Much Of Smoke About Nothing


A vast majority of developing countries rely on burning solid fuel for everyday energy needs. The current practices significantly impact climate change and result in a dramatically detrimental impact on health, causing more than two million deaths per year through indoor air pollution (IAP). This paper reviews the work of Hanna, Duflo and Greenstone (2016), who studied an intervention by the Indian NGO Gram Vikas, which promoted and distributed improved stoves among 2575 households in rural, low-income villages over a period of four years, but, in line with similar research in this field, failed to find significant effects. The discussed study shows that stove adoption and proper use were low, likely due to a failure to consider the goals, needs and practices of participants. Study shortcomings are discussed in respect to Goal Framing Theory, Maslow’s Hierarchy of Needs and Self-Determination Theory. Two improvements are suggested, 1. to maximise stove adoption by adapting it to the needs and usage patterns of the target group, and 2. to focus on Social Norms and Self-Perception Theory in promoting the intervention.  An evaluation plan is proposed that emphasies studying participant needs prior to implementation and strongly involving the target group in conducting the study, as well as using more objective measures for data collection. This essay underlines that interventions in developing countries should not be imposed but rather be adapted to the needs and context of the target group.

Keywords: Stove, India, village, rural, indoor air pollution, IAP, goal framing, hierarchy of needs, self-determination theory, social norms, self-perception theory



Much of Smoke About Nothing: Improving Improved Cooking Stove-Interventions


p to 95% of the population in developing countries relies on solid fuel for everyday energy needs (Hanna, Duflo, Greenstone, 2016). The result is a dramatic increase in indoor air pollution (IAP) that causes more than 2 million deaths per year (Bruce, Prez-Padilla, Albalak, 2000), as well as significant contributions to climate change. In response, interventions have endeavoured for more than thirty years to create alternatives to traditional stoves (Hanna et al., 2016). Laboratory studies on improved cooking stoves do show decreases in fuel requirements and IAP, yet results have been underwhelming in real-world settings: Adoption was low (Romieu, et al., 2009), outcomes insignificant (Hanna et al., 2016), or even detrimental to subjects (Beltramo & Levine, 2013). Based on the paper of Hanna et al. (2016), this essay aims to find possible causes for the given results. I will (1) describe the intervention, results, and theory, (2) criticise the intervention based on psychological theory, (3) propose two improvements, and (4) suggest a monitoring and evaluation plan.

The Intervention

Hanna et al. (2016) describe an intervention by the Indian NGO Gram Vikas (GV), which promoted and distributed improved stoves with an enclosed cooking chamber, two pot openings and a chimney via a rollout design among 2575 households in 44 rural, low-income villages. Figure 1 in the appendix depicts a result chain of the intervention. Stoves were distributed in 3 waves over the four years the study was conducted, and data on smoke exposure, health, stove usage and socio-demographic information was collected. The study’s goal was to decrease fuel usage and smoke exposure and thus increase smoke-related health.

Theory of Change

            Figure 2 in the appendix shows the proposed effect of the intervention. The study presumed that provided improved stoves would be adopted and their use and proper maintenance would change both the amount of fuel consumed and the amount of IAP subjects were exposed to. This was expected to result in increased health and decreased greenhouse gas emission.


Results show that the stove was not generally adopted. 70% of treatment-households built an improved stove, but household interest decreased over time and by year 4, 32% of the households had destroyed their stove. Overall, only 36% of treatment households properly maintained the stove and only 19% cooked the majority of their meals on it. On average, only one meal was cooked inside (where the stove was installed), 7.5 meals were cooked outside. The authors report that CO2 breath concentration was significantly reduced for primary cooks in the first year, but only at the α=0.1 level, which is a more liberal testing criteria than the more common testing at an α=0.05 level. The effect turned insignificant in the second year and no other significant effects were found, neither for primary cooks, nor for children. No positive effect was found for health outcomes and greenhouse gas emission. Fuel use was unchanged, cooking time increased and maintenance increased significantly. The results are in line with previous findings on similar interventions (Romieu et al., 2009, Beltramo & Levine, 2013).


Goal framing theory (Lindenberg & Steg, 2007) distinguishes between hedonic-, gain- and normative goals, which can be compatible, but oftentimes are not. In the given study, the improved stove was installed in the wrong place, not used properly and ultimately destroyed. Participants likely did not find the stoves useful, but inconvenient. As data shows, the new stoves made both cooking and maintenance more complicated and therefore the stoves were not compatible with participant’s hedonic- and gain goals, As health concerns were not a priority (Hanna et al., 2016), they were not part of the gain goals, and the normative goal of reducing IAP was likely outweighed by hedonic goals of convenience and ease, and gain goals of practicality, which resulted in a return to the old stove.

Interestingly, research thus far has not investigated the needs, goals and priorities of low income, rural villages, and the little research that does brush this topic finds improved stoves to be drastically outranked by more urgent needs such as sanitation, education, and electricity (Mobarak, Dwivedi, Bailis, Hildemann, Miller, 2012). As Hanna et al. (2016) report children in the given sample to score well below malnourishment, it is save to assume that global warming and IAP were not a priority concern for the participants. This relates to the hierarchy of needs as proposed by Maslow (1943), which states that most fundamental physiological needs such as food need to be secured before individuals can transcend to higher order needs such as fighting air pollution or global warming. Researchers could have investigated those needs. Instead, they imposed their stoves on participants. The stove benefits prioritized by researchers likely did not overlap with the priorities, goals and needs of the subjects (Mobarak et al., 2012, Romieu, et al., 2009). Self-Determination Theory (Ryan & Deci, 2000) proposes a basic need for competence, autonomy and relatedness, yet this intervention did not include or ask participants for their competent insights but imposed upon them and was enacted by strangers rather than the relationally highly-valued collectivistic ingroup (Markus & Kitayama, 1991). As a result, participants likely did not adopt the stoves based on a strong intrinsic motivation to use them, i.e. inherent interest and enjoyment, but rather on an extrinsic, superficial motivation that was based on compliance (Ryan & Deci, 2000).

Improving the intervention

Research in this field indicates that especially low-income populations (Kremer, Leino, Miguel, Zwane, 2011) do not consider the health-threats associated with traditional stoves. As Mobarak and colleagues (2012) conclude, basing interventions on educating about the health hazards is thus unlikely to change existing cooking practices. In line with Beltramo and Levine (2013), who emphasise that stoves should be designed based on the needs of its users, I thus propose two improvements; (1) To maximise usefulness of the stove by adapting it to the specific usage patterns and needs of the target group, and (2) to not endeavour to increase appraisal by providing information, but by using social influence.

Based on Goal Framing Theory (Lindenberg & Steg, 2007), I propose to increase convenience and practicality of the stove and thus accommodate for hedonic and gain goals, by basing the stove design on traditional stoves and thus not requiring users to change usage habits. For the particular target group, which cooks outside, I also propose to build the stove outside rather than inside. As Hansen, Postmes, Tovote, and Bos (2014) indicate, it is the interaction with others that brings change, not the product itself. Interventions must thus not be imposed but rise from within and be owned by the target group (Kelly, 1999), and we need to involve the whole village, especially in a sample like ours, that has a collectivistic background (Markus & Kitayama, 1991), to increase stove-adoption.

Placing the stove outside will not only facilitate gain goals, it will also enhance social norms. Goldstein and Mortensen (2012) argue that, when uncertain, e.g. about a new stove, individuals attend to their social environment for cues. The focus theory of normative conduct (Goldstein & Mortensen, 2012) distinguishes between descriptive (how things are) and injunctive (how things should be) norms, and each norm influences individual behaviour based on its salience. By placing stoves outsides, individuals consistently observe their peers cooking with the improved stove and are thus exposed to a new descriptive norm, which strengthens stove adoption. This was not the case in the study by Hanna et al. (2016), in which each household cooked inside, and as such, no norms could be observed.

All these approaches are based on the premise that the intervention indeed is owned by the community, as Kelly (1999) emphasises. To facilitate this, researchers need to involve villages to a maximum extent. Pretests should determine the needs and expectations of potential participants, and agreements should be reached with village leaders, not only to implement the study, but to gain their support. The intervention should be promoted, stoves be built and training provided by the village itself, not by outgroup-researchers. Romieu and colleagues (2009) argue that stove interventions require repeated support and training throughout the intervention. This further emphasises the need for a community-driven intervention. To increase the appeal of the stove and gain community support, it is thus important to increase intrinsic motivation. In line with Goldstein and Mortensen (2012), I thus propose to advertise the intervention differently: Researchers should not ask for permission to conduct a study, but depict a high demand in the improved stove that has resulted in a shortage. Villages are given the chance to apply for the stoves and are asked to provide reasons for why they need the improved stove, thus increasing intrinsic motivation (Ryan & Deci, 2000).

Stoves will still be distributed randomly, but receiving villages will believe to have been chosen because their intrinsic need for the stove was higher than the need of other villages. Lastly, villages should be asked to replace their traditional stoves with the improved stove rather than adding a second stove. Based on self-perception Theory, which proposes that individuals derive personal attitudes from their own behaviour (Bem, 1967), participants who applied for the stove and gave up on their old one are likely to conclude that the new stove must be important to them, and thus will likely value it more and exert more efforts on maintaining it.

Monitoring and Evaluation

Prior to the intervention, a survey should gather information on needs, goals and priorities of representative villagers and their expectations in an improved stove. Based on these findings, the intervention may need to be adjusted or a different sample must be chosen (e.g. one that cooks outside). To evaluate the effectiveness of the intervention, a randomized controlled field trial should be conducted. Hanna et al. (2016) used a randomized rollout design within villages, but in a collectivistic sample such as ours, spill-over (e.g. the treatment group sharing the stove with the control group) is likely. We should tackle this issue by selecting approximately 40 villages of similar income and cultural background, asking villages to apply for the stove (see above), and then randomly assigning stoves per village, rather than per household, to stove-treatment and no-stove control-conditions. At the end of the study, stoves will also be distributed among control-villages. Data will be collected by village leaders under supervision of researchers prior to implementation (baseline) and subsequently every 3 months for the first year, and once more after the second year, to investigate long-term effects.

Smoke exposure, health, and fuel usage will be measured. Smoke exposure and health data are assessed with a Micro Medical CO Monitor and self-reports respectively. Fuel and usage will be self-reported, and, to reduce desirability bias, a stove use monitor (Ruiz-Mercado, Lam, Canuz, Davila, Smith. 2008) should be installed in addition, as previously used by Beltramo and Levine (2013). Self-reports and monitor measures can then be compared. Maintenance will be organised by village leaders and frequency of repairs reported to researchers. Improved stoves have a history of insignificant results (Hanna et al., 2016). The success of my proposed improvements is thus uncertain. However, social norms have been found to be very effective (Collins & Spelman, 2013), and adapting stoves to subject needs will likely increase adoption. I thus expect an increase in adoption and proper use, and a significant decrease in IAP and increase in reported health.


In this essay, I endeavoured to improve upon the research of Hanna et al. (2016). I highlighted the importance of studying the needs and goals of the target group prior to implementing interventions, and the necessity of involving the target group in the intervention. By changing stove properties, using norms, strengthening intrinsic goals, and utilising self-perception theory, the intervention has the potential to be more successful than its predecessors. The results could establish a new approach to stove-interventions and provide first evidence for the power of firm collaboration with villages on developing effective interventions.


Beltramo, T., Levine, D. I. (2013). The effect of solar ovens on fuel use, emissions and health: results from a randomised controlled trial, Journal of Development   Effectiveness, 5:2, 178-207.

Bem, D. J. (1967). Self-perception: An alternative interpretation of cognitive dissonance phenomena. Psychological Review, 74(3), 183-200. doi:10.1037/h0024835

Bruce, N, Perez-Padilla, R., Albalak, R. (2000). Indoor air pollution in developing countries: A major environmental and public health challenge. Bulletin of the World Health Organization, 78 (9): 1078–92.

Collins, S. E., & Spelman, P. J. (2013). Associations of descriptive and reflective injunctive norms with risky college drinking. Psychology Of Addictive Behaviors, 27(4), 1175-doi:10.1037/a0032828

Goldstein N, Mortensen C. Social norms: A how-to (and how-not-to) guide. Six degrees of social influence: Science, application, and the psychology of Robert Cialdini [e-book]. New York, NY, US: Oxford University Press; 2012:68-78.

Hanna, R., Duflo, E., Greenstone, M. (2016). “Up in Smoke: The Influence of Household Behavior on the Long-Run Impact of Improved Cooking Stoves.” American Economic Journal: Economic Policy, 8(1): 80-114.

Hansen, N., Postmes, T., Tovote, K. A., & Bos, A. (2014). How modernization instigates social change: Laptop usage as a driver of cultural value change and gender equality in a developing country. Journal Of Cross-Cultural Psychology, 45(8), 1229-1248.

Kelly, J. A. (1999). Community-level interventions are needed to prevent new HIV infections. American Journal of Public Health, 89(3), 299-301.

Kremer, M., Leino, J., Miguel, E., Zwane, A. P. (2011). Spring Cleaning: Rural Water Impacts, Valuation, and Property Rights Institutions. The Quarterly Journal of Economics, 126(1), 145-205.

Lindenberg, S., & Steg, L. (2007). Normative, gain and hedonic goal frames guiding environmental behavior. Journal of Social Issues, 63(1), 117-137. DOI: 10.1111/j.1540-4560.2007.00499.x

Romieu, I,, Riojas-Rodríguez, H., Marrón-Mares, A. T., Schilmann, A., Perez-Padilla, R., Masera, O. 2009. “Improved Biomass Stove Intervention in Rural Mexico: Impact on the Respiratory Health of Women.” American Journal of Respiratory Critical Care   Medicine 180 (7): 649–56.

Ruiz-Mercado, I., N. Lam, E. Canuz, G. Davila, and K. Smith. 2008. “Low-Cost Temperature Loggers as Stove Use Monitors (SUMs).” Boiling Point 55: 16–19.

Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68-78.

Markus, H. R., Kitayama, S. (1991). Culture and the Self: Implications for Cognition, Emotion, and Motivation. Psychological Review, 98(2), 224-253.

Maslow, A.H. (1943). “A theory of human motivation”. Psychological Review. 50 (4): 370–doi:10.1037/h0054346

Mobarak, A. M., Dwivedi, P., Bailis, R., Hildemann, L., Miller. G. (2012). Low Demand for Nontraditional Cookstove Technologies. Proceedings of the National Academy of Sciences 109 (27): 10815–20.