Optimising balance using covariate balancing propensity score: The case of South African child support grant

ABSTRACT

In this paper, we explore the use of covariate balancing propensity scores (CBPS) in estimating the impact of the South African child support grant (CSG) on the height-for-age score of benefiting children. CBPS is a different approach to estimating propensity score, under CBPS the scores are estimated such that the estimation incorporates covariate balancing condition. This approach is therefore relatively robust to misspecification of the propensity score model which makes it ideal for this case study. We show that utilising the CBPS leads to treatment effect estimate that is larger and more precisely estimated than estimates that have been reported in the literature because the method exploits the dual function of propensity score. The effect of CSG under CBPS is as large as 44% of standard deviation on average. This implies that the effect of the grant cannot be regarded as small as previously reported in the literature.     

The performance of the Child Support Grant: Review and research priorities

This article reports the findings of a review on the performance of the Child Support Grant (CSG) programme. It suggests that the CSG impact theory is sound, the programme has in general been implemented well and the limited research on impact is suggestive of it achieving its ultimate objectives of reducing child deprivation and promoting human capital development. It points to the vastness of the child poverty that remains to be addressed and raises the concern that in spite of young children being prioritised in the roll out of the CSG, child poverty incidence may be highest in children aged zero to four. Weaknesses in programme implementation are identified, as well as research priorities in three areas: the child poverty profile, CSG programme implementation and CSG impact.     

Social assistance and dignity: South African women's experiences of the child support grant

Many women interact with the South African social security system in relation to the Child Support Grant (CSG), which is social assistance payable for children living with low-income caregivers. This paper explores women's accounts of how the CSG serves to protect and respect dignity, a foundational value in the South African Constitution. Drawing from focus groups and in-depth interviews with female CSG recipients of working age, it is argued that whilst the experience of using the CSG does protect dignity in certain important respects, other aspects including the application process, the small amount of the grant and negative discourses associated with the status of being a CSG recipient were experienced by many as erosive of dignity.     

THE IMPLICATIONS OF SOUTH AFRICAN ECONOMIC GROWTH FOR THE REST OF AFRICA

This paper measures the extent to which South African economic growth is an engine of growth in sub‐Saharan Africa. Results based on panel data estimation for 47 African countries over four decades suggest that South African growth has a substantial positive impact on growth in the rest of Africa, even after controlling for other growth determinants. The estimates are robust to the effects of global and regional shocks, changes in model specification, and sample period.     

IMPLEMENTING A ROBUST RISK MODEL FOR SOUTH AFRICAN EQUITY MARKETS: A PEAK‐OVER‐THRESHOLD APPROACH

This paper implements a market risk model for the South African equity market using daily returns of the Johannesburg Stock Exchange All Share Index. Firstly, we separate positive returns from negative returns and model them using the peak‐over‐threshold (POT) method in order to compute the downside as well as upside risk measures separately. We thereafter compute the value‐at‐risk (VAR) and the expected shortfall (ES) estimates corresponding to upside and downside risks. We bootstrap these risk measures and compute their standard errors and confidence intervals (CIs) to see whether they fall inside these CIs. Secondly, we compute out‐sample forecasts of VAR estimates using the POT method and the generalised autogressive conditional heteroscedasticity process. Three backtesting methodologies are employed: the unconditional and conditional coverage tests and the counting of number of exceptions according to Basel II green zone. We find that all our VAR and ES estimates are well inside their CIs and that at lower quantiles, parametric ES estimates are equal to POT‐ES estimates, although the difference between the two is more pronounced at higher quantiles (99% or higher). Furthermore, our market risk model falls into the Basel II green zone, as it produces fewer exceptions in out‐sample space.     

The Food (In)Security Impact of Land Redistribution in South Africa: Microeconometric Evidence from National Data

The South African land-reform program has been widely criticized for its slow pace as well as its apparent lack of contribution to poverty reduction. No econometric evidence of the impact of land transfers has been provided to date and this paper attempts to fill this gap by considering the impact of receiving a land grant on households’ food insecurity. Propensity score matching and univariate probit estimates using two national household surveys indicate that, on average, land grant recipients are more food insecure than comparable non-participants. Recursive bivariate probit estimates suggest that selection bias is not driving this result.     

CIGARETTE DEMAND IN SOUTH AFRICA OVER 1996-2006: THE ROLE OF PRICE, INCOME AND HEALTH AWARENESS

The study seeks to re‐investigate the role of price, income and health awareness in the evolution of South African cigarette demand over the period 1996 to 2006. At first glance, rising cigarette prices appear to have played an important role in reducing cigarette consumption over the sample period, especially during the late 90s. But how dependent is the impact of price increases on general economic conditions and overall health awareness among smokers? Health awareness, in particular, has not received sufficient attention in the South African context, due to a lack of data. Previous estimates of price and income elasticities of cigarette demand are based on long annual time series data, which do not allow for changes in underlying tastes and preferences. The paper attempts to disentangle the forces of price, income, health awareness and policy intervention using a quarterly data set. In addition, the study also cautions against the upward bias in estimates derived from formal cigarette sales data – in the light of increasing illicit cigarette volumes in South Africa.     

SOUTH AFRICA'S TRANSITION TO A CONSOLIDATED BUDGET

The size of South Africa's fiscal stimuli, whether intended or not, has helped to avert negative consequences of the global financial downturn. With the economic cycle turning again, consolidation of deficits and a reduction of the level of debt are again the focus of policymakers. These outcomes are generally achieved by either increasing tax rates or cutting spending (discretionary fiscal policy), whereas an alternative option is to allow automatic stabilisers to consolidate budgets. This study attempts to answer whether cyclical factors or discretionary policy minimise output volatility and which one of the two presents a better policy option regarding uncertainty in real economic recovery. For this purpose, a small open‐economy gap model is built using South African data, where the budget deficit is endogenised by way of a fiscal policy “rule.” Sensitivity analyses and robustness checks are carried out using a structural vector autoregression. Given the estimates of both the automatic stabilisers as well as the components of discretionary fiscal policy, we are able to obtain impact multipliers on output and conduct scenario testing for optimal fiscal policy response towards fiscal consolidation as well as debt sustainability.     

CYCLICAL UNEMPLOYMENT AND CYCLICAL OUTPUT: AN ESTIMATION OF OKUN'S COEFFICIENT FOR SOUTH AFRICA

Persistently high unemployment in South Africa, especially in the face of improved economic conditions since 1994, begs the question: Does unemployment in South Africa respond to changes in output? Okun's law refers to the inverse relationship that exists between cyclical output and cyclical unemployment. This paper estimates Okun's coefficient for the South African economy, using annual data from 1970‐2005. Output and unemployment are decomposed into their trend and cyclical components, using a variety of detrending methods. The presence of structural breaks in Okun's relationship is also investigated, while cointegration analysis was also considered. Evidence of a statistically significant relationship between cyclical output and cyclical unemployment are found in both symmetric (estimates range from ?0.77 to ?0.16) and asymmetric (estimates range from ?0.77 to ?0.18) specifications of Okun's law, irrespective of the detrending technique. However, cyclical unemployment constitutes only a relatively small fraction of total (observed) unemployment, which implies that a more expansionary macroeconomic policy stance might only have a limited impact on total unemployment in South Africa.     

The South African–United States sovereign bond spread and its association with macroeconomic fundamentals

The yield spread of South African to United States 10-year government bonds over the last 5 years has increased substantially to levels approaching those last seen during the mid-1980s. This paper examines the association between the spread and macroeconomic fundamentals over the 1960–2019 sample period, under the GARCH and GARCH-M class of estimators. We find that higher South African economic growth, lower inflation, public and private debt, as well as rand–dollar appreciation are all associated with a statistically significantly lower South African–United States yield spread. The strongest impact is associated with the public debt-to-GDP ratio. Mean spread levels do not appear to be influenced by yield volatility. Finally, while there is no evidence of sign bias in the impact of shocks on yield volatility (negative shock impacts are no different than positive), there is evidence of size bias for both positive and negative shocks: larger shocks have a larger impact on volatility than small, regardless of their sign. Collectively, and even ignoring the impact of private sector leveraging, South Africa’s performance in these macroeconomic fundamentals is associated with an increase in the South African–United States yield spread of 363 basis points (since 2012).     

LIQUIDITY CREATION IN SOUTH AFRICAN BANKS UNDER STRESSED ECONOMIC CONDITIONS

The financial crisis placed severe pressure on global bank liquidity. Many banks were unable to create sufficient liquidity and had to receive government support or face default. This paper attempts to determine the impact of the financial crisis on liquidity creation within South African banks using a model previously applied to US banks. Four measures of liquidity creation are discussed and applied to data spanning 2004‐2009. Although created, liquidity decreases steeply from 2007, liquidity levels in 2009 remain about 45% higher than those of 2004. The four large South African banks created about 80% of the total market liquidity. The authors are aware that the liquidity measures discussed in this paper may not accurately reflect the manner in which South African banks created liquidity before and during the financial crisis as the period under review was characterised by various other factors. These factors include the concurrent rise in securitisation, considerable changes in the South African Reserve Bank's (SARB) stance on monetary policy, historic low‐interest rates during the period and severe curtailing of local inter‐bank dealing. It remains possible that these factors may have had an impact on the way in which South African banks created and disseminated liquidity. This possibility will be the subject of future research.     

LIQUIDITY AND VALUATION IN EAST AFRICAN SECURITIES MARKETS

This study estimates liquidity premiums using the recently developed Liu measure within a multifactor capital asset pricing model including size premiums and a time‐varying parameter model for the East African emerging markets of Uganda, Tanzania and Kenya together with London and South Africa. The evidence suggests that while size and liquidity effects are significant in the smaller emerging markets of Uganda and Kenya, they are less important in explaining returns in South Africa and London. Costs of equity are highest in Uganda followed by Kenya, with industrial and consumer non‐cyclical sectors being lowest, and then South Africa and London.     

Impact of Macroeconomic Announcements on Foreign Exchange Volatility: Evidence from South Africa

This study focuses on scheduled macroeconomic news announcements and evaluates their impact on the volatility of the South African rand (ZAR) and US dollar (USD) exchange rate using high frequency data. The following asymmetries are studied: news items by geographical location, no‐news vs. surprise news announcements and positive vs. negative news announcements. We make the following findings in our empirical study: (i) After the release of a news announcement, the level of foreign exchange volatility rises. This is independent of whether the news item surprised the market or not. (ii) Both South African and US news items significantly impact USD/ZAR volatility, suggesting that the news items are being used to formulate investor expectations regarding the future prospects of the currency pair. (iii) Negative news appears to have a greater impact on exchange rate volatility relative to positive news. This result is also state dependent, as investors tend to behave differently to news depending on the economic climate at that point in time. Investor cognitive biases give rise to the asymmetric news effects on exchange rate volatility. Finally, investors do not always act in rational manner, especially when faced with multiple news items that are contradictory to each other.     

Global Risk Factors and South African Equity Indices

South African equity is frequently portrayed as a market requiring a high degree of local expertise – to appropriately understand its many idiosyncratic features – as well as intimate knowledge of its unique drivers – to prudently invest in the same. This claim is evidenced by the amount of research and effort devoted to understanding South African‐specific economics, interest rates and risks. The aim of this research is to debunk this perception with a simple yet robust and highly replicable statistical model (best‐subsets regression) for the majority of the traded South African equity indices. We show how the South African equity market is mostly a one‐way mirror of a confluence of international factors, all arguable largely unrelated to South Africa. We discuss why these models are currently less useful than their longer‐term predictive averages and note the current relevance of including implied volatility and interest rates as predictors.     

Are Determinants of Portfolio Flows Always the Same? ‐ South African Results from a Time Varying Parameter Var Model

The literature on determinants of cross‐border capital flows has consistently assumed the determinants of such flows to be constant throughout the sample. This paper investigates this notion by estimating the time varying relationship between portfolio flows to South Africa and two widely accepted determinants of such flows: the sovereign spread and global risk (measured by the CBOE Volatility Index, henceforth VIX). The results show that the time variation is highly significant and a constant parameter model will give biased estimates of the effects of risk on capital flows. The paper also gives important insights to South African policy makers and financial practitioners: Bond flows (non‐resident purchases of South African bonds) have become more sensitive to the VIX after 2010. Share flows were particularly sensitive at the peak of the 2008 global financial crisis, but have at other times not responded in a statistically significant manner to changes in global risk. The relationships are estimated using a time varying parameter vector autoregressive (TVP VAR) model with stochastic volatility.     

External Shocks and the HIPC Initiative: Impacts on Growth and Poverty in Africa

Abstract: The paper uses a simple macroeconomic model to estimate the impact of debt relief and terms of trade shocks on growth and poverty in African countries. For the 18 heavily indebted poor countries (HIPCs) that reached the enhanced HIPC decision point by end‐December 2000, the basic quantitative findings are that HIPC debt relief has boosted economic growth in these countries by an average of 2.9 percent per annum and that the computed result of this increase in growth is a reduction in poverty by an average of 2.2 percent per annum. However, the paper shows that recent deteriorations in the terms of trade have counter‐balanced these positive effects by lowering growth by an average of 2.0 percent per annum and by increasing poverty by an average of 1.3 percent per annum. Clearly, much of the positive impact emanating from the HIPC Initiative has been eroded due to recent deteriorations in the terms of trade. The paper also estimates the net effect on growth and poverty of the recently agreed 100 percent multilateral debt relief. This is predicted to boost economic growth by an average of 5 percent per annum and reduce poverty by about 5.3 percent per annum for the group of all African HIPCs. The paper concludes that 100 percent debt relief is crucial for Africa, but that more aid and policies need to be focused on a long‐term development strategy that fosters the necessary structural transformation.     

Financial distress and non‐executive director compensation: Evidence from state‐owned enterprises in South Africa post King III

Using a sample of South African state‐owned enterprises (SOEs), we examine the pre‐ and post‐period impact of King III on non‐executive director (NED) compensation with emphasis on financially distressed SOEs. This paper adopts a difference‐in‐differences analysis technique with repeated measures as the basis for testing the hypotheses. The revised Altman Z‐score model which incorporates features unique to emerging markets is used to measure financial distress. Our findings indicate that SOEs that adopted King III will increase NED compensation when the firm has a positive performance and will severely penalize NED when the firm faces financial distress. This study highlights the importance of well‐crafted corporate governance policies. It further sheds light on the importance of King III and how its implementation may prove vital for the success of an enterprise.     

WHAT PRICE‐LEVEL DATA CAN TELL US ABOUT PRICING CONDUCT IN SOUTH AFRICA

This article reports the findings of a study of two unique microdata sets, which are the unit level basis of South Africa's measured Consumer Price Index and Producer Price Index, over the period December 2001‐December 2007. In particular, based on techniques that have been used in comparable international studies, findings are reported on the frequency and magnitude of price changes, the duration of prices, and heterogeneity in pricing. The results for South Africa are compared with the stylised facts for pricing conduct, which have been presented in recent international studies. The article offers an illustration of how microdata‐based findings on pricing conduct may impact on the modelling of monetary policy by introducing micro‐founded results into an open economy dynamic stochastic general equilibrium model for the South African economy. The article concludes by identifying areas for further research, where it has not as yet been determined how South African pricing conduct compares with certain stylised pricing facts identified in the international literature.     

FISCAL SUSTAINABILITY AND THE FISCAL REACTION FUNCTION FOR SOUTH AFRICA: ASSESSMENT OF THE PAST AND FUTURE POLICY APPLICATIONS

How does the South African government react to changes in its debt position? In investigating this question, this paper estimates fiscal reaction functions using various methods (ordinary least squares, threshold autoregressive, state‐space modelling and vector error‐correction model). This paper finds that since 1946, the South African government has run sustainable fiscal policy by reducing the primary deficit or increasing the surplus in response to rising debt. Looking ahead, this paper considers the use of fiscal reaction functions to forecast the debt/gross domestic product (GDP) ratio and gauging the likelihood of achieving policy goals with the aid of probabilistic simulations and fan charts.     

Income and Price Elasticities of Demand in South Africa: An Application of the Linear Expenditure System

This paper investigates the expenditure patterns of South African households using detailed cross‐sectional expenditure and price data that varies across region and time. Linear expenditure system parameter estimates are used to calculate income and price elasticities for a number of product categories at different points of the income distribution. We find substantial variation in the price and income elasticities of demand for items across the income distribution, with the bottom quartile being extremely sensitive to increases in the price of food and clothing items, and the top quartile being as sensitive as households in developed countries.