Policy Research Working Paper                    10182




Increasing Tax Collection in African Countries
              The Role of Information Technology

                             Oyebola Okunogbe
                              Fabrizio Santoro




Development Economics
Development Research Group
September 2022
Policy Research Working Paper 10182


  Abstract
 Many African countries struggle to collect an adequate                             income taxes. It then describes the ways in which technol-
 amount of tax revenue to support needed investments in                             ogy solutions may be deployed to address these challenges
 public services. This paper examines how African coun-                             by helping to identify the tax base, monitor compliance,
 tries may take advantage of recent advances in technology                          and facilitate compliance. Lastly, it provides insights from
 to improve tax administration. It provides an overview of                          interviews with senior tax administrators across the conti-
 the potential and challenges of different tax categories in                        nent on their practical experiences in adopting technology
 Africa: consumption taxes, real estate taxes, trade taxes, and                     for taxation.




 This paper is a product of the Development Research Group, Development Economics. It is part of a larger effort by the
 World Bank to provide open access to its research and make a contribution to development policy discussions around the
 world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may
 be contacted at ookunogbe@worldbank.org.




         The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development
         issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the
         names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those
         of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and
         its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.


                                                       Produced by the Research Support Team
       Increasing Tax Collection in African Countries: The
                 Role of Information Technology∗
                        Oyebola Okunogbe, World Bank
        Fabrizio Santoro, International Centre for Tax and Development




Keywords: Africa, public finance, technology, taxation, tax administration, tax compliance
JEL codes: H20, O33




   ∗
     Okunogbe ookunogbe@worldbank.org; Santoro F.Santoro@ids.ac.uk; We thank Moyo Arewa, Christine
Awiti, Kathleen Beegle, Anne Brockmeyer, Matthew Collin, Justine Knebelmann, Giulia Mascagni, Laura
Munoz, Giovanni Occhiali, Dina Pomeranz and Abebe Shimeles for helpful comments. This paper was
supported by the African Economic Research Consortium (AERC) and an early version of this paper was
presented at the December and June 2021 AERC Plenary under the theme “COVID-19 and Public Finance
in Africa: Challenges and Opportunities.” Santoro acknowledges support from the Gates Foundation through
the International Center for Tax and Development DIGITAX program. Kairat Umargaliyev provided in-
valuable research assistance. The findings, interpretations, and conclusions expressed in this paper are those
of the authors and do not necessarily represent the views of the World Bank Group, its Executive Directors
or the countries they represent.
1     Introduction

Taxes are important for economic development. Taxation lies at the crucial intersection
between the provision of public goods, income redistribution, social safety nets, and govern-
ment accountability. Tax revenues also allow countries to be less reliant on foreign aid and
natural resource revenues. However, many African countries struggle to collect an adequate
amount of taxes.
    This paper examines how African countries may use recent advances in technology to
improve tax administration. While digital transformation across different sectors has been
occurring over the last few decades, the surge in remote work experiences and the need for
social distancing precipitated by COVID-19 has accelerated the adoption of many technolo-
gies that already existed and spurred the development of others. As different sectors of
government adopt technology solutions to increase efficiency and improve service delivery to
citizens, tax administration presents an important application.
    The paper begins with an overview of the challenges faced by African economies in
maximizing key tax categories: consumption taxes, real estate taxes, trade taxes, and income
taxes. It then describes the ways in which technology may be deployed to address these
challenges. In particular, it applies the framework in Okunogbe and Santoro (forthcoming)
to each of these tax categories to examine how technology may be used to define compliance,
that is to identify the tax base; to monitor compliance, that is, to detect evasion when it
occurs; and to facilitate compliance, that is, to make it easier for people to comply with
their tax obligations. It combines an overview of the literature with case studies highlighting
specific country examples. As a unique feature, the paper concludes with insights from
interviews with senior tax administrators across the continent on their practical experiences
and challenges in adopting technology for tax administration.




                                              2
1.1      Features of Taxation in Africa

Tax collection in Africa is low, but similar to other regions at a similar income level (Figure 1).
In 2018, the most recent year with wide data coverage, Sub-Saharan African countries col-
lected 14 percent in taxes as a share of GDP (UNU-WIDER Government Revenue Dataset,
2021). This continent-wide average masks significant variation across countries. High- and
upper-middle-income countries like Seychelles, Namibia and South Africa have rates as high
as 28 - 33 percent whereas low-income countries like Chad, Democratic Republic of Congo
and Ethiopia have rates as low as 7 percent. These numbers have remained stagnant over
the last 3 decades, with African countries collecting an average of 12 - 15 percent of GDP as
taxes from 1990 to 2020. Both the South Asia region and the Middle East and North Africa
region have similarly low rates as Sub-Saharan Africa of about 14 percent. In contrast, the
Europe and Central Asia region has the highest rate of 32 percent (UNU-WIDER Govern-
ment Revenue Dataset, 2021). Across all countries, on average, higher income countries
collect a higher share of GDP as taxes.
      The composition of taxes in Africa is also roughly similar to that of countries at similar
income levels. Figure 2 presents evidence on the reliance of countries on different tax instru-
ments: taxes on consumption, personal income, corporations, payroll, trade and property.
As in almost all regions, the greatest component of taxes in Sub-Saharan Africa is consump-
tion taxes (49 percent), which consist of sales tax, value added taxes and excise taxes. Some
striking differences of African countries, even compared to countries at a similar income level,
are that African countries collect a minimal amount in payroll taxes, possibly a reflection of
the low share of the economy employed in the formal sector.1 In contrast, African countries
collect more trade taxes than countries at a similar income level.
      With regards to tax rates, the tax rate on income, profits and capital gains in African
countries range from 3 percent to 48 percent. On average, African countries have a higher
  1
   According to Medina and Schneider (2018), the shadow economy amounted to about 40 percent of
Sub-Saharan Africa’s GDP in 1991-2015, compared to 17 percent of OECD countries’ GDP.




                                                3
tax rate on income, profits and capital gains, compared to other countries, even those at
similar income levels (Figure 3a). However, Value Added Tax (VAT) rates, which range
from 5 percent to 20 percent across the continent, are comparable to other countries with
an average rate of 15 percent (Figure 3b).
      Taxpayers in Africa face higher tax compliance costs, even compared to countries at
similar income levels. Figure 4 presents how much time, measured in hours per year, firms
spend preparing and paying three major types of taxes: the corporate income tax, the value
added or sales tax, and labor taxes, including payroll taxes and social security contributions.
African countries spend on average 273 hours per year on taxes, a greater burden than the
average of 220 hours in other regions (World Development Indicators, 2018).2
      Tax administration patterns in African countries feature lower reliance on technology
and greater reliance on manual systems and in-person interactions between taxpayers and
tax collectors. In African countries, 72 percent of firms report being required to meet with
tax officials, and for those affected, 3.2 meetings are held on average each year (World Bank
Enterprise Surveys, 2010-2020). In contrast, 37 percent of firms in Europe and Central Asia
are required to meet with tax officials and of those affected, 2.1 meetings are held on average.
The high frequency of meetings is also correlated with corruption indicators in taxation. On
average, 17 percent of firms in Sub-Saharan Africa report that they are expected to give
gifts in meetings with tax officials. However only 7 percent report the same in Europe and
Central Asia. Similarly, 28 percent of firms in Sub-Saharan Africa cite tax administration
as a major constraint, compared to 16 percent in Europe and Central Asia (World Bank
Enterprise Surveys, 2010-2020).
      Along the same lines, the use of information technology in tax administration, such as
e-filing, increases with national income. For example, as of 2016, 85 percent of high-income
countries, 65 percent of middle-income countries and 32 percent of low-income countries had
adopted e-filing (World Bank, 2016). Its prevalence has since continued to grow rapidly.
  2
      Data from the most recent year available between 2010 and 2020 is used to calculate the regional averages.



                                                        4
1.2       Information Technology Adoption in Africa

The use of information technology continues to rise substantially across the continent. In
2008, less than 5 percent of Africans were using the internet. That fraction had increased
fivefold to 25 percent by 2018 (World Development Indicators, 2018) and continues to grow
(Figure 5a). Mobile cellular access, in particular, has large penetration across the continent
with about 90 subscriptions per 100 people across the continent, more than double the
prevalence rate a decade earlier (Figure 5b).
       Africa has achieved striking successes in some aspects of technology innovation, particu-
larly in the adoption of mobile money with services like M-Pesa being able to leapfrog the
technology frontier, with users bypassing traditional financial service providers. Kenya has
been a leader with more than half of the adult population having a mobile money account.
Other countries like Uganda, Zimbabwe, Gabon, Namibia and Ghana also have high levels of
mobile transactions (Figure 6a). While the most common use of mobile money is for sending
remittances, commercial uses such as paying utility bills and receiving wages are starting
to emerge (Figure 6b). This holds promise for Africans using technology services for tax
purposes as well.3
       The use of information technology is also growing across different sectors of government
in Africa. In many countries, tax administrations often lead technology adoption and digi-
talization, compared to other government bodies (World Bank, 2016).


1.3       Role of Information Technology in Increasing Tax Mobiliza-

          tion in Africa

Robust adoption of technology is designed to move tax administration from manual systems
characterized by tax official discretion across taxpayers, tedious and error-prone data entry,
   3
    In recent times, a number of countries such as Uganda, Ghana, Malawi, Tanzania, Cameroon, and
Republic of Congo have introduced or implemented legislation to impose taxes on mobile money transac-
tions. These taxes may have negative economic and financial inclusion impacts, especially on disadvantaged
individuals with limited access to the formal banking system.


                                                    5
and case-by case detection of evasion to a reliance on electronic systems, where there is a more
consistent and predictable experience across taxpayers, timely data for decision making and
automated detection of suspicious activity. Using the framework developed in Okunogbe and
Santoro (forthcoming), this paper examines how technology may help to improve core tax
administration functions for the main tax types: consumption taxes, property (real estate)
taxes, trade taxes, and income taxes (personal income, corporate income and payroll taxes).
   Define Compliance: In order to tax, the government must be able to identify the
tax base. Tax authorities can use technology-based tools to collect information to identify
taxable entities (such as individuals or property during registration drives) as well as to
collect information on the tax liability that may otherwise be concealed by the taxpayer.
For example, tax authorities may require firms to use electronic billing machines that record
sales transactions, or they may collect information from third-party sources like employers,
vendors, customers or financial institutions.
   Monitor Compliance: Technology can also help the tax authority detect evasion when
it occurs. Technology provides tools for collecting and analyzing large amounts of data to
automatically detect inconsistencies, such as mismatches between self-reported and third-
party reported tax liability. Analyzing different indicators of evasion allows a tax authority
to have a data-centric approach to targeting audits by building a risk profile for each taxpayer
and prioritizing those with higher risk of evasion.
   Facilitate Compliance: Technology can be used to simplify procedures and improve
service delivery to taxpayers. Services like electronic filing and payment can make the taxpay-
ing experience less time consuming and more consistent across taxpayers. Electronic billing
machines reduce the costs of compiling and submitting information. Electronic modes of
communication such as email and SMS provide a timely and cost-effective way of providing
information to taxpayers. These technologies also reduce the level of in-person interactions
between taxpayers and tax officials, thus reducing opportunities for extortion and collusion.
   Besides these three core functions, technology may also improve the ability of tax admin-


                                                6
istrations to make timely and data-based decisions such as forecasting revenues, measuring
progress, and monitoring staff performance. However, these internal functions have limited
empirical evidence in the literature, and we do not include them in the scope of this paper.
Instead, we focus on technology applications that directly affect interactions between tax-
payers and the tax administration. The rest of this paper proceeds as follows. Sections 2,
3, 4 and 5 provide an overview of consumption taxes, real estate taxes, taxes on trade and
income taxes, respectively, in Africa and discuss the possible role of technology in improving
the administration of these taxes. Section 6 provides insights from interviews with rev-
enue administrators in Africa on their practical experiences and challenges in implementing
technology solutions. Section 7 concludes.



2         Technology and Consumption Taxes

Consumption taxes include sales tax, value added tax (VAT) and excise taxes. VAT is
becoming increasingly common in Africa, as in the rest of the world, accounting for about
40 percent of total tax revenue (ATAF, 2019). Albeit quite sophisticated and complex
to administer (Slemrod and Velayudhan, 2022), the VAT holds great potential to improve
compliance, mostly due to its in-built self-enforcing mechanism as buyers must report their
input purchases to be able to deduct the VAT they paid from their own VAT liability
from their sales. This generates opposite incentives for buyers and sellers in reporting the
same transaction: buyers have an incentive to maximize the amount paid, whereas sellers
have an incentive to minimize the amount received, thereby creating less room for evasion.
Importantly, the VAT generates a paper trail or a record of transactions from the different
trading partners in the value chain. Especially when digitized, such a trail provides a wealth
of information to tax administration which they can use in their monitoring function.4
        Technology can significantly help in the correct administration of such a complex tax
    4
    Another crucial use of VAT data is that it can be very informative about the state of the economy and
help government measure trends and shocks, as done in Mascagni and Lees (2022) for Rwanda.



                                                   7
as the VAT. In most African countries, the introduction of VAT has been accompanied by
the launch of electronic fiscal devices (EFDs), such as electronic billing machines (Mascagni,
Dom and Santoro, 2022). EFDs are machines that automatically record transactions as they
are performed and communicate this information electronically to the tax administration.
Such machines hold potential to boost tax compliance in different ways. First, they help in
defining compliance, in the sense that they provide information about the amount of sales
made in a given period, that may then be cross-checked with self-reported sales amounts
from both VAT and income tax returns.
   Second and quite interrelated, this information may then be used for monitoring compli-
ance by having data-based, automatic flags of irregularities and discrepancies, on which the
authority can build its audit strategy to ensure proper enforcement.
   Third, EFDs are also thought to facilitate compliance, helping tax-paying firms with
their record-keeping. Especially in its more sophisticated version (see the case of Rwanda
in Mascagni, Dom and Santoro, 2022), EFDs can store valuable information on sales, pur-
chases and inventory which taxpayers can access when filing their returns. Likewise, more
sophisticated, software-based EFDs, especially when integrated with e-filing platforms (as in
Rwanda), can provide additional information to taxpayers around deadlines, VAT rates and
the like. Ongoing research is evaluating whether these benefits then translate into higher
VAT compliance in Rwanda (Hakizimana and Santoro, forthcoming).
   Against these important theoretical benefits from EFDs, the existing evidence on their
impacts in Africa is rather mixed. On the one hand, it is true that such technology improved
VAT revenue collection as also shown in other contexts beyond Africa (e.g. Fan et al., 2018).
Some positive evidence comes from Ethiopia, where two studies evaluate the impact of
the electronic sales register machines (ESRMs) – as EFDs were named there – on VAT
compliance. Box 1 summarizes the key lessons learned from these evaluations.




                                              8
  BOX 1: Electronic Sales Registration Machines in Ethiopia (Ali et al., 2021;
  Mascagni, Mengistu and Woldeyes, 2021)

  Problem: Despite its in-built self-enforcing mechanism that creates opposite incentives for
  buyers and sellers, the collection of VAT is thought to be largely underperforming in low and
  middle - income countries (Keen and Simone, 2004; Gordon and Li, 2009; Besley and Persson,
  2013; Pomeranz, 2015). One of the main reasons, apart from the administrative complexity,
  is widespread evasion as the lack of information on sales severely undermines the enforcement
  capacity of revenue authorities.

  Technology intervention: The Ethiopian Revenue and Customs Authority (ERCA) started
  the rollout of electronic sales register machines (ESRMs) in 2008 with the goal of moving from
  paper-based receipts, which make monitoring business transactions on a daily basis extremely
  expensive and unfeasible, to a digital solution. ESRMs, on top of printing out receipts, auto-
  matically report transactions via a network to an ERCA server – with the agency being able
  to monitor sales and eventually use that information to enforce compliance.

  Research design: The studies of Ali et al. (2021) and Mascagni, Mengistu and Woldeyes
  (2021) heavily rely on administrative tax data provided by the revenue authority. The authors
  have access to tax returns and payments data over a large span of years, as well as to the date
  of machine adoption. Both studies employ a difference-in-difference design. In Ali et al. (2021),
  the design is reinforced by a matching technique making adopters and non-adopters more com-
  parable. In Mascagni, Mengistu and Woldeyes (2021), several restrictions are imposed on the
  sample (for instance, never adopters are excluded), and a triple difference-in-difference is per-
  formed to reduce imbalance between treatment and control groups.

  Impact: Impacts from ESRM adoption are largely positive. On the one hand, Ali et al. (2021)
  find that the adoption leads to increase in reported sales and VAT liability. They also show
  that this improvement happened without the erosion of the tax base, as there was also an
  increase in employment rates. On the other, Mascagni, Mengistu and Woldeyes (2021) find
  a positive impact on tax revenue, increasing by 48 percent for VAT. Considering also income
  taxes, they rise by 12 percent after adoption, indicating spillover effects. Further, they mea-
  sure positive impacts on the accuracy of reporting. However, such positive effect is partially
  muted by strategic taxpayer response, which will be discussed in more detail in Section 6. In
  Ethiopia, taxpayers respond by simultaneously adjusting both reported sales and costs, thus
  yielding net revenue gains that are proportionally lower than the increase in sales.

  The key lesson from the two studies is that technology can improve compliance, but perverse
  responses from taxpayers can partially mute its impact.




   In a different context, Rwanda, Eissa and Zeitlin (2014) find that the first version of
electronic billing machines (EBM v1) that automatically records a firm’s sales leads to an


                                                 9
increase in VAT payments by an average of 8 percent – an effect highly variable by sector
and firm size, with smaller firms as well as firms in computing/printing, construction, and
restaurant sectors experiencing larger impacts.
       On the other hand, increasing qualitative evidence indicates how taxpayers struggle to
optimally use the technology, often with concerning repercussion on equity. For instance,
and again in Rwanda, Mascagni, Dom and Santoro (2022) show that the expected benefits
of EBM do not always materialize, especially for small taxpayers. The authors first show
that the incidence of poor record keeping (as proxied by inconsistencies between VAT returns
they submit and data automatically recorded in the EBM) is larger for smaller firms. They
then use focus group discussions to demonstrate that taxpayer confusion, complexity and
technical problems with the EBM, for example, issues with topping up SIM cards, and with
the necessary equipment such as computers and internet connection are particularly severe
for smaller taxpayers. The more advanced version EBM2 seems still expensive – especially
in terms of the necessary equipment (computers, connection, etc.) - for smaller taxpayers
who already face budget constraints in using the traditional EBM.5
       This disproportionate effect brought about by more and more sophisticated technologies
(as the ever-evolving versions of EBMs) could exacerbate existing inequalities in the taxpayer
population. Similar impediments are witnessed in the deployment process of EBMs in Kenya
and Tanzania. Eilu (2018) reviews the evidence from seven studies based on the adoption
experiences of these two countries, to conclude that the most-cited challenge with the use of
EBMs was the lack of training of VAT-collecting enterprises, mentioned as a hindrance in
six of the studies. The second most-cited challenge was the high sunk costs associated with
technology adoption.
       With regards to the potential of in-built mechanisms of the VAT to help identify the
tax base and monitor compliance, it is still unclear whether African revenue authorities
   5
   At least in Rwanda, EBMs need airtime as they transmit information through the mobile network.
Taxpayers, especially the smallest ones, easily forget (or are unaware they have) to top up airtime, thereby
impeding the EBM from communicating with the authority. In addition, businesses have to buy the machine
themselves without any subsidy (Mascagni, Dom and Santoro, 2022).


                                                    10
are fully exploiting the potential of the VAT-generated paper trail which has proved to be
essential in other contexts (Pomeranz, 2015; Carrillo, Pomeranz and Singhal, 2017). Recent
evidence indicates that, despite the adoption of EBMs, there are widespread discrepancies in
data reported by buyers and sellers for the same transactions, both in Rwanda (Mascagni,
Mukama and Santoro, 2019) and Uganda (Almunia et al., 2022). Such studies suggest that
tax administrations are not fully equipped to unlock the potential of VAT cross-checks,
mostly due to structural challenges with software and skills (see Section 6). More broadly,
such challenge suggests that much more can be done in investing in automated data-based
monitoring systems, which have shown to be impactful in more developed contexts – such
                                                                urkiye (Dogan, 2011).
as India (Mittal and Mahajan, 2017), Pakistan (Shah, 2020) and T¨



3     Technology and Real Estate Taxes

Although real estate taxes make up only about 3 percent of taxes in African countries
(Figure 3b), they are often the primary source of revenue available to local governments.
Increases in property taxes have also been shown to be more closely linked to increases in
spending on public services, compared to other revenue sources such as transfers from the
central government (Gadenne, 2017; Hoffman and Gibson, 2005). As many countries adopt
fiscal decentralization programs, it will be important to have the appropriate information
technology systems in place to support the implementation of real estate taxes.
    At first glance, given the visibility and immobility of the tax base, real estate taxes would
seemingly not suffer from the core information problem in taxation: the ability of authorities
to identify the tax base and determine tax liabilities. Yet, in practice, many tax authorities
in Africa struggle to keep accurate and comprehensive cadasters and other databases for
real estate taxation purposes. Cadasters, when they exist, are often outdated, and property
valuations often do not reflect existing market prices. Further, administrative processes to
collect tax revenues and identify defaulters may be inefficient and open to manipulation.



                                               11
Technology may help with these issues.
   First, technology can help to identify and value properties. In cases where there is
no cadaster, satellite images may be used to produce a map of buildings, with additional
ground-truthing to populate information on property characteristics. Traditional valuation
approaches such as those based on construction costs are costly and time consuming. In
other settings, property owners are asked to self-declare their property value but this is
subject to manipulation. Alternatively, a fixed fee may be imposed on each dwelling, but
this approach is likely to be regressive. Computer assisted mass appraisal methods that use
statistical models to assign an estimated market value to property using data on property
characteristics (location, size, amenities etc.) and sale price from recent transactions are
widely used in middle- and high-income countries, but in many African contexts with weak
property markets, the required data on recent purchase prices are often not available.
   An alternative approach aided by technology is a simplified “points-based” system re-
cently adopted by the Freetown City Council in Sierra Leone (Fish, 2018), with support
from the International Growth Center, the International Centre for Tax and Development
and the UK government (Grieco et al., 2019). The project involved the city’s authorities
using satellite imagery to identify and measure properties in Freetown. They then deployed
teams to collect data on easily observable characteristics such as the quality of walls, roofs
and windows. Following this, they used existing data on rental values to build a simple model
for calculating the taxable value of each property. This approach almost doubled the number
of properties on the tax register from 57,000 to about 110,000. Crucially, the new system
produced much higher tax bills for the most valuable properties, owned by the wealthiest
taxpayers, which have been historically under-taxed, thereby increasing the progressivity of
the tax system (Kamara, Meriggi and Prichard, 2020).
   A similar example comes from Senegal where Knebelmann, Pouliquen and Sarr (2021)
report that, at most, a third of cadastral plots in Dakar are on the valuation roll. In addition,
rental values listed on the roll are on average lower than values obtained through other sources


                                               12
such as surveying owners and real estate experts, likely due to poor and biased valuation
practices. Furthermore, cadastral information on plots in the tax database is not linked to
address information that is visible and known by taxpayers. This lack of data harmonization
makes it impossible to send tax notifications to most property owners. As a final challenge,
manual practices in collecting new information on properties imply a cumbersome, time-
consuming and error-prone data collection process by tax agents. Ongoing research examines
the impact of a new mobile application that helps to determine property values and process
billing and collection on tax collection. Box 2 presents a case study from Liberia on how
using information from a newly developed electronic tax database increased tax collection.
   Second, technology is important for facilitating and monitoring compliance with real
estate taxes. The ideal technology system will not only support property identification using
a geospatial database and valuation, but will also integrate billing, payments, appeals and
enforcement. Dzansi et al. (2022) report from a census of local governments in Ghana that
only 17 percent of them have a digital database of properties or a revenue software of any
kind to assist with billing, and find that use of technology is associated with higher revenue
collection. In addition, using a randomized experiment in a large municipal government, they
find that revenue collectors using electronic tablets with geospatial data and navigational
directions to more easily locate properties collected twice as much revenue as control group
collectors.
   Technology solutions can also be used to reduce compliance costs and revenue leakage.
                     an et al. (2022) describe the use of a payment technology in door-to-
Weigel (2020) and Bal´
door property tax campaigns in Kananga, Democratic Republic of Congo. Taxpayers made
cash payments to tax collectors instead of traveling to the tax office, thereby reducing their
compliance costs. The tax collectors used handheld printers to issue receipts to taxpayers and
each transaction was recorded in the device’s memory and downloaded to the government
database on a regular basis. To monitor leakages, supervisors confirmed that the revenues
deposited were equal to the amounts on the receipts that were issued.


                                             13
BOX 2: Real Estate Tax in Liberia (Okunogbe, 2021)

Problem: Real estate taxes in Liberia, like many low-income countries, is characterized by
high non-compliance rates. One significant hurdle in administering the real estate tax is the
lack of a property cadastral. In the absence of comprehensive information on property location,
ownership and values, tax enforcement cannot be conducted in a systematic way and many
properties remain out of the tax net. Only about 5 percent of residential property owners were
estimated to be on the tax roll (Olabisi, 2013).

Technology intervention: The Liberia Revenue Authority (LRA) initiated a low-cost tech-
nology investment to create a new property database. The tax authority recruited and trained
young people to use open-source software on tablets to capture the location, ownership, photo
and GPS of properties. The “foot soldiers,” as they were called, went door-to-door, conducting
a brief survey with the resident of each property to collect ownership information and record
property characteristics, including a photograph of the property. This resulted in the creation
of a new property database that the tax authority could then use for property tax administra-
tion.

Research design: The tax authority used the information from this new database to send
notices to taxpayers to request payment of real estate taxes. Four types of notices were sent
out, with property owners randomly assigned to the different notices (Okunogbe, 2021). The
first group received a “Plain Notice” that provided information on the property tax require-
ment and the procedures to pay. The second group received a “Detection Notice” that used
information from the new tax database to personalize the notice: it was addressed to the prop-
erty owner by name and included a photograph of the property. The third group received a
“Penalty Notice” that provided information on the legal consequences of not complying. The
fourth group received a “Detection and Penalty Notice” that combined the features of those
two notices.

Impact: The Detection and Penalty notice tripled the property registration and tax payment
rate among recipients, relative to the baseline registration rate of 4 percent and payment rate
of 2 percent among those who received the Plain notice. The Detection notice and the Penalty
notice alone did not have a statistically significant impact on registration or payment.

This intervention shows the effect of using technology to organize the publicly visible infor-
mation on property tax liabilities to boost tax compliance. It highlights the complementary
roles of communicating government detection capacity to taxpayers and alerting them of conse-
quences for non-compliance. Nevertheless, it highlights the fact that additional factors, such as
enforcement, may play an important role because even with perfect detection and information
on penalties, less than 15 percent of those who received notices registered or paid.




 Given the rapid increases in open-source software for geospatial database creation and


                                               14
simple record keeping, many of the technologies described in this section can be adopted
at reasonable costs by local governments.6 It is important to pilot and test software to be
sure it meets the specific needs in a given context. Knebelmann (2022) provides detailed
illustrations from the experiences from over 15 property tax digitalization projects in different
countries to show how satellite imagery tools and georeferenced data systems can help in
detecting, registering, characterizing and valuing properties.



4         Technology and Trade Taxes

Trade and customs taxes account for a notable share of total tax revenues in Africa where they
account for 14 percent of tax revenues, relative to 4 percent in other regions (Figure 3b). Also,
the majority of VAT revenues are collected at the border in many countries (World Customs
Organization, 2017). For this reason, national customs departments play a crucial role in
revenue mobilization, in addition to being instrumental in facilitating trade and exchange
with foreign investors. At the same time, customs departments are often characterized by
high levels of corruption – mostly due to the fact that tax officials enjoy strong discretionary
powers in their profession, and risk-based systems of control and accountability are often
missing or easily manipulated.
        Technological innovations can significantly improve the administration of trade taxes.
Similar to VAT, automated customs systems produce high frequency transaction level data
that can be used for tax monitoring. It is not surprising that, particularly in developing
countries, the initial, experimental attempts to digitalize and automate tax administration
took off from customs offices, usually in parallel with broader reorganization reforms of cus-
toms bureaucracies. According to the World Bank Global e-Government Systems database,
by 2014 low-income countries have invested much more in automating customs (97%), than
tax management (57%), filing of taxes (32%), or online service portals (11%). For instance,
    6
   Relatedly, geolocalized address information and a more accurate taxpayer register can improve enforce-
ment on other tax types as well, such as income taxes.


                                                   15
the first “pilot” examples of integration and automation of data management systems are
given by the Asycuda platform for trade taxes, which then laid out the foundations for
introducing automated systems at the level of domestic tax administration as well.
   Technology can significantly boost all core functions of custom tax administration – es-
pecially if through an integrated custom management system as Asycuda that processes all
necessary customs documentation such as cargo manifests, customs declaration, payments,
and other regulatory requirements in a timely fashion (UNCTAD, 2020). First, technology
could help to more clearly identify the taxable base, such as the value and quantity of im-
ported goods – for instance, cargo information can be captured by scanning barcodes on
shipments at the port of departure and communicating the information in real time to Asy-
cuda. Likewise, integrated systems requiring a unique taxpayer identification number from
traders would significantly help the tax administration identify the tax base and seamlessly
connect traders with other tax databases (World Customs Organization, 2014).
   Second, in relation to the above, technology could greatly improve the enforcement of
tax collection and monitoring of performance. Using the information on tax liabilities and
taxable entities, a sophisticated IT system could embed independent valuation reports from
third parties and introduce explicit monitoring tags informing inspectors of high-risk dec-
larations. In the most advanced versions of Asycuda, statistical tools and algorithms are
included to detect fraud. Artificial Intelligence systems such as pattern recognition can be of
great help in supporting risk assessment and profiling policy (World Customs Organization,
2014).
   Lastly, technology could, more broadly, speed up the customs processes of importing
and exporting of goods, incentivizing foreign traders through a smoother and less burden-
some cross-country trading experience. In this sense, Single Windows are one-stop-shop
electronic systems, usually integrated with Asycuda, that allows traders to submit infor-
mation at a single-entry point to fulfill all import and export regulatory requirements. By
reducing interactions with inspectors, Single Windows streamline and standardize traders’


                                              16
experience, significantly curbing application processing times and increasing transparency
and predictability (UNCTAD, 2020). In addition, e-payment modules embedded in Asycuda
dramatically decrease customs clearance time and traders’ compliance costs, improving their
experience and at the same time speeding-up the transfer of funds to government accounts
(UNCTAD, 2020).
   However, existing evidence on the impacts of technology on customs in Africa is rather
scarce. A first set of qualitative studies indicate that integration between customs and do-
mestic departments is still partial, when it comes to data sharing and automatic cross-checks.
In Uganda, as described in Mayega et al. (2021), the URA e-tax still needs improvements
towards integration. For instance, the system is not able to validate some of the submitted
information with information that is already available within different sub databases in e-tax
and, more importantly, with the custom platform, Asycuda World. A similar challenge is
documented in Malawi (Ligomeka, 2019). Some promising suggestive evidence comes from
Zambia, where, in 2018, the tax administration implemented a successful data matching
exercise cross-checking information from domestic tax returns and data from the customs
information management platform (ATAF, 2019).
   A second strand of research indicates how technology in customs departments could easily
backfire due to conscious sabotaging responses from customs officials. A particular case study
from Madagascar well illustrates this point and provides novel evidence on the impacts of
IT-based information on customs performance in an African country, and is summarized in
Box 3.




                                             17
BOX 3: Technology and Corruption in Madagascar Customs (Chalendard et al.,
2020)

Problem: Madagascar is characterized by a very low tax-to-GDP ratio (below 10 percent)
and almost half of tax revenues in 2019 came from customs. Corruption by customs officials,
particularly undervaluation of imports, was an important concern especially given the con-
centration of revenue monitoring among a small number of officials. In Toamasina, the main
seaport in the country, only 16 inspectors clear declarations for about two-thirds of non-oil im-
ports and three-quarters of non-oil tax revenue. Across the country, each inspector on average
oversees the collection of approximately 10 million USD in terms of import revenues each year.

Technology intervention: The customs department in Toamasina worked together with an
information service provider specialized in assisting customs with information technology (IT),
equipment and risk analysis, including the provision of valuation advice on traded goods. Such
advice comprises a detailed, high-quality report on the price and quality of declared goods.
The advice is produced only for high-risk declarations more likely to evade tariffs. Chalendard
et al. (2020) evaluated the impacts of such intervention on customs outcomes and inspectors’
behavior.

Research design: The methodology consists of two steps. First, the authors estimate the
impact of the provision of third-party valuation advice information on inspector actions and
customs outcomes, through a regression framework. Second, in a pioneering randomized con-
trolled trial in customs, the study randomly assigns high-risk customs declarations to improved
information provision by the customs risk management unit and also randomly tags declara-
tions for increased monitoring.

Impact: First, the authors show that third-party valuation advice significantly increased the
probability of goods being scanned by 10 percentage points and the probability of a declaration
being deemed fraudulent by 22 percentage points. Third-party data availability also increased
tax revenue by 5.2 percentage points on average. However, third-party valuation advice is least
effective in improving customs outcomes for declarations for which opportunities for corruption
are largest. Such evidence is confirmed in the RCT - where stakes are higher, improved infor-
mation remains less effective. Likewise, increased monitoring did not improve most customs
outcomes.

A key lesson from this study is that customs officials may not react optimally to new IT-based
information if their incentives for graft are largely unaltered. Also, higher monitoring is largely
ineffective in a context of low fear of enforcement and an enduring culture of impunity.




                                                18
5     Technology and Income Taxes

Income taxes paid by individuals, companies as well as employees are an important revenue
source for African economies, representing about 33 percent of total revenues (UNU-WIDER
Government Revenue Dataset, 2021). Unlike Pay-As-You-Earn (PAYE) taxes that are de-
ducted at source on wages and salary, corporate income and other personal income taxes are
typically self-declared, providing more opportunity for evasion (Kleven et al., 2011). This
makes them particularly vulnerable to broader structural issues like high informality and
low tax morale that to an extent also affect other tax categories.
    One key obstacle to optimizing income taxes is informality: many individuals and compa-
nies operate completely outside the purview of the tax authority. The existence of an informal
economy generates horizontal inequities and economic distortions that potentially undermine
tax perceptions and the moral fiber of a society (Joshi, Prichard and Heady, 2012). Within
the informal sector, most harm to revenue is produced not by micro-enterprises such as street
vendors, but by hard-to-tax professionals such doctors, lawyers, architects, etc. (Ogembo,
2020). Even among those formally registered, tax minimizing schemes by professionals and
high net worth individuals determine large revenue losses and damage to the fairness of the
tax system (Keen, 2012; Kangave et al., 2016; Santoro and Waiswa, 2022).
    A second major challenge for income taxes is that even among taxpayers that are formally
registered, there is a high level of non-compliance: the vast majority of registered taxpayers
in Africa either do not file their returns or zero-file (Santoro, 2021; Mascagni, Dom and
Santoro, 2022; Almunia et al., 2022). In addition, African taxpayers’ morale and willingness
to pay are remarkably low. Different factors associated with low willingness to pay include:
distrust of the state (Bratton and Gyimah-Boadi, 2016; Isbell, 2017), dissatisfaction with
public services (Blimpo et al., 2018), perceived unfairness of the tax system (D’Arcy, 2011;
Okunogbe and Edjigu, 2022) and the complexity of navigating obscure tax regulations (Aiko
and Logan, 2014; Mascagni and Santoro, 2018).
    A third constraint is that African tax administrations experience important limitations in

                                              19
their traditional enforcement strategies and find it difficult to optimally monitor compliance
with self-reported income taxes.
   In this context, technology may be instrumental in boosting the performance of income
tax collection in many ways. On the taxpayer side, technology can be used to simplify
tax registration processes, such as the creation of one-stop shops for firms. In addition,
taxpayer e-services such as e-filing and e-payment facilities, as well as web tax portals, can
significantly ease the burden to comply, increase transparency and predictability of taxes,
and reduce scope for corruption, factors which may increase willingness to comply. On the
tax administration side, IT solutions such as Electronic Fiscal Devices, can provide crucial
data on sales which can then be cross-checked automatically with self-reported income tax
declarations. Similarly, the tax administration can obtain and crosscheck data from third-
party reporting such as VAT data from clients and vendors and rental income withholding
with self-reported income. Against these promises, the existing evidence is, once again,
rather mixed, and we summarize it below.


5.1    Pay As You Earn

In theory, given the at-source deduction of PAYE tax by the employer, there is less scope for
evasion with PAYE compared to self-declared income (Kleven et al., 2011), however, there is
very limited empirical evidence from Africa on this. Rather, the existing descriptive evidence
suggests that large gaps persist in the quality of PAYE returns data, thus undermining the
capacity of revenue authorities to make best use of it and opening loopholes for tax evasion.
   One issue is lack of proper tax identification of individuals. Descriptive evidence from
both Uganda (Mayega et al., 2021) and Ethiopia (Mascagni and Mengistu, 2019) reveals that
tax identification numbers (TINs) for employees are largely missing or incorrect in employers’
PAYE returns. This means that the tax authority is unable to track payments by individuals
over time and match individuals’ records with other databases. Reporting aggregated PAYE
taxes at the employer level and not linking to individuals also creates an opportunity for


                                             20
inaccuracy in the amount of taxes remitted. Relatedly, the inability to match taxpayer-level
data from PAYE to other data sets (such as income tax returns) makes it difficult to add up
individuals’ total income and enforce progressive income tax rates.
    A second concern revealed in the Ethiopian context, but common in many others, is
that PAYE data is not fully digitized. Many returns are kept in paper form, prone to loss or
deterioration with time and making it very unlikely that this data will be analyzed (Mascagni
and Mengistu, 2019).7


5.2     Corporate and Personal Income Taxes

Different technological interventions, such as EFDs and e-filing/e-payment services, have
been introduced in Africa with the goal of facilitating compliance of self-declaring business
income earners who are remitting corporate or personal income taxes. In Section 2, we
discussed the potential benefits and key challenges for VAT compliance brought by the
introduction of EFDs, which apply to compliance with income taxes as well. Furthermore, it
is still unclear whether the revenue gains on VAT attributable to EFDs would then produce a
parallel increase in income taxes. If anything, in Ethiopia, EFD users only partially improve
their filing of income taxes, despite significant rises in VAT reporting (Mascagni, Mengistu
and Woldeyes, 2021). Such gaps in compliance also depend on the sophistication of the
IT system within the tax administration, which would limit discrepancies. In Rwanda, as
discussed in more depth in Section 6, EFD data is used to benchmark self-reports from
income tax returns.
    Regarding the potential role of IT to facilitate compliance of income taxpayers, a first
strand of the literature considers factors associated with taxpayers’ adoption of and practical
experiences with tax e-services. Box 4 summarizes this evidence.




   7
     Statistical analyses would be a rather frustrating task if based on analog filings, which make impossible
to precisely trace and identify employees across periods and firms.


                                                     21
  BOX 4: Adoption of Tax E-services in Africa

  Problem: Traditional tax processes involve paper-based filing and extensive in-person inter-
  actions between taxpayers and tax officials. This system creates several challenges such as
  large compliance costs, collusion with tax officials or extortion by them, and poor tax records
  that limit the potential of tax agencies to conduct analyses for compliance monitoring and
  forecasting. Whereas new technology to digitize taxpayer services may hold a lot of promise,
  important barriers to adoption of these technologies may severely curb the potential for rev-
  enue mobilization.

  Technology intervention: Many African governments have been heavily investing in e-
  services as a way to move from paper-based systems to fully digital platforms. As an example
  of such e-services, e-filing and e-payment facilities enable digital filing of returns and digital
  payment of tax liabilities. These e-services often live within integrated and automated tax
  administration systems. The associated web portals also provide a range of benefits in terms
  of clear information on deadlines, access to assistance, and better record-keeping.

  Research design: A number of studies from Africa attempt to understand the barriers to
  access of such e-services. These studies are typically based on surveys of representative samples
  of taxpayers, collecting detailed information around taxpayer experiences with e-services. In
  the case of Rwanda, Santoro et al. (2022) originally link survey data with tax administration
  data, in order to gain more information around taxpayers and their filing behavior.

  Results: Lack of awareness and lack of training are key factors that have been identified in
  both Nigeria (Mas’ud, 2019) and Zimbabwe (Obert et al., 2018) as barriers to adoption of e-
  services by taxpayers. In Nigeria, Efobi et al. (2019) conclude that internal firm characteristics
  like owner’s education, age of the company, and use of an external auditor and computerized
  accounting system, are most strongly correlated with the use of e-filing and e-payment by firms.

  Partial take-up can persist even when e-services are mandatory, like in Rwanda. Santoro et al.
  (2022) uncover important digital divides among users and non-users. First, they show that
  female, less educated taxpayers, with less sophisticated businesses, are significantly less aware
  of e-services and less likely to use them.

  In sum, for broad adoption of e-services to occur, tax administrators must ensure that the
  introduction of new technology is accompanied with extensive sensitization, training, and as-
  sistance, especially for disadvantaged taxpayers that are more likely to be excluded.




   A second strand of the literature considers the impacts from technology adoption. While
the impact of e-services like e-filing on tax payments, compliance costs and corruption
have been rigorously studied in other similar contexts, such as Tajikistan (Okunogbe and


                                                 22
Pouliquen, 2022), there is some emerging evidence from Africa as well. Santoro, Amine and
Magongo (2022) evaluate the effect of the Eswatini government mandate to adopt e-filing
and e-payment services for all income taxpayers as a response to the pandemic, introduced
in September 2020. Using extensive administrative data from the tax agency, the authors
examine the impact of IT adoption by exploiting the partial take-up of the mandate. A first
key finding is that, indeed, more than 8 months after the mandate came into force, only a
minority (41 percent) of taxpayers moved to filing through the e-Tax system. In line with
other research from Africa described in Box 4, qualitative evidence from interviews with the
tax agency reveals that most small taxpayers lack the equipment, an adequate connection,
and the knowledge to navigate e-Tax. These taxpayers, mostly from rural areas where inter-
net coverage is poor, then resort to imperfect solutions such as bringing their manual returns
to the revenue office or emailing them and asking staff to help enter them in e-Tax.
       In terms of impacts, the evidence from Eswatini is rather positive, as e-Tax uptake
significantly improved filing behavior, as e-Tax registered taxpayers are almost 60 percent
less likely to file nil, and they declare 32 percent more turnover and 50 percent more taxable
income. However, surprisingly, final tax liability remains unchanged due to corresponding
increases in expenses and deductions. For payment behavior, the authors find a strong
impact of e-Tax registration on the probability to pay conditional on filing, as well as on
the final tax amount paid (70 percent and 64 percent increases, respectively).8 In terms of
mechanisms, the study suggests that technology is facilitating compliance through a salience
or reminder effect and helping taxpayers to track their records and fully report their expenses
and deductions.
       As a cautionary note to these positive impacts of technology on income taxes, in Rwanda,
Santoro et al. (2022) do not find a positive relationship between adoption of E-tax or M-
declaration solution and self-reported taxpayer outcomes such as tax compliance, perceived
   8
    The divergent impact on tax declared and tax paid is not surprising in a context where filing and payment
behavior are often disconnected, and where revenue authorities often prioritize payment over filing (Santoro
and Waiswa, 2022).



                                                     23
fairness, or the intrinsic motivation to comply with taxes. From the qualitative information
gathered from focus group discussions, the concrete benefits from e-services such as time
savings and more transparent access to information are balanced out by a multitude of
practical challenges in usage, such as connectivity problems and a slow IT system in the tax
agency, on top of inadequate assistance from RRA when technical issues arise. Such issues
seem to exacerbate the frustration and perceived unfairness in usage, and call for sustained
investments in strengthening the e-services.



6         Insights from Interviews with Tax Administrators

In this section, we complement the evidence gathered from the desk review with qualitative
insights from eleven in-depth interviews with tax officials from seven African revenue au-
thorities, in a unique exercise carried out in 2021. These interviews tell us about the direct
experience of practitioners, experts and implementers, which is particularly important to
complement the limited information available in the literature. The countries included in
this exercise are: Eswatini, Ethiopia, Malawi, Nigeria, Sierra Leone, Rwanda, and Uganda.9
In-depth interviews were mostly conducted remotely (due to COVID-related restrictions),
lasted on average two hours and followed a specified protocol (see Appendix B for the inter-
view form). First, we ask our respondents about practices and challenges with technology,
focusing on those IT solutions more relevant to them. Then, we discuss general aspects on
constraints to digitalization, ranging from infrastructure, funding and staff to the regulatory
background and political will.
        The sample of countries we cover is not representative of the African continent but rather
selected out of convenience to introduce practical policy insights. Appendix Table A1 reports
the details of the revenue authorities and officials we contacted in this study, as well as the
date when the interview took place. The seven selected countries show varying degrees of
technological maturity. We recur to the latest GovTech Maturity Index (GMTI) of the World
    9
        In two cases (Uganda and Rwanda), we interviewed multiple tax officials in a revenue authority.


                                                       24
Bank (Dener et al., 2021), which assigns a 0-1 score and a grouping across four categories
(A, B, C and D) to each country as an indicator for the state of its public sector digital
transformation. The GMTI for our sample ranges from 0.28 to 0.62. In comparison, the
average of low income countries is 0.27 and that of lower-middle income countries is 0.46.
We summarize the novel evidence around practices and challenges with technology below,
to reflect the direct experiences of tax officials from the continent.


6.1    Issues within the Tax Administration

The success of an IT intervention relies heavily on a range of considerations related to the tax
administration itself. First, a very practical factor is funding. Governments need to secure
enough funding for implementation, upgrading, and maintenance of any technology. From
the interviews, budgetary constraints emerge consistently as a major challenge across the
board. In Malawi, for instance, there is a dedicated budget for IT implementation which is,
however, not sufficient. Nigeria’s revenue authority also heavily invested in the development
of the new automated system TaxProMax. Often, the revenue authority is reaching out to
the government and donors for help.
   Second, the tax administration needs to have adequate equipment and enjoy a solid infras-
tructure to fully benefit from any technology. In Malawi, for instance, internet connectivity
is a major issue within the tax administration. Sometimes, the very same MRA’s systems
are down and there is often congestion. Traffic management of these systems continues to
be an issue, especially with the integrated tax administration system (ITAS), as it often
crashes. But the system has not been upgraded because it is quite expensive, consistent
with the funding obstacle discussed above. In Sierra Leone, our respondent envisages that
hard infrastructural barriers like poor internet connectivity and electricity disruptions could
severely undermine the success of the ongoing ITAS project. The same risk is acknowledged
by the FIRS respondent in Nigeria, who expects that TaxProMax and e-services for tax-
payers are much likely to be impacted by infrastructural constraints. At the same time, in


                                               25
Rwanda, interviewees point out that congestion in online systems at peak times is likely to
increase, without adequate (and costly) investments in storage capacity and server upgrades,
as more and more taxpayers are now brought into the tax net (also see Mascagni, Dom and
Santoro, 2022).
    On the bright side of availability of tax infrastructure and systems, the tax administration
in Rwanda set up automatic cross-checks of different data sets, such as VAT, income taxes
and EBMs, which imply that the self-reported turnover for income taxes is now benchmarked
to the total of sales as transmitted by the EFD. Much more can be done in this direction,
eventually enabling the abundant pre-filling of income tax returns now common in high
income countries.
    Third, another important precondition is the preparedness and incentives of revenue au-
thority staff that interface with a new technology. In Sierra Leone, concerns exist around
change management, which will determine whether the upcoming ITAS turns out to be op-
timal or not – since tax officials are used to one specific way of operating which could be
difficult to change in a short time. A similar story comes from Malawi, where, as technology
evolves, it closes loopholes previously used by some staff for illegal practices. As a result,
there was significant resistance and delay in the launch of ITAS. Eventually, the top man-
agement provided clear directions of change and started imposing strict deadlines on the
staff.
    In the case of Uganda, in line with Mayega et al. (2019) and the challenges discussed
in section 4, there seems to be weak incentives for staff to change. Since adapting to the
new technology is inconvenient to them, they tend to resist it. Staff are very familiar with
Excel, where they have more discretion on what to input, while ITAS modules are more
complicated, with a variety of often-sophisticated checks and controls, and require more
training and commitment. Similarly in Ethiopia, manual practices co-exist with digital ones
and are difficult to erase. There are many returns which are not captured into SITGAS, but
inputted manually by tax officials, especially outside the capital. This is mostly due to poor


                                               26
understanding from staff, with the result that many declarations are lost. Against this issue,
the revenue authority is providing sustained training to staff. Extensive training takes place
at RRA in Rwanda as well, where even minute changes (for instance, small modifications in
a single module) are discussed at length with the staff, which get fully trained to deal with
these changes. The same pattern is confirmed in Nigeria, where technology is ubiquitous
and staff seem to be embracing the change.
       These conversations with tax administrators highlight the importance of ensuring that
digitalization is accompanied by a broad shift in work culture and structural changes in
how staff are organized. This means not having an isolated IT department in charge of
digitalization. Instead, there need to be deep changes in how operational departments are
organized: hiring, reallocating and retraining staff as needed to maximize the impact of
newly acquired data and monitoring capacity on compliance, for example, shifting from
case-by-case enforcement to large-scale, data-driven interventions. These structural changes
are especially important when digitalization is accompanied by an expansion in the number of
registered taxpayers. Relatedly, since digitalization programs are typically very expensive,
it is crucial to ensure a right match between the needs of the tax authority and the IT
solutions procured. The procurement process should involve extensive consultations with
staff as well as maintenance contracts that allow for some flexibility and future adjustments
in the software.10


6.2       Issues among Taxpayers

Our African tax administrator respondents’ description of factors affecting effective adoption
of technology by taxpayers echoes some of the findings from the desk review, such as reliable
infrastructure and connectivity, adequate awareness and knowledge around tax e-services,
and the deliberate misuse of an IT solution to avoid taxes.
  10
     Occhiali, Akol and Kargbo (forthcoming) discuss procurement lessons from in-depth case studies of the
adoption process of Integrated Tax Administration System (ITAS) in Uganda and Sierra Leone, based on a
series of semi-structured interviews with revenue authority and ministry of finance officials.



                                                   27
   First, the infrastructural limits and connectivity hiccups within a given tax administra-
tion inevitably affect the capacity of taxpayers to use a technology. As a recurring theme
across the selected countries, when connectivity is poor, taxpayers are often forced to revert
back to manual practices. For instance, when EFDs are not functioning, firms end up record-
ing transactions manually, limiting the benefits coming from this technology. In Ethiopia,
up to recent times, due to poor internet connectivity, taxpayers were forced to go to the local
tax office to file electronically.
   The same pattern emerges in Eswatini, where e-filing became mandatory in 2021, but
taxpayers in remote, poorly connected areas still have to visit the local offices or internet
kiosks to access the web portal and file their returns (see also Santoro, Amine and Magongo,
2022). In Rwanda, on the other hand, the RRA introduced the USSD-based M-declaration
application that works on simple feature phones as a simpler tool for tax declaration and
payment for presumptive taxpayers (Section 5) instead of the e-tax platform that requires
internet connection, with the clear policy intent of reaching less connected taxpayers and
facilitating their compliance. The app also easily enables tax payments through a direct
link to mobile money platforms, which function smoothly on USSD phones. Second, IT
literacy and high adoption costs must be factored in when implementing a new technology.
In Uganda, for instance, we learned that lack of sufficient consultation with taxpayers on
the new e-registration system, integrated with the national ID database, resulted in over-
whelming feedback from taxpayers that URA’s registration process was too cumbersome.
As another example from Uganda, the biggest challenge with e-filing is that the average
taxpayer does not have adequate knowledge of the tax system and familiarity with IT tools
to fill in online forms. Also, language barriers play a role, as e-returns are only in English.
This means that most taxpayers need a middleman or tax agent to register or file returns,
a situation that the interviewees fear may lead to agents preying on ignorance of taxpayers
and overcharging them.
   In Malawi, taxpayer education is a key priority of the MRA. However, due to COVID-19,


                                              28
MRA had to suspend all tax training and educational activities. There is a major concern
that lack of awareness and training may result in low take-up of the upcoming e-filing system.
Finally, in Rwanda, most taxpayers are familiar with basic IT solutions (such as the national
e-government platform, Irembo) and that makes them less resistant to the use of technology
for taxation. However, compliance costs remain high and disproportionately affect smaller
taxpayers, as mentioned in Section 5. The authority is aware that it is common practice
for less educated taxpayers to visit tax centers to get assistance in e-filing by staff, or use
                                              es.11 The launch of the M-declaration app,
tax advisors, or visit designated internet caf´
mentioned above, also shows the intent of the RRA to ease the burden of smaller taxpayers
as the app solution effectively eliminated the need for small business to hire accountants and
thereby reduced their compliance costs (Schreiber, 2018).
       Lastly, our conversations with tax officials confirmed the desk review evidence that tax-
payers may consciously misuse an IT solution with the aim of reducing their tax liabilities
(see, for instance, Mascagni, Mengistu and Woldeyes, 2021). All interviewees are well aware
of such risk, acknowledging that, especially for self-declared income taxes, the quality of
incoming data from new IT options and their broader potential for improving compliance
mostly depend on the taxpayer’s decision on what to declare and how to use the tools. All
interviewees agree that traditional enforcement measures are essential to combat any poten-
tial strategic responses by taxpayers to the new reporting requirements associated with the
use of new technologies.


6.3       Issues with Other Parts of the Government

A key lesson from the country experiences is that technology for taxation works best when
it is embedded within a broad, coherent national transformational vision. In this sense,
the case of Rwanda exemplifies how technology in tax administration belongs to a broader
  11
     A similar pattern is observed in Eswatini, where the ERS provided internet kiosks in all tax centers to
assist less knowledgeable and less equipped taxpayers.




                                                    29
discourse at the national level.12 The sustained investment in infrastructure described in
Section 5 reflects this process. As confirmed by the interviewees, the revenue authority
participates in the national-level reform strategy, which is affecting the digitalization of all
public institutions and putting technology as one of the main priorities. A similar example of
a long-term (2006-2021) digitalization journey is offered by the Mauritius Revenue Authority,
which sequentially introduced more and more sophisticated technologies, from e-filing in 2007
to ITAS in 2010, mobile payments in 2014 and a digitally-enabled COVID-19 assistance
scheme in 2020.13
    Additional evidence on how the institutional vision can influence the success of an IT
strategy comes from the in-depth interviews in Uganda. For the launch of the new taxpayer
e-registration system and integration with the national ID system, interviewees state that
public buy-in will be crucial for the initiative to be successful. They believe that it is not
enough for only a few government players to be supportive, but whole administrative and
institutional bodies need to be seen to embrace these initiatives, through a collaborative
effort between public institutions and constant sensitization of the public. More specifically,
they think the government should coherently incentivize citizens to register with the national
identification systems and with URA. However, from the perspective of respondents, it is
not clear that there is sufficient political will to embark on such an endeavor now.
    Relatedly, the lack of a broad strategy for reform implies that often the tax adminis-
tration, historically more IT-developed, coordinates with external public institutions which
are at a lower IT development stage. This creates hiccups and delays in key functions. An
example of this comes from Sierra Leone, where smooth integration, data sharing or commu-
nication between the NRA and the bureau responsible for registration of traders (OARG) is
hampered by the lack of technical capacity of the latter (due to financial and HR constraints).
  12
     Rwanda is undergoing a successful move towards fiber-optic network and the elimination of duties on
the import of computers and other equipment. These policies, combined with the rapid surge of internet
penetration country-wide, significantly facilitated the introduction of modern technologies (Schreiber, 2018).
  13
     Presentation by Soobhash Sonah, Director of Information System Department at the Nigeria Governors’
Forum Technology and Tax Event on April 19, 2021.



                                                     30
On top of structural differences in IT development across institutions, the persistence of a
silo mentality which hampers inter-institutional data sharing, is another key consequence
of the lack of a uniform, whole-government, IT strategy as reported in Ethiopia (also see
Mascagni and Mengistu, 2019).
    Another key element to consider refers to the legal background in which a given tech-
nology lands. For instance, it is unclear how the tax code adapts to an ever-evolving tech-
nological landscape. In Malawi, the revenue authority could not introduce e-filing solutions
because of the existing tax code, which did not contain specific instructions around the tech-
nology. The authority experienced significant delays with e-filing rollout and is now playing
a more prominent role in informing the tax amendment bills. As a similar challenge, the
tax code is often unclear about the mandatory or voluntary nature of a technology, such as
e-filing, an aspect which can be relevant for its effectiveness (Section 5).
    In addition, regulatory confusion could arise when it comes to data sharing or co-using
a given technology between institutions. Interviewees from Malawi and Uganda believe that
the law stipulates that the tax administration may access any data required to perform its
functions. However, this does not necessarily happen on the ground. In Uganda, private
partners, especially banks, often refrain from sharing data due to confidentiality and political
reasons (see also Santoro and Waiswa, 2022).14 Similar patterns arise in Ethiopia, where
many banking institutions are reluctant to share customer data with the Ministry even
when there are legal obligations on them to do so.
  14
     In 2018, for instance, the URA lost an important battle to access financial transaction information from
banks after vehement protests from the industry. In that case, the government quickly blocked this attempt,
stressing the political element around inter-institutional data sharing. See Santoro and Waiswa (2022) and
the article from Mumbere (2018) ‘Museveni blocks tax body from accessing Ugandans’ bank details’ on
Africa News, (10 April) accessible at https://www.africanews.com/2018/04/10/museveni-blocks-tax-
body-from-accessing-ugandans-bank-details/.




                                                     31
7     Conclusion

Increasing tax revenues is an urgent policy goal in many African countries. This paper
examines the ways in which technology may be used to increase revenue mobilization by
addressing tax administration challenges across four major tax categories: consumption
taxes, real estate taxes, trade taxes, and income taxes. It focuses on the role of information
technology to define compliance by helping the tax administration verify the extent of the
tax base, to monitor compliance by leveraging multiple sources to uncover evasion, and to
facilitate compliance by simplifying processes and reducing tax compliance costs. While
this paper focuses on relatively established technologies being widely implemented in Africa,
such as electronic fiscal devices, we believe that the same considerations and concerns apply
to newer technologies, such as digital merchant payments and digital IDs, that are gaining
momentum in Africa (World Bank, 2016).
    The findings from the literature review as well as the interviews with senior tax admin-
istration officials highlight key lessons for African countries (and other countries at similar
income level) to consider in the process of technology adoption, three of which we highlight
here: equity concerns, taxpayer strategic responses and tax official training and incentives.
New tax technology has the potential to exacerbate inequities among taxpayers. Smaller tax-
payers, those with lower education or other disadvantages may find it more difficult to adopt
new technologies and may bear disproportionate tax compliance costs. Not only should
the introduction of new technology be accompanied by extensive awareness campaigns and
trainings, but governments should also proactively consider the technical, infrastructure and
financial limitations of less privileged taxpayers and develop programs and structures to
support them.
    A different concern with taxpayers is that they may respond strategically to undo the
effects of the new technology on increasing their tax compliance. For example, if a technol-
ogy application provides third-party information on one source of their revenues, they may
compensate by misreporting on other aspects of their tax declaration. This implies that tech-

                                              32
nology does not replace traditional tax enforcement strategies such as audits to expose tax
evasion. Rather, technology can complement traditional enforcement by helping to analyze
large amounts of data across firms to identify taxpayers at higher risk of non-compliance.
   Tax officials are a crucial factor in the successful deployment of technology for taxation.
If they lack adequate training or if their incentives are not well-aligned, they may avoid
or circumvent the new technology. Tax authorities need to provide ongoing training to
tax officials to be able to effectively use IT systems, as well as specialized training to use
the electronic tax data generated for sophisticated analyses to help monitor compliance.
Further, the highest levels of leadership need to establish and enforce standards of conduct
and performance for tax officials to comply with the deployment of the new technology.
   Whereas this paper has focused on the use of technology to administer existing taxes, an
interesting area for further research is to consider how advancements in the use of technology
may change the mix of tax instruments used across African countries. For example, with
greater access to third-party information and employer withholding on individuals’ incomes,
countries may shift to a greater reliance on income and payroll taxes, and a lower reliance
on trade taxes compared to higher income countries.
   Lastly, an important concern with the deployment of technology systems in taxation is
the need to protect citizens from unintended consequences in terms of privacy, confidential-
ity, data leakages and cybersecurity. This aspect is particularly important in Africa. On
the one hand, weaknesses of the legal system and rule of law present a risk of government
surveillance and violation of citizens’ privacy. On the other hand, the nascent nature of data
management systems within tax administration implies that such systems may be vulner-
able to cyberattacks in which confidential data are accessed. As such, countries looking to
expand the use of technology for taxation need to also focus on strengthening the regulatory
framework to protect citizens’ privacy rights and the cybersecurity capacity of all technology
systems deployed.




                                             33
References
Aiko, Rose, and Carolyn Logan. 2014. “Africa’s willing taxpayers thwarted by opaque
 tax systems, corruption.” Policy Paper 7, Afrobarometer.
Ali, Merima, Abdulaziz B. Shifa, Abebe Shimeles, and Firew Woldeyes. 2021.
 “Building Fiscal Capacity in Developing Countries: Evidence on the Role of Information
 Technology.” National Tax Journal, 74(3): 591–620.
Almunia, Miguel, Jonas Hjort, Justine Knebelmann, and Lin Tian. 2022. “Strategic
 or Confused Firms? Evidence from “Missing” Transactions in Uganda.” The Review of
 Economics and Statistics, 1–35.
ATAF. 2019. “African Tax Outlook.” Ataf Publication, African Tax Administration Forum.
   an, Pablo, Augustin Bergeron, Gabriel Tourek, and Jonathan L. Weigel. 2022.
Bal´
 “Local Elites as State Capacity: How City Chiefs Use Local Information to Increase
 Tax Compliance in the Democratic Republic of the Congo.” American Economic Review,
 112(3): 762–97.
Besley, Timothy, and Torsten Persson. 2013. “Taxation and Development.” In Hand-
 book of public economics. Vol. 5, 51–110. Elsevier.
Blimpo, Moussa, Justice Tei Mensah, Ken Ochieng’ Opalo, and Ruifan Shi. 2018.
  “Electricity provision and tax mobilization in Africa.” Working Paper 179, Afrobarometer.
Bratton, Michael, and E Gyimah-Boadi. 2016. “Do trustworthy institutions matter for
 development? Corruption, trust and government performance in Africa.” Dispatch 112,
 Afrobarometer.
Carrillo, Paul, Dina Pomeranz, and Monica Singhal. 2017. “Dodging the Taxman:
 Firm Misreporting and Limits to Tax Enforcement.” American Economic Journal: Applied
 Economics, 9(2): 144–64.
Chalendard, Cyril, Alice Duhaut, Ana M Fernandes, Aaditya Mattoo, Gael Ra-
 balland, and Bob Rijkers. 2020. “Does Better Information Curb Customs Fraud?”
 World Bank Policy Research Working Paper 9254, World Bank Group.
D’Arcy, Michelle. 2011. “Why Do Citizens Assent to Pay Tax? Legitimacy, Taxation and
 the African State?” Working Papers (126), Afrobarometer.
Dener, Cem, Hubert Nii-Aponsah, Love E Ghunney, and Kimberly D Johns.
 2021. “GovTech Maturity Index: The State of Public Sector Digital Transformation.”
 International Development in Focus; Washington, DC, World Bank. ©World Bank,
 https://openknowledge.worldbank.org/handle/10986/36233.
Dogan, Ugur. 2011. “Tools for Risk Management: The Case of Turkey.” In Risk-Based
 Tax Audits: Approaches and Country Experiences. , ed. Sultan Khwaja Munawer, Awasthi
 Rajul and Jan Loeprick. Directions in Development Series. Washington, DC:The World
 Bank.

                                            34
Dzansi, James, Anders Jensen, David Lagakos, and Henry Telli. 2022. “Technology
 and Local State Capacity: Evidence from Ghana.” NBER Working Paper 29923, National
 Bureau of Economic Research.

Efobi, Uchenna, Ibukun Beecroft, Tanankem Belmondo, and Amelia Katan. 2019.
  “Small Business Use of the Integrated Tax Administration System in Nigeria.” African Tax
  Administration Paper 8, International Centre for Tax and Development. Brighton: ICTD.

Eilu, Emmanuel. 2018. “Adoption of Electronic Fiscal Devices (EFDs) for Value-Added
  Tax (VAT) Collection in Kenya and Tanzania: a Systematic Review.” The African Journal
  of Information and Communication, 22: 111–134.

Eissa, Nada, and Andrew Zeitlin. 2014. “Using Mobile Technologies to Increase VAT
  Compliance in Rwanda.” Mimeo, McCourt School of Public Policy, Georgetown University.

Fan, Haichao, Yu Liu, Nancy Qian, and Jaya Wen. 2018. “Computerizing VAT
  invoices in China.” NBER Working Paper 24414, National Bureau of Economic Research.

Fish, Paul. 2018. “Practical Guidance Note: Training Manual for Implementing Property
  Tax Reform with a Points-Based Valuation.” African Tax Administration Paper 2, Inter-
  national Centre for Tax and Development. Brighton: ICTD.

Gadenne, Lucie. 2017. “Tax Me, but Spend Wisely? Sources of Public Finance and
 Government Accountability.” American Economic Journal: Applied Economics, 9(1): 274–
 314.

Gordon, Roger, and Wei Li. 2009. “Tax structures in developing countries: Many puzzles
 and a possible explanation.” Journal of Public Economics, 93(7): 855–866.

Grieco, Kevin, Amou Bakarr Kamara, Niccol´         o F Meriggi, Julian Michel, Wilson
 Prichard, and Graeme Stewart-Wilson. 2019. “Simplifying Property Tax Adminis-
 tration in Africa: Piloting a Points-Based Valuation in Freetown, Sierra Leone.” Summary
 Brief 19, International Centre for Tax and Development. Brighton: ICTD.

Hakizimana, Napthal, and Fabrizio Santoro. forthcoming. “Technology Evolution and
 Fiscal Capacity: Evidence from Rwanda.” Working Paper Series, International Centre for
 Tax and Development. Brighton: ICTD.

Hoffman, Barak D, and Clark C Gibson. 2005. “Fiscal governance and public ser-
 vices: evidence from Tanzania and Zambia.” San Diego: Department of Political Science,
 University of California, San Diego.

International Monetary Fund. 2022. “Tax Policy Assessment Framework (TPAF), VAT
  module.” https: // www .imf .org/ en/ Data/ TPAF .

Isbell, Thomas. 2017. “Tax compliance: Africans affirm civic duty but lack trust in tax
  department.” Policy Paper 43, Afrobarometer.



                                           35
Joshi, Anuradha, Wilson Prichard, and Christopher Heady. 2012. “Taxing the In-
  formal Economy: Challenges, Possibilities and Remaining Questions.” Working Paper 4,
  International Centre for Tax and Development.

Kamara, Abou Bakarr, Niccol´     o F Meriggi, and Wilson Prichard. 2020. “Free-
 town just implemented a new property tax system that could quintuple revenue.” https:
 // www .ictd .ac/ blog/ freetown-new-property-tax-system-quintuple-revenue/ .

Kangave, Jalia, Suzan Nakato, Ronald Waiswa, and Patrick Zzimbe. 2016. “Boost-
 ing Revenue Collection through Taxing High Net Worth Individuals: The Case of Uganda.”
 Working Paper 45, International Centre for Tax and Development.

Keen, Michael. 2012. “Taxation and Development – Again.” Working Paper 12/220, In-
 ternational Monetary Fund.

Keen, Michael, and Alejandro Simone. 2004. “Tax Policy in Developing Countries:
 Some Lessons from the 1990s and Some Challenges Ahead.” In Helping Countries Develop:
 The Role of Fiscal Policy. , ed. Sanjeev Gupta, Benedict Clements and Gabriela Inchauste,
 302–52. Washington: International Monetary Fund.

Kleven, Henrik Jacobsen, Martin B Knudsen, Claus Thustrup Kreiner, Søren
 Pedersen, and Emmanuel Saez. 2011. “Unwilling or Unable to Cheat? Evidence from
 a Tax Audit Experiment in Denmark.” Econometrica, 79(3): 651–692.

Knebelmann, Justine. 2022. “Digitalisation of property taxation in developing countries:
 recent advances and remaining challenges.” ODI Report. London: ODI.

Knebelmann, Justine, Victor Pouliquen, and Bassirou Sarr. 2021. “Developing
 and adopting new digital tools for property tax reform in Dakar, Senegal.” Blog. Inter-
 national Center for Tax and Development https: // www .ictd .ac/ blog/ developing-
 adopting-digital-tools-property-tax-reform-dakar-senegal/ .

Ligomeka, Waziona. 2019. “Assessing the Performance of African Tax Administrations:
  A Malawian Puzzle.” African Tax Administration Paper 14, Brighton: ICTD.

Mascagni, Giulia, and Adrienne Lees. 2022. “The Economic Impact of the Pandemic
 in Rwanda: An Analysis of Firm-Level VAT Data.” Journal of African Economies.

Mascagni, Giulia, and Andualem Mengistu. 2019. “Effective tax rates and firm size in
 Ethiopia.” Development Policy Review, 37(S2): O248–O273.

Mascagni, Giulia, and Fabrizio Santoro. 2018. “What is the Role of Taxpayer Educa-
 tion in Africa?” African Tax Administration Paper 1, International Centre for Tax and
 Development. Brighton: ICTD.

Mascagni, Giulia, Andualem T Mengistu, and Firew B Woldeyes. 2021. “Can ICTs
 Increase Tax Compliance? Evidence on Taxpayer Responses to Technological Innovation
 in Ethiopia.” Journal of Economic Behavior & Organization, 189: 172–193.


                                           36
Mascagni, Giulia, Denis Mukama, and Fabrizio Santoro. 2019. “An Analysis of Dis-
 crepancies in Taxpayers’ VAT Declarations in Rwanda.” Working Paper 92, International
 Centre for Tax and Development. Brighton: ICTD.

Mascagni, Giulia, Roel Dom, and Fabrizio Santoro. 2022. “The VAT in Practice:
 Equity, Enforcement and Complexity.” International Tax and Public Finance.

Mas’ud, Abdulsalam. 2019. “Acceptability of e-Filing of Taxes by Micro-Entrepreneurs in
 Northwestern Nigeria.” Working Paper 96, International Centre for Tax and Development.
 Brighton: ICTD.

Mayega, Jova, Robert Ssuuna, Muhammad Mubajje, Milly I Nalukwago, and
 Lawrence Muwonge. 2019. “How Clean is our Taxpayer Register? Data Management in
 the Uganda Revenue Authority.” African Tax Administration Paper 12, Brighton: ICTD.

Mayega, Jova, Ronald Waiswa, Jane Nabuyondo, and Milly Nalukwago Isin-
 goma. 2021. “How Clean is our Taxpayer Register? Data Management in the Uganda
 Revenue Authority.” African Tax Administration Paper 24, Brighton: ICTD.

Medina, Leandro, and Friedrich Schneider. 2018. “Shadow Economies Around the
 World: What Did We Learn Over the Last 20 Years?” IMF Working Paper No. 2018/017.

Mittal, Shekhar, and Aprajit Mahajan. 2017. “VAT in Emerging Economies: Does
 Third Party Verification Matter?” Working Paper, Anderson School of Management, Los
 Angeles.

Mumbere, Daniel. 2018. “Museveni blocks tax body from accessing Ugandans’ bank de-
 tails.” on Africa News, (10 April) accessible at https: // www .africanews .com/ 2018/
 04/ 10/ museveni-blocks-tax-body-from-accessing-ugandans-bank-details/ .

Obert, Sifile, Kotsai Rodgers, Mabvure Joseph Tendai, and Chavunduka Des-
 derio. 2018. “Effect of E-Tax Filing on Tax Compliance: A Case of Clients in Harare,
 Zimbabwe.” African Journal of Business Management, 12(11): 338–342.

Occhiali, Giovanni, Doris Akol, and Philip M. Kargbo. forthcoming. “ICT and Tax
 Administration in Sub-Saharan Africa: Adopting ITAS in Uganda and Sierra Leone.”
 ICTD African Tax Administration Paper, Brighton: Institute of Development Studies.

Ogembo, Daisy. 2020. “Taxation of Self-Employed Professionals in Africa: Three Lessons
 from a Kenyan Case Study.” ICTD African Tax Administration Paper 17.

Okunogbe, Oyebola. 2021. “Becoming Legible to the State: The Role of Detection and
 Enforcement Capacity on Tax Compliance.” World Bank Policy Research Working Paper.

Okunogbe, Oyebola, and Fabrizio Santoro. forthcoming. “The Promise and Limitations
 of Information Technology for Tax Mobilization.” World Bank Research Observer.

Okunogbe, Oyebola, and Habtamu Edjigu. 2022. “Citizens’ Support for Taxation in
 Africa: The Role of the Social Contract.” Working Paper.

                                          37
Okunogbe, Oyebola, and Victor Pouliquen. 2022. “Technology, Taxation, and Cor-
 ruption: Evidence from the Introduction of Electronic Tax Filing.” American Economic
 Journal: Economic Policy, 14(1): 341–372.

Olabisi, Oyebola. 2013. “Optimising Real Estate Tax in Liberia: Implications for Revenue
 Performance and Economic Growth.” Working Paper, The International Growth Centre
 (IGC).

Pomeranz, Dina. 2015. “No Taxation without Information: Deterrence and Self-
 Enforcement in the Value Added Tax.” American Economic Review, 105(8): 2539–69.

Santoro, Fabrizio. 2021. “To file or not to file? Another dimension of tax compliance
  - the Eswatini Taxpayers’ survey.” Journal of Behavioral and Experimental Economics,
  95: 101760.

Santoro, Fabrizio, Adrienne Lees, M. Carreras, and T Mukamana. 2022. “E-Tax
  Services in Rwanda: Uptake and Impacts.” Working Paper Series. International Centre
  for Tax and Development. Brighton: ICTD.

Santoro, Fabrizio, and Ronald Waiswa. 2022. “Small Nets for Big Fish: Taxing the
  Richest in Uganda.” ICTD African Tax Administration Paper, Brighton: Institute of
  Development Studies.

Santoro, Fabrizio, R. Amine, and T Magongo. 2022. “Mandating E-filing and E-
  payment and Tax Compliance: Evidence from Eswatini.” Working Paper Series 140. In-
  ternational Centre for Tax and Development. Brighton: ICTD.

Schreiber, L. 2018. “A Foundation for Reconstruction: Building the Rwanda Revenue
  Authority, 2001-2017.” Innovations for Successful Societies, Princeton University, 29.

Shah, Jawad A. 2020. “Using Computerized Information to Enforce VAT: Evidence from
  Pakistan.” Working Paper.

Slemrod, Joel, and Tejaswi Velayudhan. 2022. “The VAT at 100: A Retrospective
  Survey and Agenda for Future Research.” Public Finance Review, 50(1): 4–32.

UNCTAD. 2020. “ASYCUDA Partnerships for Sustainable Development – Compendium
 2020.” UNCTAD.

UNU-WIDER Government Revenue Dataset. 2021. “Version 2021.” https: //
 doi .org/ 10 .35188/ UNU-WIDER/ GRD-2021 .

Weigel, Jonathan L. 2020. “The participation dividend of taxation: How citizens in Congo
 engage more with the state when it tries to tax them.” The Quarterly Journal of Eco-
 nomics, 135(4): 1849–1903.

World Bank. 2016. World Development Report 2016: Digital Dividends. World Bank
 Group.


                                          38
World Bank. 2018. “Global Financial Inclusion.” Accessible            at   https: //
 databank .worldbank .org/ source/ global-financial-inclusion .

World    Bank    Enterprise      Surveys.   2010-2020.   Accessible   at    http: //
 www .enterprisesurveys .org .

World Customs Organization. 2014. “Application Of Information And Communication
 Technology – Kyoto Convention Annex 7.” Brussels, Belgium: WCO.

World Customs Organization. 2017. “Annual Report 2016–2017.” Brussels, Belgium:
 WCO.

World Development Indicators. 2018. Washington, D.C.: The World Bank. Accessible
 at https: // datatopics .worldbank .org/ world-development-indicators/ .




                                       39
Figures

                      Figure 1: Taxation Levels by Country and Region
                              Panel A: Taxation Levels by Country




                              Panel B: Taxation Levels by Region




Notes: Tax revenue data comes from UNU-WIDER Government Revenue Dataset (2021) for 2018, the most
recent year with comprehensive tax data across a wide range of countries. GDP data comes from World
Development Indicators (2018).




                                                40
                  Figure 2: Composition of Taxes by Country Income Level
                                 Panel A: Taxes as Share of GDP




                           Panel B: Tax Types as Share of Total Taxes




Notes: Notes: Tax revenue data comes from UNU-WIDER Government Revenue Dataset (2021) for 2018,
the most recent year with comprehensive tax data across a wide range of countries.
- Low income SSA: Rwanda, the Gambia, Togo, Sierra Leone, Somalia, Chad, Mali, Madagascar, Malawi,
Sudan, Guinea-Bissau, Burkina Faso, Democratic Republic of the Congo, Niger, Liberia, Guinea, Uganda,
Central African Republic, Mozambique.
-Lower-Middle Income SSA: Cote d’Ivoire, Zambia, Eswatini, Comoros, Ghana, Lesotho, Sao Tome and
Principe, Benin, Kenya, Senegal, Tanzania, Mauritania, Republic of the Congo, Angola, Cape Verde, Zim-
babwe.
-Upper-Middle Income SSA: Botswana, South Africa, Equatorial Guinea, Mauritius, Namibia.
-High Income SSA: Seychelles



                                                 41
                             Figure 3: Tax Rates across Countries
               Panel A: Taxes on income, profits and capital gains (% of revenue)




                           Panel B: VAT Tax Rates across Countries




Notes: Panel A: The data on tax rates is from World Development Indicators (2018). Panel B: The data
is from International Monetary Fund (2022), Tax Policy Assessment Framework (TPAF) in 2018 over 214
countries. IMF country groups:
-Low income SSA: Burkina Faso, Burundi, Central African Republic, Chad, Comoros Islands, Democratic
Republic of the Congo, Eritrea, Ethiopia, Gambia, Guinea, Guinea-Bissau, Liberia, Madagascar, Malawi,
Mali, Mozambique, Niger, Rwanda, Sierra Leone, Somalia, South Sudan, Tanzania, Togo, Uganda, Zim-
babwe
-Lower-Middle Income SSA: Cameroon, Cape Verde, Republic of the Congo, Cote d’Ivoire, Djibouti, Eswa-
tini, Ghana, Kenya, Lesotho, Mauritania, Nigeria, Sao Tome and Principe, Senegal, Zambia
-Upper-middle Income SSA: Angola, Botswana, Gabon, Mauritius, Namibia, South Africa
-High Income SSA: Seychelles, Equatorial Guinea


                                                 42
        Figure 4: Compliance Costs: Time to prepare and pay taxes (hours per year)




Note: The data is from World Development Indicators (2018). Figure 4 presents how much time, measured
in hours per year, firms spend preparing and paying three major types of taxes: the corporate income tax,
the value added or sales tax, and labor taxes, including payroll taxes and social security contributions.
-Low income SSA: Burkina Faso, Burundi, Central African Republic, Chad, Democratic Republic of the
Congo, Eritrea, Ethiopia, The Gambia, Guinea, Guinea-Bissau, Liberia, Madagascar, Malawi, Mali, Mozam-
bique, Niger, Rwanda, Sierra Leone, Somalia, South Sudan, Sudan, Togo, Uganda
-Lower-Middle Income SSA: Benin, Cabo Verde, Cameroon, Comoros, Republic of the Congo, Cˆ         ote d’Ivoire,
Eswatini, Ghana, Kenya, Lesotho, Mauritania, Nigeria, Senegal, S˜  ao Tom´e and Principe, Tanzania, Zambia,
Zimbabwe
-Upper-Middle Income SSA: Botswana, Equatorial Guinea, Gabon, Mauritius, Namibia, South Africa
-High Income SSA: Seychelles




                                                      43
           Figure 5: Use of Internet and Mobile Phones
        Panel A: Percentage of Population using the Internet




          Panel B: Number of Mobile Phone Subscriptions




Note: Panels A and B: Data from the World Development Indicators (2018).



                                  44
                          Figure 6: Use of Mobile Money
             Panel A: Ownership and Usage of Mobile Money Accounts




  Panel B: Usage of Mobile Money Accounts and the Internet for Bill Payments




Note: Panels A and B: Data from the World Bank Global Financial Inclusion data for 2017.



                                          45
Appendix
Appendix A: In-depth interviews conducted for this study


                                  Tech
                         GovTech          Revenue
 N       Country                 Maturity                                Date
                          Index           Authority
                                  Group
 1                                                                    05/05/21
 2        Uganda            0,62             B            URA         06/05/21
 3                                                                    02/06/21
 4        Malawi            0,47             C          MRA           17/05/21
 5                                                    Ministry of     03/06/21
          Ethiopia          0,33             C
 6                                                     Revenue        22/06/21
 7                                                                    10/06/21
          Rwanda            0,53             B            RRA
 8                                                                    08/07/21
 9     Sierra Leone         0,37             C            NRA         25/06/21
 10       Nigeria           0,49             C            FIRS        22/07/21
 11      Eswatini           0,28             C            ERS         04/11/21
     Roles of interviewees: Manager of Data Warehouse, Head of Business In-
     telligence Computer Centre, Head of Software applications, Directors of ICT,
     ICT Support Experts, Data specialists and business analysts, Heads of research
     departments.




                                        46
Appendix B: In-depth Interview Form
Section I – Technologies
  1. CORE: Broad, introductory

   • What do you think are the current key challenges with technology and priorities in
     technology development your RA has?

   In this session, we want to discuss both more structural, inward-looking technologies, like
the integrated tax management system, and more piece-meal, transactional technologies, like
EFDs and e-filing/payment. Let’s start with the first.

  2. CORE: Integrated system

   • Which IT system or systems does your RA currently use to manage its data and
     operations?

        a. [if not an ITAS] Which are the biggest obstacles to adopt an ITAS?
       b. [if not an ITAS] If you are planning to get an ITAS, when this will happen?
        c. [with or without ITAS] Do departments managing different datasets and systems
           communicate with each other and share information?

   • [if using an ITAS] What do you think are the most significant impacts from ITAS?

        ◦ PROBE I: how does that affect data practices and data quality?
        ◦ PROBE II: what are the impacts on tax administration efficiency?

   • DIVE-IN, BACK-UP QUESTIONS: More broadly, how does the technology used in-
     fluence the data quality? What issues in data quality are due to the technology used
     and/or could be solved with a different technology?

  3. CORE: Technologies for taxpayers
     For each of the following technologies (i) EBM/EFDs, (ii) e-filing and e-payments, if
     they are in place:

   • What do you think are the impacts on Domestic Revenue Mobilization and taxpayer
     compliance?

   • Was facilitating compliance a policy goal of this technology? How is it performing on
     that side? Are you noticing challenges, complaints, or technical difficulties with using
     the technology?

   • Is the new technology improving the quality of data within the RA?

        ◦ PROBE I: how is the new data used for enforcement?



                                             47
Section II – Broader considerations
  1. Staff and IT: Let’s explore the linkages between RA staff and technology: How do you
     think the RA staff is coping with the increasing complexity produced by evolving data
     practices and technology?

       a. PROBE I: What measures did you put in place to encouraging staff to accept
          new technologies and data practices?
      b. PROBE II: Who is in charge of dealing with raw data, cleaning and managing
         data?

  2. Infrastructure: are you facing connectivity or hard infrastructure issues in the imple-
     mentation of technology?

  3. Funding: do you have a sufficient budget allocated to technological improvement?

  4. Legal framework: how do you think the regulatory/legal framework in your country
     helps the uptake of technology by taxpayers? Is the tax code adapting to evolving
     technology? Could you give an example?




                                            48