Faculty of Economic and Management Sciences
School of Economic Sciences
Department of Economics
Selected Highlights from Research Findings
A series of recent studies examined the different ways (ranging from welfare cost of inflation to the effect of inflation targeting on average levels and volatility of inflation) of evaluating the inflation-targeting regime in South Africa. The research found that welfare cost estimates, based on partial equilibrium approaches that involve estimating the area under the money-demand function, are sensitive to specifications, econometric methodologies and data aggregation. More importantly, the research indicates that the partial equilibrium approach severely underestimates the sizes of welfare cost of inflation. The welfare cost of inflation for the current target band of 3–6% is between 0.8% of gross domestic product (GDP) to around 2% of GDP, figures that are more than double when compared to the partial equilibrium approaches. Using a model of dynamic time inconsistency in monetary policy-making and the cosine-squared cepstrum, the average levels of inflation and inflation volatility respectively would have been lower if the old regime of so-called eclectic monetary policy-making had continued. The research points to two conclusions. Firstly, there are huge welfare gains from reducing the target below the 3% level. Secondly, the poor performance of the inflation-targeting regime in terms of average levels and volatility of inflation could be associated with the width of the target, which does not allow inflation expectations to converge to a focal point. This line of thinking is in accordance with the current theoretical literature, as well as international evidence of economies pursuing much stricter (point) targets. In sum, these cumulative research studies suggest that the best policy would be not only lowering the target, but also making it narrower.
Contact person: Prof R Gupta.
This research investigates a number of issues concerning the modelling of monetary policy in South Africa. Recently, researchers have questioned the linear specification of monetary policy, which assumes that the response of interest rates to inflation and output is constant. Furthermore, recent economic events have turned attention to the behaviour of asset prices (stock prices, house prices and commodity prices) and the concern by central banks over the maintenance of financial stability.
Using a variety of nonlinear parametric, nonparametric and semi-parametric regression estimation techniques, the findings of this research are that the response of the South African Reserve Bank (SARB) of monetary policy to inflation is asymmetric and greater during business cycle recessions and seems to place high importance on inflationary pressures of output during boom periods. The 2007–2009 financial crisis witnesses an overall decreased reaction to inflation, output and financial conditions, amidst increased economic uncertainty, with a shift from an asymmetric response to financial conditions over recessions to a more symmetric response irrespective of the state of the economy.
Semi-parametric models, which are data-driven and do not impose an often clumsy linear/nonlinear parametric structure on the data, consistently outperform linear and nonlinear parametric models and are best suited in forecasting the rate of interest.
The response of the SARB’s policy-makers to financial conditions arguably has important policy implications as it might shed some light on why the current downturn in South Africa and the Eurozone area (where similar practices have been reported) is less severe than in the USA where financial conditions do not feature in the Federal Reserve Bank’s reaction function. Secondly, the forecasting superiority of semi-parametric models implies that there are new types of complex and subtle monetary policy responses that have hitherto been obscured by the use of restrictive parametric models.
Contact person: Dr R Naraidoo.
Water is a limiting factor constraining economic development, a reality that is often overlooked. South Africa’s unallocated water resources (that is, water that has not been dedicated to a specific sector) have dwindled to precariously low levels. It has been estimated that the country’s surplus water in 2000 (the latest available official figure) was 1.4% of its national supply. With an economy that is growing at 3% per annum, and with a population growth in excess of 2%, the surplus will soon be exhausted. The authorities acknowledge this and estimate that the country as a whole will be in a water deficit within the next 10 years. Currently, 12 of the 19 water management areas are already in a water deficit. These shortages are managed through an intricate web of interbasin water transfer schemes, but options are dwindling at an alarming rate.
At a macroeconomic and strategic level, the question therefore is how to make best use of the country’s available water resources. This research applied a macro-economic computable general equilibrium (CGE) model using an integrated database comprising South Africa’s social accounting matrix (SAM) and sectoral water use balances. The research references the South African government’s Accelerated and Shared Growth Initiative for South Africa (AsgiSA) and concludes that introducing the proposed programmes in a business-as-usual and water-intensive manner would strengthen the current growth in the demand for water, accelerating the time of a national water deficit. The analysis demonstrated that a hypothetical injection into the economy of R1 billion to each of nine targeted AsgiSA industries would lead to a deficit in the available amount of water. It would, therefore, be physically impossible to stimulate the nine industries as planned, unless the necessary water supplies were reallocated from other sectors. An alternative scenario is proposed that incorporates a hypothetical water tax. This tax would decrease the total water demand sufficiently enough to provide for the AsgiSA initiatives and have some savings left over. If the revenues from the water tax were recycled towards the nine industries, the negative impact of the water tax would be diminished in terms of GDP and employment effects, while a large net saving of water would remain. This analysis shows that macro-economic planning and the design of economic development strategies cannot be done in isolation from consideration of natural resource constraints.
Contact person: Prof JH van Heerden.
Two research projects conducted with Dr Olufunke Alaba, a postdoctoral fellow in the department, focused on health issues and their effects on households. The first project examined the timing of treatment for ill children. Using a variant of a matching estimator, the research estimated the probability that ill children received either immediate, delayed or no treatment. Consideration was given to the impact of immediacy on the choice of facility (public or private) in which treatment was sought. The results suggest that immediate treatment (over delayed or no treatment) is less likely for children in larger households, but more likely for children with multiple symptoms. Access to insurance does not appear to affect timing. Delayed treatment (over no treatment) is more likely for children from the wealthiest households. Multiple symptoms also appear to result in immediate treatment for these households. Immediate treatment in a public facility is less likely for children covered by insurance. However, delayed treatment in a public facility is not affected by insurance coverage, although it is affected by the number of observed illness symptoms. In total, the research suggests that children without access to insurance may not be receiving immediate care, despite free public health care. This may possibly result in a greater number of symptoms that are then examined in private facilities, presumably, at a high cost to the household.
The second study postulated that mandatory contributions to a national insurance pool could have detrimental effects on the households most likely to be affected by the mandates. The analysis used matching estimators to examine differences in household budget allocations across various subgroups of the population. The analysis revealed that the mandates would impact on budget allocations in expected ways. Firstly, households would be required to allocate a greater share of their budget towards health (health insurance). Secondly, in order to offset the increase in their health budgets, households would need to make cuts in other areas. The results suggest that reallocations towards the health share of the budget would most likely arise from reductions in the share of the budget allocated to food, transportation and communication. When placed in the context of policy, although the analysis cannot indicate for sure, whether or not the reallocations result in social improvements, the results are strongly suggestive of negative household level welfare effects.
Contact person: Prof SF Koch.
|
