Household electricity consumption accounts for 40% of global energy related CO2 emissions and this is expected to grow globally by 58% by 2030 if no measures are introduced (Rasul & Hollywood 2012). Residential energy consumption is also made worse by a supply-demand mismatch in developing countries.
Bangladesh is rated the 6th most vulnerable nation to climate change in the 2015 Global Climate Risk Index. Not only are there supply-side constraints, but also deep inequality in electricity access where a large number of households do not have adequate access and face blackouts in the summer. Setting a carbon price or emission trading scheme is not politically feasible and subsidies to encourage energy efficient technology are costly. That leaves us with considering non-price energy conservation schemes.
Among non-price mechanisms, there are two alternative approaches to motivate individuals effectively. The first and more traditional approach is through information campaigns, moral exhortation, and fear inducing messages. A second approach that has emerged as an alternative is social norms marketing that acts as an exogenous factor in order to diffuse certain behaviours (Donaldson et al. 1995, Schultz et al. 2007).
We use a randomised controlled field experiment to examine the relative effectiveness of information in influencing residential energy consumption in Bangladesh. We assess the role of information on energy consumption by testing the effects of the following information:
- Expert advice on electricity conservation
- Average electricity consumption of others in the city
- Own electricity consumption relative to neighbours.
To our knowledge, this is the first RCT on energy saving tips and information provision using social norms in a developing country. The experimental design targets those households for the project that have air-conditioners at home. We first identify the 2,400 households with air-conditioners in three cities in Bangladesh – Dhaka, Khulna, and Jessore. We then randomly assign the 2,400 households into four groups. Three of the four groups are treated with the aforementioned three different treatments, and the remaining group is the control. The sample size in each group is almost equally divided. Randomisation took place at the household level.
Starting on 4 April, 2017, we provided three treatment groups with the corresponding interventions in the months of April-May, May-June, and August. The final round collected data in October without any intervention provided. The first treatment group received information related to the production and consumption of electricity and its relation to the climate change shown in a flow chart. The enumerators visited the households once every month from April-May, June, and August to provide and explain the information using a flow chart. This group was also provided with the 10-point tips, with advice from the city electricity distribution company experts, to conserve electricity and hence reduce expenditure on electricity use. In addition, for the information to be reminded and emphasised to the households, most importantly six of 10 tips were provided in the form of stickers and displayed at obvious places in the house.
In each of the months mentioned above, the enumerators also visited the second treatment group to provide a flyer to each household. Households are divided into three sub-groups based on their electricity usage compared to that of the city’s average household.
The three sub-groups were:
- Good/efficient users whose electricity bills are one standard deviation lower than the average household's bill
- Bad/inefficient users whose electricity bill is one standard deviation higher than the average household's bill
- Average users whose electricity bills are within one standard deviation of the average household's bill.
The good/efficient households were provided flyers with a norm sign which indicates appreciation for their good use, while the bad users were provided with flyers with a thumb-down sign indicating dis-appreciation for their extravagant electricity use. The average users were provided with flyers showing the city average. Each sub-group was explained about their standing of electricity use compared to the city average households' use of electricity.
The last treatment differs from the second one mainly in terms how comparison is made. Instead of city-wide average dweller, a household electricity bill was compared with a neighbour's electricity bill. If a household's bill is above the neighbour's bill, we posted a sad sign, but if the bill is below we posted a smile sign and so on. Surveys before, during and after the field experiment have and will further generate data on values, attitudes, and household characteristics of the participants and will also help examine the impact on awareness of energy conservation potentials.
The strongest effects were observed for households that were given knowledge and tips treatment. The reduction in electricity consumption gets larger over time – there is reduction of 8% after two months, followed by reduction of 11% after four months and further onto reduction of 14% after seven months.
Providing advice on saving energy could thus potentially reduce households’ energy consumption significantly. Results remain similar for neighbour versus suburb comparisons, although the effect is generally stronger for above-average users. The main policy implication is that information could reduce electricity consumption when feedback is repeated and frequent.
Achieving energy efficiency through non-price energy conservation programmes is an appealing approach to combatting the threats that climate change poses, given its cost effectiveness and political feasibility (Allcott 2011). Testing the effects of non-price energy conservation strategies on residential electricity usage in a developing country context will provide important insights and policy implications. If found effective, the proposed approaches to reduce household energy use may bring the electricity supply-demand gap down in countries such as Bangladesh.
This piece draws from as well as updates an entry on ResearchGate© by the same authors on 21st October, 2017.
Editor’s Note: This blog is part of the IGC’s 10 year celebration series. This blog is linked to our work on Energy for all.
Allcott (2011), “Social norms and energy conservation”, Journal of Public Economics, 95: 1082-1095.
Ahsanuzzaman, A, L Wang and A Islam (2017), “Non-price energy conservation programme and household energy consumption in Bangladesh”, American Economic Association, https://www.socialscienceregistry.org/trials/2367
Ahsanuzzaman, A, L Wang and A Islam (2017), “Non-price energy conservation programme and household energy consumption in Bangladesh”, Research Gate,
Donaldson et al. (1995)
Rasal & Hollywood (2012)
Schultz, P, J Nolan, R Cialdini, N Goldstein and V Griskevicius (2007), “The constructive, destructive, and reconstructive power of social norms”, Association of Psychological Science, 18(5): 429-434