Author ORCID Identifier

Date of Award

Summer 8-2021

Degree Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Dr. Jorge Martinez-Vazquez

Second Advisor

Dr. Alberto Chong

Third Advisor

Dr. Charles Hankla

Fourth Advisor

Dr. Pierre Nguimkeu


This dissertation consists of three essays and fits in the broader literature of how to improve public finances in a context of increasing use of electronic means of payments and how to efficiently allocate scarce public resources to development programs. Analyzing the latter, the first essay evaluates the impact of the 2008-2009 universal primary education policy in Togo on school outcomes while the second essay investigates the effect of the same policy on child labor. Finally, the third essay examines whether the growing reliance on electronic means of payments stimulate public revenues by assessing the impact of plastic money use on value-added tax (VAT) compliance in the European Union (EU).

Entitled “A Tale of a Comprehensive Pro-Poor Education Policy?”, the first chapter sheds light on the causal effects of free primary education policies implemented in quite a few African countries. Using administrative data on the universe of school districts in Togo from 2005-2017, I leverage differences in pre-policy enrollment rates across local areas to identify the impact of the policy. I find that the policy significantly spurred primary school completion and graduation, more so in areas with low baseline enrollment rates. The graduation rate was not affected, suggesting a non-deterioration in education quality. A likely reason is that school openings and teacher hiring increased consistently with the surge in enrollment. Moreover, districts with a rising supply of schools and teachers recorded higher number of final graders and graduates. It also appears that the intervention has not crowded out private school outcomes.

As for the second chapter, titled “Does Free Primary Schooling Curb Child Labor? Evidence from Togo”, inform the empirical literature on the causal effect of a nationwide suppression of public primary school fees on child labor using as case study the 2008-2009 universal primary education policy (UPEP) in Togo. Using data from the 2006 and 2010 waves of the Multiple Indicator Cluster Survey (MICS) in a difference-in-differences approach that compares potential beneficiaries of the UPEP to non-beneficiaries before and after the policy, I find that the policy would curb the propensity to child labor by about 4-5 percentage point. This result is mainly driven by a decline in children’s engagement in non-productive activities. However, the policy has not affected children’s likelihood to perform economic activities. This finding suggests that a free primary education policy is not sufficient to reduce some types of child labor even though it effectively increases school attendance.

Finally, the third chapter analyzes the impact of credit and debit card usage on consumption tax compliance using annual national level data on 26 European Union countries from 2000 to 2016. Exploiting spatiotemporal variation in plastic money use along with an instrumental variables approach, this study finds that a 1% increase in card payments reduces VAT gap by 0.51 percentage point whereas a 1% increase in cash withdrawals increase VAT gap by 0.6 percentage point. This paper’s contribution lies in using more adequate measures of VAT compliance gap and in accounting for potential confounders including ex-ante enforcement capacity of tax administrations.