Getting Targets Right: How Much Revenue can Lower-Income Countries Raise?

Lower-income countries have huge financing needs. While tackling the climate challenge and achieving the Sustainable Development Goals requires substantial capital investments, the Covid-19 epidemic has highlighted an urgent need for spending on health and social protection. One recent report estimated the investment needs for low- and lower-middle-income countries (henceforth, lower-income countries) at US$3 trillion per year by 2030. This represents about 7 per cent of these countries’ combined GDP, and is a herculean challenge, particularly for low-income countries. While loans have long played a central role in bridging the financing gap, rising interest rates have made this even more difficult, and debt servicing costs are increasingly making up frightening proportions of government spending. And yet, on average, low-income countries raise just under 13 per cent of GDP in taxes, versus over 35 per cent on average in OECD economies, and the rate of growth of their tax ratios remains slow.


Introduction
Lower-income countries have huge financing needs.While tackling the climate challenge and achieving the Sustainable Development Goals requires substantial capital investments, the Covid-19 epidemic has highlighted an urgent need for spending on health and social protection.One recent report estimated the investment needs for low-and lower-middle-income countries (henceforth, lower-income countries) at US$3 trillion per year by 2030. 1 This represents about 7 per cent of these countries' combined GDP, and is a herculean challenge, particularly for low-income countries.While loans have long played a central role in bridging the financing gap, rising interest rates have made this even more difficult, and debt servicing costs are increasingly making up frightening proportions of government spending.And yet, on average, low-income countries raise just under 13 per cent of GDP in taxes, versus over 35 per cent on average in OECD economies, 2 and the rate of growth of their tax ratios remains slow. 3 this context, one question has taken on a renewed critical importance for the next decade of financing for development and tackling the climate challenge: 1 Independent Expert Group 2023.
2 If social contributions are included.Average tax-to-GDP for OECD economies is about 24 per cent without social contributions (UNU-WIDER 2023). 3UNU-WIDER 2023.
How much revenue can lower-income countries raise?Recent years have seen some highly ambitious estimates, including an IMF paper that suggested a potential of 9 additional percentage points of GDP.But thinking of this as achievable in the medium term is unrealistic: history suggests that a year-on-year increase of 0.5 per cent of GDP is already very ambitious.We highlight that overly zealous tax targets can be actively counterproductive to tax administration and suggest five concrete ways to set better targets.
How much more tax can developing countries raise?
Recently, a paper published by the International Monetary Fund (IMF) has captured global attention by claiming that, in developing countries, 'a staggering 9 percentage point increase in the tax-to-GDP ratio is feasible' . 4The claim has since become a common talking point in discussions on financing development. 5Building on historical trends and the practicalities of tax administration in lower-income countries, we argue that such a goal is not only misleading but actively harmful to building tax capacity.
While the IMF paper highlights a range of important priorities and a sensible agenda for building tax capacity, it is its headline claim that is more likely to raise eyebrows.
It stands in a long tradition of seeking to estimate the tax potential in lower-income countries, an area with substantial methodological disagreement and estimations ranging from as low as 2 to 3 percentage points 6 to over 107 percentage points.We argue that because high tax potential estimates do not provide a clear account of the time needed to develop tax capacity, and because they are not sufficiently sensitive to national reform agendas and global contexts, high tax potential estimates are a dangerous contribution to policy discussions.They can actively incentivise bad policies as well as misreporting.They can undermine social contracts and create the unrealistic impression that development financing needs can be fixed with tax alone.
We suggest five ways to create better targets and to have a more realistic conversation about the role of expanding tax capacity in development finance: • Make tax-to-GDP targets realistic and time-specific Asking low-income countries to increase tax-to-GDP by 9 percentage points is unrealistic Looking at recent trends in revenue collection highlights the need for a long-term horizon.Building tax capacity is rarely a process of quick and sustained increases.
Short-term changes in tax-to-GDP ratios need to be treated cautiously, as they can be driven by GDP fluctuations, commodity price changes or global pandemics.But looking at them over a longer timeframe is instructive.Figure 1 shows the trajectory of tax-to-GDP ratios8 in lower-income countries over the last two decades.On average, the picture is one of slow growth in the early 2000s, of about 2 percentage points on average, and relative stagnation in the 2010s.This perspective highlights the enormity of the task at hand: for the average lowincome country, increasing their tax-to-GDP ratio by 9 percentage points would imply an expansion of their tax take by more than two-thirds of its current level.
Unsurprisingly, in the last 20 years, only one low-income country has increased their 4 Benitez et al. 2023. 5 For instance, at a high-profile tax panel at the IMF/World Bank annual meeting in Marrakech in October 2023, Germany's Minister of Finance Christian Lindner pointed to the paper as evidence of the substantial tax potential of developing countries. 6McNabb, Danquah and Tagem 2021.tax-to-GDP ratio by at least 9 percentage points -Mozambique.A small number of low-income countries saw a more moderate 5 percentage point increase, but only two -Rwanda and Mozambique -have achieved such increases since 2010.9Furthermore, even 5 percentage point increases among the best performing low-income countries take a substantial period of time to realise: it took 12 years in Togo, eight years in Rwanda and Malawi, five years in Burkina Faso, and six years in the DRC. 10 Notably, increases have usually not been linear, but included setbacks and fluctuations.In addition, most of these countries have achieved these increases in the context of postconflict recovery or natural resource discovery or exploitation.Figure 2 summarises average year-on-year changes in tax-to-GDP ratios, both among low-and lower-middle income countries in general, and among the 'top performers' which saw an increase of at least 5 percentage points over the past three decades.
They highlight that even a 0.5 percentage point year-on-year increase, which is often used as a target in IMF support programmes, is highly ambitious.The left-hand graph shows that from 1990 to 2020, lower-income countries saw only a 0.1 percentage point increase in tax-to-GDP on average.Among the top performers, this average was 0.2 percentage points.A sustainable 9 percentage point increase, history and experience suggests, will take decades.These trends are not surprising to anyone engaging with building tax capacity in lowerincome countries.The gradual nature of revenue mobilisation reflects critical contextual and institutional factors such as the technical constraints, politics, the quality of service provision, and perceptions of government accountability, which develop along their own timeline, and do not always improve linearly.Measures of tax potential and tax gaps are victim to well-known limitations, sensitivities, and imprecisions.12Most importantly, however, they abstract away from the role of time.They quantify an aggregate tax gap, creating the illusion that it can be swiftly addressed and providing little guidance on how this should be done.This is not merely a matter of being ambitious, but can have direct negative consequences.We would argue that getting tax targets right should include a fundamental methodological shift: rather than being based on unrealistic estimates of aggregate tax gaps, they should be guided by history and experience.Doing so would result in much more realistic and achievable goals, thus mitigating some of the risks we discuss below.

Tax-to-GDP targets can do more harm than good
Targets have a direct impact on how policy is made and implemented, and on how revenue authorities work.Broad statements on tax-to-GDP ratios, like those advocated in the recent IMF publication,13 are not only methodologically dubious but can actively undermine the goals they are supposed to achieve.They can set damaging incentives by ignoring the length of time required to achieve sustained improvements and generating benchmarks that are removed from the political, economic, and administrative context.

Overly ambitious targets incentivise bad policy and administration
While some countries have indeed seen substantial year-on-year increases in their tax-to-GDP ratios over the last two decades, not all of them provide models to follow.Mozambique, for example, saw their tax-to-GDP ratio nearly double since 2013.However, a closer examination of the factors behind this performance shows that the increase was achieved through 'compromised efficiency rather than better enforcement' ,14 including a lack of VAT refunds being paid and a growing number of exemptions on business inputs which led to tax cascading and an artificial expansion of the VAT base.
High-level tax-to-GDP targets can incentivise the adoption of ill-advised policy or administrative measures.Pressed to reach unachievable tax-to-GDP targets, governments might force taxpayer registration through mass registration campaigns, bloating the tax register and hoping for revenue gains that in most cases do not materialise. 15erly ambitious targets may also increase pressure to introduce new taxes that disproportionately burden lower-income citizens, such as recent experiences with mobile money taxes in Ghana and Uganda. 16Pursuing revenue to the exclusion of other policy objectives might also contribute to an increase in poverty. 17Especially when facing short time horizons, governments striving to reach a tax-to-GDP target are likely to try and extract more revenue through instruments that also heavily affect lower-income groups. 18 addition, all governments face constraints on how many capacity-expanding reforms or innovations they reasonably can implement at a time.Ambitious targets may lead them to tackling too many at once, which may limit their effectiveness.Or they may lead to the de-prioritisation of essential reforms which may not yield large revenue increases in the beginning.For instance, at least part of the reason for Rwanda's success in sustaining their growth in tax-to-GDP has been the staggered nature of reform efforts.Key administrative measures included introducing electronic registration, filing and payment systems, enhanced staff training, improving risk management approaches, automating customs operations, introducing electronic billing machines, and expanding taxpayer education on these new tools.These administrative measures were complemented by tax policy adjustments to reduce exemptions, to raise several tax rates, and to facilitate taxpayer compliance. 19Nepal has a similar story, as policy reforms in the late 1990s and early 2000s were subsequently consolidated by modernising the tax administration, including adopting new technologies and organisational restructuring. 20A successful reform strategy often starts with less contentious policy reforms to build momentum which is then consolidated through administrative reforms which might take longer to show results.
The performance of tax administrations in many low-income countries is already assessed by their ability to deliver the revenue target set in annual budgets.Placing more emphasis on a one-dimensional indicator, which encourages short-term revenue gains, can undermine efficiency and equity.These targets do not reward tax administrations for efforts to simplify the compliance process, provide good services to taxpayers, or improve internal processes.Instead, they can incentivise counterproductive measures in pursuit of 'quick wins' on a narrow revenue objective.

Tax-to-GDP targets incentivise misreporting
Tax-to-GDP ratios are influenced by GDP figures just as much as they are influenced by revenue figures.Consequently, such targets are also sensitive to the vagaries of government accounting and GDP (mis)reporting.Governments are already known to under-or over-state GDP 21 or use real GDP rebasing episodes strategically. 22For instance, following a rebasing in 2019, the Ugandan economy was found to be '11.6 per cent larger 15 Gallien et al. 2023. 16Akua Anyidoho et al. 2022; Lees and Akol 2021. 17A recent report (World Bank 2023) found that in two-thirds of low-and middle-income countries, poor households are worse off after paying taxes and receiving transfers or subsidies. 18This point was made in a 2015 ICTD blog in response to including tax targets as part of the Sustainable Development Goals.It is equally valid almost a decade later. 19 Akitoby et al. 2019. 20 USAID 2018. 21Martinez 2022. 22Nigeria became Africa's largest economy after recalculating their GDP, and the head of the National Bureau of Statistics reportedly said that 'the only instruction the President had given him was to ensure that the recalculated GDP was larger than South Africa's' (Dercon 2022).
than earlier thought' , 23 which also reduced the country' s tax-to-GDP ratio by at least 1 percentage point, putting the achievement of their medium-term revenue target of 16 per cent of GDP at risk.Targets also incentivise other forms of creative accounting that do not actually increase tax capacity.For instance, in 2015, the Pakistan Board of Revenue was exposed for appearing to meet annual tax collection targets by borrowing large sums from big banks in the form of 'advanced taxes' . 24x-to-GDP ratios are also influenced by various aspects outside of government control, such as commodity price fluctuations or external crises like natural disasters or the COVID-19 pandemic.This is particularly true for economies that rely heavily on natural resource exploitation.Returning to the example of Mozambique's stellar tax performance between 2013 to 2020, this also coincided with a period of significant economic deceleration.Part of the tax increase was derived from CIT linked to a fledgling extractive industry, which was less responsive to economic cycles.
Consequently, external shocks can dramatically impact whether targets are achievable, or even challenging, substantially skewing the incentives of revenue authorities and other actors.Importantly, these external factors place success outside of a government's control, because they overemphasise a final, individual target -rather than assessing the process and measures taken to sustainably improve tax collection in a more comprehensive way.

Getting ahead of institutional change undermines social contracts
A substantial increase in tax collection reflects a modification of social contracts.
Along with states seeking to control a much larger chunk of resources come new public expectations of accountability, transparency, and, most importantly, improved and expanded service provision.In fact, many citizens in lower-income countries already feel that they are not being provided with services commensurate with their current tax levels. 25creasing tax takes in a way that is compatible with changing social contracts requires expanding state capacity in other areas.Governments must expand their ability to provide services, as well as make credible commitments to citizens and businesses to provide transparency and accountability.This is the underlying -but often implicit -reasoning behind the need for wider institutional change that often accompanies projections of tax potential.The recent IMF study, 26 for example, estimates that, on average, 2.3 percentage points of the potential 9 points of GDP can only be achieved if tax reforms are accompanied by wider reforms bringing low-income countries up to the average level of institutional capacity in emerging economies.However, everything that we have highlighted with respect to building tax capacity so far is similarly true for wider institutional change: it is typically slow and non-linear, and many measurable results may take longer than a government's four-or five-year term to arrive.Importantly, it is not purely technocratic, but deeply embedded in the normative politics of each country.
If tax targets are out of step with wider institutional and political development, they risk undermining social contracts, and threatening both public acquiescence to and political support for tax reforms.The recent introduction of new taxes and increased tax rates from Kenya to Ghana have been accompanied by widespread public opposition.Given the wider context of cascading economic and health crises, these dynamics are hardly surprising.

Five ways to create better targets
Poorly set targets are not only of little use to guide reform but can be actively harmful to reform goals.However, not all targets are created equal.There is an obvious need to track progress on revenue mobilisation, and targets have their place in policy making and building tax capacity.Targets can be useful when they help governments to track progress, set expectations, and identify potential reform priorities.We argue that there is scope for tax targets that improve upon the narrow tax-ratio ones that have historically dominated the debate.The key challenge then is to 'get targets right' .
How can targets be improved to be better suited to these purposes?We draw on examples from the 'top performers' in recent decades to suggest five principles that can help to improve targets for expanding tax capacity in low-income countries.They all emphasise the role of time, and a closer engagement with the realities and practicalities of expanding tax capacity in low-income contexts.Our focus here is not on proposing particular indicators and target levels -that lies beyond the scope of this brief -but suggesting guidelines and priorities in their development.

Set tax-to-GDP targets realistically
Estimating tax gaps is a bit of a holy grail amongst tax practitioners, hoping that knowing how much revenue is missing, and from where, can outline a targeted reform agenda to fill the gap.However, this is inherently difficult in practice, and estimation methods are necessarily imperfect.The utility of tax gaps in setting tax targets is limited both by their methodological flaws, and because they do not outline a timeframe for reaching that 'potential' , thereby setting utopian expectations.This does not mean that we need to drop tax-to-GDP targets altogether.Setting better tax-to-GDP targets requires them to be time-specific and realistic, based on the history and experience of lower-income countries.Based on the data we have presented above, a 0.5 percentage point year-on-year increase in the tax-to-GDP ratio is already ambitious and sustaining this over multiple years is extremely rare.Importantly, such a target should be seen as one part of a broader set of indicators (or the set of targets), as we describe below.

Prioritise sustainable revenue gains
Better targets recognise that increasing tax collection is not a linear process.Reform efforts take time before they are translated into higher tax-to-GDP ratios, and increases are not always automatically sustained.Many of the most successful interventions, such as adopting a value-added tax, eliminating unproductive exemptions, or setting up a semi-autonomous revenue authority (SARA), are associated with large, positive step-changes in tax performance.However, those increases are not always sustained over time: they need attention and 'maintenance' in the form of monitoring, tweaks, and adjustments, and continued administrative effort.Better tax targets should specifically measure the sustainability and buoyancy of revenue gains, setting clear expectations and incentives over longer time horizons.Emphasising sustainability as a key aspect of the target set -such as only considering an increase as achieved if it has been sustained over multiple years -encourages governments to prioritise reforms that yield sustainable revenue gains over time, as opposed to quick wins that might undermine economic activity, equity, or compliance culture in the longer term.

Incorporate reform sequencing and timelines
Improving tax administration and building tax capacity does not generate linear improvements.A well-sequenced reform agenda acknowledges that revenue gains from different kinds of measures have different timelines.
Policy Brief tax registration provides a good example here. 30Better tax targets should instead place tax-ratio targets firmly within a broader set of targets -some of which might not have immediate or quantifiable revenue implications.

Adjust tax targets to political contexts
Political factors have a substantial impact on the potential for expanding tax collection.
On a domestic level, this includes political budget cycles and competing reform projects, but also attention to the longer timelines for institutional and distributional processes that are required to situate tax increases within changing social contracts.
At a more global level, rising interest rates and lending arrangements have meant that large proportions of many low-income country budgets are now spent on debt service costs, substantially limiting the space for service expansion.This complicates states' ability to situate increased tax takes in direct fiscal reciprocity.In the medium term, this 27 Moore and Prichard (2017) discussed how lower-income countries can increase their tax revenue.Many of those prescriptions are still relevant in the post-pandemic world. 28The IMF have been heavily involved in many of the efforts to develop medium-term revenue strategies.Benitez et al. (2023) also highlight the value of these strategies in driving tax system reforms. 29While the number of countries working on designing a medium-term revenue strategy is growing, very few have adopted and implemented such a strategy (Papua New Guinea, Indonesia and Uganda are among the few lower-income countries with published strategy documents).This points to the challenging nature of developing these strategies.provides another argument for cautious tax targets.It also means that better tax targets need to be designed with a view not merely to the timing of a wider tax reform agenda, but also to the timing of other public reforms, especially around service provision.
Considering the political sustainability of tax targets also requires their expansion beyond total revenue, including other aspects that make them more politically acceptable, including accountability, transparency and equity.
Unrealistic targets distract from the need for other financing support Building stronger tax capacity in low-income countries is essential to foster sustainable development and strong, independent states.This is a substantial task and needs substantial support in the form of technical assistance, financial support, capacity building, and research engagement.Big and ambitious efforts come with big goals and ambitious targets -and there is value in optimism.However, as we have argued here, there are downsides to bold proclamations.
Overly broad targets and excessively optimistic expectations can create skewed incentives and be counterproductive.They also risk creating a narrative where countries and government agencies are framed as at fault for failing to meet objectives that are fundamentally unrealistic.With this comes a risk of disillusionment, as well as a risk of narrowing attention to how much is collected, at the expense of how and from whom -all of which can have substantial negative long-term effects.
Perhaps the most important case for realism, however, is not about domestic incentives, but international ones.Overstating the potential for rapid increases in tax capacity risks creating the impression that tax alone can be enough to solve the development financing challenges in low-income countries.It cannot -neither in the short, nor the medium term.Nor can it be pointed to in order to frame increasingly challenging public debt levels as almost sustainable.Any conversation about public finance in lowincome countries needs to also be a conversation about other forms of international financial assistance -be they aid, debt restructuring or cancellations.The scope of the challenges ahead and the difficult international context make these conversations more urgent, and more fundamental.Realistic targets can show that tax capacity is an important aspect of these conversations -but not an alternative.
Policy Brief

•
Prioritise the sustainability of revenue gains • Sequence reforms and consider timelines explicitly • Diversify the criteria used to assess the performance of tax administration • Adjust tax targets to political contexts.

Figure 1 :
Figure 1: Tax-to-GDP ratios (including all direct, indirect and resource taxes, but excluding social contributions) 11 for low-income countries in grey (excluding those without enough data over the time frame) plus averages for low-income and lowermiddle-income countries.Source: UNU-WIDER 2023.

Figure 2 :
Figure 2: Average year-on-year changes in tax-to-GDP ratios for low-income countries and lower-middle income countries, 1990 to 2020.The graph on the right limits the sample to the 'top performers' , countries which saw at least a 5 percentage point growth in their tax-to-GDP ratios in this time period.The dotted line represents the average across this period.Source: UNU-WIDER 2023.
International Centre for Tax and Development www.ictd.ac • Number 6 • March 2024 International Centre for Tax and Development www.ictd.acBettertargetsthereforeneed to look at past increases as imperfect predictors of immediate potential for further increases.While the menu of potential tax reforms is long, 27 many of the low-hanging fruits have already been plucked in many lower-income countries.While introducing the VAT or adopting electronic billing machines usually results in immediate revenue gains, these cannot be repeated by introducing the same measure again.Increasing revenue further requires improvements in analytical capacity (for instance, to make full use of the wealth of data coming from VAT and electronic billing machines) and traditional functions (such as well-trained auditors who can perform efficient checks of complex businesses).These improvements are harder to obtain, usually require large upfront investments, and might take longer to show up in revenue increases, but are essential to building tax capacity.Better tax targets should consider the reform history and agenda in each context, and the timeline required for those reforms to yield revenue gains.Increasing numbers of lower-income countries are already moving in this direction by developing holistic reform plans covering the whole tax system, such as medium-term revenue strategies, 28 which link targets to specific reforms, properly sequence interventions, and prioritise developing implementation plans early on.29Bettertargetsshouldtake these sequences into account, setting specific timeframes in which revenue gains can realistically be achieved.4.Diversify the criteria used to assess the performance of tax administrationBetter targets should move beyond tax-to-GDP ratios and revenue targets toward more specific criteria, including reform processes.The process of increasing revenue, as well as how that revenue is generated, are just as, if not more, important as how much revenue increases.This calls for a diversification in the criteria used to assess the performance of tax administrations.The African Tax Initiative and the • Number 6 • March 2024 International Centre for Tax and Development www.ictd.ac