Bangladesh Chars Tobacco Assessment Project (CTAP) 2018

The first intervention will provide the participants with two different posters, to be hung near their sleeping arrangement such that it is commonly visible. One of the posters will describe the negative health effects on minors of second hand smoke and the smokers themselves. The other poster will detail the harms to pregnant women and the fetus due to the consumption of smokeless tobacco products. Contrary to visuals on local cigarette packets, none of the posters will feature graphic content but rather will be designed using localized graphic styles such that it is more reflective of their lifestyle, and also contain appealing taglines, highlighting the primary message of the visual content. The posters will be designed to act as a psychological nudge through two channels. First, this intervention hopes to encourage conscious acknowledgement of negative impacts of smoking on a daily basis. Second, by highlighting the impacts on pregnant women and children, the intervention aims to motivate smokers to consider the aggregate impacts of smoking due to associated negative externalities even if they discount costs to oneself.

The second intervention will provide the respondents with a journal to keep a record of the amount of tobacco or tobacco related products consumed daily for the intervention period. This journal will require the respondents to record the number of cigarettes, bidis, and the amount of smokeless tobacco, separately, they consume daily. The data from the journals will not be used for analysis, but rather simply constitutes the intervention. This will act as a nudge by requiring a conscious acknowledgement of the individual's degree of tobacco consumption, allowing them to evaluate their own tobacco spending behavior. Moreover, this intervention will target the non-rational discounting of the aggregate effects of individual tobacco consumption events by allowing better recollection of total consumption over the intervention period.

Both interventions do not coerce individuals in any way to reduce tobacco consumption. Rather, they simply alter the incentive regime through cost-effective nudges.

2018-11-01

2018-12-01

Weekly Tobacco Expenditure

Data on the weekly consumption of tobacco will be used to construct the weekly expenditure measure using the appropriate prices. We want to check if this varies between the baseline and end-line due to the intervention.

In order to assess the effectiveness of the interventions described above, two treatment groups and one control group will be formed.Based on a two-stage cluster sampling approach, 24 chars from Gaibandha district will be randomly selected. As we plan on using a Difference-in-Difference (DID) estimator to assess the impact(s) of the treatments, the remoteness of the chars will minimize spillovers. From each of the chars, 42 households will be chosen using simple random sampling. This gives 336 respondents for each group. For each responding household, the head of the household and spouse will be administered the questionnaire.

The baseline questionnaire will follow the standard of the Household Income and Expenditure Survey (HIES) by the Bangladesh Bureau of Statistics (BBS) to collect information regarding tobacco consumption, education, household expenditure, shocks, and other socio-economic conditions. The control groups will be administered the questionnaire only, whereas each treatment group will be informed of the harmful effects of smoking on adults and children and receive the designated treatment. The treatment period will be for 4 weeks following the baseline. A shorter end-line survey will be administered at the end of the treatment period.

Two-stage cluster randomization done using Stata

First stage: Chars (cluster); Second Stage: Households

Yes

24 chars

1,008 households

336 households (8 chars) for each of the control and treatment groups.

Accounting for sample design and clustering, the minimum detectable effect size of the main outcome is 0.28 standard deviations.