#StayAtHome, but #SaveElectricity — behavioral experimentation continues in the era of COVID-19
Insights on the 2019 and 2020 behavioral experiments aimed at reducing energy consumption in Moldovan households
During the last few years, behavioral experimentation became a niche for UNDP Moldova and its Innovation Unit. The “architect” of these bold projects is Dumitru Vasilescu, Policy Specialist, who tests with enthusiasm new solutions to existing development challenges.
What does a behavioral experiment mean and when it is worth using it?
For me, the behavioral experiment is a test of a solution that is focused on improving and measuring change in a specific human behavior. In our case this is about testing specific social norms and how these work in the local context. I think this is a niche for UNDP going forward, particularly in a context in which we are increasingly facing complex systemic risks and human behavior (i.e. the way to do things). This is an important aspect that is to be taken while talking some of the complex development challenges.
Why did you conduct a behavioral experiment around electricity consumption in the middle of the pandemic? You have already conducted a similar one in 2019. Why did you come back to it? What is the value added?
We are scaling-up the behavioral experiment conducted in 2019. It means that now we are not randomly selecting a ‘treatment’ group, but we are expanding the previously successful behavioral text to the entire population of ‘less’ efficient households — consumers of electricity. Scaling-up also means that we are sending out behavioral letters to households outside the Municipality of Chisinau and we will be indeed looking on the impact of this letter produced on urban and rural households (there might be differences). We will also be looking into the impact of such letters in 2–3 months after the experiment.
Unfortunately, not so many successful innovation experiments are scaled-up for many reasons, due to scarcity of resources, implementation capacities and so forth.
I am particularly proud that, despite the current crisis, we are able to scale-up the 2019 experiment and continue nudging most “inefficient” households to improve/reduce their consumption of electricity even while staying at home.
What were the reactions of consumers?
Given the measured impact of the 2019 experiment (i.e. a reducing of 2%), we were indeed expecting an overall decrease in electricity use in July 2020 of same size. Together with Premier Energy we have started to send out the behavioral letters at the end of June 2020 and throughout July 2020; the process was complementary and in tight connecting with sending out the bills for electricity use in same households. Given the current pandemic situation we have seen that the overall consumption increased in June and July as compared to similar period of the 2019.
Consumption in targeted households grew by 5%, whereas the overall increase is/was of about 34%.
We will closely monitor and evaluate the evolution in the consumption and share final conclusions. Overall, the evaluation shows that consumers are well receiving the tailor-made behavioral letters and that social norms continue to work.
What challenges and unexpected turnouts occurred during the experiment?
My view is that COVID-19 delayed the scaling-up of the experiment. As robust impact evaluation was required, with the pandemic, it was difficult to get access to real consumption by households, only an estimated consumption figures was available initially. The estimated consumption figures for April and May were not sufficient and so we decided to postpone the scaling-up to the moment in which real consumption figures became available, so as we are able to evaluate the impact of the scaling-up. No other major impediments were encountered, the process is smooth and well coordinated, while the technical and informational support from Premier Energy is extremely timely and of high quality.
What are the results revealing?
The major output is that the overall increase in the consumption in households that use over 140kWh per months and received behavioral letters is lower that in the general population. We worked with granular data for June and July 2020, having a solid historical data set of previous 60 months to compare with.
Do you consider that the ‘magic effect’ of the letters could stop? What needs to be done to keep the consumers engaged and responsible even in cases when they will no longer receive letters comparing their data to neighbors?
From the 2019 experiment, we’ve see that the effect holds even after the second and the third month.
We have signs then that in the scaling-up process, the effects will be in place for a longer period, but indeed to keep consumers engaged there is need for some amendments in the current legal framework or at least some specific additional incentives for households to behave more responsibly.
How a behavioral experiment that encourages reduction of electricity consumption could translate in a permanent behavior?
The work done by UNDP and Premier Energy in Moldova aims to set a precedent that will be further used systematically and will contribute to the gradual change in consumers’ behavior towards sustainable energy consumption. We understand that this is a long-term process and multiple factors are involved, but we count on both the financial incentive and the environmental one to educate a new type of behavior among customers. Results have proven that even small energy savings benefits the environment.
All it requires is a bit more discipline and attention towards our daily electricity consumption habits. When all that is in place, the savings will become automatic. This is especially relevant in a context in which the operational load on existing networks is growing (particularly in urban settlements), consumers are not particularly efficient, and almost all electricity is being imported from abroad.
The behavioral experiments were conducted by UNDP Moldova in partnership with Premier Energy and funded through the UNDP’s Global Innovation Facility. The experiment was inspired by the Opower program in the United States of America.