What Would Happen to Poverty in Africa if Most Aid Were Delivered as Social Cash Transfers? A Case Study of Uganda

Abstract

This article explores the potentially game-changing idea of repurposing a major part of overseas development assistance (ODA) to fund the up-scaling of social cash transfers (SCTs) in Africa. The analysis contributes to the debate about how best to use global development aid by considering the idea of simply bypassing the aid industry and instead giving most funds directly to the poor. With Uganda as a case study, we use the UGAMOD microsimulation model to analyse the socio-economic impact of using a major part of the current aid envelope, which for 2017–2020 was on average 2.37 billion USD, to provide two types of SCTs, namely old-age pensions and child grants. The scenarios analysed include both universal and means-tested SCTs at different cost levels, ranging between 4 and 115 per cent of current annual ODA in Uganda. We demonstrate that allocating a major part of ODA to SCTs would, seen in isolation, lead to very considerable reductions in poverty. In one tested scenario, where most ODA is allocated for universal child support, it is predicted that about two-thirds of current poverty would be eliminated. Importantly, however, large-scale SCTs would also come with significant socio-economic allocation costs, as aid is shifted away from its current uses, much of which goes to the social sectors. We discuss what this may imply for government revenue generation and the timing of any potential scaling of SCTs. Lastly, we note that scaling of SCTs would have political economy implications, which would need to be better understood.

Keywords:

• social cash transfers
• aid policy
• economic modelling
• innovative aid
• Africa
• Uganda
• scaling of social cash transfers

1. Introduction

Simply giving money to the poor in the form of social cash transfers (SCTs) has shown itself to be an effective¹ and relatively cheap² tool for creating socio-economic impacts in the Global South (Bastagli et al., 2016; Beegle et al., 2018). SCTs proved especially workable in the COVID pandemic and have seen what Gentilini (2022) calls ‘the largest scale-up in history’. Financed by governments as well as donors, SCTs such as child support, old age-pensions, disability benefits and COVID relief have spread like wildfire across the Global South³ (Beegle et al., 2018; Gentilini, 2022; Hickey et al., 2020; Lavers and Hickey, 2021). SCTs already make up a sizeable 20 per cent of global humanitarian aid (Cash Learning Partnership, 2020) and a smaller but growing share of development aid.

SCTs have also been critiqued. Torkelson (2020) shows how the bundling of social cash transfers with financial services in South Africa can worsen indebtedness and de Sardan and Piccoli (2018) question the uniform applicability of SCTs. Some stakeholders suggest that SCTs are less effective on their own and advocate a Cash Plus approach.⁴

At the same time, current aid seems to be having a hard time making a sizeable dent in poverty. A number of relatively stable countries politically, such as Uganda, Rwanda, Mozambique, Tanzania, Ghana, Malawi and Zambia, have for decades received significant levels of development aid and experienced sizeable economic growth but still struggle with persistently high poverty levels of 20–54 per cent as shown in Table 1.

Table 1: Aid, GNI and poverty indicators for selected countries^a.

	Aid per capita	GNI per capita	Poverty^b
Ghana	68	2280	23%
Malawi	75	620	51%
Mozambique	82	480	46%
Rwanda	124	840	38%
Tanzania	37	1100	26%
Uganda	69	760	20%
Zambia	54	1030	54%

^a All data from https://data.worldbank.org/indicator.

^b Poverty headcount ratio at national poverty lines (% of population).

This begs the question of whether SCTs, which are often enabled by new technologies in the form of smartphones and mobile banking, could be a cheaper and more effective way of delivering poverty alleviation than current aid. What would happen to poverty if a much larger share of ODA were to be allocated to SCTs? This article contributes to answering this question by examining the expected poverty-reduction impact of various SCT large-scaling⁵ scenarios in the case of Uganda.

Donor funding for SCTs has typically been granted for two purposes: (1) to provide immediate relief for recipients, in particular as part of humanitarian aid; and (2) to help governments in the Global South build their own social protection systems, which are seen as an integral part of how a responsible state takes care of its vulnerable citizens. However, ambitious ‘out of the box’ thinking is emerging which goes beyond those two purposes to discuss large-scaling SCTs in the Global South as a tool for eradicating global poverty. Globally, for example, in 2016 the Brookings Institution made a very rough estimate that $100 billion in cash transfers annually would be sufficient to eliminate the poverty gap (Brookings, 2016). Nationally, GiveDirectly, a leading provider of SCTs, is currently at an early stage of assessing what amount of SCTs it would take to lift a single country entirely out of poverty.⁶ A recent paper (Ulrich et al., 2022) explores the thought experiment of allocating a major share – for example, 50 per cent or more – of all aid to SCTs over time, and suggesting that this has the potential to lead to immediate large-scale socio-economic impacts.

Against that backdrop, this paper explores the potential socio-economic impacts of allocating large proportions of aid to SCTs in Uganda. We consider Uganda an interesting case. It has been characterised as authoritarian and a hybrid regime (Tripp, 2010) but also has relative political stability, a growing economy with a significant share of its people still living in poverty, relatively stable relations with multiple donors and sizable support, and emerging but nascent social protection systems. The lessons learned in Uganda may be relevant for roughly comparable countries such as Ghana, Malawi, Mozambique, Rwanda, Tanzania and Zambia, as shown in Table 1. We are not assessing potential partner arrangements or implementation mechanisms⁷ for any scaling. Instead, our economic modelling assumes implementation by governments, and the scenarios we analyse are based on current and specific government SCT schemes or proposals. This means that we are focusing on two types of SCTs, namely child support and old-age pensions, both of which are support aimed at people (children and pensioners) who, because of their age, are by definition outside the labour force. We could also have focused on other types of potential recipients, such as the disabled, the unemployed, the poorest of the poor, refugees, or a particular area. However, we have prioritised SCTs, which would have a large reach given the national demography (child support) and/or are as close as possible to current government schemes or thinking (old age pension).

Using UGAMOD, a static tax-benefit simulation model,⁸ we show how ten specific SCT scenarios would have varying impacts on both rural and urban poverty and inequality and would cover both male- and female-headed households. This gives us an indication of the effectiveness and potential reach of scaling ODA-funded cash transfers in Uganda. We explore and discuss the cost of these SCTs in relation to the ODA resource envelope available for Uganda, including the administrative costs.

The key finding of our analysis is that repurposing substantial parts of current ODA to scale up SCTs in Uganda is predicted to lead to relatively large reductions in poverty levels. It is a banal and trivial point that giving poor people money results in reduced poverty, but the scale at which this could be achieved within current aid budgets in Uganda is far from trivial. Our data suggest that, seen in isolation, the direct impact of allocating different (large) levels of the current ODA envelope to SCTs would reduce national poverty by close to one third or two thirds. This is obviously, seen in isolation, a remarkable potential impact.

However, there are also important potential socio-economic costs that our model does not address, as well as other political and economic costs, uncertainties and risks that would need to be better understood. The socio-economic reallocation cost stands out as perhaps the most important and is not captured by our UGAMOD-based modelling. By socio-economic reallocation cost, we mean that, if the ODA received by Uganda were to be shifted to SCTs, it would be allocated away from its current purposes, 55 per cent of which goes into social infrastructure and services,⁹ with a significant share of this percentage being spent on medicines and activities related to HIV/AIDS. A reduction of this funding for social sectors could have negative socio-economic impacts. We discuss the limitations this places on large-scaling SCTs, as well as overall potential pathways for managing the challenge, including the links to government revenue and taxation.

The rest of the article is structured as follows. Section 2 discusses the key literature and positions our analysis and theory of change within relevant academic debates. We then move on to set the scene in Section 3, which introduces the existing national SCT schemes, as well as the ODA that Uganda currently receives. This is followed by our analysis in Section 4, which uses the UGAMOD model to assess the potential impact of ten selected scenarios of scaled SCTs. We relate this cost to the current ODA resource envelope and consider the administrative costs. Section 5 discusses the socio-economic allocation costs and the links to government revenue and taxation. Section 6 provides a conclusion.

2. Literature review and positioning of research

This article, with its focus on a potential major shift of aid to SCTs, naturally falls within at least two debates. The first is a long-standing discussion about the overall role of development aid, with some arguing that aid works (Doucouliagos and Paldam, 2015) and that the key problem is that there is not enough of it (Sachs, 2005). Others, such as Easterly and Pfutze (2008) and Deaton (2015), claim the exact opposite and suggest aid produces negative incentives and that it mainly ends up in expensive and ineffective bureaucracies. More recently, Glennie (2021) has suggested that the global architecture needs a complete overhaul, a new balance of power and more long-term financial commitments, potentially modelled on EU’s structural funds. It would go beyond the scope of this article to unpack the wider debate about aid. Instead, we focus on the other key debate, which is more specifically about SCTs, including the role of donors.

In recent decades, SCTs in the Global South have been documented and analysed extensively and from many different perspectives, several of which are important in informing our analysis of ODA-financed scaling. Bastagli et al. (2016) and Beegle et al. (2018) document the positive socio-economic impacts of SCTs. Bastagli et al. (2016) include a review of 165 studies covering 56 cash-transfer programs, primarily in Africa and Latin America, and document reductions in monetary poverty, increases in food expenditure and increases in school attendance by children, as well as improved health outcomes and increased use of health services. Social cash transfers have also resulted in overall increases in participation in work and its intensity among adults of working age, coupled with reductions among the elderly, those caring for dependants and those in casual work, as well as reductions in child labour.

de Sardan and Piccoli (2018) and Fisher et al. (2017) document the contextuality and very diverse livelihood impacts and beneficiaries’ perspectives of SCTs, while Araujo and Macours (2021) present evidence of intergenerational sustainability of the poverty-reduction effect. Warwick et al. (2022) compare the redistributive powers of SCTs and VAT exemptions in Africa and suggest that SCTs have better effects. Barrientos (2012), Egger et al. (2019) and Beegle et al. (2018) provide analysis and documentation of the links between SCTs, economic growth and multipliers. Egger et al. (2019) suggest an economic multiplier of 2.6 for SCTs based on empirical data from Kenya.

There has also been criticism of SCTs for potentially creating an idle dependency culture, but, although this view appears to be prevalent among citizens, elites and decision-makers in Africa and elsewhere (as discussed by Seekings, 2016, 2019), it not supported by much, if any, evidence (Noble et al., 2007). Overall, the literature points to the positive socio-economic effects of SCTs in terms of both incomes and broader socio-economic indicators, albeit with different results in different contexts.

This current paper is about the socio-economics of scaled SCTs, not about the politics. A separate paper (Ulrich, 2024, unpublished manuscript) explores the political economy of scaled SCTs in Uganda in more depth. Here, however, it is worth highlighting that the politics behind the introduction and roll-out of current social protection and SCTs in Africa are important and have been extensively analysed. Bender et al. (2021) and Hickey et al. (2020) document how a combination of international policy alliances and competitive domestic politics, as well as crisis situations (Lavers and Hickey, 2021), have been key factors in introducing and/or institutionalising SCTs. Bukenya and Hickey (2020) describe how political settlements have influenced social protection (and health insurance) in Uganda. Seekings (2016, 2019, 2021) analyses how domestic politics and ideology, including a widespread belief in productivism and a scepticism of the efficacy of hand-outs, may work either for or against increased social protection (including SCTs) in Africa. The politics and influence of private capital has also been analysed. Mader (2016) and Duvendack and Mader (2017) expose the very large economic interests and profits that international financial corporations can earn from micro-credits (which have many similarities to SCTs) in the global South. Torkelson (2020) shows how SCTs in South Africa got caught up in scandal when an implementing tech/finance corporation took undue advantage of its involvement.

Some studies discuss the current or future scaling of SCTs, including as a potential key modality for ODA. As mentioned in the introduction, the Brookings Institute analysed the potential impact of the large-scaling of non-donor-funded SCTs in eradicating poverty globally. Hanlon et al. (2010) suggested distributing SCTs as an effective alternative to other forms of ODA. More recently Gentilini (2022) has described the current COVID-related proliferation of SCTs as ‘the largest scale up in history’¹⁰ (ibid.: title page). However, few if any academic studies have systematically examined what the impacts and feasibility in a given country would be of spending a major part of the available ODA envelope on SCTs. Given the potential positive impacts, this seems to be an important gap in the literature which our analysis aims to address.

We do so by exploring a theory of change which comes in two parts. The first part is simple and suggests that shifting a major proportion of aid to SCTs in a given ODA-receiving country, in our case Uganda, would potentially lead to immediate and widespread socio-economic impacts, as sizeable groups of the poor would receive cash infusions. Our theory would suggest that the impact would continue for at least as long as the cash transfers are given, but possibly also for longer.¹¹ The interesting part for us is not whether SCTs will have a positive socio-economic impact: on the contrary, the literature discussed above is clear that they will. Instead, in our analysis, the interesting element is the reach, i.e. how big a dent in national poverty could be achieved by scaling SCTs within the existing aid envelope.

The second part of our theory of change is more multi-facetted and relates to the socio-economic reallocation costs. What happens to the social sectors that are currently largely funded by ODA if a major part of ODA is shifted to SCTs? We suggest that there are three potential pathways, which we label ‘the good’, ‘the bad’ and ‘the moderate’. The ‘good’ pathway is tax-financed and involves reallocating ODA, including ODA currently spent on the social sectors, but with the government gradually taking over social-sector expenditure financed through increased taxation. It would, of course, always be the long-term vision for government to fully take over funding from the donors for the social sectors, but there would be short- and medium-term challenges in doing so related to political and economic feasibility. The ‘bad’ pathway would involve re-allocating a significant share of ODA, including the share currently spent on the social sectors, but without the government picking up the cost. This would make significant funding available for SCTs, but would predictably also lead to substantial parts of social services collapsing. The ‘moderate’ pathway would involve ring-fencing the share of ODA that is currently allocated to the social sectors and only consider reallocating non-social sector ODA. This would result in a smaller part of the ODA envelope being available for scaled SCTs, which would lower the potential reach, but it would also reduce the socio-economic risks. We have not modelled these three pathways in UGAMOD, but we do include a discussion of their feasibility in Section 5.

We are under no illusion that the ODA-financed scaling of SCTs is a uniformly applicable miracle cure. On the contrary, any major shift to a new ODA modality (in this case, large-scaled SCTs) comes with complexities and specifics that will have to be contextualised and understood. With a tone of scepticism, de Sardan and Piccoli (2018) label cash transfers ‘a travelling model’ due to their widespread application and the trust (presumably too optimistic) that the aid industry shows in the concept. In our analysis, we have aimed to explore a corner of the complexity by showing how the socio-economic impacts of scaled SCTs might potentially unfold in one specific country: Uganda.

We continue below by setting the scene and presenting Ugandan SCTs as they are today and the ODA that the country receives.

3. Setting the scene: current SCTs and ODA in Uganda

3.1. Current SCTs in Uganda

Currently SCTs in Uganda are delivered primarily by a wide range of mostly international NGOs and UN agencies (Nystrand et al., 2024, forthcoming). They provide SCTs to different people and groups in society, but mostly in refugee areas and areas of extreme poverty, such as Karamoja. The only government SCT program currently falls within the framework of the Social Assistance Grant for Empowerment (SAGE), which is part of the Social Protection Programme being implemented by the Ministry of Gender, Labour and Social Development (MoGLSD). Within this framework, a Senior Citizen Grant (SCG) provides cash transfers to the elderly and primarily targets citizens over the age of 80, who receive a small monthly grant of UGX 25,000, equivalent to less than 7 USD.

Despite the hitherto limited public-sector involvement in SCTs in Uganda, there seems to be an emerging government interest in them (Bukenya and Hickey, 2020).¹² A recent analysis of direct income support by Save the Children (2021) involved an extensive dialogue with the government, in particular MoGLSD, and argues for the gradual implementation of a universal child benefit in Uganda (Save the Children, 2021). This analysis is complemented by a comprehensive Government of Uganda (GoU) review of specific potential child-benefit schemes in Uganda (GoU, 2016) and a corresponding UNICEF Policy Brief (2017). The positive interest of MoGLSD is balanced by a more sceptical and cautious position by the Ministry of Finance, Planning and Economic Development (Bukenya and Hickey, 2020).

Nystrand et al. (2024, forthcoming) argue that SCTs are playing an increasing role in national Ugandan politics, with the government seeing an opportunity in the political gains and patronage that SCTs can provide. For example, the government’s old-age pension scheme, mentioned above, was expanded due to direct intervention by President Museveni in response to a request from Ugandan citizens. In addition, the recent government suspension of the work of GiveDirectly, the NGO with the largest cash-transfer portfolio in Uganda, indicates that cash transfers are no longer flying under the political radar. The suspension was formally based on technical justifications, but it probably reflected the government’s calculations that cash transfers outside its control would play an adversarial role in the election cycle (ibid.). The emerging but still relatively nascent status of SCTs in Uganda is discussed by Lavers and Hickey (2021), whose assessment is that SCTs in Uganda are still much less firmly established institutionally than in other comparable African countries.

3.2. Current ODA in Uganda

Where exactly would the ODA for scaled SCTs come from? A starting point for any discussion about repurposing aid is the aid itself; this section accordingly provides a brief overview.

Uganda has for many years been a significant recipient of aid. In 2019, aid from OECD countries was equivalent to about 6.1 per cent of Gross National income (GNI).¹³ Total annual aid (loans plus grants) is well on the plus side of USD 2 billion, as shown in Table 2 for 2017–2020.¹⁴ The average for this period is 2.37 billion USD, with 2020 as an outlier primarily because of a large IMF package. Aid is given mainly as grants (about three quarters) and the remainder in the form of subsidised loans, as shown in Table 2.

Table 2: Total OECD country aid for Uganda (USD millions).

Year	Total loans from official donors (US dollars, millions)	Total grants from official donors (US dollars, millions)	Total aid for Uganda from all official donors (US dollars, millions)
2017	489	1584	2073
2018	512	1533	2045
2019	597	1539	2140
2020	1479	1729	3208

As a likely reflection of Uganda’s strategic role as a relatively stable cooperation partner in a volatile region, its biggest donors are the World Bank Group and the US, which have each spent around 500 million USD annually as aid to Uganda in recent years. These are followed by donors such as the Global Fund, EU institutions, and large bilateral donors like the UK, Sweden, Germany and Japan.¹⁵

As shown in Figure 1, about half of the aid (53.6 per cent on average in 2017–2020) goes to the public sector, although not all of it is necessarily ‘on budget’, i.e. recognised by the government as flowing through the government.

Figure 1: Total aid for Uganda from all official donors by channel (USD millions).

Figure 2 provides an overview of the aid using OECD/DAC sectoral codes and shows that most aid goes to the social sector, i.e. social infrastructure and services (OECD/DAC code 100). Cumulatively, the social sector made up 55 per cent of total aid in 2017–2020 and on average amounted to almost 1.3 billion USD annually. A large part of this was for HIV/AIDS-related activities, expenditure on which amounted to an average of 413.24 million USD annually.¹⁶

Figure 2: Total Aid for Uganda from all official donors using OECD/DAC sectoral codes (USD millions).

Humanitarian aid is included in the overall ODA figures and averaged 277 million USD annually in 2017–2020.¹⁷ This relatively large amount reflects the fact that Uganda is hosting large numbers of refugees from conflicts in its volatile neighbourhood. As shown in Figure 2, the average annual share of humanitarian aid is high (about 11.71 per cent of total ODA).

4. Analysis of the potential socio-economic impacts of ten SCT scenarios in Uganda

Based on the above introduction of, on the one hand, emerging cash transfers and, on the other, sizable ODA, this section moves on to assess the impact of allocating some of that aid to specific SCTs.

4.1. UGAMOD microsimulation methodology

The socio-economic impact of different SCT schemes was analysed using a microsimulation model called UGAMOD version 1.8 (Waiswa et al., 2022). The UGAMOD model was built using the EUROMOD framework (Sutherland and Francesco, 2013). This is a micro-simulation model that simulates individual and household tax liabilities and benefit entitlements based on the country’s tax and benefit rules. It is a static model that does not account for behavioural changes due to tax or benefit reforms, and it assumes full compliance by taxpayers and beneficiaries. Using nationally representative micro-data, the model calculates the static effects of a tax-benefit system on household incomes for the national population. The model is useful in assessing the effect of consolidated national tax-benefit policies and how tax-benefit policy reforms affect income distribution, work incentives and government budgets. UGAMOD uses the Uganda National Household Survey (UNHS) data for 2016/17 as the underpinning data which includes 15,721 households and 74,422 individuals (Uganda Bureau of Statistics, 2018). Although the underpinning data come from the UNHS 2016/17 survey, they have been updated by uprating factors (national consumer price index, (CPI)) in each subsequent year. The data provide a nationally representative sample collected by the Uganda National Bureau of Statistics, a statutory government institution with the mandate to collect national statistics. The survey contains data on household demographics, consumption expenditure, income sources and labour market characteristics, which are used to simulate tax incidence and changes in household welfare. The model allows the alternative financing scenarios for the SCT to be analysed by calculating the effect of each SCT on income poverty and inequality using a hypothetical family structure defined by the user based on the underpinning data and the rules for each SCT.

We test the economic impact of allocating between 6 and 115 per cent of current development and humanitarian aid to Uganda (worth 2.37 billion USD on average over the 2017–2020 period, as discussed in Section 3) to SCTs considering different scenarios targeting older persons and children. The scenarios differ by age eligibility, criterion, grant amount and targeting as outlined specifically in Table 3.

Table 3: Simulated scenarios.

Reform name	Target group	Age eligibility	Eligibility criterion	Amount of benefit (UGX monthly)
SCT 1	Older Persons	60	Universal	25,000
SCT 2	Older Persons	60	Universal	35,000
SCT 3	Older Persons	65	Universal	25,000
SCT 4	Older Persons	65	Universal	35,000
SCT 5	Children	0–5	Universal	67,852.6
SCT 6	Children	0–5	Universal	101,778.9
SCT 7	Children	0–5	Means tested below poverty line	67,852.6
SCT 8	Children	0–5	Means tested below poverty line	101,778.9
SCT 9	Children	0–5	Means tested below twice poverty line	67,852.6
SCT 10	Children	0–5	Means tested below twice poverty line	101,778.9

We simulate all ten scenarios and assess their respective financial cost and impact on poverty and income inequality. We also compare the impact on rural and urban households.

The proposed older person’s schemes are variations of the existing scheme known as the senior citizen grant (SCG), which targets individuals aged 80 years and above, who receive a grant of UGX 25,000 per month, as discussed in Section 3. During the pilot phase (before 2021), the SCG targeted individuals aged 65 years (60 years for the Karamoja region) in a few districts. Currently, there are plans to increase the grant amount to UGX 35,000, which we simulate in this study.

The child benefits were informed by an analysis commissioned by the Ministry of Gender, Labour and Social Development and conducted by UNICEF that proposed various such benefits (GoU, 2016), as discussed in Section 3. The study proposed child support benefits for 0–2 years and 0–8 years at different amounts: high (30 percent of the average monthly household expenditure, equivalent to UGX 101,778.9) and low (20 percent of the average monthly household expenditure, equivalent to UGX 67,852.6), both of which we adopt in this study. The figures are obtained from the estimated household monthly income of UGX 339,263 from the Uganda National Household Survey (UNHS) for 2016/17.

Prior to selecting the 0–5 year-old child benefits for our modelling, we simulated the impact of benefits for 0–8 years (both high and low amounts) as well as for 0–2 years (also for both high and low amounts). These were then dropped from our analysis because they were at either too high or too low a cost to implement relative to the aid envelope that is our analytical starting point. The target group of 0–5 years of age came out as the most suitable scenario for assessment, given the size of the ODA.

The proposed child support schemes are motivated by Uganda’s young population, where those aged 0–14 constitute 46 percent of the total population (Uganda Bureau of Statistics, 2020). Besides, Uganda still has a significant proportion of its population living in poverty (21.8 per cent, PPP),¹⁸ which exposes the children living in such households to multiple deprivations related to health, education, nutrition, water and sanitation.

The alternative SCT schemes are summarized in Table 3.

4.2. UGAMOD results

Using the UGAMOD Model with the 2016/17 underpinning dataset, we simulate the static effects of alternative SCT programmes in targeting older persons and children on poverty, income inequality and the cost of the programme.

4.3. Targeting older persons

In this section, we analyse the economic impact of SCT schemes that target older persons with varying eligibility and benefits. Specifically, we consider the impact of the different SCT schemes on poverty and income inequality, the cost of the cash transfer, and the impact of the SCT by rural and urban populations.

4.4. Impact of older-person SCT schemes on poverty and inequality

We analyse the impact of the different older-person SCT schemes on poverty and inequality, the results being presented in Table 4. The results reveal that the SCT2 scheme has the greatest reduction in headcount poverty of 1.08 percentage points (from 21.80 per cent in the baseline system to 20.72 per cent). Comparably, the poverty gap (which measures how far poor households fall below the poverty line) falls from an index of 5.77–5.36, mainly driven by the large reduction in the poverty gap of poor households with older persons. The effect of the SCT2 scheme is more concentrated in households with older persons, followed by households headed by females. Perhaps this could signal that women live longer than men. Similarly, SCT 2 results in the greatest reductions in income inequality, reducing the Gini coefficient from 0.3983 to 0.3946. The relatively high impact of the SCT2 on poverty and inequality is driven by the lower eligibility age of 60 years and above, combined with a high benefit of UGX 35,000. As expected, SCT3 has the lowest economic impact of reducing poverty by 0.45 percentage points and reducing the Gini coefficient from 0.3983 to 0.3970. The low impact is a result of the higher eligibility age of 65 years and above combined with a low benefit of UGX 25,000.

Table 4: Impact of older persons’ SCT Schemes on poverty and inequality.

	Baseline	SCT 1	SCT 2	SCT 3	SCT 4
Share of poor population, in %
All	21.80	21.07	20.72	21.35	21.12
Poor households out of . …
… male-headed households	21.75	21.10	20.82	21.33	21.14
… female-headed households	21.95	20.97	20.46	21.43	21.07
… households with children	22.94	22.24	21.90	22.51	22.29
… households with older persons	23.88	19.78	17.76	20.16	18.20
Poverty gap (average normalised poverty gap, FGT (1))
All	5.77	5.50	5.36	5.62	5.51
Poor households out of …
… male-headed households	5.63	5.41	5.29	5.51	5.42
… female-headed households	6.15	5.74	5.54	5.92	5.76
… households with children	6.06	5.79	5.65	5.90	5.80
… households with older persons	6.48	4.99	4.08	5.18	4.32
Gini (household income)	0.3983	0.3946	0.3959	0.3970	0.3960
P80/P20	2.98	2.94	2.95	2.97	2.95

4.5. Cost of older person’s SCT schemes

As can be seen in Table 5, SCT 2 is the most expensive SCT scheme because of its wide coverage and high level of benefit. A comparison with the baseline implies that the cost would be an additional UGX 566,849.27 million, which would increase the current budget for older person’s SCT by 365 per cent. The cheapest SCT scheme is SCT 3, which increases the current budget for older person’s SCT by 136 percent. As discussed in Section 3, on average Uganda receives US$ 2.37 billion worth of ODA annually, which is equivalent to UGX 8,718 billion.¹⁹ Therefore, depending on the proportion of the ODA that will be allocated to SCT beneficiaries, all the simulated SCT scenarios for older persons fall financially well within the existing aid envelope.

Table 5: Cost of older persons’ SCT schemes (excluding administrative costs).

	Baseline	SCT 1	SCT 2	SCT 3	SCT 4
Social assistance (million UGX)	155,226.38	515,768.32	722,075.65	365,761.30	512,065.82

As shown in Table 6, all old-age SCT schemes will have more beneficiaries in rural than in urban populations because of a higher incidence of poverty in rural than in urban areas. Indeed, over 80 per cent of the beneficiaries (84 per cent for SCT 1 &2 and 86 per cent for SCT 3 & 4) will be drawn from the rural population. Interestingly, we observe slightly lower levels of income inequality for rural than for urban populations, which could signal a higher population of older persons and a relatively more equal distribution of income in rural than in urban areas.

Table 6. Economic impact of older person’s SCT by rural/urban populations.

Variable name	SCT 1			SCT 2			SCT 3			SCT 4
	All	Rural	Urban	All	Rural	Urban	All	Rural	Urban	All	Rural	Urban
Cost (million UGX)	515,768	431,022	84,746	722,075	603,430	118,645	365,761	311,285	54,476	512,126	435,860	76,266
Beneficiaries	1,719,228	1,436,740	282,488	1,719,228	1,436,740	282,488	1,219,204	1,037,618	181,586	1,219,204	1,037,618	181,586
Poverty	21.07	24.78	9.88	20.72	24.35	9.77	21.35	25.12	9.99	21.12	24.83	9.91
Inequality	0.3959	0.3545	0.3961	0.3946	0.3536	0.3951	0.3970	0.3554	0.3967	0.3960	0.3546	0.3961

4.6. Targeting children

This section analyses the welfare effects of different SCT schemes targeting children using two different benefit amounts, both universal and means-tested, as presented in Table 3. Specifically, we consider the impact of the different SCT schemes on poverty and income inequality, the cost of the cash transfer, and the impact of the SCT on rural and urban populations.

4.7. Impact of child SCT schemes on poverty and inequality

We simulate the impact of six (6) child SCT schemes on poverty and inequality, the results being presented in Table 7. The first two schemes, SCT 5 and SCT 6, are universal in targeting all children aged 0–5 years, while SCT 7 to SCT 10 are means-tested. The SCT 6 scheme has the greatest impact on poverty of 15.43 percentage points (from 21.80 to 6.37 per cent) and on inequality, which drops from 0.3983 to 0.3322 because it is universal and offers a high benefit of UGX 101,778.90. Interestingly, the impact of SCT 6 on poverty is more concentrated in male-headed households than in households with children. SCT 7 has the least impact on poverty of 5.40 percentage points (from 21.80 to 16.40 per cent) and the lowest reduction in inequality (from 0.3983 to 0.3770). Overall, the universal schemes have a greater economic impact because of their coverage. Poverty will fall by 13.62 percentage points for SCT 5, 15.43 percentage points for SCT 6, 5.40 percentage points for SCT 7, 6.17 percentage points for SCT 8, 7.62 percentage points for SCT 9 and 8.66 percentage points for SCT 10. Our results are consistent with a similar study by Meng and Donald Pfau (2011) that simulated the impact of cash transfers on school-age children from poor households in Cambodia and found the greatest impact of the cash transfer to be on universal programs. Furthermore, the results are consistent with a study by Acosta et al. (2011) of countries in Latin America that found the impact of cash transfers to be significantly greater for children than for the elderly for an equivalent amount of benefit. The greater impact on child benefits than on benefits for the elderly is possibly in part related to the higher poverty rates for households with children than for households with the elderly in developing countries that have high rates of population growth such as Uganda.

Table 7: Impact of child SCT schemes on poverty and inequality.

	Baseline	SCT 5	SCT 6	SCT 7	SCT 8	SCT 9	SCT 10
Share of poor population, in %
All	21.80	8.18	6.37	16.40	15.63	14.18	13.14
Poor households out of …
… male-headed households	21.75	7.41	5.54	16.07	15.33	13.89	12.89
… female-headed households	21.95	10.27	8.62	17.27	16.46	14.97	13.82
… households with children	22.94	8.30	6.35	17.13	16.31	14.75	13.63
… households with older persons	23.88	14.75	12.45	20.99	20.40	19.22	18.21
Poverty gap (average normalised poverty gap, FGT (1))
All	5.77	1.86	1.50	3.95	3.76	3.39	3.16
Poor households out of …
… male-headed households	5.63	1.61	1.25	3.81	3.64	3.27	3.05
… female-headed households	6.15	2.56	2.18	4.34	4.08	3.74	3.43
… households with children	6.06	1.86	1.47	4.10	3.89	3.50	3.25
… households with older persons	6.48	3.49	3.13	5.34	5.17	4.70	4.47
Gini (household income)	0.3983	0.3458	0.3322	0.3770	0.3725	0.3675	0.3607
P80/P20	2.98	2.51	2.43	2.77	2.76	2.66	2.66

4.7.1. Cost of child SCT schemes

Child support is considerably more expensive than old-age support, and as shown in Table 8, the universal schemes (SCT 5 and SCT 6) are more expensive than the means-tested schemes (SCT 7–10) due to their wider coverage. At the same time, however, the cost of the means-tested schemes may increase because of the administrative costs, which can be large due to bureaucratic requirements and may also carry a risk of bias in identifying the poor (Meng and Donald Pfau, 2011). The feasibility of each of the SCTs 5–10 will depend on how much of the ODA funding is earmarked to be allocated to the beneficiaries of SCTs, but certainly SCT 6 is unattainable given the current resource envelop for Uganda’s ODA (as further illustrated in Table 11).

Table 8: Cost of child SCT schemes (excluding administrative costs).

	SCT 5	SCT 6	SCT 7	SCT 8	SCT 9	SCT 10
Child benefit (Billion UGX)	6,595.67	10,048.74	1,821.08	2,654.01	2,664.13	3,918.58

4.8. Economic impact by rural/urban population

Table 9 shows the economic impact of implementing universal child SCT schemes by rural and urban populations. The results indicate that rural populations benefit more than urban populations by a large margin. For instance, 78 percent of beneficiaries are drawn from the rural population, which is consistent with a higher incidence of poverty in rural than in urban areas. Consistent with the results for the elderly, inequality is lower in rural than in urban areas, which suggests that there will be greater economic gains in poverty reduction in the case of rural areas.

Table 9: Economic impact of universal child SCT by rural/urban population.

Variable name	SCT 5			SCT 6
	All	Rural	Urban	All	Rural	Urban
Cost (million UGX)	6,595,673	5,116,187	1,479,486	9,893,511	7,674,281	2,219,230
Beneficiaries	8,100,493	6,283,458	1,817,035	8,100,493	6,283,458	1,817,035
Poverty	8.18	9.32	4.75	6.37	7.17	3.93
Inequality	0.3458	0.3047	0.3627	0.3322	0.2950	0.3508

Table 10 shows the economic impact of implementing means-tested child SCT schemes by rural and urban populations. Consistently, rural populations benefit more from SCT than urban populations by a greater margin than previous scenarios of the elderly and the universal child SCT. For instance, 87 per cent of the beneficiaries will be drawn from the rural population, although in terms of the overall economic impact the universal child SCTs have a greater economic impact than means-tested child SCTs.

Table 10. Economic impact of means-tested child SCT by rural/urban population.

Variable name	SCT 7			SCT 8			SCT 9			SCT 10
	All	Rural	Urban	All	Rural	Urban	All	Rural	Urban	All	Rural	Urban
Cost (million UGX)	1,665,858	1,447,755	218,103	2,498,788	2,171,633	327,155	2,508,903	2,182,316	326,587	3,763,354	3,273,473	489,881
Beneficiaries	2,045,928	1,778,064	267,864	2,045,928	1,778,064	267,864	3,081,315	2,680,216	401,099	3,081,315	2,680,216	401,099
Poverty	16.40	18.90	8.57	15.63	18.05	8.35	14.18	16.32	7.72	13.14	15.02	7.45
Inequality	0.3770	0.3344	0.3877	0.3725	0.3326	0.3844	0.3675	0.3252	0.3828	0.3607	0.3220	0.3778

4.9. Comparing the costs of SCT scenarios with the ODA resource envelope

Table 11 compares the calculated cost of each SCT scenario with the current total aid for Uganda and current domestic revenue and includes the projected poverty rates for reference. Two scenarios seem to stand out in terms of impact and affordability. SCT 5, universal child support at the low monthly benefit level, would cost 76 per cent of current aid (excluding administration²⁰) and is the scenario which in financial terms comes closest to allocating the major part of aid to SCTs. It results in a very noticeable reduction of poverty from 21.8 to 8.18 per cent.

Table 11: Share of ODA needed to finance each SCT scenario.

	Share of poor population in %	Cost annually in billion UGX	Approximate % of development aid allocated (ex admin cost)	Current government domestic revenue in billion UGX (21/22)^a	Average ODA 2017–2020 in billion UGX^b^,^c
Baseline	21.80			22,425	8,718
SCT 1	21.07	515.8	6%
SCT 2	20.72	722.1	8%
SCT 3	21.35	365.8	4%
SCT 4	21.12	512.1	6%
SCT 5	8.18	6,595.7	76%
SCT 6	6.37	10,048.7	115%
SCT 7	16.4	1,821.1	21%
SCT 8	15.63	2,654.0	30%
SCT 9	14.18	2,664.1	31%
SCT 10	13.14	3,918.6	45%

^a Ministry of Finance Planning and Economic Development: Background to the Budget 2022/23 (p. 198).

^b This total ODA figure includes both grants and loans. See also Section 3 and Table 2. We are not suggesting that Uganda should consider taking out expensive loans on the commercial markets and then give the money to its people. However, by definition the loan part of ODA comprises subsidised loans, and it could be argued that allocating these to build social or human capital through provisions of SCTs may, under some circumstances, be within the realm of the advisable and feasible. Thus, we use total ODA rather than only the grant part of ODA as reference point.

^c Using exchange rate 1 USD = 3,678.48 UGX accessed on xe.com 3rd Feb 2023.

SCT 9, child support means-tested below twice the poverty line at the low monthly level, would cost 31 per cent (plus administration) and is a scenario which comes relatively close to the current global allocation of 20 per cent of humanitarian funding for SCTs. It also results in a very noticeable reduction of poverty to 14.18 per cent and would, in principle, be possible to finance without touching the 55 per cent of donor aid which is currently allocated to the social sectors.

SCTs 1–4, which are all focused on old-age pensions, are much cheaper (between 4 and 8 per cent of ODA plus administration), but also result in more marginal poverty reductions of up to 1.08 percentage points only.

The results generated using UGAMOD are consistent with the findings of Asiimwe et al. (2023), where a dynamic Computable General Equilibrium (CGE) model was used to assess the effects of repurposing donor aid funds from government and Non-Profit Institutions Servicing Households (NPISH) to financing household cash transfers in Uganda. The findings reveal that shifting aid funds to local actors (households) provides macroeconomic benefits to the economy and also generates spill-over effects to non-recipient households and other economic agents. Financing the cash transfers by reducing allocations to government and NPISH increases tax revenue collection, household incomes and savings.

4.10. Administrative costs and overheads

A key observation would be that, administratively, SCTs appear to be relatively cheap and to require few international overheads. One of the leading NGOs involved in cash transfers, GiveDirectly, claims a 90 per cent cash delivery ratio, i.e. only 10 per cent of the donations they receive are spent on administration and overheads.²¹ A World Bank study (Beegle et al., 2018, 230) concluded that the average cost of running SCTs in Africa through government systems is 17 per cent of program spending, but with a large variation, and typically declining with scale. For example, the government-run Productive Safety Net Programme (PSNP) in Ethiopia, with close to 7 million recipients, had a cost of around 7 per cent (ibid.: 358). As an outlier, the old-age pension scheme in Lesotho had a cost of only 2 per cent (Oliver, 2013, 102). At the other end of the spectrum, a small cash program focused on children in Uganda reached fewer than 100,000 recipients at a cost of about 21 per cent (Beegle et al., 2018, 360).

Based on the above, we would suggest that indicative figures for the administrative cost of running scaled SCTs at 7–10 per cent would not be unrealistic, as this has been achieved by NGOs and governments elsewhere in Africa.

5. Discussion of socio-economic allocation costs and taxation

Our analysis above explores the potential socio-economic impacts and costs that are directly related to the scaling of SCTs and suggests poverty-reduction results, which, seen in isolation, are remarkable at least for SCTs aimed at children. However, the analysis also leaves aside a number of unanswered questions, the most important of which may be the reallocation costs, which are linked to government revenue and taxation. It would go beyond the scope of this article to make a full analysis, but in the following we discuss what we consider the key issues.

First and foremost, the allocation costs of shifting ODA to SCTs at scale would affect the areas where aid is now focused. The bulk of current aid is dispersed to the social sectors through public channels (as shown in section 3). With 55 per cent of aid allocated to the social sectors, there is a real risk of a full or partial collapse of donor-funded interventions in education, health and HIV/AIDS, or other activities currently funded by aid if ODA finance were to be abruptly moved to SCTs.

This has potential implications for taxation and domestic revenue generation. If social-sector interventions hitherto funded by ODA were to be continued without introducing additional user payments, then one obvious potential solution would be increased domestic revenue generation. This would enable governments to finance the continued provision of services themselves. In other words, to make up for lost ODA revenue, taxes would need to go up. While this appears a tall order, and is certainly never likely to be a popular proposition, in the long term it may not be entirely far-fetched. Over the last decade, total revenue generation relative to GDP has grown moderately but consistently in Uganda from 7.95 per cent of GDP in 2010 to 13.09 per cent of GDP in 2021.²² This trend is expected to continue. The Ministry of Finance, Planning and Economic Development projects aid to become significantly less important relative to other sources of public finance in Uganda. The overall resource envelope is expected to increase from 42.8 trillion UGX in 2021/2022 to almost 63 trillion in 2026/27, which is equivalent to an increase of about 47 per cent. Over the same period, aid is projected to remain stable, hovering at about UGX 11 billion, while, in contrast, domestic tax revenues are expected to more than double from 20.8 trillion to UGX 43.7 trillion (MFPED, 2022, 198).

This trend is similar to the experiences of other African economies in recent decades. Moore et al. (2018, 29) show that Africa is experiencing a tectonic shift, from an ‘aid era’ into a ‘tax era’. The key point is that a combination of growth in African economies, declining aid and an increase in tax-to-GDP ratios meant that by 2014 average national tax revenues were more than twice average aid revenues, as opposed to the period from 1986 to 1996, when aid exceeded tax (ibid.: 30).

A shift of ODA funds to SCTs in Uganda, in parallel with the government further increasing its tax base and taking over the financing of social-sector activities, would accelerate the ongoing move from aid reliance to tax reliance. Yet, at the same time it would create a significant extra tax burden, which appears unrealistic, at least in the short term. As discussed in Section 3, current donor funding of social sectors was on average 1.3 billion USD in 2017–2020 and about 1.2 billion USD for three of the years from 2017 to 2019. That is equivalent to more than 20 per cent of total domestic revenue generation and reflects the fact that donors are the biggest funders of the social sectors. This potential additional financial and taxation burden would arise in the current situation, where the government, despite the medium-term projections for increased domestic revenue, is concerned about tax saturation and wary about increasing any tax rates.²³

Based on the above, we would suggest that, in the short to medium terms, a moderate and gradual pathway to any scaling of SCTs in Uganda would be the most realistic. Such a pathway would include an element of ringfencing current donor funding for the social sectors and reallocating other ODA funds instead.

6. Conclusion

In both Uganda and elsewhere, investing in scaled SCTs would be a relatively new modality for donors who have hitherto seen the ODA funding of SCTs either as a tool for immediate humanitarian relief or as support for aid-receiving countries to build and finance their own social protection systems. And yet large-scaled, externally funded SCTs in the Global South are not an entirely new or alien idea. Organisations such as the Brookings Institute and GiveDirectly are considering how scaled SCTs can potentially lift individual countries and the entire Global South out of poverty.

Whether the donor-funded super-sizing of SCTs in Uganda is a good idea depends greatly on the objective of ODA as prioritised by both government and donors. If poverty reduction and actually reaching the poor is key, then our modelling would suggest that repurposing aid for SCTs could be very impactful. In particular, two of the child-support scenarios, SCT 5 and SCT 9, both focused on supporting children between 0 and 5, appear interesting and would, seen in isolation, go a long way to reducing poverty levels in Uganda. With implementation of the two scenarios, the model suggests that poverty would fall from the current 21.8 per cent to respectively 8.18 and 14.18 per cent, which are very sizeable reductions. This would come at a cost of respectively 76 and 31 per cent (plus administration) of current aid. Based on benchmarking from elsewhere, we suggest that efficient and scaled SCT schemes would have to bear a cost of 7–10 per cent in administration.

Allocation of ODA for comprehensive old-age pension schemes would be much cheaper (4–8 per cent of ODA, plus administrative costs), reflecting the demographics of the country, which has few elderly but many children, but it would also have a much lower poverty-reduction effect (up to 1.08 percentage points). For both older persons’ and children’s SCT schemes, the modelling predicts that rural populations would benefit more than urban populations by a large margin.

In a nutshell, when it comes to socio-economic impact, our analysis suggests that old-age pensions are unlikely to be a game-changer, but that scaled child support might very well be. The modelling predicts that allocating a large chunk of ODA to child support could go a long way towards reducing national poverty and, importantly, would be affordable within Uganda’s existing ODA envelope.

However, this analysis comes with important caveats. Most importantly, a shift to ODA-financed, large-scaled SCTs would come with allocation costs which our modelling does not capture. As aid is diverted from current allocations, 55 per cent of which is for the social sectors, there is a risk that these donor-funded interventions would fall apart if ODA were to be redirected abruptly without alternative financing. In the long term, a successful shift of ODA to SCTs therefore assumes that the government would step in and pick up the tab to replace currently ODA-financed expenditure, in particular in the social sectors. The medium-term government budget in Uganda already predicts a declining role of aid as domestic revenue generation grows. But a shift of ODA to scaled SCTs would necessitate an even faster transition, the feasibility of which would be challenging. Consequently, a phased and more moderate pathway for scaling could more realistically be considered. This could involve at least initial ring-fencing of all or large parts of current ODA allocations for the social sectors while mainly re-allocating to scaled SCTs ODA that is currently spent on the non-social sectors.

Overall, we note that, while a transition to scaled SCTs may bring large rewards in terms of poverty reduction, it would also be a risky process with many unknowns. This calls for a negotiated, gradual, careful and monitored shift, rather than any abrupt changes.

The findings from one case, Uganda, may be relevant for other countries in Sub-Saharan Africa that are comparable in terms of aid per capita, GDP per capita, persistently high poverty levels, political stability and donor relations. Such countries include, for example, Tanzania, Mozambique, Malawi, Rwanda, Ghana and Zambia. In terms of policy recommendations for ODA in Uganda and comparable countries, our analysis suggests that donors and governments can think of SCTs from a different perspective and consider funding in particular for scaled child support. The primary aim would be to do something dramatic about poverty reduction. As a minimum, the very positive poverty reduction results associated with scaled SCTs can be used to benchmark current aid. Do current ODA-funded interventions deliver results which are better or worse than what scaled SCTs would be predicted to deliver?

As a concluding remark, we point to the many political economy issues which need to be better understood. The (major) donors provide aid not only for altruistic purposes, such as poverty reduction, but also to maintain regional stability and influence and to satisfy their own domestic political constituencies and broader political agendas. It remains to be seen what weight poverty and inequality arguments in favour of scaled SCTs would have in competition with donors’ other political agendas. As part of this, the politics of SCTs may include a battle for overheads. An attempt in Lebanon to streamline the delivery of cash transfers showed that the aid industry is not necessarily going to give up its overhead funding without a fight.²⁴ In terms of domestic politics, current aid in Uganda and elsewhere also serves a role in financing the current political regime and in ensuring national political stability. Linked to this, elites and public opinion in sub-Saharan Africa may have stronger confidence in productivism and development investments rather than allocating scarce finance for redistribution and hand-outs, which SCTs are often perceived as (Seekings, 2016, 2019, 2021). While there might be an emerging political interest in the patronage and goodwill that SCTs can buy, considering the domestic politics of large-scaling would lead us into uncharted territory beyond the scope of this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Danish International Development Agency.

Notes on contributors

Jacob Ulrich

Mr. Jacob Ulrich is a PhD fellow at Roskilde University and has a master’s degree in International Affairs from Columbia University. He is a practitioner with +25 years of experience working with development and aid management in more than 30 countries worldwide.

Susan Kavuma

Dr. Susan Namirembe Kavuma is a Senior Lecturer in the Department of Policy and Development Economic of the School of Economics at the College of Business and Management Science (COBAMS), Makerere University. She has served the University since 2006, lecturing and supervising research for undergraduate and post-graduate students. She has a PhD in Economics from the University of Nottingham, United Kingdom (2015) and was a Postdoctoral Research Fellow at the University of Cambridge under the Cambridge Africa Partnership for Research Excellence (CAPREx) programme in 2017. She is a Research Fellow at the Environment for Development Initiative Makerere University. Susan has authored and co-authored several research papers in peer reviewed Journals. She has peer-reviewed publications in the African Development Review Journal.

Wilson Asiimwe

Mr. Wilson Asiimwe is a Senior Economic Modeller at the Ministry of Finance, Planning and Economic Development Uganda. He is also a Research Fellow at the Macroeconomic and Financial Management Institute for Eastern and Southern Africa (MEFMI, Zimbabwe); where he tutors short-term courses in Macroeconomic Modelling. He holds a Master’s degree in Development and Natural Resource Economics from the Norwegian University of Life Sciences, Norway. He also holds an Msc in Financial Engineering from WorldQuant University, USA and a Bachelor of Economics degree from Makerere University (Uganda). He has 13+ years of experience in CGE and macroeconomic modelling of social protection, climate change and fiscal policy.

Bruno Yave

Dr. Bruno L. Yawe is an Associate Professor at the College of Business and Management Science (COBAMS), Makerere University. He holds a PhD in Economics and Master’s degree in Economics from the University of Dar es Salaam, Tanzania. He has 20+ years in teaching economics at Makerere University. He has also served as a Senior Research Fellow at Environment for Development and also capacity building of institutions like Uganda Management Institute (UMI), National University of Rwanda, Kenya School of Monetary Studies, Uganda Institute of Banking and Financial Services and many more.

Notes

1 ‘Effective’ here refers to the ability of SCTs to impact positively on indicators such as reduction of monetary poverty, increases in food expenditure, increases in school attendance by children, health outcomes and the use of health services, increases in livestock ownership and purchases of agricultural inputs as well as overall increases in participation in and the intensity of work among adults of working age (Bastagli et al., 2016).

2 ‘Cheap’ refers to overhead costs of SCTs as further discussed in section 4 of this article.

3 Although Gronbach (2023) suggests that the actual roll-out of COVID related SCTs has been much less expansive than the policy and program declarations would have led to believe.

4 See for example https://www.unicef-irc.org/research/cash-plus/.

5 We use the terms ‘scaling’ and ‘large-scaling’ of SCTs in the sense of expanding from a small group of recipients to a larger group of recipients.

6 At the time of writing, two African countries, Malawi and Rwanda, are being assessed for the feasibility and cost of lifting these countries out of poverty by using SCTs. Interview with GiveDirectly, 20th October 2022.

7 There are many implementation options. See, for example, Gronbach (2020) for systematic documentation of the great variety of current cash-transfer payment systems in Sub-Saharan Africa.

8 For a discussion of UGAMOD, see Waiswa et al. (2022). In addition, see Sutherland (2013) and Hufkens et al. (2019) on the EUROMOD framework based on which UGAMOD and other African country models were developed.

9 OECD/DAC sectoral code 100. See also Section 3.

10 However, a more recent paper (Gronbach, 2023 forthcoming) discusses the nature of the reforms triggered by the pandemic and argues that the pace and extent of cash-transfer expansion in the region (Africa) have lagged behind global expectations.

11 If the SCTs contribute to building human, social or physical capital, the impact may also continue beyond the lifespan of the SCT provision. That would be a key part of the theory of change for ‘productivism’. We do not disagree with the notion (see, for example, Araujo and Macours (2021) on the intergenerational impact of SCTs), but we note that our static modelling captures the immediate impacts only.

12 Confirmed in an interview held in June 2022 with the Country Director of a Danish NGO working with cash transfers in Uganda.

13 https://data.worldbank.org/indicator/DT.ODA.ODAT.GN.ZS?locations=UG.

14 All ODA data in this section have been generated from the OECD/DAC Creditor Reporting System database unless otherwise indicated. We note that data for 2021 and 2022 are not fully available in the OECD/DAC Creditor Reporting System at the time of writing.

15 https://www.oecd.org/countries/uganda/aid-at-a-glance.htm#recipients

16 OECD/DAC codes: Infectious disease control (12250), STD control including HIV/AIDS (13040), Social mitigation of HIV/AIDS (16064).

17 Using OECD/DAC sectoral code 700.

18 Using US$1.25 as the poverty line.

19 Using exchange rate 1 USD = 3683.79 UGX accessed 3rd Feb 2023 at xe.com.

20 For benchmarking of what administrative rates might be achievable for scaled cash transfers, our analysis would suggest between 7 and 10 per cent, as discussed in the next sub-section.

21 This excludes any costs for larger research studies, e.g. randomized control trials, which are budgeted and financed separately and typically carried out independently by outside partners according to an interview with GiveDirectly, 20th October 2022. It also corresponds to the US Government’s approved overhead rate (NICRA) of GiveDirectly, which is 8 per cent according to a communication from GiveDirectly dated 17th November 2022

22 All figures from UNU-WIDER at UNU WIDER – GRD Dataset.

23 According to one of the authors of this article, who is employed in the Ministry of Finance in Uganda.

24 www.thenewhumanitarian.org/investigations/2017/02/20/unconventional-cash-project-challenges-aid-statusquo-lebanon