, and information on wage and material costs along with the sociodemographic characteristics of the sampled households.
Our work builds on a handful of earlier studies of household energy consumption behavior, most of which draw on data obtained from surveys in the U.S. With respect to home retrofits, Cameron (1985) was among the first to analyze household choice behavior using a nested logit model. She finds that income, relative energy prices, and retrofit prices are the key determinants of demand for conservation retrofits. Subsequent studies using U.S. household survey data have extended this line of inquiry in a number of directions, including analyses that address the effectiveness of energy conservation programs (Hartman 1988), the effects of changes in energy prices on the consumption of housing, residential energy, and other goods (Quigley and Rubinfeld 1989), and the extent to which homeowners apply high discount rates to home-improvement opportunities (Metcalf and Hassett 1999). Among the few studies of this issue from the European context, Banfi et al. (2008) estimate household's marginal Willingness to Pay using an innovative stated choice experiment conducted among a sample of Swiss apartment tenants and homeowners. Their estimates, obtained from a multinominal logit model, suggest the importance of both energy savings as well as comfort benefits as determinants of retrofit choices.
Although publicly financed programs to encourage energy conservation are increasingly common in industrialized countries, only a few studies have investigated the magnitude of free rider effects. Joskow and Marron (1992) and Eto et al. (1995) conduct a meta-analysis of free ridership by surveying evaluations of demand-side management (DSM) programs conducted by U.S. utilities. With respect to residential programs, the authors uncover a wide range of estimates, varying from zero to up to 50% of free riders. However, most of the reviewed evaluations are based on simple survey questions that ask the respondents whether they would have hypothetically reached the same decision in absence of the DSM program. Due to the nature of these questions, the calculated free rider share may therefore be susceptible to a hypothetical- or response bias. (1) Malm (1996) circumvents these difficulties by analyzing the revealed choice of high efficiency heating system purchases among different clusters of consumers. He derives an impressive share of 89% of households that would have bought the efficient equipment even in the absence of a subsidy.
The present paper illustrates an alternative approach for quantifying free-riding by combining revealed preference data with cost estimates derived from engineering calculations. Our method is similar to Cameron's (1985) in that nests are imposed to capture correlation of the utility across alternatives, but, rather than using the nested logit model, we employ an analog thereof that involves the specification of an error-components structure (Brownstone and Train 1999). We additionally allow for heterogeneous preferences by specifying household specific random parameters, closely following Revelt and Train's (1998) analysis of the Willingness to Pay for lower operating costs of household appliances. Our investigation uncovers a potential free-rider share of up to 50% of the sampled households, substantially lower than Malm's (1996) estimates but still sufficiently high to warrant scrutiny of financial support for renovations.
The remainder of the paper is outlined as follows. After a brief description of the data, Section 2 discusses the challenges of accommodating unobserved heterogenity in a discrete choice framework and describes alternative models derived from random utility theory for addressing them. Section 3 catalogues the empirical results and uses these to derive household-specific estimates of marginal Willingness to Pay. These results are used to draw policy implications with respect to free-rider problems in the context of Germany's current grants scheme. Section 4 concludes.
2. METHODOLOGY
Our data are drawn from a sample of 2530 single-family home owners, surveyed in 2005 as part of the German Residential Energy Consumption Survey. Four different retrofit measures (and their combinations) are surveyed: roof insulation, facade insulation, windows replacement, and heating-equipment replacement. These measures, along with the option not to undertake a retrofit, yield a total of 16 different combinations from which...
|
