Abstract:
This paper contains a micro econometric analysis of household electricity and natural gas demand for a cross section of 2,885 Danish households observed in 1996. The sample includes fulltime employed couples in single family houses. The specification of the model is guided by an explorative nonparametric data analysis. The analysis reveals, among other things, the fairly surprising result that demand for heating is unaffected by the number of children in the household. The dependence between demand for gas and demand for electricity is examined in the paper. This is done by testing for separability of demand for gas from demand for electricity, and vice versa. Separability of electricity (gas) from gas (electricity) is tested by estimating demand for electricity (gas) conditional on demand for gas (electricity). The model allows for endogeneity of the conditional variable. Building regulations and individual time variation, that is panel data, identify the test. The test indicates that demand for electricity is separable from demand for natural gas, and that demand for natural gas is separable from demand for electricity. The result of the test is evidence in favour of single equation modelling of household energy demand in this context.

Abstract:
In this paper we ask if consumption of energy for space heating by households is habit forming. A model of intertemporal consumption allocation allowing for habit-forming preferences is estimated on a register-based panel data set with high quality information about consumption of natural gas for a sample of Danish households. Results indicate that preferences are weakly habit forming.

Abstract:
This paper analyzes the effect of supplying feedback by text messages (SMS) and email about electricity consumption on the level of total household electricity consumption. An experiment was conducted in which 1,452 households were randomly allocated to three experimental groups and two control groups. Feedback was supplied throughout 2007 to members of the experiment groups who accepted the invitation, and data on consumption of electricity for 2006 and 2007 collected for all participants and control group members. 30% of the households invited to receive feedback accepted the invitation. Results suggest that email and SMS messaging that communicated timely information about a household�s �exceptional� consumption periods (e.g. highest week of electricity use in past quarter) produced average reductions in total annual electricity use of about 3%. The feedback technology is cheap to implement and therefore likely to be cost-effective.