Reforming Botswana’s Social Security System
Author:
International Monetary Fund. African Dept.
Search for other papers by International Monetary Fund. African Dept. in
Current site
Google Scholar Close
Publication Date:
10 Sep 2024
eISBN:
9798400288258
Language:
English
Keywords:
protection system; program coordination; poverty eradication program; program adaptability; expenditure program; Social protection spending; Income distribution
Download PDF (580.8 KB)

Despite its strong economic performance and sound policy management, Botswana remains one of the most unequal countries globally. The country's complex social protection system lacks efficiency, with funds often allocated towards regressive programs and burdened by high administrative costs. To make growth more inclusive, there is a pressing need for more targeted fiscal interventions and a comprehensive overhaul of the system. This includes simplifying the framework, improving its adaptability, modernizing delivery and targeting methods, reallocating funds away from regressive programs, and relying more on progressive financing sources.

Reforming Botswana’s Social Security System1

Despite its strong economic performance and sound policy management, Botswana remains one of the most unequal countries globally. The country's complex social protection system lacks efficiency, with funds often allocated towards regressive programs and burdened by high administrative costs. To make growth more inclusive, there is a pressing need for more targeted fiscal interventions and a comprehensive overhaul of the system. This includes simplifying the framework, improving its adaptability, modernizing delivery and targeting methods, reallocating funds away from regressive programs, and relying more on progressive financing sources.

A. Introduction

1. Several decades of growth have substantially boosted average income and alleviated poverty in Botswana. The discovery of diamonds in late 1960s sparked an unparalleled economic expansion over the subsequent five decades. As a result, real per capita income increased by a factor of 17 between 1960 and 2022, ranking as the fourth highest growth rate globally (Figure 1, Panel A).2 The share of people living below the international poverty line3 declined from over 40 percent to 14 percent in 2022 (Figure 1, Panel B).

2. Despite this strong economic performance, Botswana’s income inequality was one of the highest in the world in 2015. In 1985, Botswana’s income inequality, as measured by the Gini coefficient, was just below the median of upper-middle income countries (based on today’s sample). Over the following two decades, inequality in Botswana increased, in contrast to a decline among its peers (Figure 1 Panel C). The 2000s saw some improvement, with inequality levels returning to 1980s levels. Yet, by 2015 (latest year with available data), Botswana was one of the most unequal countries in the world.4 Within sub-Saharan Africa (SSA), Botswana is the 4th most unequal country, according to World Bank Gini index. The level of inequality is particularly notable given the country’s current income per capita level (Figure 1, Panel D).

3. Although recent data is lacking, there are reasons to think that the situation has not improved and may even have deteriorated in the past decade. The unemployment rate, one of the leading causes of poverty and inequality in Botswana, increased from 17.6 to 25.9 percent between 2015 and 2023 (and is as high as one third among the youth), while labor force participation declined from 61.5 to 59.7 percent of the working-age population. The public sector wage bill has remained elevated at nearly 13 percent of GDP (as of FY2023), perpetuating the income gap with the private sector and the unemployed. A series of large shocks, including the Covid-19 pandemic and droughts in 2018/19 and 2023/24, have negatively impacted the livelihood of the most vulnerable, potentially aggravating further inequalities (UNEDP 2021).

Figure 1.
Figure 1.

Real Income per Capita, Poverty, and Inequality

Citation: IMF Staff Country Reports 2024, 287; 10.5089/9798400288258.002.A002

Source: World Development Indicators and IMF staff calculations.Note: Panel A: Real income per capita defined as aggregate real income in constant 2017 PPP US$ divided by population.Panel B: Poverty headcount is defined as the share of population living on less than $2.15 a day at 2017 US$ PPP rates.Panel C: Gini coefficient measures income inequality, data for upper-middle-income countries is a 5-year moving average.Panel D: Real income per capita definition similar to Panel A; Gini coefficient and real income measured in the same year.

4. High and persistent inequality in Botswana is primarily explained by structural factors related to economy and geography:5

  • Mining production has a low labor content. The country’s economy is largely driven by the diamond sector, which is very capital-intensive. While the mining industry accounts for close to a quarter of GDP, it accounts for less than 2 percent of total employment (Statistics Botswana, 2023).

  • Private sector development remains limited. High mineral revenues have supported the expansion of the public sector (central government and SOEs), where wages tend to display an elevated premium relative to the private sector (see IMF 2023a, and IMF 2023b).6 Such disparity has led to the concentration of wealth and opportunities within a relatively small segment of the society. Furthermore, the dominance of SOEs in several sectors of the economy has further contributed to sluggish private sector growth (IFC 2022, and World Bank 2023).

  • Inequality is further aggravated by the urban-rural divide and disparities in land ownership. Botswana is a large country similar in landmass to Ukraine but with a small population equivalent to Moldova, with a large share of people scattered in sparsely-populated rural areas. Urban centers are hubs of resources, infrastructure, and economic opportunities, contrasting with rural and remote areas, where limited access to productive land and essential services perpetuates poverty and hinders development (UNDP 2021). This divide is further widened by uneven land distribution, with a minority of individuals controlling vast tracts of land, leaving many rural dwellers without the means to secure their livelihoods or invest in agricultural productivity. Finally low population density in some areas of the country makes it difficult to supply all the necessary services (health, education, finance, connectivity), further aggravating the disparity in opportunities.

5. Beyond these structural factors, the education system tends to perpetuate existing inequalities. Although universal primary education enrollment was achieved in the late 1990s, the secondary education system faces challenges including shortages of learning materials and classrooms, alongside limited teachers’ training. Additionally, the vocational education system is fragmented and suffers from limited organizational capacity (World Bank, 2023). Skill mismatches drive unemployment and create significant barriers to private sector development (Stepanyan and others 2013; and IMF 2017b). Moreover, while obtaining a tertiary education degree usually results in a significant wage premium, tertiary education is disproportionately accessed by wealthier individuals, thereby exacerbating social inequities.

6. In most countries in the world, fiscal policy is a powerful tool to affect income distribution. There is extensive literature showing that fiscal policy instruments (both taxation and expenditure programs) have significant distributional effects and can either mitigate or aggravate preexisting trends.7 On the revenue side, progressive income and consumption taxes can be used to mitigate income disparities. On the spending side, social protection systems offer a bulwark against inequality, extending a lifeline at the lower levels of the income pyramid through cash disbursement and in-kind benefits. Health and education spending exert a subtler influence on income distribution by influencing the incentive to work. Empirical analysis suggests that, in advanced economies, taxation paired with transfers reduces income inequality by approximately one third (see Gaspar and García-Escribano, 2017).

7. The Botswanan authorities have launched important initiatives to reduce inequality and reform their social protection system. Botswana Vision 2036, unveiled in 2016, aims to “promote equal opportunities for all, and ensure that prosperity is widely shared,” partly by enhancing the social protection system. In 2020, the Cabinet approved a new National Social Protection Framework (NSPF), accompanied by an implementation plan (MoLGRD 2020a and 2020b). Since then, the authorities have collaborated with development partners and external consultants to chart out a roadmap of actionable measures aimed at improving social protection (see, for example, MoLGRD 2022). The NSPF includes proposals to consolidate and integrate programs for individuals at different life stages. It also envisages the creation of a Single Social Registry, efficiency-enhancing measures like more accurate targeting, and transitioning away from in-kind to cash transfers. Nonetheless, despite the anticipated improvements that the new framework is expected to bring, its implementation has progressed very slowly (UNICEF 2022).

8. This chapter analyzes Botswana’s social protection system, estimates its impact on poverty and inequality, and discusses reform options to strengthen its effectiveness. The chapter primarily examines the programs that fall under the umbrella of the NSPF (see Box 1).8 Section B evaluates existing policies, covering the level of spending, impact on inequality, and significant barriers that limit the framework’s effectiveness. Section C discusses the financing of social protection and how it impacts inequality, focusing more specifically on the personal income tax (PIT) and value added tax (VAT). Section D discusses international experience and best practices, while suggesting policy reform options for Botswana. Section E concludes.

Botswana Social Protection System

Botswana social protection system encompasses nearly 30 distinct social assistance, labor market, and social insurance programs, all supervised by an array of 10 ministries and government bodies (Text Table 1). Social assistance programs deliver non-contributory support to beneficiaries through both cash and in-kind transfers. Labor market programs include a range of interventions aimed at fostering relevant job skills. The Public Officers Pension Fund stands as the only public social insurance scheme. Text Table 2 presents a comprehensive breakdown of key programs across these categories, detailing eligibility criteria and the nature of benefits.

Box 1. Table 1.

Botswana: Social Protection Programs by Type and Supervising Ministry (as of 2023)

article image
Source: National Social Protection Framework (2018). Notes: MLRGD -- Ministry of Local Government and Rural Development; MYESCD -- Ministry of Youth Empowerment, Sport, and Culture Development; MOHW -- Ministry of Health and Wellness, MOBE -- Ministry of Basic Education, MOTE -Ministry of Tertiary Education, MELSD -- Ministry of Employment, Labor Productivity and Skills Development, MNIG -Ministry of Nationality, Immigration, and Gender Affairs, MOA -- Ministry of Agricultural Development and Food Security, MSP -- Ministry for State President, MOF -- Ministry of Finance, NBFIRA -- Non-Bank Financial Institutions Regulatory Authority
Box 1. Table 2.

Botswana: Select Social Protection Programs: Eligibility and Benefits (as of 2023)

article image
Source: MoLGRD, WB, 2022

B. Botswana’s Social Protection System: Generous but Lacking Effectiveness

9. In this section, we evaluate the main features of the social protection system, focusing more specifically on its distributional effects and implementation challenges. To assess the effect on income distribution, we use data provided by the Ministry of Local Government and Rural Development as well as the 2015–16 Multi-Topic Household Survey (MTHSS).9 Data from the Ministry allows us to construct aggregate measures of social protection spending across different programs, while tracking the number of beneficiaries of each program. Using survey data, we estimate program coverage and transfer amounts received by individuals at different consumption levels. We then infer the distribution of pre-transfer consumption, which allows us to estimate the impact of each program on inequality.10 Combining this analysis with aggregate measures of spending enables us to determine whether social protection spending is used efficiently to reduce inequality.

A Generous Social Assistance System

10. Botswana’s social assistance budget is relatively large by international standards. Botswana’s allocation of 2 percent of GDP to social assistance programs in FY202211 positions it in the upper quartile among SSA countries and beyond the median for upper-middle income peers (Figure 2, Panel A). In contrast to many SSA countries, the funding for Botswana’s social safety nets comes predominantly from domestic sources, underscoring its self-reliant approach to social welfare (World Bank 2015, and UNICEF 2019).

11. The bulk of social assistance spending is concentrated in a few key initiatives. Above 90 percent of the budget dedicated to social assistance is absorbed by only 5 programs: tertiary education scholarships and sponsorships, which alone consume over 45 percent of the FY2022 social assistance budget or 1 percent of GDP, followed by old age pensions (20 percent of the budget, equivalent to 0.4 percent of GDP), Ipelegeng (public works), the Secondary School Feeding Program, and the Destitute Persons Program (Figure 2, Panel B and Table 1).

12. There has been a discernible decline in social assistance spending in recent years. Since FY2019, the allocation of social assistance (as a share of GDP) has been on a downward trend (Figure 2, Panel C). The allocation fell sharply from 3.7 percent of GDP in FY2019 to just above 2 percent in FY2022, with a decline of 20 percent in nominal terms, as the government prioritized other types of spending.

13. Compared to social assistance, spending on labor market programs and social insurance is more limited. Between FY2018 and FY2022, the nation earmarked, on average, merely 0.5 percent of its GDP to labor market programs. This allocation is modest when compared to the OECD median of 1.7 percent of GDP,12 a gap that becomes even more apparent given Botswana’s unemployment rate of 26 percent (as of September 2023) and the pressing need for skill enhancement across the workforce. Moreover, the country’s social insurance spending, at about 1 percent of GDP, is half the level of neighboring countries like Namibia and South Africa (World Bank ASPIRE Database).

Figure 2.
Figure 2.

Botswana Social Protection Spending and Coverage

(Percent of GDP)

Citation: IMF Staff Country Reports 2024, 287; 10.5089/9798400288258.002.A002

Note: Panel A: data for Botswana covers FY2022. Data for other countries is latest available from the World Bank ASPIRE Database. Botswana Social Assistance programs are defined in Box 1. Panel B: spending across programs in FY 2022/23. Panel C: coverage is defined as the share of total population who are direct or indirect beneficiaries of the program. Direct beneficiaries are those who receive the benefit; indirect beneficiaries are those who reside in the same household as program direct beneficiaries.

A Broad Safety Net Covering a Large Part of the Population....

14. The coverage of social assistance and labor market programs is extensive. More than half of the total population are beneficiaries, directly or through household association, 13 of at least one social assistance scheme—a figure that is double the SSA average and much higher than in upper-middle income countries (Figure 2, Panel C and Table 2). In terms of reach, the main programs are old age pensions, Ipelegeng program, and the universal secondary school feeding program, collectively casting a wide net over the nation’s recipients and beneficiaries (Tables 1 and 2). However, tertiary scholarships and sponsorships, despite being the most generously funded, benefit only 2½ percent of the population. Overall, the combination of all social assistance programs manages to cover nearly four-fifths of Botswana’s poor population. Furthermore, labor market programs cover 6 percent of total population, which is comparable to the average of uppermiddle income countries.14

15. On the other hand, social insurance covers a much smaller segment of the population. Approximately 10,000 pensioners receive benefits from the Public Officers Pension Fund, representing only 15 percent of those over the age of 65 (Table 1). There are 150,000 active contributors– representing 30 percent of formal employment and 18 percent of total (formal and informal) employment. When combined with non-public pension funds, slightly below 60 percent of formal sector employees are enrolled in a pension plan, thereby excluding a considerable portion of the formal and the majority of informal sector workers from contributory pension schemes (IOPS 2017).

Table 1.

Botswana: Social Protection Programs: Spending and Number of Recipients

article image
Source: MoLGRD, WB, 2022, and BPOPF annual report Note: The number of recipients is equal to the number of direct beneficiaries, and hence does not necessarily correspond to the number of direct and indirect beneficiaries that define the coverage of the programs in Table 2. Hence, the shares may differ between the two tables.
Table 2.

Botswana: Social Protection Programs Coverage (2016)

article image
Source: WB ASPIRE Database based on the 2015/16 multi-topic household survey Note: Coverage is defined as the share of population (total, within each decile of consumption, or by poverty status) receiving the transfer directly or indirectly (residing in the same household as the transfer recipient). Poverty is defined using the international poverty line of $2.15 2017 US$ PPP per day.

….Including Rich Households.

16. A significant share of social assistance benefits is not poverty-targeted.15 A large portion of social assistance funds benefits non-poor people, defined as those above the lowest quintile consumption bracket.16 In fact, more than 70 percent of beneficiaries are from the upper four quintiles of consumption (Table 2).17 Nonetheless, there is some heterogeneity across programs when it comes to benefit distribution: old age pensions and school feeding schemes display more progressivity compared to the allocation for tertiary scholarships.18

17. Compared to social assistance, labor market programs tend to better direct aid towards those most in need. Almost 40 percent of the individuals served by these initiatives fall within the poorest quintile of the population (Table 2). It is important to highlight, however, that, due to data limitations, the current analysis excludes the ISPAAD program, which represents the largest financial commitment among all labor market programs.

18. Finally, Botswana's social insurance scheme tend to benefit a more narrow and wealthier segment of the population. The recipients predominantly hail from higher income quintiles (Table 2), reflecting the comparatively generous remuneration of public sector employees, as noted by IMF (2023b).

Complex Design and Costly Administration

19. The large number of programs and oversight structures complicates the operation of the social protection system. As shown in Text Table 2, social protection encompasses 30 programs. These programs fall under the jurisdiction of 10 different ministries and government entities. Moreover, there is currently no single fully operational electronic registry of beneficiaries or joint information management system across programs to ensure that benefits are provided in a consistent manner and without duplication or omission. As a result, the targeting and selection of beneficiaries can be time-consuming, frequently requiring extensive paperwork. Beneficiaries may need to repeatedly submit the same information when they apply for different programs. Social protection programs are usually implemented by local authorities (e.g., city/town/district councils or district commissioners’ offices), but coordination with the relevant supervising ministry is often weak (MoLGRD, 2020). The implementation of these programs exhibits significant variation across districts (World Bank, 2022). Finally, the large number of delivery methods (Text Table 2), coupled with the absence of single social registry, limits the ability of the system to be rapidly expanded in times of shocks.

20. The social assistance budget, in particular, is burdened by high administrative costs. The lack of a unified registry of beneficiaries, coupled with the diversity of benefits, complex eligibility criteria, and labor-intensive registration processes, complicate the profiling of beneficiaries and the general management of the programs. This complexity results in substantial overhead costs, as highlighted by World Bank (2022). Specifically, administrative expenses constitute approximately 12 to 14 percent of the social assistance budget. In comparison, average administrative costs associated with social protection programs in the European Union amount to about 2½ percent of total spending.

21. The lack of coordination between programs also leads to considerable redundancy. For instance, two thirds of the destitute persons program beneficiaries also benefit from the old age pension (World Bank, 2022). While this overlap can sometimes provide necessary additional support, duplication more frequently creates the possibility that beneficiaries receive multiple forms of assistance, potentially beyond what they are eligible for (World Bank 2022). This leads to financial leakages and makes the system less efficient.

Limited Effects on Poverty and Inequality

22. Overall, the poor targeting and complexity described in previous paragraphs reduce the ability of social protection to markedly reduce poverty and inequality. In particular:

  • A large share of financing for social protection is allocated to regressive programs. Expenditures on tertiary education yield limited benefits in terms of poverty alleviation and inequality reduction,19 primarily due to their coverage that predominantly benefits individuals in the higher income quintiles. In a hypothetical scenario where all tertiary sponsorships and scholarships were discontinued, the impact on the poverty gap would be limited.20 However, such a measure would result in a 50 percent reduction in spending on social assistance, which could increase the benefit-cost ratio21 by more than a third without adversely affecting the Gini inequality coefficient.

  • At the same time, progressive social assistance programs receive significantly lower allocations. Old age pensions, destitute persons, and secondary school feeding collectively reach more than half of the poor population. But, the average spending per recipient of such programs is 9-11 times lower than the average spending per recipient of tertiary education scholarships and sponsorships, considerably limiting the distributional impact.

  • Administrative costs inflate total social assistance spending without providing benefits to the poor. By reducing administrative costs to levels observed in the European Union, the benefitcost ratio could improve by 10 percent. This improvement is observed even without considering the possible benefits from reallocating spending towards programs that have greater impact.

23. The limited effect on poverty and inequality is confirmed by an empirical analysis focused on social assistance.22 Based on the 2015–16 Multi-Topic Household Survey, the World Bank ASPIRE database provides various estimates of the effects of social assistance on different poverty and inequality metrics. The results are presented in Figure 3. Panel A plots the benefit-cost ratio and social assistance spending as a share of GDP. For Botswana, the benefit-cost ratio is 0.15, meaning that for every 100 pula spent on social assistance, only 15 accrue to the individuals below the poverty line. This outcome is significantly lower than the average benefit-cost ratio of approximately 0.3 and much lower than the frontier ratio of 0.5 for countries with comparable levels of spending. As discussed above, even though social assistance programs reach most of the poor population, actual spending is tilted towards richer individuals through regressive programs. Consequently, this leads to only modest improvements in the Gini coefficient (Figure 3, Panel B).

Figure 3.
Figure 3.

Effects of Social Assistance Spending on Poverty and Inequality

Citation: IMF Staff Country Reports 2024, 287; 10.5089/9798400288258.002.A002

Source: World Bank ASPIRE Database based on the 2015/16 Multi-Topic Household Survey and FAD Social Protection and Labor Toolkit. EME – emerging market economies, SSA – sub-Saharan Africa countries.

C. Financing Social Protection: The Other Side of the Coin

24. To have a more complete picture of its distributional effects, it is important to also consider the financing side of social protection. A progressive social protection system could, for instance, be funded through regressive taxes, which might negate its beneficial effects on inequality. Conversely, a broadly neutral system of transfers could still have beneficial effects if it is financed through highly progressive taxation.

25. This section focuses on the effects of the PIT and VAT, which have the most direct effects on income and consumption inequality.23 An in-depth examination of the entire budget, encompassing both expenditures and revenues, is out of the scope of this paper. The objective of this short section is not to provide a comprehensive analysis, but rather to offer a preliminary assessment of the consolidated effects of social protection and its financing.

26. PIT and VAT taxes have limited impact on inequality in Botswana. Regarding PIT, Botswana has five income tax brackets with progressive rates varying from 0 to 25 percent for top earners. The highest marginal tax rate is low by international standards and is the lowest in the SACU region.24 Consequently, PIT has limited effect on inequality, reducing Gini coefficient by 3.5 percentage points, a figure that is lower than simulated for other countries in the region (Figure 4). Regarding VAT, Botswana applies a uniform rate of 14 percent, albeit with some exceptions. Most notably, some staple food items25 and fuel are taxed at a zero rate. The VAT impact on the Gini coefficient is assessed by analyzing consumption shares across the income distribution using data from the 2015-16 Multi-Topic Household Survey.26 Given that wealthier households tend to consume more fuel and transport services, while poorer households consume more food, the overall effect of VAT on inequality metrics is marginally negative, with an estimated reduction in the Gini coefficient by approximately 1 percentage point.

27. The limited evidence presented in this section suggests that the tax system is insufficiently progressive to markedly add to the effect of social protection on income inequality. With the relatively flat PIT rate schedule, individuals with higher earnings do not contribute enough towards a more equitable distribution of income. Furthermore, although consumption taxes represent a smaller share of total taxes in Botswana than in other African countries,27 VAT exemptions for certain goods may benefit richer consumers more than poorer consumers, aggravating income disparities. For social protection to have a more pronounced effect on income equality, there is a clear need for a more progressive taxation approach, ensuring that the tax system complements rather than undermines the goals of social protection.

Figure 4.
Figure 4.

Redistributive Capacity of the PIT, 2023

(Percent)

Citation: IMF Staff Country Reports 2024, 287; 10.5089/9798400288258.002.A002

Sources: World Inequality Database, IBFD, IMF-staff computationsNote: The results are based on the IMF FAD Personal Income Tax tool based on the survey microdata and simulations. Redistributive capacity is calculated as the difference between pre-tax and post-tax Gini coefficients in percentage points.

D. Reforms: International Experiences and Options for Botswana

28. This section explores international best practices in the management of social protection systems and proposes reform options for Botswana. It examines four key areas: (1) overall design, (2) choice of programs, (3) delivery, and targeting technologies, and (4) financing. General principles are discussed in IMF (2014), Clements and others (2015), World Bank (2022b), and ILO (2023). Country experiences are discussed in greater details in Soares and others (2010), World Bank (2014), ILO (2016), and World Bank (2019). Some policy options discussed below are priority areas of the authorities’ NSPF.

Simplifying the Social Protection System

29. Best practice 1: Several developing and emerging countries have rethought their social protection system towards more simplicity and adaptability. Examples include Brazil, Indonesia, and China (World Bank, 2019). This process has been achieved by:

  • Consolidating programs. The consolidation of numerous fragmented programs into fewer programs with broader coverage enables countries to achieve enhanced coherence in policy objectives, aligning social protection initiatives more closely with broader national development goals. Additionally, a more unified and streamlined system improves accessibility for beneficiaries, simplified the process to obtain assistance, and minimized bureaucratic barriers.

  • Introducing a single social registry. A single social registry centralizes the collection, management, and use of data regarding potential beneficiaries. This enhances data collection and analysis capability, thereby improving the targeting of programs. It also supports the continuous evaluation and adaptation of programs to address the evolving needs of the population effectively.

  • Making the system more ”adaptive.”28 Adaptive social protection systems are designed to be flexible, allowing for rapid adjustments—both vertical (the level of benefits) or horizontal (the number of beneficiaries)—in response to economic shocks, natural disasters, or demographic shifts. This responsiveness not only provides immediate relief to those affected by adverse events in the near term, but also contributes to long-term poverty reduction and social equity by preventing the deepening of existing vulnerabilities.

30. The Bolsa Familia Program in Brazil is a good example of effective streamlining of the social protection system. This program, which was introduced in 2003, has integrated multiple preexisting cash transfer initiatives into a unified conditional cash transfer program, supported by a new single social registry. Thanks to its flexible design and solid underlying infrastructure, the Bolsa Familia program can expand or adjust its coverage to align with socio-economic developments and adapt its structure and conditions to address new challenges, such as those encountered during the COVID-19 pandemic (World Bank, 2020).

31. Application to Botswana: The authorities could consider the following measures to streamline their social protection system:

  • Reducing the number of programs. The current landscape, featuring nearly 30 programs, results in a fragmented and operationally expensive system. There is potential to enhance efficiency by merging smaller, lower-coverage programs into broader ones with wider eligibility criteria.

  • Accelerating the rollout of the digital Single Social Registry (SSR). Initiated in 2016, the creation of the SSR has encountered delays due to the pandemic and various technical hurdles.29 Nonetheless, the SSR remains vital for improving the precision of targeting, reducing administrative expenses, fostering program coordination, and enabling more effective monitoring and evaluation. Furthermore, it would support the adaptability of the social protection framework.

  • Improving program adaptability. Given Botswana's reliance on mining activities and vulnerability to climate-related challenges like droughts, the economy is particularly prone to shocks. Thus, social protection programs that can be swiftly scaled up in response to economic or environmental crises are essential. Such programs should be capable of adjusting their benefits and beneficiaries to meet the specific needs arising from the materialization of risks, thereby delivering more targeted and efficient aid.30

Introducing More Effective Programs

32. Best practice 2: Two types of social protection programs have proven to be very effective in reducing inequality and boosting employment:

  • Conditional cash transfers (CCT). CCT provide financial assistance directly to families, often the poorest, under the condition that they meet certain criteria, such as ensuring that their children attend school or receive regular health check-ups. This approach not only helps to alleviate immediate financial hardship for the most vulnerable, but also encourages behaviors that can break the cycle of poverty in the long term (hence, facilitating timely graduation from means-tested programs). For example, the Prospera program in Mexico (previously known as Opportunidades) has played a key role in reducing poverty in some regions and resulted in significant improvements in education, consumption, and labor market outcomes in the medium and long term (Parker and Todd 2017, Parker and Todd 2020, World Bank 2014).

  • Comprehensive active labor market programs that enhance broad-based skills and address other labor market bottlenecks. On the labor supply side, programs focused on workforce development, technical and vocational education, and training have been effectively implemented in various emerging economies (e.g., Ghana, Uruguay, India). ILO (2023) offers a review of successful initiatives. These programs aim to enhance employability by equipping individuals with skills in demand by the private sector. Such strategies are known to address employment barriers and promote social mobility. The effectiveness of these policies in low- and middle-income settings is well documented, with outcomes heavily dependent on the design and execution of the programs (Kluve and others, 2019). In addition to supply-side programs, other policies can be used to facilitate the functioning of the job market and foster entrepreneurship. These include employment incentives, public employment services, and start-up incentives, among others. Ample evidence shows that comprehensive approaches (not just focused on boosting supply) are more effective in tackling the multifaceted nature of unemployment, particularly for the youth (ILO 2022).

33. Application to Botswana: The authorities might contemplate improvements to certain social protection programs by:

  • Moving from in-kind delivery to means-tested conditional cash transfers.31 Transitioning to this approach for some programs could eliminate inefficiencies associated with procurement and delivery and enhance targeting accuracy. Implementing such a change necessitates sufficient administrative capacity to conduct means-testing and design and monitor the conditionality of transfers.32 Additionally, the success of this transition is dependent on the effective operation of the SSR.

  • Reforming labor-market programs with greater focus on skills development and market functioning. To achieve a meaningful reduction in inequality and poverty, Botswana authorities need to foster conditions for job creation, especially among the youth. In this regard, there is scope to consolidate and reform the multiple labor-market programs. A primary objective should be to enhance life-long learning and the acquisition of both technical and soft skills, which could significantly enhance employability and adaptability among the workforce.33 In this context, cooperation with the private sector will be pivotal, as such collaboration can ensure that training programs are aligned with industry needs. Furthermore, complementary policies can include job-search assistance, career guidance, and intermediation services allowing young workers to signal their skills. Entrepreneurship promotion can also facilitate youth-owned businesses.34

Exploiting New Technologies for Targeting and Delivery

34. Best practice 3: Digital technology can be leveraged to improve the efficiency of social protection.35 Governments in some emerging countries (Turkey, Chile) have developed sophisticated electronic social registries that allow them to exchange data across various databases (such as land, vehicle, and health registries, among others). Such integration can be used to proactively target potential beneficiaries. Additionally, the application of modern machine-learning techniques to analyze registry data allows for the construction of need-based indicators, as seen in Colombia and Togo, and can be supplemented with non-traditional ‘big data’ sources in contexts where government data are sparse. Furthermore, the adoption of modern payment methods, like mobile money, has proven to significantly reduce the delivery costs of social assistance and insurance programs—a strategy employed by Ghana, Kenya, Rwanda, and Uganda.

35. Application to Botswana: The authorities could build capacity to harness the potential of modern technology, with a view to improving the efficiency of their social programs. This includes enhancing the digitization of beneficiary data, and regularly updating the dataset. By integrating information from this registry with other data sources, such as satellite imagery or mobile phone usage, the authorities can identify opportunities for expanding social protection programs. For example, satellite data on crop performance in drought conditions may be utilized to scale up the programs in the most affected areas, as done in Kenya, Uganda, or Niger (World Bank, 2022b). Moreover, Botswana can capitalize on innovative delivery mechanisms, like digital or mobile payments, supported by the nation’s comprehensive identity card system.

Ensuring Sufficient and Progressive Financing

36. Best practice 4: Optimizing the financing of social programs is crucial to strengthen the progressivity of fiscal policy and ensure that scarce resources are used in the most efficient way, given limited fiscal space. Insights from global practices highlight two key strategies:

  • Redirecting funds away from less effective programs. This approach, which could be implemented in a budget-neutral way, ensures that limited resources are directed to interventions that yield the highest impact, particularly in improving the lives of the most vulnerable. By prioritizing funding for programs with proven outcomes, governments can achieve greater poverty reduction, enhanced social equity, and improved human capital development. Such reallocation often involves rigorous evaluation of existing programs to identify areas where funds may be underutilized or not delivering the desired results. This approach, for example, was successfully implemented in Indonesia in the early 2000s, when costly and regressive fuel subsidies were cut to finance a scheme supporting low-income families (ILO, 2016).

  • Improving the progressivity of income taxation and minimizing the use of reduced VAT rates. PIT, when designed with progressive rate structures (higher income brackets taxed at higher rates) directly targets income disparities by redistributing wealth from richer segments of society to fund public services and social protection programs that benefit the poor. Consumption taxes, such as VAT are, in general, a less preferable tool to combat inequality. Exemptions or reduced VAT rates are blunt redistributive instruments, as the poor usually consume less than the rich in absolute terms.36

37. Application to Botswana: On the financing side, the priority is to maintain or increase the budget allocated to social protection, while enhancing progressivity. The authorities could consider:

  • Reallocating existing funding across programs. Introducing means-tested eligibility criteria for tertiary education scholarships and sponsorships could reduce the number of beneficiaries (excluding the richest ones) and generate savings. Alternatively, the government could evaluate replacing most of the scholarships and sponsorships with a student loan program. The funds saved from such reforms could be redirected to more targeted programs that better address the needs of economically disadvantaged groups, including school feeding schemes, old age pensions, and the destitute persons program.

  • Enhancing tax system progressivity. While the social protection system plays a key role in mitigating inequality, there is room for refining the tax framework to make it more progressive. This could involve increasing the top marginal PIT rates, which are currently among the lowest in SSA. Furthermore, reconsidering the application of zero-rated VAT on items that disproportionately benefit higher-income groups, such as petrol and diesel, could also be considered (see IMF 2017b).

E. Conclusions

38. Botswana stands as one of the most unequal countries globally. The primary driver is the economic structure, which heavily relies on capital-intensive mining industries. Additionally, geographic characteristics (with the country being both vast and sparsely populated) further exacerbate the situation. These elements create a challenging environment for addressing economic disparities and fostering a more inclusive growth pattern. In this context, targeted fiscal interventions can play a key role to mitigate these trends.

39. The social protection system, despite being designed to support the neediest, suffers from some inefficiency. While the country has a wide array of programs aimed at reaching its poor population, the allocation of funds is tilted towards regressive programs. Furthermore, the system is characterized by fragmentation and incurs high administrative costs, compounding its inefficiency.

40. Improvements in Botswana’s social protection system are both necessary and achievable. Simplifying the existing framework, making it more adaptive, and introducing a single social registry, could significantly enhance the system’s efficiency and effectiveness. By strengthening the role of conditional transfers and active labor market policies, the system could be better aligned with the needs of the population it aims to serve. Modernizing delivery and targeting methods is another critical step towards ensuring that administrative costs are reduced, and assistance reaches those who need it most. Finally, securing sufficient and progressively-sourced financing is paramount to sustaining improvements over time.

References

  • Botswana Public Officers Pension Fund. 2023. “Annual Report 2022/23.”

  • Clements, B. J., R. A. de Mooij, S. Gupta, and M. Keen. 2015. Inequality and Fiscal Policy, (Washington: International Monetary Fund).

  • Cust, J., D. Shantayanan, and P. Mandon. 2022. “Dutch disease and the public sector: How natural resources can undermine competitiveness in Africa,” Journal of African Economies, Vol. 31, No. Supplement_1, pp. i10-i32.

  • Gaspar, V., and M. García-Escribano. 2017. “Inequality: Fiscal Policy Can Make the Difference,” https://www.imf.org/en/Blogs/Articles/2017/10/11/inequality-fiscal-policy-can-make-the-difference.

  • International Labor Organization (ILO). 2016. “Financing Social Protection Through Contributions and the Removal of Fuel Subsidy,” https://www.social-protection.org/gimi/gess/RessourcePDF.action?ressource.ressourceId=53849.

  • International Labor Organization (ILO). 2022. “The Impact of Active Labor Market Programmes on Youth: Systematic Review and Meta Analysis 1990-2022.” (International Labor Organization: Geneva).

  • ______________________________. 2023. “Public employment services and active labour market policies for transitions,” (Geneva: International Labour Organization).

  • International Finance Corporation (IFC). 2022. “Creating Markets in Botswana: Country Private Sector Diagnostic” (Washington: World Bank).

  • International Monetary Fund (IMF). 2014. “Fiscal Policy and Inequality,” IMF Policy Paper 2014 (Washington: International Monetary Fund).

  • ______________________________, 2017a. “Tackling inequality,” in IMF Fiscal Monitor (Washington: International Monetary Fund).

  • ______________________________. 2017b. “Botswana: Selected Issues,” IMF Country Report No. 2017/250 (Washington: International Monetary Fund).

  • ______________________________. 2023a. “Botswana: 2023 Article IV Consultation Staff Report,” IMF Country Report No. 2023/117, (Washington: International Monetary Fund).

  • ______________________________. 2023b. “Botswana: Selected Issues,” IMF Country Report No. 2023/318 (Washington: International Monetary Fund).

  • Kluve, J., S., S. Puerto, D. Robalino, J. Romero, F. Rother, J. Stoterau, F. Weidenkaff, M. Witte. 2019. “Do youth employment programs improve labor market outcomes? A quantitative review,” World Development, Vol. 114, pp. 237-253.

  • Lowe, C., J. P. Rigolini, L. Solbes Castro, and F. A. Bastagli. 2023. “Pathways toward digitalization in Social Protection and Labor (SPL) service delivery,” World Bank Working Paper No. 185112 (Washington: World Bank).

  • International Organization of Pension Supervisors (IOPS), “IOPS country profile: Botswana,” https://www.iopsweb.org/resources/46896420.pdf

  • Ministry of Local Government and Rural Development (MoLGRD). 2020a. “National Social Protection Framework.”

  • ______________________________. 2020b. “National Social Protection Framework: Implementation Plan.”

  • ______________________________. 2022. “Botswana Social Protection Consolidation: Options Paper.”

  • Organization for Economic Co-operation and Development (OECD). 2022. “Reassessing the regressivity of the VAT,” OECD Taxation Working Paper No. 49 (Paris: OECD).

  • ______________________________. 2023. “Revenue Statistics in Africa 2023,” (Paris: OECD).

  • Parker, S., and P. Todd. 2017. “Conditional Cash Transfers: The Case of Progresa/ Oportunidades.” Journal of Economic Literature 55(3): 866-915.

  • Parker, S., and T. Vogl. 2023, “Do Conditional Cash Transfers Improve Economic Outcomes in the Next Generation? Evidence from Mexico.” Economic Journal, October 2023

  • Statistics Botswana. 2023. “Quarterly Multi-Topic Labour Force Module Q3, 2023.”

  • Stepanyan, A., R. Garcia Verdu, A. David, F.F. Fleermuys, and P. Gitton. 2013. “Enhancing Inclusive Growth and Employment in Small Middle Income Countries”, in Selected Issues Paper on The Challenges of Small Middle-Income Countries in sub-Saharan Africa (Washington: International Monetary Fund).

  • Soares, V. F., R. Perez Ribas, R. Guerreiro Osorio. 2022. “Evaluating the Impact Of Brazil’s Bolsa Familia: Cash Transfer Programs in Comparative Perspective,” Latin American Research Review, Vol. 45, No. 2, pp. 173-190.

  • United Nations Development Programme (UNDP). 2021. “Inequality in Botswana: An analysis of the Drivers and District-level Mapping of Select Dimensions of Inequality.”

  • United Nations International Children’s Emergency Fund (UNICEF). 2019. “Botswana Budget Brief: Social Protection,” https://www.unicef.org/esa/media/5786/file/UNICEF-Botswana-2019-Social-Protection-Budget-Brief.pdf.

  • ______________________________. 2022. “Country Office Annual Report.”

  • Vision 2036 Presidential Task Team. 2016. “Vision 2036 – Achieving prosperity for all,” Botswana.

  • Warwick, R., Harris, T., Phillips, D., Goldman, M., Jellema, J., Inchauste, G., and K. Goraus-Tańska. 2022. The redistributive power of cash transfers vs VAT exemptions: A multi-country study. World Development, Vol. 151, 105742.

  • World Bank. 2009. “Conditional Cash Transfers. Reducing Present and Future Poverty,” (Washington: World Bank).

  • ______________________________. 2014. “A Model from Mexico for the World,” https://www.worldbank.org/en/news/feature/2014/11/19/un-modelo-de-mexico-para-el-mundo

  • ______________________________. 2015. “Botswana Poverty Assessment,” (Washington: World Bank).

  • ______________________________ . 2017. “Indonesia’s Social Assistance System: Praising Reforms But More Work Ahead,” https://blogs.worldbank.org/en/eastasiapacific/indonesia-social-assistance-system-praising-reforms-more-work-ahead.

  • ______________________________. 2019. “Schemes to Systems: The Future of Social Protection in India,” https://www.worldbank.org/en/news/feature/2019/11/20/schemes-to-systems-future-social-protection-india.

  • ______________________________. 2020. “World Bank approves loan supporting the expansion of the Bolsa Familia Program,” https://www.worldbank.org/en/news/press-release/2020/10/29/world-bank-approves-loan-supporting-expansion-of-bolsa-familia-program

  • ______________________________. 2022a. “Botswana Social Protection Programs and Systems Review,” (Washington: World Bank).

  • ______________________________. 2022b. “Charting a Course Towards Universal Social Protection,” (Washington: World Bank).

  • ______________________________. 2023. “Botswana – Systematic Country Diagnostic Update: At a Crossroads – Reigniting Efficient and Inclusive Growth,“ (Washington: World Bank).

  • ______________________________.2024a. “ASPIRE: The Atlas of Social Protection – Indicators of Resilience and Equity.” Retrieved from: https://www.worldbank.org/en/data/datatopics/aspire/about

1

Prepared by Sergii Meleshchuk (AFR). The author wants to thank, for their very helpful comments and suggestions, Luc Eyraud, Andrea Richter Hume, BoB Deputy Governor Kealeboga Masalila, BoB Director Innocent Molalapata, Carolina Diaz-Bonilla, Liang Wang, the World Bank social protection team, Samuel Phiri, Balazs Horvath, the UNDP and UNICEF teams in Botswana, Claudia Rodriguez, Usama Zafar, and the participants of the workshop at the Bank of Botswana.

2

Real income per capita is measured in 2017 US$ using purchasing power parity (PPP) rates.

3

The share international poverty line is defined as $2.15 per day in terms of 2017 PPP rates.

4

The last household survey that can be used to measure Gini coefficient was conducted in 2015-16. Income and consumption inequality may have further deteriorated during the Covid-19 pandemic (UNDP, 2021).

5

Gender inequality does not seem to be one of the main drivers of inequality in Botswana, which fares better than sub-Saharan African countries’ average along various indicators (e.g., Gender Development Index (UNDP), Gender Inequality Index (UNDP), Global Gender Gap Index (World Economic Forum), secondary and tertiary education enrollment rates, maternal mortality rate). However, there is still room for improvement, as Botswana is among the countries with the highest gender-based violence incidence.

6

Cust and others (2022) document this pattern for many countries: resource revenues windfalls often lead to elevated public wage bill.

7

See IMF (2014), Clements and others (2015), and IMF (2017a) for a discussion on how fiscal policy can reduce inequality.

8

Other types of public spending and policy interventions that may have direct effects on poverty and inequality, such as health, education, infrastructure, and regulated utility prices, are outside the scope of this chapter, which focuses, more narrowly, on social protection programs.

9

The 2015–16 MTHHS contains responses from more than 7,000 households that constitute a representative sample of the Botswanan population. The survey has information on household consumption and income, with detailed breakdown of social protection benefits across various programs. This is the main source of information to conduct an analysis of programs’ incidence and effects on inequality. It has, for instance, been used to construct the indicators underlying the World Bank Atlas of Social Protection Indicators of the Resilience and Equity (ASPIRE) cross-country database. The survey was also used in the recent Social Protection Review by the World Bank (2022). This recent study differs from the previous one as it analyzes the joint effects on inequality of social protection and tax systems and uses more complete information on spending across different programs.

10

To calculate pre-transfer consumption, we subtract the value of transfers from total consumption. Pre-transfer consumption is a measure of welfare commonly used in the social protection literature (see, e.g., WB (2022) and the World Bank Aspire Dataset).

11

The fiscal year in Botswana runs from April 1 to March 31.

12

OECD countries implement active and passive labor market policies. The second type does not exist in Botswana. Spending on passive labor market policies represents, on average, 0.6 percent of GDP in OECD countries.

13

Beneficiaries include recipients (direct beneficiaries) and those who reside in the same household with the recipients (indirect beneficiaries).

14

This is a conservative estimate of the coverage of labor market programs, since our analysis does not include ISPAAD due to insufficient data.

15

The 2022 Bank of Botswana Annual Report (Chapter 2) reaches similar conclusions.

16

Using the 2015–16 MTHHS, households are ranked based on per-capita consumption. Consumption is defined in pre-transfer terms by subtracting from total consumption the amount of all identified social protection transfers that the household receives. The analysis omits some social protection programs, notably social insurance and some labor market programs due to survey questionnaire limitations.

17

Some social protection programs (e.g., tertiary education scholarships and sponsorships) are not poverty-targeted with eligibility criteria that are not means-tested.

18

As shown in Table 2, the share of beneficiaries receiving old age pensions declines with the level of consumption, from almost 38 percent in the poorest quintile to less than 5 percent in the richest. On the other hand, the share of beneficiaries receiving tertiary education increases from 1.8 percent (poorest quintile) to 3.4 percent (richest quintile).

19

While poverty eradication is not an explicit goal of tertiary education subsidies and scholarships, these programs fall under the umbrella of the NSPF, whose purpose is “to prevent, address, and reduce the risks of poverty and vulnerability for Batswana throughout their lives.”

20

Tertiary education subsidies may have positive effects on human capital accumulation in the long run. This chapter focuses on distributional impacts. It does not take into account the effects of tertiary education subsidies on human capital accumulation and future earnings of the recipients. As access to tertiary education is skewed towards wealthier individuals, these earnings effects may only lift a limited number of people out of poverty, while aggravating inequalities.

21

The benefit-cost ratio is calculated as the reduction in poverty gap for every unit of local currency spent. Poverty gap is defined as the difference between the poverty line and actual income, or zero, if income is above the poverty line. Hence, a benefit-cost ratio of 1 implies that every dollar spent goes to individuals with incomes below the poverty line, while a benefit-cost ratio of 0.5 implies that only 50 cents per dollar spent reach poor individuals.

22

Due to data limitations, it is not possible to conduct the same analysis for labor market and social insurance programs.

23

Social assistance and labor market programs are financed directly from the budget. VAT, personal income, or other taxes are not specifically earmarked for spending on social assistance and labor market programs.

24

Based on the most recent PWC Worldwide Tax Summaries.

25

Zero-rated food items include, among others, some sorghum, rice, and maize products, brown breads, as well as fresh vegetables and fruits.

26

We estimate the amount of VAT paid by each individual in the survey. To do so, we first compute ꚍg, the average VAT rates for 12 categories of consumption goods and services, denoted by g, like food or transportation (by aggregating individual VAT rates for the items within each category). Then, the VAT paid by individual i is computed as VATi = ∑g ꚍgCig where Cig is consumption by individual i of consumption category g.

27

See comparative analysis in OECD (2023). Consumption taxes, levied at a flat rate, are generally regressive.

28

For the purpose of this paper, we define “adaptive” social protection systems as those capable of rapidly scaling up coverage or benefits in response to shocks.

29

In the past decade, the World Bank has provided substantial support to Botswana in developing and operationalizing the SSR, including during its pilot phase in 4 districts.

30

Adaptive programs should be able to scale down benefits to normal levels and allow beneficiaries to graduate when the shock is over.

31

The introduction of conditional cash transfers should be based on a rigorous analysis to determine which social sector services (education, health, nutrition, etc.) should be incentivized. Moreover, given the level of poverty in Botswana, some segments of the society should still receive direct income support through unconditional cash transfers.

32

In the 2010s, the Government of Botswana started to reform the targeting system through a proxy means test (PMT) mechanism, aimed at objectively identifying a household’s poverty status based on empirically derived poverty scoring coefficients. This system was pilot-tested to assess its performance in 2015/16, and then fine-tuned for nationwide use. However, it was never scaled to full implementation.

33

The launch of the Ipelegeng Skills Development Component in 2022 is a positive development in this regard.

34

The newly established Chema Chema Fund is another step in the right direction. It aims to provide credit to young people, stimulate entrepreneurship, and offer access to credit and business development services for the youth.

35

The examples provided in this paragraph are based on Lowe and others (2023).

36

As discussed in Warwick and others (2022) while reduced VAT rates and exemptions may reduce poverty, they are usually quite expensive in terms of forgone tax revenues, and rich households benefit more in nominal terms.

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover IMF Staff Country Reports
Botswana: Selected Issues
Author:
International Monetary Fund. African Dept.
Volume/Issue:
Volume 2024: Issue 287
Publisher:
International Monetary Fund
ISBN:
9798400288258
ISSN:
1934-7685
Pages:
49
DOI:
https://doi.org/10.5089/9798400288258.002
Figure 1.

Real Income per Capita, Poverty, and Inequality
Figure 2.

Botswana Social Protection Spending and Coverage

(Percent of GDP)
Figure 3.

Effects of Social Assistance Spending on Poverty and Inequality
Figure 4.

Redistributive Capacity of the PIT, 2023

(Percent)