Economics serving society
Justine Knebelmann

Justine Knebelmann

Phd Student

Paris School of Economics, EHESS

Campus Jourdan – 48, boulevard Jourdan, 75014 Paris

5th floor, office 32

Phone +33(0)1 80 52 13 53

Thesis Supervisor: COGNEAU Denis

Academic year of registration: 2017/2018

Thesis title: Fiscal Capacity and State Construction in Sub-Saharan Africa


  • Bringing property owners into the tax net: avenues of fiscal capacity and local accountability. Evidence from Dakar, Senegal with Victor Pouliquen and Bassirou Sarr

Ongoing field experiment. 

Property taxes are levied for local governments and often represent an important component in their budget funding. As such they play a crucial role in the face of increasing needs for public services in rapidly growing cities. In the context of developing countries, with cadaster shortcomings, weak administrative information and IT systems and poor enforcement tools, most local administra- tions experience substantial shortfall in property tax revenues. We partnered with the Senegalese tax administration (Direction Générale des Impôts et des Domaines) to develop a new property tax man- agement system in Dakar, including an intensive fiscal census, a new data collection and management application, and the incorporation of modernized cadastral information. It’s implementation through a randomized controlled trial will shed light on four questions: i) the extent and mechanisms by which this administrative investment increases fiscal capacity; ii) the respective advantages of a rule compared to discretion in the assessment of tax liability by tax officials; iii) the effects of increased local taxation on local governance dynamics and on the activities of neighborhood chiefs; iv) the extent to which the tax burden is shifted to tenants .



  • Strategic or Confused Firms? Evidence from “Missing” Transactions in Uganda with Miguel Almunia, Jonas Hjort, and Lin Tian

Working paper

Are firms sophisticated maximizers, or do they consistently make errors? We study this question in Uganda. We show that sellers and buyers report different amounts in 79 percent of transactions subject to value-added tax (VAT), despite invoices being easily comparable. We estimate that 29 percent of firms misreport own sales and purchases such that their liability increases. However, 71 percent are self-advantageous misreporters. Only such firms misreport less when exposed to tighter enforcement (at customs, where exchange rate-variation-induced imports pass through). Despite the gain from firm errors, overall, unilateral VAT misreporting cost Uganda USD 446 million in revenue from 2013-2016.


  • Information, fiscal capacity and tax enforcement: an experimental evaluation with Miguel Almunia, Jonas Hjort and Lin Tian

We conduct a randomized controlled trial to study the impact of a tax compliance
program on the behavior of VAT-registered firms in Uganda. A subset of seller-buyer
pairs are randomly assigned to receive a letter notifying them that the revenue authority
has found discrepancies in their VAT declarations and that they are requested to file
amended returns. We find a large increase in the amendment of past returns and
a significant reduction in measures of tax misreporting in the months following the
intervention. However, we do not find significant effects on declared tax liabilities.
The randomization procedure is designed to capture spillover effects on the trading
partners of treated firms. We find significant, although short-lived, spillover effects
on the amendment rates and tax misreporting of trading partners going in the same
direction as the direct effects.


  • Taxation in Africa from colonial times to present with Denis Cogneau, Yannick Dupraz and Sandrine Mesplé-Somps

 This project aims to shed light on the trajectories of African states, in terms of taxation and expenditure, from colonial times to the present. Building on previous work focused on colonial public finance in former French colonies (Cogneau, Dupraz and Mesplé-Somps 2018), we compile a novel dataset by combining and comparing previously available data with recently digitalized data from historical archives (“Zone Franc” and BCEAO reports), to produce continuous and comparable taxation data series, and thus contribute to filling the data gap on African public finances for the 1960-1980 period. This allows us to study the evolutions of the level and composition of tax revenues during post-independence decades, with a special focus on the critical juncture of independence, on socialist experiences, on the impact of commodity bonanzas, and on the consequences of structural adjustment policies, for a set of “Francophone” countries and a few “Anglophone” comparators.


  • Natural Resources' Impact on Government Revenues

UNU-WIDER Working Paper 2017/10

Motivated by the fact that the taxation of natural resources is both crucial and particularly challenging for developing countries, this paper draws on a unique dataset to produce empirical evidence on two issues pertaining to the fiscal impact of oil. On a sample of 31 countries during the 2000s oil price boom, we first assess which country and sector characteristics are correlated with the effective tax on oil, i.e. the share of oil income collected by the government. Secondly, we test whether oil revenue evicts traditional tax revenues. We propose a new methodology to address this question and we conclude to the absence of such an eviction effect: we observe no effect of oil revenue on non-oil taxes through taxation channels, and linkages with the non-oil economy seem to yield additional non-oil tax revenues. These econometric analyses are complemented by six comparative case studies of countries observed before and after oil production begins. Historical, institutional and oil sector-specific information allows to account for differences observed in the evolution of the effective tax on oil and of non-oil taxes.