Experimental Design
Our study is part of a larger survey, which is conducted among c. 4,500 individuals. Out of these individuals, only those households with natural gas or oil heating systems currently in place will take part in the experiment. In 2022, gas boilers accounted for about 49% of heating systems in German households, while oil was used in about 25% of households, such that we expect the sample for the experiment to contain more than 3,000 of the 4,500 households that participate in the survey.
Our study investigates the effect of the provision of cost information on long-term investment decisions for heating systems of private households. Additionally, we assess how knowledge about the uncertainty associated with cost information influences the investment decision. While we primarily study investment decisions between a fossil-fuel based heating system and a heat pump, the findings might be applicable to other investment decisions under uncertainty, where two investment choices differ in terms of their cost profile (lower up-front cost and higher operating costs, and vice versa).
For this purpose, we conduct an experiment among German households using natural gas or oil heating and retrieve their investment choices between two types of heating systems for ten different pairs of up-front investment costs. We include both tenants and homeowners in the experiment and ask them to imagine that the current heating system of their building is damaged and needs to be replaced on short notice. They are presented with two options:
(i) An equivalent replacement of their current, fossil heating system.
(ii) An air-to-water heat pump.
This introductory information to the experiment differs between tenants and homeowners. While for owners it is obvious that they make the decision and investment themselves, this is usually not the case for tenants. Therefore, we ask tenants to imagine that they can advise their landlord on the decision. We also remind them that, as a result of the replacement of the heating system, their annual rent may be increased by up to 8% of investment costs according to German law.
Before receiving any further information, and before making a choice between the two options, respondents are asked to provide an estimate of the anticipated expenses for purchasing and installing the two solutions presented above. They are also asked to indicate the expected annual (owners) and monthly (tenants) heating costs of both options. Subsequently, we randomly assign the study participants to one of three experimental groups: the treatment group ‘Information’, the treatment group ‘Uncertainty’ and the control group.
Next, participants in both the ‘Information’ and the ‘Uncertainty’ group receive information about the average costs associated with the purchase and installation of the two heating alternatives. The information is taken from a study commissioned by the German Federal Ministry for Economic Affairs and Climate Action (BMWK). They further receive the information that, according to this study, an air-to-water heat pump would be the cheaper alternative over the whole lifecycle of the heating system. Respondents are presented with the assumptions about investment and operation costs, useful life, and the annual development of electricity and fossil fuel prices underlying this assessment. In addition, the ‘Uncertainty’ group is informed that the future development of electricity and oil and gas prices is based on projections, such that the numbers stated above are subject to uncertainty. Hence, the actual prices could turn out to be lower or higher, such that the presented savings when investing into a heat pump are not guaranteed. Participants in the control group do not receive further information on the price of the two options or the uncertainty associated with it.
Afterwards, we show the respondents a Multiple Price List, which comprises ten choices between two options: an oil or gas heating system and a heat pump. We ask respondents to assume that all characteristics of the two heating options (apart from the investment costs) are the same. Each option is associated with a price that corresponds to the investment costs, not including the future cost of operation. While the investment costs for the fossil fuel system are held constant at the actual average investment costs as provided by the BMWK, the investment costs for the heat pump are lowered by 10 percentage points in each step, thereby coming closer to, and eventually being lower than, the price of the fossil fuel solution. For each of the ten price pairs, we ask respondents to indicate their choice between fossil fuel system and heat pump. After making their choices, respondents also indicate what aspects they have considered when making their decisions, how reliable they deemed the cost information provided, and whether they have taken the information about the heating systems and their costs into account.
In a later part of the survey, we also elicit the risk and time preferences of respondents as well as their financial literacy, which can be used for heterogeneity analyses. To assess financial literacy, following Lusardi and Mitchell (2011), we ask respondents to solve an arithmetical problem associated with interest rates and compound interest. The exercises to elicit risk and time preferences are incentivised by offering respondents vouchers, the value of which is based on the respondents’ choices in the respective exercises. Elicitation of time preferences is based on Meier and Sprenger (2010): respondents make several choices between a voucher that they receive one month after the survey and a voucher they receive six months after the survey. The latter voucher always has a value of 30 euros, while the value of the former voucher varies between 29 and 9 euros. The assessment of risk preferences is based on Holt and Laury (2002). Respondents need to make repeated choices between two lotteries that determine the value of the voucher. Each lottery has two potential outcomes, which are assigned varying probabilities.