We are estimating the effect of a probabilistic incentive (essentially a lottery ticket) on household willingness to request and retain receipts and formal tax invoices for their retail purchases.
Intervention
We randomize whether a financial incentive is provided to households in Lusaka, where the expected value of the incentive is increasing in the value of formal VAT invoices that they retain from retail purchases made over a two-week period.
Experimental design
This is a household-level RCT implemented in one residential area of Lusaka, Zambia. Treated households are provided with a financial incentive to retain formal VAT invoices from their retail purchases over a two-week period, control households receive no incentive.
Design details
Randomization method
Randomization is embedded in the survey, and will be completed by the survey software at the time of the baseline survey and enrollment.
Randomization unit
Household.
Clusters
We are targeting a minimum sample size of 400 (expected 200 in treatment and 200 in control), but will increase the sample size above that if the observed enrollment rate is high enough.
Observations
We are targeting a minimum sample size of 400 (expected 200 in treatment and 200 in control), but will increase the sample size above that if the observed enrollment rate is high enough.
Arms
No clustering.
Primary outcomes
The primary outcome is the total value of purchases made by the household over two weeks that are reflected on machine-printed VAT invoices (formal receipts from VAT-registered sellers).
Secondary outcomes
The first secondary outcome is the total value of purchases made by the household over two weeks that are reflected on any receipt or invoice (not limited to machine-printed VAT invoices).
The second secondary outcome is the household's self-reported tendency to inquire about the VAT registration status of sellers before competing a purchase.
The third secondary outcome is the average value of monthly VAT remittances from firms visited by treated households compared to control households. This will be observed in administrative tax data linked to the invoices provided by respondents.
Cited by
No citation data available for this registration yet.