Summary: |
This paper investigates the effect of housing companies’ firm size on the outcome of energetic refurbishment. We argue that economies of scale, economies of scope and effects of learning should have an impact on the production of energy efficient housing services. To test our hypothesis, we use a unique dataset derived from energy consumption bills. The data contains information on the exact energy requirement of a dwelling, its size, age and the level of refurbishment. In particular, we rely on information on roughly 100,000 multifamily houses. Besides owner characteristics and refurbishment effort, we introduce several control variables to capture vintage, size and spatial effects. We find strong evidence for the presence of firm-specific, in particular size effects on the energetic outcome of refurbishment. We find that professional housing companies (a company owning more than 1000 flats in their portfolio) reduce the energy requirement by about 37.1% when major parts (roof, facade, windows, basement ceiling and heating system) of the dwelling were refurbished within the last 15 years. In contrast, single unit owner (owning up to 20 flats) reduce energy requirements only by 12.3 %, even when for dwelling attributes like size, age and spatial location is controlled for. Moreover, differences between firm types increase with refurbishment effort. This inquiry contributes to the literature on housing service supply as well as to the ongoing political debate on the reduction of CO2 emissions. For example, the introduction of Germanys’ federal governmental concept for climate protection in autumn 2010 let to a fierce controversy on the feasibility of the introduced targets on energy efficiency improvements in the existing housing stock in particular because of the plans to introduce redevelopment obligations for all landlords. After intense protests, this strategy was withdrawn, predominantly because the law did not distinguish between real estate characteristics. The present paper adds the aspect of firm size to the discussion. |