Abstract
Untapped improvements in energy efficiency in the residential sector may deliver large savings in energy use and the CO2 associated emissions. Yet empirical assessments have been difficult and controversial.
We collect monthly natural gas meter readings from a sample of homes in Transcarpathia, in Western Ukraine, an early adopter of the country's trend away from district heating, from January 2013 to April 2017, a period over which the residential natural gas tariffs rose by over 700%. We combine the monthly meter readings with documentation about each household's heating-related energy efficiency upgrades to the home (wall, attic or basement insulation; new windows; boiler replacement, and insulation around pipes) to form a panel dataset.
We estimate the effect of the energy efficiency renovations on natural gas consumption, controlling for weather, income and government energy assistance. The decision to do the renovations and natural gas consumption are likely endogenous (people do the renovations because they hope to consume less), so we instrument for the renovations by creating a cross-validation instrument based on a supply-side argument.
Even for a given type of energy efficiency upgrades, the estimated effect of the renovations varies dramatically in magnitude, depending on whether the renovations are instrumented for and on how detailed the fixed effects are. The coefficients on the renovations are almost always negative in our regressions, but practically and statistically significant only when we instrument for the renovations.
This is in agreement with our respondents' difficulty assessing whether the renovations had saved them gas or money. The IV estimates indicate that insulation delivers 13-24% reductions in natural gas usage, and up to a 5% internal rate of return (IRR) to the investment over 20 years.
Judicious use of an existing government program can yield positive IRRs and make energy efficiency upgrades a good investment in a generally poor-performing housing market.