Prof. Dr. Frank Krysiak & Joëlle Velvart

Umweltökonomie 

Research Objectives
In the political and societal discussion of climate policies, distributional effects
are at the forefront of the debate. It is often argued that climate policies incur
high costs on households with low income, the rural population, and families.
In Switzerland, such arguments may have been important in the rejection of the
revised CO2 act in 2021. This revision would have introduced the possibility of
higher CO2-taxes on fossil fuels used in stationary installations as well as a new
CO2-tax in the mobility sector. The argument against these changes have been
that households with low and middle incomes as well as people living in the
countryside would be disadvantaged (Abstimmungsbüchlein Volksabstimmung
13. Juni 2021, p. 63).

Given the per-capita reimbursement of the tax revenue as well as many
technological and behavioral options for reducing individual CO2-emissions,
the distributional effects are not obvious. The aim of this research project is
to calculate the actual distributional effects of climate policies. The analyses
include different CO2-prices and redistribution-schemes and are conducted in
multiple steps that cover short-run and long-run responses to climate policies.
The analysis is based on microsimulations using survey data for Swiss households
(SHEDS) and covers CO2-emissions from space heating and mobility.
In particular, the project separates direct distributional effects (caused by
tax payments and the reimbursement), distributional effects after short-run
responses (changing usage behavior, such as less traveling or altered room
temperatures) and distributional effects after some long-run responses (e.g.,
changing the heating system).

Completed Steps
The calculations for this project are based on the Swiss Household Energy
Demand Survey (SHEDS). The survey was developed by researchers from the
Swiss Competence Center for Research in Energy, Society and Transition (SCCER
CREST) and contains information on Swiss peoples’ energy usage in the domains
of electricity, heating and mobility for the years 2016 to 2021. In each year, about
5’000 households participated in the survey. The data is representative for the Swiss
population (excluding Ticino).
As a supplement, we used data from the Household Budget Survey (HABE) from
the Federal Statistical Office. In HABE, detailed information on expenditures and
amounts for different goods as well as household characteristics are listed for the
years 2006 to 2017, providing ca 90’000 observations in total. While HABE provides
much information, it does not have the level of detail that the analysis requires.
Therefore, we could not use the dataset as extensively as we had planned.
The first stage of the analysis is the calculation of the tax basis for the households
represented in SHEDS. Based on SHEDS information about the heat source and
usage and using an online-tool (2000 Watt wohnen), we have calculated the primary
energy consumption for the households for space heating. To get the most accurate
estimates for the households’ heating energy consumption, we relied on information
about the building (e.g., year of construction, renovation of windows) as well as on
the living unit itself (e.g., house or apartment, size of the dwelling). Next, we have
converted the amount of energy a household uses into CO2-emissions, based on the
prevailing heating system in the dwelling (CO2online, energie-umwelt.ch).
For mobility, we have calculated the households’ CO2-emissions resulting from car
usage. We first calculated the fuel consumed based on SHEDS data regarding the
type of a household’s most often used car and the kilometers are driven with this
car per year. With values from the literature (Hecking 2019, Izzi 2022), we have
converted those liter-values into resulting CO2-emissions.

The second stage of the project has consisted of calculating the direct distributional
effects of climate policy. To this end, we have conducted microsimulations to find
the distributional effects for different taxes and redistributions in the heating-area.
To do that, we have defined the following scenarios:
• S0: baseline scenario, using the actual CO2-prices of CHF 96 per ton (2021-level)
and CHF 120 per ton (2022 level) as well as a per capita redistribution of CHF 87
(2021) and CHF 88.20 (2022), respectively
• S1: rejected CO2-law scenario, tax of CHF 210 per ton of CO2 with a per capita
redistribution using 2/3 of the generated tax-revenue
• S2: high-cost scenario, CO2-price of CHF 600 per ton and per capita redistribution
using 2/3 of the generated tax-revenue

To include the mobility sector, the costs that the consumers are charged at the gas
stations are relevant. Fuel-importers have to compensate for pollution resulting
from fuel-usage. Under the current law, they can pass on max. 5 Rp. per liter of
gasoline or diesel to consumers. This value would have been increased to 12 Rp.
per liter under the rejected CO2-law. Thus, the additional financial burden for
households would have increased by 7 Rp. per liter of car fuel without having any
impact on the amount redistributed over the Swiss population.
With this data, we have conducted a first analysis, assuming that people do not
adapt their consumption behavior to the higher prices in the mobility or the
heating sector.

In the third project stage, we have introduced price-elasticities for short-run
adjustments of usage behavior, which were taken from the literature and (as far as
was possible) from an analysis of the HABE data. We have distinguished elasticities
for home owners and tenants to account for the restricted influence tenants have
on their heating demand compared to owners. Using this description of behavioral
change, we have repeated the microsimulations.

The final stage of the project has aimed at covering long-run adjustments, that is,
changes in investment patterns in space heating and mobility. Unfortunately, the
available data was not sufficient to estimate long-run elasticities. Therefore (and in
light of the results gained so far, see below), we have switched towards verifying our
results by introducing long-run adjustments in the microsimulations that capture
changes in long-run behavior that amplify distributional effects.

Results
The results of all stages show that the climate policies in the scenarios S1 and even S2
do not induce substantial distributional changes, neither between different income
groups nor between the rural and urban population. Of course, the additional taxes
and the higher reimbursement results in some redistribution. But, in the more
plausible scenario S1, this effect is small compared to the average income level
and, in both scenarios, the effect follows none of the patters (“poor vs rich”, “city
vs countryside”) that are commonly used in the political and public discourse. In
particular, we have not found a robust pattern of redistribution that would provide
a sound basis for an argument against a tax-based climate policy.
Given that several studies show that other policies (such as technology mandates or
subsidies) are likely to cause substantial efficiency losses, our results indicate that
the emphasis on distributional consequences of Swiss climate policy is misplaced.

Publikationen
A working paper and later a publication is planned to result from the project.