Department of Economics
Permanent URI for this community
Browse
Browsing Department of Economics by browse.metadata.advisor "Burger, Rulof"
Now showing 1 - 6 of 6
Results Per Page
Sort Options
- ItemA comparison between existing mortality risk algorithms and machine learning techniques(Stellenbosch : Stellenbosch University, 2022-12) Scholtz, Jenny; Burger, Rulof; Retief, Riani; Stellenbosch University. Faculty of Economic and Management Sciences. Dept. of Economics.ENGLISH SUMMARY: This thesis assesses the feasibility and benefits of using the patient data of a large private South African hospital group to estimate a model of mortality risk using flexible machine learning techniques. Specifically, I investigate whether such a model would have been able to outperform a commonly used medical scoring system, SAPS 3, in predicting mortality during the second half of the Covid-19 pandemic. A LightGBM machine learning model is shown to be much more accurate in predicting mortality (76.15% accuracy, compared to 56.58% for SAPS 3) for the Covid-19 positive sample. Roughly half of this gain in predictive accuracy is obtained from using the most recent and relevant data to train the model, while the remaining lift is attributable to allowing the model to find patient symptoms and attributes that are measured but ignored by SAPS 3. Interestingly, the flexible functional form of the machine learning models, which allow the predictors to affect mortality through non-linearities and interactions, has a negligible effect on predictive accuracy. The same method is also found to produce more accurate forecasts for patients who tested negative for Covid-19, but this improvement is smaller than for Covid-19 positive sample. The results of this thesis illustrate that machine learning methods are valuable tools to predict patient outcomes, particularly when there are unexpected shifts in the relationship between patient features and patient outcomes. Large hospital groups can obtain more accurate forecasts from a dynamic scoring system which is frequently frequently retrained on their own patient data.
- ItemGraduate unemployment, higher education access and success, and teacher production in South Africa(Stellenbosch : Stellenbosch University, 2016-03) Van Broekhuizen, Hendrik; Van der Berg, Servaas; Burger, Rulof; Stellenbosch University. Faculty of Economic and Management Sciences. Dept. of Economics.ENGLISH ABSTRACT: In the context of South Africa’s socio-economic challenges, Higher Education (HE) has a key role to play, not just in terms of producing su cient numbers of graduates and the scarce skills that are required for economic development and growth, but also in terms of providing opportunities for social mobility and restitution. This dissertation examines the extent to which the public HE system ful ls these roles by investigating three topics within the nexus between the secondary schooling system, HE, and the labour market in South Africa. Chapter 2 of the dissertation investigates graduate unemployment by focussing on the associations between HE institutions and the expected employment and unemployment outcomes for graduates from di erent race groups. Using a probabilistic approach to link labour force and HE data, the chapter estimates the associations between the probability of having graduated from a particular type of HE institution and the likelihood of being unemployed or employed. The results indicate that graduate unemployment in South Africa is neither high, nor rising rapidly over time and that much of the observed inter-racial variation in graduate unemployment rates may be explained by di erences in the types of HE institutions that di erent race groups have historically been likely to attend. Chapter 3 investigates HE access and success in the Western Cape, with speci c emphasis on the roles that demographics, academic performance, and school-level factors play in explaining the extent of, and the differentials in, HE participation and throughput among matric learners in the province. By explicitly linking learner records from matric examination data with student records from HE data, the chapter estimates the marginal contributions and relative importance of various pre-entry correlates and HE-level factors for observed HE outcomes among learners in the Western Cape and illustrates the respective roles that HE access rate and HE throughput rate di erentials play in explaining observed racial di erentials in HE graduations. The ndings reveal that HE access, throughput, and dropout rates are strongly correlated with matric performance and that much of the observed racial di erentials in HE access and dropout in theWestern Cape can be explained by di erences in matric performance levels between race groups. It is argued that the persistent HE completion rate premiums for White students may partly be driven by di erential conditional selection into HE. Lastly, Chapter 4 focusses on the production of Initial Teacher Education (ITE) graduates by the public HE system between 2004 and 2013 and its implications for teacher supply in South Africa. Using aggregate Higher Education Management Information System (HEMIS) data, the chapter provides a comprehensive descriptive analysis of the trends and underlying correlates of rst-time enrolments and graduations in ITE programmes. Despite the fact that enrolments and graduations in ITE programmes have risen signi cantly since 2004, the ndings suggest that South Africa is currently not producing su cient numbers of teacher graduates. Projections indicate that the system could begin to produce su cient numbers of graduates to satisfy projected teacher demand within the next decade, conditional on current enrolment growth and programme throughput rates. The chapter concludes that, in order to address South Africa’s teacher supply shortfall, greater emphasis is needed on ensuring that ITE students complete their programmes, specialise in high-demand subject areas and phases, and transition into the teaching profession with minimal delay.
- ItemAn investigation of the labour market determinants of income dynamics for a highly unequal society: The South African case(Stellenbosch : Stellenbosch University, 2019-04) Malindi, Kholekile Nicholas; Burger, Rulof; Stellenbosch University. Faculty of Economic and Management Sciences. Dept. of Economics.ENGLISH SUMMARY : South Africa ranks as the country with the highest income inequality in the world. Inequality of labour market outcomes drives most of this inequality. Labour market success (or failure) is a crucial determinant of where an individual or household is positioned on the income distribution. Furthermore, labour market outcomes in South Africa are characterised by a strong racial and gender hierarchy. Black women are on many levels the most disadvantaged with the lowest average earnings, highest unemployment, lower level of skill attainment, etc. They are consequently located at the bottom of this hierarchy regarding labour market outcomes. White men, on the other hand, are the most advantaged and are thus located at the top of this hierarchy. Differences in labour market outcomes in South Africa have spawned a large body of literature that identifies pre-labour and labour market differences in the accumulation of and returns to human capital as the key determinants of labour market inequality. A smaller strand of the literature points to labour market discrimination and barriers to entry into wage employment as contributing factors to the inequality of labour market outcomes in South Africa. This dissertation contributes to both strands of the literature. It contributes to the first strand of the literature by investigating the two critical components of the dynamic structure of wages. This includes the wage returns to labour market experience and job tenure for different demographic groups. On-the-job training as a means of human capital investment and a source of inequality is mostly ignored in the South African literature on differences in labour market outcomes. The dissertation adds theoretical and empirical evidence of the importance of information asymmetry and statistical discrimination in the barriers to entry and labour market discrimination literature, respectively. The empirical evidence presented in this dissertation is based on rigorous implementation and adaption of micro-econometric techniques to a nationally representative household South African panel dataset. The overall result points to better labour market outcomes for black workers regarding higher wage growth. This is due to the accumulation of on-the-job training and subsequent resolving of uncertainty regarding their expected productivity. This result is contrary to the stereotypical racial and gender hierarchy that sees black workers having inferior labour market outcomes. Additionally, this motivates the observed decline in inter-racial income inequality and the rise in intra-racial income inequality, especially amongst the black population.
- ItemMeasuring social and economic mobility in South Africa: new approaches to well-known survey data concerns(Stellenbosch : Stellenbosch University, 2016-12) Zoch, Asmus; Burger, Rulof; Van der Berg, Servaas; Stellenbosch University. Faculty of Economic and Management Sciences. Dept. of Economics.ENGLISH SUMMARY : The aim of this dissertation is measuring economic and social mobility in South Africa. The work from this thesis shows that various problems with survey responses can produce misleading impressions of the South African labour market and of income dynamics. This highlights the importance of measuring variables of interest accurately and to carefully consider the ways in which unreliable responses can bias the results of conventional estimators. It will be demonstrated that even the most appropriate estimator and identification strategy can fail to yield unbiased estimates if important measurement issues are ignored. To address these shortcomings this thesis applies and adapts new approaches to remedy well-known survey data reliability concerns. The most important findings of the three chapters are as follows: First, in the context of high unemployment and weak labour market attachment for many South African youth, the formulation of survey questions matters and approaches commonly used to elicit reliable responses in developed countries cannot be assumed to work equally well. This is particularly true for subjective measures. Hence, answers to the traditional question on reservation wages may fail to provide meaningful answers. It appears that different formulations and ordering of the reservation wage question can trigger different cognitive processes in the respondent that elicit different answers. However, using a series of questions intended to elicit a more accurate response, the new reservation wage measure seems to be more internally consistent and the regression results to be in line with labour market search models. Second, this thesis shows that estimating the speed of convergence between the poorest and richest households using micro growth regressions without controlling for measurement error would overestimate income mobility significantly. Therefore, a newly developed GMM estimator was applied to four large national panel studies to obtain less biased β estimates. The findings of four large representative national panel studies from the USA, South Africa, Chile and Tanzania show that naïve OLS regression coefficients would overestimate the extent of income mobility by a factor of about 4-6. The hypothesis of no measurement error can be rejected for all the countries observed. While the data reliability for the US, Chile and Tanzania correspond to their levels of economic development, South Africa’s data reliability appears to be unexpectedly high. The nonparametric estimates also show that the speed of convergence varies over the income distribution and that income is more reliably captured for richer than for poorer households. Third, the relative importance of family, neighbourhood and school quality in explaining variation in socio- economic outcomes are evaluated. Using spatial merging techniques to combine different data sets, new school wealth quintiles have been created that predict individual learner and school outcomes more accurately than the old school quintiles. This chapter provides evidence of the importance of quality education in explaining university enrolment. In addition, there seem to be a significant premium for quality education in labour markets earnings regressions, which confirms the long-term importance of schooling.
- ItemSocial mobility and cohesion in post-apartheid South Africa(Stellenbosch : Stellenbosch University, 2015-04) Von Fintel, Marisa; Van der Berg, Servaas; Burger, Rulof; Stellenbosch University. Faculty of Economic and Management Sciences. Dept. of Economics.ENGLISH ABSTRACT: Twenty years after the end of apartheid, South Africa remains one of the most unequal countries in the world. Socio-economic polarisation is entrenched by the lack of social capital and interactions across racial and economic divides, blocking pathways out of poverty. This dissertation examines social mobility and cohesion in post-apartheid South Africa by considering three related topics. Chapter 2 of the dissertation examines the impact of school quality on the academic performance of disadvantaged learners as one of the most important enforcing factors perpetuating the social and economic divides. Given the historic racial and economic stratification of the South African public school system, many black children are sent to historically white public schools as a way to escape poverty. Using longitudinal data, this chapter estimates the effect of attending a historically white school on the numeracy and literacy scores of black children. The main challenge is to address the selection bias in the estimates, for which a value-added approach is implemented in order to control for unobserved child-specific heterogeneity. In addition, various household covariates are used to control for household-level differences among children. The results indicate that the attendance of a former white school has a large and statistically significant impact on academic performance in both literacy and numeracy which translates into more than a year’s worth of learning. The main finding is robust to various robustness checks. In Chapter 3 the dissertation examines social cohesion by considering the concept of reference groups used in the evaluation of relative standing in utility functions. The chapter develops a model in which various parameters are allowed to enter the utility function without linearity constraints in order to determine the weight placed on the well-being of individuals in the same race group as the respondent versus all the other race groups living in one of three specified geographic areas. The findings suggest that reference groups have shifted away from a purely racial delineation to a more inclusive one subsequent to the country’s first democratic elections in 1994. Although most of the weight is still placed on same-race relative standing, the estimates suggest that individuals from other race groups also enter the utility function. The chapter also examines the spatial variation of reference groups and finds evidence that the relative standing of close others (such as neighbours) enter the utility function positively while individuals who live further away (strangers) enter the utility function negatively. Finally, Chapter 4 provides a summary of the dynamics of income in South Africa, using longitudinal household data. Chapter 4 is aimed at separating structural trends in income from stochastic shocks and measurement error, and makes use of an asset-based approach. It first estimates the percentage of individuals who were in chronic poverty between 2010 and 2012 and then estimates the shape of structural income dynamics in order to test for the existence of one or more dynamic equilibrium points, which would be indicative of the existence of a poverty trap. The findings do not provide any evidence for the existence of a poverty trap. In addition, contrary to earlier findings, the results do not provide evidence for the existence of an asset-based threshold at which the structural income accumulation paths of households bifurcate. Instead, the results seem to indicate the existence of a threshold beyond which structural income remains persistent with very little upward mobility. The robustness of the results is confirmed by making use of control functions in order to correct for any measurement error which may exist in the data on assets.
- ItemSpatial heterogeneity, generational change and childhood socioeconomic status : microeconometric solutions to South African labour market questions(Stellenbosch : Stellenbosch University, 2014-12) Von Fintel, Dieter; Burger, Rulof; Van der Berg, Servaas; Stellenbosch University. Faculty of Economic and Management Sciences. Dept. of Economics.ENGLISH ABSTRACT: Microeconometric techniques have improved understanding of South Africa’s labour market substantially in the last two decades. This dissertation adds to this evidence by considering three separate labour market questions, with particular attention to data quality and the application of credible methodology. Firstly, wage flexibility is investigated. Whereas selected previous microeconometric evidence suggests that wage setters in South Africa are highly responsive to external local labour market circumstances, it is not corroborated by macroeconomic and other microeconometric studies. This question is interrogated again, with particular attention to methodological issues in wage curve estimation. The latter is a robust negative relationship between individual wages and local unemployment rates, found in most countries, except where bargaining is highly centralized. Adding time variation to the data allows controls for spatial heterogeneity to be introduced, leading to the conclusion that wages are really inflexible in the short-run. Rather, the trade-off between wages and local unemployment that previous work has found represents a long-run spatial equilibrium. This finding is robust to instrumentation for reverse causality and the measurement error that is associated with choosing incorrect labour market demarcations. Secondly, the reliability of retrospective data related to childhood is investigated, with the view of estimating the long-run influence that early life circumstances have on adult outcomes. Two indicators, parental education and subjective rankings of childhood socioeconomic status, are evaluated. The first set of indicators has poor response rates, as many South African children live without their parents. Where respondents do volunteer this information, they answer consistently across waves. Subjective rankings have higher response rates, as they require respondents to provide information about their own past, and not about those of their parents. However, individuals’ assessments are inconsistent over time, despite being asked about the same point in the life cycle. They tend to change their view of the past in line with adjustments to perceptions of their position in the village income distribution and subjective well-being, providing clear evidence of anchoring. Instrumental variables analysis has been used in previous studies to account for measurement error in subjective data. However, if anchoring affects all assessments of the past and potential outcome variables (such as employment), microeconometric techniques will yield biased estimates of the effects of childhood on long-run outcomes. Finally, age-period-cohort models for South African labour force participation are estimated. This chapter is the first contribution to relax the assumption that cohort differences must remain permanent over the life cycle. Monte-Carlo simulation studies show that highly interactive specifications can partially recover the true underlying process. Using a variety of techniques (imposing behavioural restrictions and atheoretical approaches), this study shows that cohort effects in labour force participation can be temporary in South Africa, though more data is required to verify this conclusively. Regardless of technique, a distinct surge in labour force participation is noted for the group born after 1975. Pertinently, the combination of testable assumptions and highly flexible estimation can yield credible age-period-cohort profiles, despite the many disputes noted in the literature. Previous evidence of a surge in participation for the post-1975 cohort can now be shown to be temporary rather than a part of a long-run generational increase.