This thesis investigates taxpayers' compliance behaviour in response to various nudges
from the tax authority by providing empirical evidence from two large-scale randomised
controlled trials in an emerging economy. This thesis includes two empirical chapters that
address distinct but related issues.
The first empirical chapter investigated the effectiveness of behaviourally informed
letters and the standard Surat Tagihan Pajak – STP (tax penalty bill) on tax penalty
payment compliance through a randomised experiment with over 10,000 individual
taxpayers. Participants received deterrence, information, simplification, or no letters. The
simplification nudge significantly improved compliance, while the deterrence nudge was
less effective, particularly in less-developed regions where taxpayers perceived a higher
likelihood of detection. In contrast, the information letter did not significantly promote
timely tax penalty payments. Despite mentioning penalties for late submissions, an
extended analysis of the potential spillover effect on annual tax returns revealed that the
interventions did not induce increased compliance among taxpayers in their annual tax
return filing. The study recommends that policymakers integrate well-designed nudges
into their systems, considering taxpayers' backgrounds and leveraging insights from
psychology and linguistics to create more effective and accessible forms.
The second empirical chapter analyses the monthly income tax payment behaviour of
over 41,000 self-employed taxpayers who received WhatsApp reminders prior to their
due dates. Taxpayers were assigned to either the early, close, combined, deadline, or a
control group, with no reminders. The results indicate that all reminders enhance tax
compliance, with single reminders sent near the deadline being the most cost-efficient
option, despite yielding a less marginal impact than multiple reminders. The intervention
is more efficacious for taxpayers in more developed regions (e.g., Java) and those who
filed tax returns electronically, underscoring the importance of tax and digital literacy.
We observed no sustained effects beyond the duration of the intervention. This suggests
single reminders may not be sufficient to induce lasting changes in taxpayer behaviour.
The findings of this chapter have several significant implications. Tax authorities should
consider the message content by integrating behavioural economics into the design and
select the most optimal timing and frequency. Utilising personalised digital messaging
can be an efficient choice of communication with taxpayers. Sending messages through
the tax authorities' official accounts is essential to maintain credibility.
The findings have significant implications for behavioural economists, social
scientists, policymakers, and tax authorities. Despite its limitations, it warrants further
research to enhance existing knowledge. Ultimately, this thesis advances tax compliance
nudges in developing countries, building on behavioural economics, public finance, and
development economics.
