������������������������������������Asymmetric response of poverty to growth and inequality in South Africa: implications for current and future shocks

Nicholas Ngepah, PhD

Professor of Economics, University of Johannesburg

nngepah@uj.ac.za; nnnbal@yahoo.fr.

AERC-Funded research

Africa Evidence Summit

19-20 June 2023

Nairobi, Kenya

Introduction

• Most development stakeholders have recognized poverty reduction as the starting point and the most fundamental aspect in kick-starting development.
• (UNGA, 2000), adopted the MDGS
• At the top was the eradication of extreme poverty and hunger, with three associated targets:
• Halve, between 1990 and 2015, the proportion of people living on less than $1.25 a day;
• Achieve Decent Employment for Women, Men, and Young People;
• and Halve, between 1990 and 2015, the proportion of people who suffer from hunger.

Introduction

• Post-apartheid South Africa inherited a very high level of poverty, that was predominantly among the black race.
• Bhorat and Westhuizen (2012) estimated a headcount poverty index to be 52.54 in 1995.
• Successive waves of policy measures
• RDP - to decrease poverty and redress past racial economic injustices, lacked a focus on economic growth strategies.
• GEAR in 1996 - to address the problem of unemployment though economic growth, despite positive economic growth, employment did not grow commensurately, hence it did not result in significant poverty reduction (Adelzadeh et al, 1998; Changunda, 2006)
• ASGISA for 6% growth rate and poverty reduction through supporting small businesses and tackling inefficiencies in the service delivery mechanism, addressing skills shortages, corruption and lack of inter-ministerial policy coordination. This also fell short of its stated objectives.
• NDP 2012 - to generate labour absorptive economic growth rate of at least 7%, inclusive enough to reduce poverty significantly

Introduction

• The role of economic growth in poverty reduction has generated intense debate in the development economics
• Dollar and Kraay (2002) that the average income of the poorest fifth rises proportionately with societal average hence growth generally benefits the poor as much as others; Chinese economic growth in reducing poverty since early 1990
• Those who subscribe to this view assert the necessity and sufficiency of economic growth in poverty reduction.
• At the same time others argue that economic growth tends to raise both income and asset inequality, making it less beneficial to the poor than the rich.
• Emerging consensus seems to suggest that focusing on growth may be necessary but not sufficient in alleviating poverty for certain countries or regions in the developing world (Besley and Cord, 2007).

Introduction

• The post-1994 economic growth in South Africa has been described as relatively robust (Bhorat et al., 2020)
• from 1990-95 average of 0.9 to 2.8 in 1995-2000, 3.8 in 2000-2005, 3.1 in 2006-2010, to 2.2 percent in 2011-2015
• From 2016, growth rate rose from 0.7 to 1.5 in 2018, falling to -0.2 in 2019
• With the advent of the COVID-19 crises, the GDP fell by 6.5 in 2020.

Source: Author’s using Statistics South Africa (2017) WDI (2021)

Introduction

• The growth-sufficiency paradigm appears not to tally in South Africa like in rest of Africa
• In terms of $1.90 poverty line, World Bank (2018) shows 50% global poverty reduction between 1990 and 2015 globally, but only 28% in SSA
• It becomes clear that not all economic growths are not equal in terms of their poverty reduction abilities

Introduction

• Inequality is a key determinant of both the ability of growth to reduce poverty and the level of growth itself (Espoir and Ngepah, 2021; Fosu, 2018; Ngepah, 2016; Ravallion, 2012)
• Lakner et al. (2020) show that 1% decrease in Gini index will bring about a greater measure of poverty reduction compared to a percentage point increase in economic growth.
• It is not only economic growth that matters for poverty, the type and source of the growth equally matters
• Most of the growth in South Africa has significant episodes of negative growth and the inability to rebound from global crises.
• For example, since the global financial crises, most African countries manage to navigate relatively safely, South Africa had remained under the weight of the crises in terms of its low to negative GDP growth for some time after the crises.

RESEARCH OBJECTIVES

• This study proposes to investigate how growth and inequality affect poverty during times of recessions and depressions versus times of economic expansion in South Africa and what factors can assist to cushion the poor from economic downturns, with the following guiding questions:
• What is the effect of economic growth on poverty reduction given inequality at the micro-level?
• Do the poor suffer more losses of welfare during economic recessions and depressions than they gain during expansions?
• What, if any, are the factors that can assist the poor to stay afloat during times of economic shocks?

Methodology

• The theoretical foundation relies on the ‘identity’ model first derived by Bourguignon (2003)
• The model is based on an approximation of a log normal distribution of income allowing for inter-country and intertemporal heterogeneity (Epaulard, 2003)
• Many authors have applied this model (e.g. g., Fosu, 2009, 2015, 2018; Kalwij and Verschoor, 2007)
• The identity model links poverty to mean income, inequality and the ratio of poverty line to mean income

Methodology: The functional forms

Absolute poverty analysis

Growth and quintile mean income framework

Variables and data

• our poverty data is computed from the National incomes Dynamics Survey database (NIDS)
• The NIDS project covers a panel of individuals and households followed over five waves spanning2008 to 2017
• The municipality-level mean income growth and inequality data will be sourced from quantec spanning 1993 to 2020
• All the individual and household level variables are sourced from the NIDS

Variables and data

RESULTS: Descriptive statistics

• The sample comprises 182784 individuals across the five waves with 54% females
• The average absolute positive growth over the period is 0.27 percentage points and the corresponding negative growth in GDP is 0.22 percentage points
• The headcount poverty is 41% according to the lower bound poverty line. By the same token, the poverty gap is 19%.

RESULTS: correlations

• The correlations are in line with theoretical expectations
• Growth in income pc is negatively correlated with all poverty indicators but positively with the measures of welfare
• Initial inequality is also negatively (positively) correlated with current poverty (welfare)
• Growth in inequality is positively correlated with levels of all poverty measures but also positively correlated with overall welfare
• This seems to suggest that inequality rises with levels of welfare.
• This is also supported by the positive association of growth in inequality with mean income growth.
• This association is not surprising in South Africa as inequality tends to reduce only the welfare of the poor, and leaves that of the rich either unchanged on enhances it
• This will be supported later in the regression modes
• Positive change in per capita mean income is associated with lower poverty and higher levels of welfare. On the contrary, negative economic growth rate is associated with higher poverty and lower levels of welfare

*** p<0.01, ** p<0.05, * p<0.1

RESULTS: growth and absolute poverty

• Judging from the LR Chi2 and the Chi2 p-values, the models fit significantly better
• The coefficients accord with the theoretical predictions and are significant
• initial inequality is +ve and significant
• One % point higher growth rate in inequality raises the probability of poverty by 0.87% and the relative distance to the PL by 0. 23%
• One % point increase in mean income growth reduces the probability of poverty by 0.30% and the relative gap from the poverty line by 0.08%

*** p<0.01, ** p<0.05, * p<0.1

RESULTS: growth and absolute poverty

• +ve growth reduces poverty more in magnitude than -ve growth increases poverty
• This result could be explained by:
• The fact that although South Africa’s unemployment is high, due to unionization and regulated labour market, it difficult to quickly shed jobs.
• The second possible explanation are the safety nets like the social security programs and free education and health, which we will analyse further
• inequality erodes poverty reduction gains far more than positive economic growth reduces poverty

​

*** p<0.01, ** p<0.05, * p<0.1

• A combination of increasing inequality and negative economic shocks is a dangerous mix for the poor in SA
• Pro-economic growth policies that also reduces inequality, or at least stops its increase will prove beneficial to the poor in SA

RESULTS : Growth and B40 and T20 incomes

• the pooled, FE, IV and SUR tend to concur
• A % increase in +ve growth rate raises incomes by about 0.82% for the T20, but only by about 0.36% for the B40.
• -ve growth rate is associated with 0.28% reduction in T20 incomes compared with 0.62% for B40
• Evidently, the poor bear a greater burden of -ve shocks than the rich, but benefit less from +ve growth than the rich
• =>inequality in SA would grow worse with either positive or negative economic shocks

RESULTS: growth and welfare distribution

• A % point rise in inequality growth reduces the incomes of those at the 20^th, 40^th, 60^th and 80^th xtiles by 0.25%, 0.21%, 0.17%, and 0.10% respectively.
• The effects on are 0.46%, 0.36%, 0.21% and 0.07% consumption welfare
• The incomes of the top 20% are not affected by inequality.
• A % point increase in inequality growth increases the consumption of the rich by 0.24%.
• Therefore, in SA, inequality begets even more inequality.

RESULTS: growth and welfare distribution

• A % point increase in growth raises the incomes of those at the 20^th, 40^th, 60^th and 80^th xtiles by 0.17%, 0.17%, 0.18%, 0.19% respectively.
• The magnitudes are slightly higher for consumption
• However, growth raises the incomes-consump. of the top 20^th xtile by a lower percentage (0.06% and 0.14% respectively)

RESULTS: growth and welfare distribution

• +ve growth rates increase:
• the incomes the 20^th, 40^th, 60^th and the 80^th by 0.20%, 0.23%, 0.23%, 0.29% and 0.15%(top 20^th xtile)
• 0.28%, 0.29%, 0.34%, 0.45% and 0.14%(top 20^th xtile) for consumptions.
• A % point increase in economic decline reduces incomes in the 20^th, 40^th, and 60^th quantiles by 0.14%, 0.10% and 0.12% respectively
• The coefs of economic decline on incomes are not significant in the last two top quintiles
• We get a similar picture with consumption welfare

RESULTS: growth and welfare distribution

• The upper-bound poverty line is lies at 57th percentile of incomes and 60th percentile of consumptions.
• The lower-bound poverty line lies at the 44th and 48th percentiles for incomes and consumptions respectively.
• This is consistent with Statistics South Africa (2017) which showed that about 55% of the South African population are poor. This shifts the middle class further to the right of the distributions.
• This analysis is to show that right up to the 60th percentile, we are still dealing mostly with the poor, especially when using household survey data which generally excludes the ultra-rich.

Summary of findings

• So far, we have established that economic growth reduces poverty, but not enough to compensate for the poverty raising effects of inequality;
• Economic decline raises poverty, but economic prosperity more than compensates, by attenuating poverty levels by a higher magnitude;
• inequality poses a welfare penalty on the poor, but leaves the welfares of those at the top untouched.
• Positive economic growth enhances welfare for the poor in % terms relatively more than at the top tier of the distribution;
• economic decline reduces welfare for the lower tier of the distribution without affecting those at the top;
• positive economic growth trumps economic decline in terms of effects on the welfares of the those at the lower end of the distribution;
• the combination of high and rising inequality and negative economic growth spells present a significant jeopardy in poverty reduction efforts, increase of the welfares of the poor and reduction in future inequality.

Policy options

• Some SA’s pro-poor measures traceable in the household survey data relate to social grants, free healthcare services and free education, also assessed by Kirii et al. (2020) for Kenya
• Up to 80% of SA’s poor access social grants with an average monthly amount of 1790 Rands
• compared with 40% accessing labour incomes averaging 3104 Rands per month.
• For the non-poor 47% access social grants averaging 1343 Rands
• compare with 57% of labour incomes averaging 8061 Rands.
• a number of health and educational policy programs undertaken are accessible through the public health and education system

Policy options: Poverty models

Policy options: Bottom40 incomes

Discussion: Health & Education

• While health and educational interventions are poverty reducing, their respective interactions with negative economic shocks, though negative, are not significant.
• Both education and health policies have the potential, but falls short of easing the effects of negative shocks.
• However, the instrumental variable regressions suggest that health and educational policies do raise the incomes of the poor including in times of negative economic shocks.
• Combining the results with those of poverty models suggests that the respites that these policies give to the incomes of the poor are not enough to have significant effects on poverty during economic downturns

​

​

Discussion: Free Public Health

• The provision of free public health services in South Africa can have both positive and negative effects on the poor.
• On one hand, it has increased access to healthcare for the poor, who are often unable to afford private healthcare.
• On the other hand, it has been hampered by a number of challenges, including inadequate funding, poor infrastructure, and human resource shortages, which have limited its effectiveness.
• One of the positive effects of free public health in South Africa is increased access to healthcare services for the poor.
• According to Health Systems Trust, the percentage of South Africans who visited a public health facility increased from 54.6% in 2002 to 70.4% in 2015, with the poorest quintile of the population showing the greatest increase (from 42.2% to 66.8%).
• The study found that free public health services have reduced the financial burden on the poor, as they no longer have to pay out of pocket for healthcare services.

Discussion: Free Public Health

• However, the provision of free public health services in South Africa has also been hampered by a number of challenges.
• inadequate funding, which has resulted in shortages of essential medicines, equipment, and personnel.
• The Department of Health, highlights that the public health sector is underfunded by approximately R22 billion, which has led to a range of challenges, including equipment and medicine shortages, staff shortages, and poor infrastructure.
• The public health sector has also been affected by human resource shortages, with many health professionals leaving the country to work abroad or moving to the private sector.

Discussion: Free Education

• Free education in South Africa has had a significant impact on the poor, particularly in terms of increasing access to education and reducing financial barriers to education.
• However, there are also concerns about the quality of education and the sustainability of the funding model.
• Spaull and Kotze (2020) have established that the introduction of free primary education in South Africa in 1994 led to a significant increase in school enrolment rates among the poorest households, particularly in rural areas.
• The study also found that the policy led to a reduction in the poverty gap among households with school-aged children.

Discussion: Free Education

• However, there are concerns about the quality of education in South Africa, particularly in public schools.
• A report by the Department of Basic Education (2021) found that only 37% of Grade 5 learners in public schools achieved the expected level of proficiency in reading in 2019.
• The report also found significant disparities in educational outcomes between different provinces and socioeconomic groups.
• Moreover, there are concerns about the sustainability of the funding model for free education. A report by the National Treasury (2019) found that the cost of free higher education was projected to increase significantly over the next decade, placing a strain on government finances.

Discussion: Social Grants

• Social grants do help the poor in times of economic downturns, however it does not go far enough.
• The results show that social grants assist in reducing poverty severity during negative economic shocks, but not significantly so for poverty incidence and intensity.
• Like free health and education, social grants do raise the incomes of the poor within the bottom 40 percentile especially in times of economic downturns.
• Again, the respite is not significant enough to be detected in poverty models.
• Woolard and Leibbrandt (2016) corroborates this finding that the social grant system played a critical role in reducing extreme poverty in the country, with grants covering over 50% of the food poverty line for the poorest households

Discussion: Social Grants

• However, concerns have been raised about the sustainability of the social grant system:
• particularly given the significant cost of the program.
• According to the National Treasury (2021), social grant spending accounts for around 12% of government expenditure, and is projected to increase further in the coming years.
• The impact of the grants on work incentives, with some critics arguing that the grants discourage recipients from seeking employment.
• Tregenna (2017) found that while the social grant system had a positive impact on poverty and inequality, it also had a negative impact on labor force participation rates.

Discussion: Levels of Education & Labour Incomes

• The coefficients of levels of education and labour income, with their respective interaction terms suggest that:
• while all forms of education from secondary reduces poverty and raises the incomes of the poor,
• only tertiary education can keep the poor afloat in times of economic shocks.
• Labour incomes reduce all forms of poverty and raises the incomes of the poor including in times of negative economic shock.

Discussion: Key policy message

• The implication is that:
• while various social programs like free health, education and social grants do help, they do not go far enough.
• There is room to improve efficiency of these programs to better assist the poor in times of economic crises and shocks like during the Covid-19.
• However, what can sustainably keep the poor afloat remains programs that give the poor good education up to tertiary and also give them access to the labour market.

Conclusion

• This research provides important insights into the relationship between economic growth, inequality, and poverty in South Africa.
• The findings highlight the negative impact of high and persistent inequality on poverty reduction efforts, and the crucial role of positive economic growth in reducing poverty.
• Economic growth, while beneficial to reducing poverty, is not enough to compensate for the poverty-raising effects of inequality in SA
• The study finds that initial inequality is a stronger hindrance to poverty reduction than initial income levels, and that rising inequality erodes poverty reduction gains far more than positive economic growth reduces poverty.

Conclusion

• The poor suffer more losses of welfare during economic recessions and depressions than they gain during expansions
• the factors that can assist the poor to stay afloat during times of economic shocks are good education up to tertiary and access to the labor market.
• The results suggest that policies that reduce inequality and promote economic growth would be beneficial to the poor
• While social programs like free healthcare, education, and social grants do help, they do not go far enough to cushion the poor during times of significant economic decline

Conclusion

• The study highlights the need for the government:
• to formulate policy measures to curb or reverse the effects of rising inequality and negative economic growth
• help propose a path to balanced and equitable development in South Africa.
• Moreover, the study underscores the need for pro-growth policies that also reduce inequality, as well as effective social safety net programs that can provide respite during times of economic downturns.
• It also suggests that sustained poverty reduction efforts in SA will require a combination of pro-growth policies that reduce inequality, access to tertiary education, and better access to the labor market for the poor.

Thank You

​

​

Nicholas Ngepah, PhD

nnnbal@yahoo.fr; nngepah@uj.ac.za

​