Social Spending and Economic Growth in the Sudan. An Econometric Perspective

ABSTRACT

This paper takes into account the recent role of social spending on development. Social spending outcome in Sudan is influenced by limited targeting actions with very low interventions between results in economic growth and accesses to basic social services.

An econometric models results show that more social spending in the short run increase output which enhances GDP per capita growth by 0.5% with 3.1% towards convergence equilibrium in the long run. Sudan should universal and expanded cover to social protection services which aim at building a social protection as a productive factor may have contributed to enhancing income security, education and health outcomes, reducing poverty, inequality, and enhancing economic growth lead to sustainable development.

Key words: Economic Growth, Social Spending, Econometric model, Sudan JEL code: B22, B23, C50. C51, C52, D63, E61, E62, E65, F43

1. INTRODUCTION

There is no generally accepted definition of the social protection. It is sometimes used interchangeably with social security, social safety nets, and social assistance¹. Sudan classified social protection under social development and included central government contributions to the pension fund and to the social security fund, social subsidies that directly benefit the poor, which are mainly directed to subsidizing electricity, free medication in emergencies only, free medicines for kidney dialysis and heart disease. In addition, direct support (with limited) to poor students in higher education and primary and secondary education, medical staff for all health units.

The challenge of universalizing socioeconomic security for the Sudanese poor (majority 71%) and improving social protection has become a subject in the Sudan during the last three decades (1990s). Many people in the Sudan are overwhelmed by uncertainty regarding future education, health care, employment, and social security coverage households. An empirical evidences show that the crucial role of social protection for pro-poor growth, education, health and employment creation the core element in development context.

The role of social protection and economic growth has attracted recently much attention; a crucial issue in this regards is the role of social spending in helping counties foster human capital. Education, health and employment are the main dimensions of social protection and core element in development contexts; the role of public sector to build human capital by providing and universalising the education and health services.

The main aim of this paper is to examine the impact of social protection on the Sudanese economy growth by clarify the empirical evidence about the trad-off between social protection and economic growth (in the short and long runs) and its effects on human capital components’ education, and health capitals; based on assumption that social protection mechanism is not only protective factor it also productive factor, enhancing economic growth and socio-political stability for the Sudan.

This paper organised in seven sections, following the introduction the second section discusses briefly the theories and empirical evidences about the effect of social protection as proxies by social expenditures on economic growth. The third section addresses the Sudan social protection trends and perspectives over the period (1970-2007). The fourth section describes the methodology and data used in the estimation and presents the empirical results. Finally offers some concluding remarks.

2. THEORETICAL AND LITERATURE REVIEW

There are several theories that refer to the trade-off between social protection and economic growth and do directly relate social spending with growth. The link between social protection understood as expenditure on basic social services and growth has attracted much attention recently.

There are several reasons to believe that social protection and growth may be related. One of the important arguments in this context is social assets argument “high transfer cause high growth” through institutional assurance individuals, and hence, social protection may lead to cohesion society better able to take more risks in their economic decisions because they are insured against failure through social protection system and this may foster growth².

A number of additional considerations suggest that social protection can be good for economic growth; Korpi³ who have tended to highlight that greater social protection expenditure not only generates more equal and cohesive societies, but also greater economic growth. Korpi mentioned that “in a glaring contrast to the predictions of the market liberal hypothesis, the Golden Age of economic growth coincided with the extension of the welfare state, with decreasing income inequality, and with increasing political and organizational intervention into market processes”.

Social protection for developing countries is an important dimension in the reduction of poverty and multidimensional deprivation⁴ ; it aims to enhance the capacity of poor and vulnerable people to manage socioeconomic risks, such as unemployment, exclusion, sickness, disability and old age. Policy interventions can improve their well-being by, among other things, moderating the impact of shocks causing sharp reductions in their income or consumption. Social protection can also enhance the productive capabilities of poor, reducing poverty and inequality and supports economic growth⁵. The numbers of economists have also become increasingly influenced by this argument.

Krzyszto, et al.⁶ for example pointed out the importance of social protection for low income countries: through it can achieve sustainable development; moreover by provision of basic social security is an investment in country’s development giving not only reduced poverty but also increased demand and expanded domestic markets, healthier, better educated, empowerment and more productive workforce as well as peace, stability and social cohesion, less conflict and politically more stable societies and hence increasing economic growth.

On the other hand, the study on promoting pro-poor through social protection recommended that the best way towards achieving pro-poor growth is social protection, in which poor participate directly, as both agents and beneficiaries, is essential directly reduces poverty through improved health outcomes, increased school attendance, hunger reduction and livelihoods promotion. Social protection can provide essential support and recurring crises expose the vulnerability of poor individuals and families as well as their jobs and livelihoods. Moreover, ongoing challenges of population growth, price volatility, food insecurity, highlight the need for more effective social protection⁷.

An alternative set of arguments revolved around the idea of the relation between social protection and growth for example Arjona et al.⁸ for example point out that if benefit system (social protection) discourage people from working, therefore, the amount of labour supplied in the economy is lowered, so reducing the level of output and the level of capital investment and hence economic growth. On the other hand if social provisions discourage people from savings then, there is a reduction in the capital available for reinvestment unless public savings rises by the equivalent amount; and they suggested that a bit more passive spending bad for growth.

Fan and Rao⁹ for example analysed the public spending in developing countries, their main finding results indicated that the impact of various types of government spending on economic growth is mixed; they found that In Africa, government spending on agriculture and health was particularly strong in promoting economic growth. Asia’s investments in agriculture, education, and defence had positive growth-promoting effects. However, all types of government spending except health were insignificant impact on growth in Latin America.

Moreover, using panel data from 118 developing countries in 1971-2000, Emanuele and et al¹⁰ explored the channels linking social spending, human capital, and growth and compares the effects of alternative economic policy interventions. With separated modelling for education and health capital, explicit control for governance, and incorporation of nonlinearity, they found that both education and health spending have a positive and significant impact on education and health capital, and thus support higher growth. Also, other policy interventions, such as improving governance and taming inflation, achieved similar results.

Herce et al.¹¹ used data for European Union (1970-1994) and panel data techniques and following production function approach they found a positive growth effect of social protection expenditure on growth. When they analysed the effects of the different categories of social protection benefits, they found a significant and positive effect for the health, old age and family programmes. In contrast, such significant effect was not found for the employment and housing programmes. Moreover, in other study by Herce et al.¹² they find that a positive correlation between welfare state and economic performance, their results points towards statistically significance Granger causality running from social protection expenditure towards growth.

Moreover, McCallun and Blais¹³ find that social expenditure plays a positive role towards economic growth below a certain level and a negative one beyond it - as long as the welfare is not too large -; one possible interpretation of these result runs as follows: along welfare state may related economic growth by reducing the incentive to work, to save, to move, and to change. On the other hand, in a situation where special interest groups have a required significant power to block change if they so desire, the welfare state which offers assistance to those who are the victims of change may play a growth-enhancing role in reducing the incentive to block change.

An alternative evidence; for example Gwartney et al. (1998) indicate that social protection expenditure is bad for growth and social protection expenditure may trigger a trade-off between equity and efficiency and contribute to an overall loss of economic, innovative, and entrepreneurial capacity¹⁴.

In summary most of studies find that social protection can have a positive impact on growth in developing countries in a number of ways¹⁵. It can reduce poverty through financing investment in health and education, protecting assets that help people earn an income, encouraging risk taking, promoting participation in the labour market, and ease the pain of economic transaction.

Moreover social protection can lead to greater social integration (inclusion), political stability, human right objectives, and stable environment for individual to work, save and invest. In the other hand government must be careful to strike an appropriate balance between economic incentives and greater provision of social protection (if taxes are raised to pay for spending on social protection, tax payers may have less incentive to work and save or if government with limited revenues is not able to distribute between direct productive sectors and social sectors). Indeed government must altering the balance between apply passive (pure cash transfer of consumption) and active polices in order to encourage increased employment by the beneficences of such spending.

3. THE MODEL AND THE METHOD

The theories attempt to test empirically links between social protection and growth, in practice estimation has nearly used a simple model of the causes of economic growth and augmenting it with measures of social protection, and have used empirical model proposed by Solow and Swan (1956) with two factors: labour and capital others add human capital as a third variable of production as proposed by Romer and Weil (1992) pointed by Benank and Reft¹⁶. Bassanini and Scarpetta¹⁷ determine the growth in GDP per capita modelled as a function of: investment in physical capital (more investment means more capital assets per capita, so more growth); growth rate of the population (more population growth means slower growth in income per capita, given the level of physical capital); the level of human capital (more human capital means greater efficiency in using physical capital; here we have been divided into: education capital and health capital), and income.

Based on the above discussion, the model to investigate the interaction of social protection on economic growth is assumed taking the following forms:

illustration not visible in this excerpt

where, (Y) denotes GDP per capita economic growth in percentage; (Yc) denotes the lagged real GDP per capita (PPP$) its coefficient is expected to be negative, because it expected that the population increase at a faster rate than total income and the capital did not grow as fast ; (Se) denotes social spending proxies by the government expenditure on social services as a percentage of total expenditure, its coefficient is expected to be positive, social protection enhance economic growth through different channels; (In) denotes the investment ratio, measured in terms of gross fixed capital formation to GDP, to captures an increase in the physical capital its coefficient is expected to be positive; (Pg) denotes the annual average rate of growth of the population in percentage its coefficient is expected to be negative; (Ec) refers to the Education Capital (human capital), proxies by primary education enrolment rate, human capital promote growth its coefficient is expected to be positive; (Hec) denotes health capital and the logarithm of under five child mortality rate is used to proxy the stock of health capital as proposed by Gyimah, Wilson and Emanuele et al to facilitate interpretation, the sings of the coefficients on mortality rates are reversed so that the positive coefficients correspond to improvement in health status; (Po) denotes working age population 15-64 years of total population age structure can affect labour force and enhance growth its coefficient is expected to be positive; and ( ) refers to time dummy is used to know time shock that affect the social protection during the study period, there is incident in one year (turning point) 1992 where Sudan reform the economy by adopted liberalization and free market its coefficient is expected to be positive for the second period.

It is expected that the impact of the GDP per capita (YC), and social expenditures (SE), will be distributed over one year, which here used lagged variables. The coefficients of the model can be estimated by the Generalized Method of Moments (GMM).

The specification of above system is consistent with previous studies and it can help us for the identification of the channels through which social expenditures and other variables affect growth in Sudan. For more elaboration for the relation between social protection, human capital and growth we consider to use Granger causality as proposed by Engle and Granger (1969)¹⁸, and check the stationarity and if there is presence of unit root in the series, the most famous of the unit root tests are the ones derived by Dickey and Fuller and described in Fuller (1976)¹⁹, also Augmented Dickey-Fuller (ADF) has been mostly used within a Vector autoregression (VAR) model which is an econometric model used to capture the evolution and the interdependencies between variables, generalizing the univariate AR models. Sims advocates the use of VAR models as a theory-free method to estimate economic relationships, thus being an alternative to the "incredible identification restrictions" in structural models²⁰.

For examining the cointegartion apply (ECM) (Engle and Granger, 1987)²¹ we can rewrite the long-term relationship between Y, SE and HC as follow:

illustration not visible in this excerpt

Here Δ is the first difference operator.

4. DATA

All variables over the period cover (1970-2011) are from World Economic Development Database, World Africa Database, and UN statistics; published by IMF, WB, and UN. The data of government social expenditure as a percentages of current total expenditures from annul reports; Central Bank of Sudan, Ministry of Finance and National Economy (MoFNE), and Central Bureau of Statistics (CBS) Sudan; for the period 1970-1990; social services namely spending on education and health services only, while over the period 1991-2011 MoFNE classified social spending under Social Development and included central government contributions to the pension fund and to the social security fund. In addition, it includes social subsidies that directly benefit the poor, which are mainly directed to subsidizing electricity, free medication in emergencies, free medicines for kidney dialysis and heart disease, support to poor students in higher education and primary and secondary education teachers, medical staff for all health units, except specialized hospitals, and water supply employees.

5. EMPIRICAL ESTIMATES

5.1 SOCIAL SPENDING AND GROWTH

Table 1 and 2 present the regression results of different equations estimated to explain the effects of social spending on the growth during the period 1970- 2011. In most cases the coefficients are statistically significant; all equations have tested of over-identifications using J-statists test indicated all models have a good fit.

Table 1, Column[1]presents the estimated coefficients when the equation augmented by social spending, column[2]shows the results using the same measures, exclude social spending, column[3]exclude dummy variables for economic reforms, columns[4]and[5]are exclude health capital on the ground that its insignificant and may affect the growth equation, moreover, to see whether the affect of education capital is more or less than the affect of health capital on egrowth. The augmented model presents in table 2; here we introduced the working age population instead of population growth which it appears not statistically significant for all equations.

The results show that the levels of education capital and social spending have positive effects on the Sudan’s economic growth. The impact of health capital on growth differ from that of education capital, health capital indicator negatively and insignificant affect to growth, this seems consistent with a high rate of under five mortality in Sudan during the period under consideration. (Findings are same as in Emanuele).

TABLE 1 THE EFFECTS OF SOCIAL EXPENDITURE ON GROWTH

DEPENDENT VARIABLE: GROWTH RATE OF REAL GDP PER CAPITA IN 1990 PPP

illustration not visible in this excerpt

Notes: ** t-values significant at 1% and 5% level of significance

The results suggested that the Sudan’s economic reforms adopted in 1992 have a positive effect on the growth; an economic reforms raises the growth rate by 6% in the health capital effects and about 14% for education capital effects.

The results show that the social spending in Sudan has positive effect on the economic growth, for all equations the coefficients of social spending are significance with positive sign, however, the contribution of it is very limit with small impact; an increase of social spending by 1 per cent GDP growth could increase by 0. 3 per cent, to 0.5 per cent when working age population introduced into the growth equation.

Table 2 reports that the health capital is very weak with negative impacts on growth for the Sudan; results show that the under five mortality rate reduces growth, an increase in under five mortality rate by 1 percent is found to reduce growth by about 61 per cent, while education capital bolsters economic performance; an increase in the primary education enrolment by 1 percentage is found to increase the economic growth in Sudan by 0.8%; this result indicate that the Sudan education capital is still very weak in terms of contributions to the economic growth. The results indicate that the working age population and investment affects growth although education and health capital does not. An increase in working age population and investment by 1% is associated with an increase in the growth of 13% and 0.6% respectively.

TABLE 2

THE EFFECTS OF SOCIAL EXPENDITURE ON GROWTH

DEPENDENT VARIABLE: GROWTH RATE OF REAL GDP PER CAPITA IN 1990 PPP

illustration not visible in this excerpt

Notes: ** t-values significant at 1% and 5% level of significance

5.2 GROWTH, HUMAN CAPITAL AND SOCIAL SPENDING COINTEGRATION

In the first stage, the order of integration was tested using the ADF unit root test. Table 3 reports the results of the unit root tests. The ADF statistics for the GDP per capita growth, social spending and human capital do not exceed the critical values (in absolute terms). However, when we take the first difference of each of the variables, the ADF statistics are higher than their respective critical values (in absolute terms).

Therefore, we conclude that GDP per capita growth, social spending and human capital are each integrated of order one or I(1). The next step is to test whether the stationary variables are co integrated or not.

TABLE 3

ADF UNIT ROOT TEST OF STATIONARILY

illustration not visible in this excerpt

All the variables are stationary at their first differences and 5% level of significance

Using Johansen co-integration to test whether the stationary variables are cointeragted in the short run; the Eigen value at 5% show that there is one cointegratiog for GDP per capita growth, social spending and Human capital in the short run. Result of cointegrating equation show that there is positive relationship social spending and human capital and GDP per capita growth this relationship in the form:

illustration not visible in this excerpt

This show that if there is 141 per cent and 898 per cent change in GDP per capita growth due to 1 per cent change in social spending and human capital respectively in the short run. These results are significant at 5% level of significance.

TABLE 4

JOHANSEN COINTEGRATION TEST

illustration not visible in this excerpt

*(**) denotes rejection of the hypothesis at 5%(1%) significance level L.R. test indicates 1 cointegrating equation(s) at 5% significance level

Table 5 shows that the VEC model estimates and the results indicate that the error correction terms (ECM) in the long run of GDP per capita growth, social spending and human capital statistical significant. For the GDP per capita growth the ECM indicates 0.34 percent speed of convergence towards equilibrium position in the case of any disequilibrium situation. The ECM shows that for social spending the convergence speed of 3.1 per cent towards equilibrium and for human capital convergence towards equilibrium point at the speed of 0.19 percent.

TABLE 5:

THE VEC MODEL BASIC RESULTS

illustration not visible in this excerpt

Notes: ** t-values significant at 1% and 5% level of significance

Table 6 gives results on Granger causality tests. In carrying out the test of causality between GDP per capita growth, social spending and human capital the results indicate directional causality between the GDP per capita growth and social spending. This causality runs from GDP per capita growth to social spending and from social spending to human capital. We also see no causality from social spending to GDP per capita growth and from human capital to GDP per capita growth.

TABLE 6:

GRANGER CAUSALITY TEST

illustration not visible in this excerpt

Notes: ** F-values significant at 5% and 10% level of significance

6. CONCLUSION:

In this paper we have investigated the effects of social protection benefits on economic growth for the Sudan; covering the period (1970-2007). Social protection outcome in Sudan is influenced by adopted limited targeting actions and interventions between results in economic growth and accesses to basic social services and it’s contributed to the process of socioeconomic development.

However, the results show that more social spending increase output, there is positive and significant effect of social spending on economic growth however, this effect is very limited due to the different factors affecting that: the social spending received the lowest percentage ratio in relation to other items on average 2.8% for health and 1.2% for education with low levels of education and health capital, social development ranked very low as government priorities. However, the results show that the affect of social spending on GDP per capita equal on average 0.5% and an increase in primary education enrolment by 1% is associated with an increase in the growth of 0.8%, in contrast the health capital have negative and insignificant impacts.

In the long run causality runs from GDP per capita growth to social spending and from social spending to human capital. We also see no causality from social spending to GDP per capita growth and from human capital to GDP per capita growth in the Sudan.

Sudan must improve the wellbeing of the people through allocating more budgetary expenditures to health and education and more resources to pro-poor sectors in the efforts to reduce poverty and enhance economic growth.

The main challenge for the Sudanese policy makers is to rethinking into social protection as not only protective factor but also as productive factors enhance economic growth. In Sudan the policy-makers define the problem too narrowly (social spending viruses economic and political problems), and it is necessary to give due consideration to the wider context of social protection as a socio- economic-political stabilizer. Sudan Social protection policies should aim at protect human capital include better access to hospitals, universal health insurance, improved access to schools, universal primary education, employment creation with equity, promotion of rural development for reduce socio-economic inequality, improved infrastructures, reduction of exclusion by eliminate biases against vulnerable groups (disabilities, children, poor), reforms financial sectors for access to capital, implementation of employment support projects, and first of all equal distribution of the services among the states and increase the share of social spending (gains from high rate of growth) with stabilizing macroeconomic policies, this is may be the best way for Sudan to achieve sustainable development.

[...]

---

¹ Mishra, S. (2008), ‘Social Protection in Asia’ Strategic Asia preparing for the Asian century, S.A

² Ahmed et al (1991), ‘ Social Security in Developing Countries’ Oxford University Press; Oxford

³ Korpi, W. (1985), ‘Economic Growth and Welfare System Leaky Pocket or irrigation system? European Sociological Review Oxford University Press

⁴ Shephred, A. et al.(2004), ‘ Policy paper on social protection’, ODI

⁵ UNDP (2006), ‘Social protection the role of cash transfers’

⁶ ILO (2008), ‘Can Low-income Countries afford Basic Social Security?’ Social Policy Briefings, Paper No 3

⁷ OECD, (2009), ‘ promoting pro-poor growth: Social Protection’

⁸ Arjona, A. et al. (2002), ‘Social Protection and Growth’ , Economic Studies No 35, OECD

⁹ Fan, S. and R., (2003) ‘Public Spending in Developing Countries: Trends, Determination, and impact’, EPTD

¹⁰ Baldacci, E. And et al., (2008), ‘Social Spending, Human Capital, and Growth in Developing Countries’ Elsevier, WB paper 36

¹¹ Herce, Jose´ A. Et al (1998), ‘ Social Protection benefits and Growth: evidence from EU, FEDA

¹² Arjona, A. et al. (2001), ‘ Growth and the Welfare State in the Eu: A causality analysis’, Kluwer Academic

¹³ McCallum, J and Blais (1987),’ Government, Special Interest Groups and Economic Groups’ Martinus, Netherlands http://www.jstor.org/stable/30024765?seq=15

¹⁴ Ezucrra, R. and A. (2009), ‘Decentralization of Social Protection Expenditure and Economic Growth in the OECD’, The Journal of Federalism, pp 1-12

¹⁵ DFID (2006), ‘ Social protection and economic growth in poor countries’, practice paper number 4, Oxford

¹⁶ Bernank, B. And Reft (2001), ‘Is Growth Exogenous? Taking Mankiw, Romer and Weil Seriously’, NBER, paper number 8365

¹⁷ Bassanini, and Scarpetta, (2008), ‘Long-Run Growth Forecasting’, Springer Berlin Heidelberg

¹⁸ Engle, R.F. and C.W.J. Granger (1987), ‘Cointegration and Error-Correction: Representation, Estimation, and Testing’, Econometrica 55. websit: http://www.jstor.org

¹⁹ Dickey, A. and W.A. Fuller (1979), ‘Distribution of the Estimators for Autoregressive Time Series with a Unit Root’, American Statistical Association Journal, 74, Website: http://www.jstor.org/stable/2286348?seq=2

²⁰ Sim, C.A. (1980).’ Macroeconomics and Reality’, Econometrica 48: website: http://www.jstor.org/

²¹ Engle, R.F. and C.W.J. Granger (1987), ‘Cointegration and Error-Correction: Representation, Estimation, and Testing’, Econometrica 55. websit: http://www.jstor.org