Fiscal space and public spending on children in Burkina Faso

Burkina Faso has experienced consistently high economic growth over the last two decades. Despite a relatively solid economic performance, Burkina Faso remains a poor country and lags significantly in terms of infrastructure and human development. Since the beginning of the 2000s, Burkina Faso has made commendable efforts to promote development and to reduce poverty by developing a strategic framework for the fight against poverty (CSLP). However, assessments of the implementation of the CSLP indicate that it was largely undermined by the weakness of sectoral policies and insufficient budget allocations (Ministry of Economy and Finance 2010).

Child poverty results in huge social and economic losses with long-term effects on individuals and future generations. Intergenerational effects are not limited to poor children, as child poverty also significantly reduce economic growth for society as a whole. For example, child malnutrition imperils the acquisition of basic knowledge and thus productivity as an adult. Public social spending can improve child wellbeing, in the long run contributing to the economic growth of a nation. However, governments face budget constraints and need to make difficult choices concerning allocations of their limited budget between many pressing needs. Moreover, the sustainability and impacts of public spending on different programs depend on the financing mechanism(s) put into place.

The goal of this paper is to evaluate the impacts of three different scenarios that increase public spending on children under given budget constraints. We use a macro-micro framework to quantify the impacts of these scenarios on different education and poverty indicators.

The rest of the paper is organised as follows: the next section consists of an analysis of the poverty and the types of vulnerability and risks affecting children, in addition to a brief review of the existing social protection framework, in Burkina Faso. This is followed by two sections, which respectively present a literature review and the methodology used. The simulation results are then discussed before we conclude.

As in many developing countries, in Burkina Faso children are more likely to be living in poverty than adults. Batana et al. (2012) find that children are 20 percent more likely than adults to live in poverty in developing countries.

Major progress has been accomplished in child education in Burkina Faso since 2000, particularly with respect to the supply of and access to universal primary education. The gross enrollment ratio in primary school went from 40% in 2000 to 87% in 2013. However, the rapid increases in the number of registered students, following the elimination of school fees for primary education, constitute a significant challenge with respect to education quality. Rapid population growth also puts pressures on the system’s capacity. In effect, average per capita public spending has not stopped falling since 2003 and student-teacher ratios have barely budged (World Bank, 2009).

In Burkina Faso, a large share of individuals are excluded from health care, particularly in rural areas. Major improvements have nevertheless been achieved: child vaccination objectives were largely exceeded and the number of health centres and personnel have both increased rapidly. Despite significant progress in health coverage, with a reduction in the average distance to reach a Health and Social Promotion Centre (CSPS), nearly one in ten children dies before the age of five in Burkina Faso¹. Furthermore, more than one-third also suffer from delayed height and weight growth.

In terms of nutrition, stunting declined from 43.1% of children under the age of five in 2003 to 35.1% in 2009, with more noteworthy progress in rural areas. The share of children suffering from wasting also fell by nearly half, although the level still remains “severe” according to WHO norms. A strategic plan for 2010–2015 was adopted in order to reduce hunger and illness linked to nutritional deficiency diseases.

The pursuit of Burkina Faso’s Poverty Reduction Strategy and the recent global Sustainable Development Goals (SDGs) requires for more detailed study of the impacts of public policies and social reforms. The studies reviewed below evaluate the consequences of fiscal reforms and public spending on vulnerable populations (poor households, orphans and other vulnerable children, etc.).

Child wellbeing is affected by fiscal and budgetary policies, notably through public services such as health and education, and improving the household economy (Waddington, 2004). A few pioneering studies are based on evaluation methods for fiscal and budgetary reforms and policies. Ahmad and Stern (1984) use benefit incidence analysis to evaluate the impacts of fiscal reforms on wellbeing and marginal benefit incidence analysis to assess the benefits associated with tax reforms. This approach has seen numerous applications, including contributions on child poverty, e.g. Yitzhaki and Thirsk (1990), Yitzhaki and Slemrod (1991), Mayshar and Yitzhaki (1996), Ray (1997), and Makdissi and Wodon (2002).

Bibi and Duclos (2004) extend benefit incidence analysis to identify the direction of impacts of fiscal reforms when the objective is to reduce poverty. Their approach, illustrated with Tunisian data, consists in deriving the cost-benefit ratio of an increase in a consumption tax by minimizing a poverty indicator.

Recent fiscal reforms in Africa following trade liberalization have rekindled interest for analysis of the impacts of fiscal reform on social welfare (Sahn & Younger, 2003; Chen et al, 2001; Rajemison & Younger, 2000; Alderman & del Ninno, 1999). These analyses are particularly interesting as they can be used to analyse fiscal changes induced by new policy regimes, as well as their impacts on the most vulnerable populations.

Benefit incidence analysis has the advantage of requiring little data, making it relatively easy to carry out (Ray, 1997). This is particularly suitable for developing countries where little data is available. However, it is limited to measuring the direct effect of the fiscal reform and thus ignores indirect effects that the reform may have on individuals and their behavioral responses.

Some authors (Glewwe, 1991; Gertler & Van Der Gaag, 1990), aiming to address the shortcomings of this approach, developed econometric estimations of the impact of fiscal policies on wellbeing by controlling for other variables that may influence the estimations.

The two preceding approaches (marginal incidence and econometric analysis) are partial equilibrium, and thus do not capture the feedback effects induced by other sectors or actors in the economy. These effects need to be considered for major public policy reforms with economy-wide impacts.

Computable general equilibrium (CGE) models are the most comprehensive tool to study the impacts of such policy reforms. They have been used to analyze the impact of a variety of public policies (trade liberalization, fiscal reforms, increased public spending) in developing countries. In order to capture household-level impacts, some such studies combine the CGE model with a microsimulation model (Decaluwé et al., 1999; Cogneau & Robillard, 2001 and 2004; Cockburn, 2006; Bourguignon et al., 2003; Boccanfuso et al., 2003).

The literature provides few applications of CGE models to Burkina Faso (Balma et al., 2011; Sawadogo et al., 2015). Gottschalk et al. (2009) use the MAquette for MDG Simulations (MAMS) model, developed by Lofgren and Diaz-Bonilla (2006), to simulate fiscal space and how it could be used to finance the pursuit of the MDGs. They identify three mechanisms to create fiscal space: establish priorities in public spending, increase foreign borrowing, or increase government revenues. In terms of uses of this space, they analyse an increase in spending on health, education and infrastructure. The authors show that there are trade-offs, for example between impacts on monetary poverty and health/education impacts, between the different spending scenarios. There are also trade-offs between financing mechanisms in terms of the reliance they create on domestic and foreign sources of financing. Investment in infrastructure is shown to not only beneficial for growth, but also for attainment of the MDGs. Health and education spending may run into structural constraints, notably the fact that learning and the training of personnel takes time. Our study complements the literature by contributing a macro-micro analysis that specifically deals with child poverty and schooling.

5.3 Analysis of results

These different interventions, while equivalent in terms of public spending, will have significantly different impacts in terms of education outputs, GDP and children poverty.

5.3.1 Impact on education

Student behaviors, such as school participation and promotion rates, are affected by different factors (cf. Section 4.1). The different policy interventions, in turn, affect these factors differently: an increase in education spending has a direct impact on the education quality index; a school fees subsidy acts primarily by reducing the relative price of education; cash transfers essentially increases household real per capita consumption.

School fees subsidies have a larger beneficial impact on primary school entry but also increase the dropout rate, whereas increasing education spending has a smaller positive impact on entry but simultaneously reduces the dropout rate (Figures 1 and 2). Indeed, subsidizing school fees causes more six-year-old children to join the education system. However, since public spending on education is constant in this scenario, the higher number of students leads to a reduction in the quality of education services received by each student and therefore increases the dropout rate. Under the “spending” scenario, the education quality index increases, thus favouring both primary school enrollment and the duration of studies.

Cash transfers to households increase household real per capita consumption directly. However, our simulation results indicate that this has a negligible impact on education decisions. The different indicators are thus essentially the same as in the reference scenario (Figures 1 and 2).

Figure 1

Download asset Open asset

Figure 2

Download asset Open asset

5.3.2 Impact on labour market and growth

Given the reduction in dropout rates, it is not surprising to find that the supply of skilled workers increases most under the spending scenario. Despite higher entry rates, higher dropout rates in the subsidy scenario imply that the number of skilled workers – those who complete the first level of secondary school – increases by just 3 000 (Figure 3). Given the limited impacts of a cash transfer policy on education decisions, it is unsurprising that it has a negligible impact on the number of skilled workers (less than 400 new workers). As a result, the impacts on growth of these two scenarios are barely perceptible (Figure 4).

Figure 3

Download asset Open asset

Figure 4

Download asset Open asset

All of the scenarios have a similar impact on the deficit-to-GDP ratio (solid line in Figure 5): it increases during the period of increased spending and then returns near to reference values. Given that, compared to the reference scenario, the end-of-period ratios are slightly higher in the spending scenario and slight lower in the subsidy and cash transfer scenarios, the end-of-period debt-to-GDP ratios (broken line) are greater for the latter. These results can be explained by the increased economic growth in the spending scenario, which increases the denominator of the debt-to-GDP ratio, as income-to-GDP ratios (dotted line) are practically identical in all scenarios.

Figure 5

Download asset Open asset

5.3.3 Impact on poverty

While the cash transfer scenario has little macroeconomic impact, as it directly impacts household real consumption, its effects on poverty are significant. All of the poverty indicators, expressed in both monetary and caloric forms, improve markedly beginning in the first year (Figure 6): the incidence of monetary poverty in 2010 is three percentage points lower than in the reference scenario and the incidence of caloric poverty is nearly five percentage points lower. In other words, there would be three percent fewer children living under the poverty line and five per cent fewer children suffering from caloric poverty. These impacts decline as cash transfers are reduced, and are almost negligible in the long term. The poverty gap progresses similarly.

Figure 6

Download asset Open asset

These results can be explained by increased income among households with children, which directly increases their consumption. Given that poor households have, on average, more children, this type of policy benefits them more than the other two policies. The other intervention types have a smaller and more indirect impact on household consumption. Under the spending scenario, the greater demand for teachers pushes wages up, and has positive general equilibrium repercussions on the rest of the economy. Compared to the spending scenario, the school fees subsidy frees up a share of the household consumption budget for the purchase of other products, which explains why the results concerning poverty are better.

Despite high growth rates in recent years, Burkina Faso remains a poor country. The government recognizes the need for a stronger commitment to reduce poverty and improve human development. At the same time, the Burkinabe budget deficit has risen in recent years, in response to the various crises that have hit the country. There are strong pressures to rapidly reduce this budget deficit, but there are grave concerns about how this will be achieved.

In this paper, through a combined CGE-micro simulation model, we analysed three temporary public spending policies: an increase in education spending, a school fees subsidy and a cash transfer to households with children under the age of five. The amount injected into the economy is the same in each of these three cases, making it easy to compare the three scenarios. In all scenarios, the injections are reduced gradually in order to attain the same deficit-to-GDP ratios as a reference – business-as-usual – scenario, but over a slightly longer period: ten years, instead of five. Compared to the reference scenario, the results indicate that higher public education spending increases school participation and promotion rates, leading to an increased supply of skilled workers and higher economic growth, while also reducing the incidence and depth of both monetary and caloric poverty. The school fees subsidies have more differentiated impacts on education, with a larger beneficial impact on primary school entry but a partially offsetting decline in education quality and resulting promotion rates. As a result, the supply of skilled workers and economic growth are practically unaffected. By freeing up part of the household budget for other ends, this type of intervention has a slightly stronger effect on poverty than public spending. Cash transfers have a minimal impact on educational behaviours, and thus on the supply of skilled workers and economic growth, but they directly increase household income thus significantly reducing the incidence and severity of poverty.

In summary, if the objective is to achieve improved educational and economic performance, the best approach would be to increase public education spending. However, cash transfers to families would be more suitable if reducing child poverty is instead prioritized.

1. 1
   ‘The Theory of Reform and Indian Indirect Taxes’

   1. E Ahmad
   2. N Stern

   (1984)

   Journal of Public Economics 25:259–298.

   • Google Scholar
2. 2
   ‘Poverty issues for zero rating VAT in South Africa’

   1. H Alderman
   2. C Del Ninno

   (1999)

   Journal of African Economies 8:182–208.

   • Google Scholar
3. 3
   ‘Public education spending and poverty in Burkina Faso: A calculable general equilibrium approach (Dépenses publiques d’éducation et pauvreté au Burkina Faso: une approche en Équilibre Général Calculable)’. PEP Working Paper 2011-04

   1. L Balma
   2. MMR Kabore
   3. A Ouattara
   4. ST Kabore
   5. WFA Ilboudo
   6. Z Kassoum

   (2011)

   ‘Public education spending and poverty in Burkina Faso: A calculable general equilibrium approach (Dépenses publiques d’éducation et pauvreté au Burkina Faso: une approche en Équilibre Général Calculable)’. PEP Working Paper 2011-04.

   • Google Scholar
4. 4
   ‘Analyse de la situation de la pauvreté et de la vulnérabilité de l’enfant et de la femme au Burkina Faso (Analysis of the situation of poverty and vulnerability of children and women)’. Report commissioned by UNICEF-Burkina Faso

   1. Y Batana
   2. J Cockburn
   3. HJ-L Guene

   (2012)

   ‘Analyse de la situation de la pauvreté et de la vulnérabilité de l’enfant et de la femme au Burkina Faso (Analysis of the situation of poverty and vulnerability of children and women)’. Report commissioned by UNICEF-Burkina Faso.

   • Google Scholar
5. 5
   ‘Réformes Fiscales et Réduction de la Pauvreté: Application sur des Données Tunisiennes (Fiscal reforms and poverty reduction: An application to Tunisian data)’. Working paper, Université de Tunis El Manar Faculté des Sciences Economique et de Gestion de Tunis

   1. S Bibi
   2. J-Y Duclos

   (2004)

   ‘Réformes Fiscales et Réduction de la Pauvreté: Application sur des Données Tunisiennes (Fiscal reforms and poverty reduction: An application to Tunisian data)’. Working paper, Université de Tunis El Manar Faculté des Sciences Economique et de Gestion de Tunis, http://www.aed.auf.org/IMG/pdf/S.Bibi.pdf, accessed 29 January 2016.

   • Google Scholar
6. 6
   Un modèle CGE-Multi-Ménages Intégrés Appliqué à l’économie sénégalaise (An integrated multi-household CGE model applied to the Senegalese economy)

   1. D Boccanfuso
   2. F Cissé
   3. A Diagne
   4. L Savard

   (2003)

   Mimeo: CREA, Dakar.

   • Google Scholar
7. 7
   Representative versus real households in the macroeconomic modeling of inequality. Working Paper DT/2003/10

   1. F Bourguignon
   2. AS Robilliard
   3. S Robinson

   (2003)

   DIAL, Unité de recherche CIRPEE.

   • Google Scholar
8. 8
   Uganda’s Recovery: The Role of Farms, Firms, and Government

   1. D Chen
   2. J Matovu
   3. R Reinikka

   (2001)

   ‘A Quest for Revenue and Tax Incidence’, (Eds.), Uganda’s Recovery: The Role of Farms, Firms, and Government, Washington DC, World Bank.

   • Google Scholar
9. 9
   Globalization and Poverty: Channels and Policies

   1. J Cockburn

   (2006)

   ‘Trade Liberalisation and Poverty in Nepal: A Computable General Equilibrium Micro Simulation Analysis’, (Eds.), Globalization and Poverty: Channels and Policies, London, Routledge.

   • Google Scholar
10. 10
    ‘Fiscal Space and Public Spending on Children in Burkina Faso’. Working PAPER 13-08

    1. J Cockburn
    2. H Maisonnave
    3. V Robichaud
    4. L Tiberti

    (2013)

    CIRPEE, University Laval.

    • Google Scholar
11. 11
    IFPRI TMD Discussion paper No. 61 and DIAL Working Paper DT/2001-19, IFPRI TMD and DIAL

    1. D Cogneau
    2. A-S Robilliard

    (2001)

    IFPRI TMD Discussion paper No. 61 and DIAL Working Paper DT/2001-19, IFPRI TMD and DIAL.

    • Google Scholar
12. 12
    Poverty Alleviation Policy in Madagascar: A Micro-Macro Simulation Model. Working Paper, DT/2004/11

    1. D Cogneau
    2. A-S Robilliard

    (2004)

    DIAL, Unité de Recherche CIRPEE.

    • Google Scholar
13. 13
    ‘An Almost Ideal Demand System’

    1. A Deaton
    2. J Muellbauer

    (1980)

    American Economic Review 70:12–36.

    • Google Scholar
14. 14
    Measuring poverty and inequality in a Computable general equilibrium framework

    1. B Decaluwé
    2. JC Dumont
    3. L Savard

    (1999)

    Measuring poverty and inequality in a Computable general equilibrium framework, Cahier de recherche no 99-20, CREFA, Université Laval.

    • Google Scholar
15. 15
    ‘The PEP standard computable general equilibrium model single-country, recursive dynamic version PEP-1-t’

    1. B Decaluwé
    2. A Lemelin
    3. V Robichaud
    4. H Maisonnave

    (2010)

    Accessed January 29, 2016.
16. 16
    The willingness to pay for medical care: Evidence from two developing countries

    1. P Gertler
    2. J Van Der Gaag

    (1990)

    Baltimore: The John Hopkins University Press.

    • Google Scholar
17. 17
    ‘Investigating the Determinants of Household welfare in Côte d’Ivoire’

    1. P Glewwe

    (1991)

    Journal of Development Economics 35:307–337.

    • Google Scholar
18. 18
    Analyzing Fiscal Space Using MAMS: An Application to Burkina Faso. IMF working paper WP/09/227

    1. J Gottschalk
    2. V Manh Le
    3. H Lofgren
    4. K Nouve

    (2009)

    International Monetary Fund.

    • Google Scholar
19. 19
    ‘Burkina Faso: Deuxième revue de l’accord triennal au titre de la facilité élargie de crédit et demande de modification de critères de réalisation – Rapport des services du FMI (Second review of the triennial agreement on the expanded credit facility and the request to modify the performance criteria): press release relating to the proceedings of the board; and, statement of the IMF executive director for Burkina Faso’. IMF Country Report 11/226

    1. International Monetary Fund

    (2011)

    Accessed January 29, 2016.
20. 20
    MAMS: An Economy-wide Model for Analysis of MDG Country Strategies

    1. H Lofgren
    2. C Diaz-Bonilla

    (2006)

    MAMS: An Economy-wide Model for Analysis of MDG Country Strategies, Technical Documentation, DECPG, World Bank.

    • Google Scholar
21. 21
    ‘Consumption Dominance Curves: Testing for the Impact of Indirect Tax Reforms on Poverty’

    1. P Makdissi
    2. Q Wodon

    (2002)

    Economics Letters 75:227–235.

    • Google Scholar
22. 22
    ‘Dalton-improving tax reform: When households differ in ability and needs’

    1. J Mayshar
    2. S Yitzhaki

    (1996)

    Journal of Public Economics 62:399–412.

    • Google Scholar
23. 23
    ‘Progrès et acquis du développement du Burkina Faso 2000–2009 (Development progress and gains in Burkina Faso 2000–2009)

    1. Ministry of Economy and Finance

    (2010)

    ‘Progrès et acquis du développement du Burkina Faso 2000–2009 (Development progress and gains in Burkina Faso 2000–2009).

    • Google Scholar
24. 24
    Statistiques de l’éducation de base 2009-2010 (Primary education statistics, 2009–2010)

    1. Ministry of Primary Education and Literacy

    (2010)

    Directorate of studies and planning.

    • Google Scholar
25. 25
    Annuaire statistique de l’enseignement secondaire 2008–2009 (Annual report on secondary education statistics)

    1. Ministry of Secondary and Higher Education and Scientific Research

    (2009)

    Directorate of studies and planning.

    • Google Scholar
26. 26
    ‘Indirect tax in Madagascar: Estimations using the input-output table’. CFNPP Working Paper #106

    1. H Rajemison
    2. SD Younger

    (2000)

    ‘Indirect tax in Madagascar: Estimations using the input-output table’. CFNPP Working Paper #106.

    • Google Scholar
27. 27
    Development economics

    1. D Ray

    (1997)

    Princeton University Press.

    • Google Scholar
28. 28
    Evaluating the poverty and distributional impact of economic policies

    1. DE Sahn
    2. SD Younger

    (2003)

    ‘Estimating the Incidence of Indirect Taxes in Developing Countries’, Evaluating the poverty and distributional impact of economic policies, New York, Oxford University Press.

    • Google Scholar
29. 29
    ‘Impact de l’expansion économique et commerciale de la Chine sur la croissance et l’emploi au Burkina Faso: une analyse en équilibre général calculable’. PEP Working Paper 2015-03, PEP

    1. B Sawadogo
    2. F Bama
    3. TJ Nana
    4. HM Natama
    5. E Tapsoba
    6. K Zerbo

    (2015)

    ‘Impact de l’expansion économique et commerciale de la Chine sur la croissance et l’emploi au Burkina Faso: une analyse en équilibre général calculable’. PEP Working Paper 2015-03, PEP.

    • Google Scholar
30. 30
    ‘Global diagnostic of SCADD (Rapport sur le diagnostic global de la SCADD 2010)’

    1. SCADD

    (2010)

    ‘Global diagnostic of SCADD (Rapport sur le diagnostic global de la SCADD 2010)’.

    • Google Scholar
31. 31
    ‘World population prospects’

    1. United Nations Department of Economics and Social Affairs

    (2010)

    Accessed January 29, 2016.
32. 32
    ‘Linking Economic Policy to Childhood Poverty: A review of the evidence on growth, trade reform and macroeconomic policy’. CHIP Report No.7, CHIP

    1. L Waddington

    (2004)

    ‘Linking Economic Policy to Childhood Poverty: A review of the evidence on growth, trade reform and macroeconomic policy’. CHIP Report No.7, CHIP.

    • Google Scholar
33. 33
    La Revue des Dépenses Publique dans le secteur de l’éducation (Public expenditure review in the education sector)

    1. World Bank

    (2009)

    Burkina Faso: World Bank.

    • Google Scholar
34. 34
    ‘Welfare dominance: An application to commodity taxation’

    1. S Yitzhaki
    2. J Slemrod

    (1991)

    American Economic Review 81:480–496.

    • Google Scholar
35. 35
    ‘Welfare Dominance and the Design of Excise Taxation in the Côte d’Ivoire’

    1. S Yitzhaki
    2. W Thirsk

    (1990)

    Journal of Development Economics 33:1–18.

    • Google Scholar

Author details

1. John Cockburn

   Partnership for Economic Policy (PEP), and Université Laval Pavillon J.-A.-DeSève, Canada

   For correspondence

   jcoc@ecn.ulaval.ca

2. Hélène Maisonnave

   EDEHN, Department of Economics, Université du Havre, France

   For correspondence

   hmaisonnave@hotmail.fr

3. Véronique Robichaud

   Independent researcher and Resource person – PEP network, Université Laval Pavillon J.-A.-DeSève, Canada

   For correspondence

   vrob@videotron.ca

4. Luca Tiberti

   Partnership for Economic Policy (PEP), and Université Laval Pavillon J.-A.-DeSève, Canada

   For correspondence

   luca.tiberti@ecn.ulaval.ca