Mineworkers’ Perspectives Towards Participating in Retirement Planning in South Africa

Abstract

This study investigated the individuals’ perspectives towards participating in retirement planning in the mining industry in South Africa. The study employed a quantitative research approach. The study sampled 172 mineworkers from the selected mining company. A self-administered questionnaire was tested for validity and reliability and was used to collect primary data from the respondents. This study employed the logistic regression model and performed the Hosmer–Lemeshow test to evaluate the fit of the logistic regression and the Chi-square to determine the significance of the results. In this study, the data were analysed using descriptive and inferential statistics. The findings revealed that some participants are satisfied with their involvement in the retirement funds and are contributing to the retirement funds provided by the company. Furthermore, this study found that the majority of the respondents will be financially independent after retirement; however, there is still a firm belief of uncertainty about not being financially independent. The study found a significant and positive relationship between age and participation in retirement planning. Furthermore, a positive and significant link was found between marital status and participation in retirement planning as well as between employment status and participation in retirement planning. The study was limited to the selected company based in Gauteng. The practical implication of this paper informs the companies, policymakers, and government to prioritise awareness of retirement planning based on demographical factors such as age, marital status, and employment status to prepare mineworkers for retirement. The findings are expected to persuade the mining sector to pay special attention to the awareness and understanding of retirement planning.

1. Introduction

In South Africa, the savings that are sufficient for retirement remain an enormous challenge, with debt between individuals and society being consumption-driven (Dhlembeu et al., 2022). Retirement is the stage where an individual must permanently leave their workplace, and this implies that the salary/wages of such a person will stop upon retirement date (Parker, 2022). Sabri et al. (2020) assert that certain expenses of those individuals will continue, and they will have to depend on their accumulated savings to continue living to their desired standard. Individuals, households, organisations, and countries have limited financial resources that must be exploited to meet their unlimited needs (Tribe, 2020). According to the self-determination theory, it is significant that individuals have a high understanding of and balance between their financial needs, lifestyle choices, and commitments (Ryan & Deci, 2024).

Making efficient financial decisions requires the use of financial skills, information, and knowledge. Retirement planning is defined as the knowledge of financial concepts and the ability to allocate savings for retirement to fulfil financial security and lifestyles (Mata, 2021; Noone et al., 2022). As a result, retirement planning has a significant impact on people’s daily management of finances and long-term financial security in the face of shifting global economic conditions. Retirement planning is low, even among those economies with competitive financial systems. Approximately one-third of the world’s population understands basic financial principles that serve as the basis for daily financial decisions (Muthia et al., 2021; Lusardi et al., 2020).

Organisations aid employees with their retirement planning by providing them with employee benefits (Lusardi & Mitchell, 2017). Individuals must plan their retirement funding on time, as the retirement funds will replace their stream of income after retirement. Therefore, it should be managed effectively in meeting the goals of such persons (Shah & Bhatt, 2016). This is not spared in the mining industry, where the mining sector generated around ZAR 202.05 billion to the country’s Gross Domestic Product (GDP) in 2023 (Cowling, 2024). Despite the significant economic prospects that emerge from the industry, mineworkers are exposed to hazardous materials, tools, environments, materials, and conditions, and many injuries have been reported as a result of mining activities, which may result in an inability to work (Strzemecka et al., 2019; Singh et al., 2022). The South African Department of Mineral Resources published a statement on occupational health and safety and showed a decline in the number of deaths in the past decades; however, there is an impact in the mining accidents which can also be noticed in colleagues and family members after mineworkers lose their life (Department of Mineral Resources, 2019). Such an impact is significant in the monetary sense where the deceased can no longer provide for their family.

Dhlembeu et al. (2022) and Nteso (2021) assert that only 6% of South Africans will be able to maintain their standard of living after retirement. This retirement planning crisis shows that 94% of South Africans are not accumulating sufficient savings for retirement. Retirement planning in South Africa becomes worse every year owing to a lack of financial knowledge among other factors (Webb, 2021). The lack of financial knowledge about retirement planning and savings has received special attention in the country’s budget speech by the Minister of Finance (National Treasury South Africa, 2021). To ensure the financial well-being of households, good financial planning and management are necessary and should be managed effectively and encouraged in households. According to the Daily Industry News (2012), individuals in South Africa are struggling to manage their finances, and owing to a lack of financial knowledge, they tend to depend on the government and family members after retirement. Financial knowledge refers to a person’s knowledge about finances, retirement plans, employee benefits, and money management (Hanna et al., 2016). According to the Financial Sector Conduct Authority (FSCA, 2021), although most mining companies offer retirement fund services, most employees are not aware that they have such funds in the company. Furthermore, the FSCA (2021) reported that there is still an enormous number of unclaimed benefits on retirement funds in the South African mining industry. These funds stay unclaimed after the employee has retired or passed away without the knowledge of the retirement funds provided by the employer. When the fund provider cannot trace the member, the unclaimed benefits are transferred to the unclaimed retirement benefit funds (Atleha-Edu, 2023).

Researchers in South Africa have conducted studies on retirement planning (Zeka, 2017; Reyers, 2018; Antoni et al., 2020). The studies examined the impact of financial literacy on retirement planning and/or South African sectors with highly educated people. Prior studies found that mineworkers are illiterate (Raghav et al., 2020; Naveed & Ali, 2021; Dhlembeu et al., 2022). Scholars further argue that financially literate individuals make sound financial planning compared to their illiterate counterparts (Boisclair et al., 2017; Antoni et al., 2020). This was confirmed by Hutabarat and Wijaya (2020), when investigating the effect of financial literacy on retirement planning using 120 Indonesian lecturers, who found positive results between financial literacy and retirement planning. Therefore, individuals with positive attitudes towards retirement products are more likely to participate in retirement schemes.

South Africa is a developing country, with unique features on retirement planning systems and regulations. However, there is a necessity to examine the factors that improve retirement planning in the mining sector. As such, this study seeks to achieve the following research objective: to examine the mineworkers’ perspectives towards participating in retirement planning. Therefore, the research questions are as follows:

• What are the demographical factors that influence mineworkers’ participation in retirement planning?
• What are the mineworker’s perspectives towards participating in retirement planning?

Against the aforementioned background, this study focuses on South African mineworkers. Despite the high level of GDP contributions, the mining industry still portrays some concerns that necessitate the investigation of the knowledge and understanding of retirement planning in the industry. This study provides insights into how different variables, such as employment status, age, and marital status, impact participation in retirement planning. The introduction of financial education at an early stage of life and employment status could enhance contribution to retirement funds and improve financial decision-making. This paper contributes to the continuous discussions on retirement planning by illustrating the factors that give rise to retirement planning in the mining industry. The rest of the research paper is organised as follows: The next section provides the literature review that underpins the study, followed by the appropriate methodology utilised to address the objective. Furthermore, the next sections provide the data findings and discussion. The research paper concludes with a summary, policy recommendations, and suggestions for further research.

2. Literature Review

2.1. Introduction

The literature review below gives a detailed review of the factors determining the participation of individuals in retirement planning. It starts with the theoretical literature that underpins the study and is followed by the empirical literature review.

2.2. Theoretical Literature

The most relevant theories for the current study are the modern theory, life cycle theory, and economic theory. According to Dhlembeu et al. (2022), the modern theory of savings and consumption behaviours arises from Modigliani’s life cycle and Milton Friedman’s permanent income hypotheses. Modigliani (1986) observed that the life cycle hypothesis relates to savings behaviour during various stages in the lifetime of an individual which includes years of schooling, working, and time of retirement. Furthermore, it is hypothesised that the utility maximisation of individuals only acts to allocate resources to smooth their spending consumption and vary their savings over their lifetime to achieve their desired goals.

Modigliani and Brumberg developed the life cycle hypothesis in the 1950s which states that individuals plan for their lifetime using their income, they save during their working years, and they stop saving at retirement. According to Reyers et al. (2015), the theory presumes that people are financially literate, acceptable decision-makers who plan for retirement during their working days. Therefore, to have better consumption during retirement, individuals need to know their income and desired consumption during retirement. However, such arithmetic concepts can be well understood by literate individuals compared to their illiterate counterparts (Reyers et al., 2015).

Precious and Asrat (2014) found that life cycle theory affirms that the marginal propensity to consume should be exceedingly small when it is out of wealth, and spending out of collected wealth is spread over the remaining years of an individual’s life. Therefore, unnecessary spending on collected savings can have a detrimental impact after retirement as individuals will rely solely on their savings. The effect of wealth on individuals’ savings under this theory is assumed to be significant. According to Hubmer et al. (2021) and Hastings and Mitchell (2020), most individuals or households have little wealth or savings. Therefore, their savings and spending consumption behaviour are directly influenced by their income. Earlier studies such as those by Feldstein (1976) and Carroll (1997) found that the savings theory has further deviated from the life cycle hypothesis by suggesting that individuals not only save to finance and sustain consumption spending after retirement but also save for unforeseen situations whereby savings are used as security for such situations.

The economic theory has defined the savings of individuals as a behaviour in which part of their income has not been used or placed for spending purposes (Chudzian et al., 2015). However, the definition is considerably static as it only pertains to the income to be consumed in the future but does not include decisions on savings and motives of savings that are subjected to the analysis of the psychological theory of savings (Zaleśkiewicz, 2011). In addition, Nwosu et al. (2020) assert that young adults who are engaged in their studies (education) and starting their own families are perceived as people who do not save; the middle-aged working individuals are net savers, and the retired population are perceived as non-savers.

Modigliani (1986) emphasised that income growth significantly impacts individuals’ savings. The emphasis is derived from the importance of individuals’ income, as well as the relationship between individuals’ current and permanent income (Cervizli, 2022). Modigliani (1986) further highlights that the growth rate would increase the individuals’ income in the working phase when compared to individuals not earning income, and it would increase their level of savings.

2.3. Empirical Literature

Retirement planning shortcomings in the mining sector came to the fore after the FSCA reported approximately ZAR 47 billion in unclaimed retirement benefits (FSCA, 2021) and prior studies found that mineworkers are illiterate (Brown, 1988; Imbun, 2000; Raghav et al., 2020; Naveed & Ali, 2021).

Prior studies showed the importance of demographical factors that affect retirement planning such as age, income gender and educational level (Ketkaew et al., 2019; Vivel-Búa et al., 2019; Dhlembeu et al., 2022). Putra et al. (2024) sampled 67 civil servants of the Mataram University rectorate in Indonesia and employed a partial least square structural equation model to analyse data. The study found a positive link between financial literacy and old-age financial planning among the participants. The results imply that an increase in the level of financial literacy enhances the retirement planning activities of participants.

Saber (2023) investigated the relationship between the age of participants and retirement planning behaviour in the Saudi Arabian workforce. The study sampled 500 employees, employed the univariate analysis, and found a strong positive association between age and retirement planning. Zom (2024) examined the nexus between socioeconomic characteristics and retirement planning by sampling 321 Nigerian civil servants. The study employed descriptive statistics and found a positive relationship between age and retirement planning. Similarly, Harlow and Brown (2016) found that age had a positive influence on retirement planning. Tan and Wu (2024) investigated the impact of education level, income, and age on retirement planning behaviours of 25,000 American participants. The study employed the probit regression model and found that education level, income, and age have a heterogeneous impact via subjective financial literacy on the retirement planning of participants. Furthermore, the study found that financial literacy positively influences financial behaviours. A much earlier study conducted by DeVaney (1995) found that the perspectives and attitudes of people about retirement change due to age. Therefore, older people tend to be more willing to plan for their retirement than their young counterparts. The contradictory findings in the literature could be the attributes of different sample sizes and the measurement of retirement planning.

In contrast to the findings, Gutura and Chisasa (2024) examined the association between retirement planning and age of participants in Randburg (South Africa). The study randomly sampled 269 informal sector traders and employed multiple regression analysis. The study found a negative and insignificant relationship between age and retirement planning. Kofarmata and Adhama (2024) adopted the multinomial logic model and sampled 1350 in Nigeria and found an insignificant link between retirement planning and the age of respondents. An earlier study by Agabalinda and Isoh (2020) found that age was not a key factor in driving retirement preparations. Prior studies revealed conflicting results between the age of respondents and retirement planning.

Based on the theoretical and empirical literature, this study tests the following hypothesis:

H1. 

There is a positive correlation between age and participation in retirement planning.

Sarpong-Kumankoma (2023) examined the influence of marital status on retirement planning in Ghana. The study employed the probit model and found an insignificant relationship between retirement planning and marital status. The results are in line with Sharpe (2021), who reviewed 15 empirical research articles from 2010 to 2019 published in the Journal of Family and Economic Issues and found an insignificant association between marital status and retirement planning. Kim et al. (2024) employed the multivariate regression analysis sampling 2657 respondents and found an insignificant relationship between marital status and retirement planning. Contrary to the findings Saber (2023) found a positive link between marital status and retirement planning. While Gutura and Chisasa (2024) found a negative and significant relationship between marital status and retirement planning. Prior studies revealed conflicting results between marital status and retirement planning.

Based on the theoretical and empirical literature, this study tests the following hypothesis:

H2. 

There is a positive correlation between marital status and participation in retirement planning.

Hlabati (2020) conducted a study to investigate the correlation between retirement planning and employment status. The study employed the logistic regression model, sampled 2075 participants, and found a positive link between retirement planning and employment status. Moure (2016) assessed the correlation between retirement planning and employment status in Chile and employed the dynamic regression model. The study found a positive relationship between employment status and retirement planning. This is in line with earlier studies by Githui and Ngare (2014), who found that employed Kenyan participants were planning for retirement, and Van Rooij et al. (2011), who found a positive and significant correlation between employment status and retirement planning in the Netherlands. Their results imply that employment status significantly affects the respondent’s retirement planning.

Based on the theoretical and empirical literature, this study tests the following hypothesis:

H3. 

There is a positive correlation between employment status and participation in retirement planning.

The current study has several contributions to the relevant literature. The study augments the research on the effects of retirement planning. Prior studies investigated the varying levels of financial planning and financial literacy among different populations (Grohmann et al., 2015; Lusardi & Mitchell, 2017; Niu et al., 2020). Meanwhile, this study contributes to the literature by examining the mineworkers’ perspectives towards participating in retirement planning. These perspectives supplement and enhance the insights in retirement planning. The current study focuses on an underexamined sector, as prior studies focused on the retirement planning of high-income workers (Kajauchire, 2015; Niu et al., 2020; Tomar et al., 2021). Chelliah et al. (2022) focused on low-income workers in Malaysia. Meanwhile, Qian et al. (2024) focused on financial literacy and retirement planning in rural China. However, the current study only focused on the mining sector.

3. Methodology

The methodology section presents the population, sampling, and data collection of the study. Furthermore, this section presents the questionnaire design and data analysis when the study employed the logistic regression model and examines the performance of the Hosmer–Lemeshow test and Chi-square to determine the significance of the results.

3.1. Population, Sampling, and Data Collection

The study employed a positivist research paradigm, and it is a quantitative research design. The target population of the study consisted of the mineworkers in the chosen mining company. The study employed a non-probability sampling technique, namely, convenience sampling. The sampling method allows the researcher to select participants based on those who are easiest to reach. Cooper et al. (2018) described a sample size as the minimum number of respondents required from the population to conduct a study. The selection continued until the desired sample size was reached (Bell et al., 2022). The sample consisted of only mineworkers employed by the mine, and they received no compensation to participate in the study. The Raosoft sample size estimator was used to determine the sample size, based on the population of 310 mineworkers. The recommended sample size was 172, based on a confidence level of 95% and a margin of error of 5%. This study is dependent on primary data collection from the participants. The data were collected from the respondents to describe the individuals’ perspectives on retirement planning.

3.2. Questionnaire Design

The study used a closed-ended questionnaire, accompanied by a cover letter and consent form. Section A collected the demographical information such as age, gender qualifications, and ethnicity. Section B collected data on the respondent’s perspectives on retirement planning. Section A used a nominal scale while Section B used a four-point Likert scale where respondents rated each statement from one to four. Participants could simply tick the appropriate box from the range of answers that were provided for each question. As recommended by Brace (2018), the questions should be simple and short to understand and produce a high response rate. To ensure a comprehensive understanding of participating in retirement planning, the approximate time to complete the questionnaire was 12 min.

3.3. Measurement of Variables

Table 1. Measurement of variables.

Variables	Measurements
Dependent variable
Retirement planning	Contributing to the retirement funds
Independent variables
Age	Selecting the age ranges (18–25, 26–35, or 36+ years)
Marital status	Selecting the applicable marital status (married, divorced, single, or other)
Employment status	Selecting the applicable employment status (permanent, fixed term, or casual)

The variables in Table 1 were measured employing the scales established from prior studies. Source: own composition.

below provides the measurement of dependent and independent variables.

3.4. Data Analysis

Content analysis construct validity was used for all statements in the questionnaire measuring financial literacy and retirement planning. To assess content validity, a pilot study was conducted with 14 experts in the field of finance to confirm whether the statements measured individuals’ retirement planning. Statistical Package for the Social Sciences (SPSS) was used in performing the descriptive and inferential statistics. The statistical analyses of the data were conducted on the collected data from the respondents. The Hosmer–Lemeshow test was employed to evaluate the fit of the logistic regression. To analyse and process the biographical information, descriptive statistics were used; meanwhile, inferential statistics were employed to test the significance level of the respondent’s answered questionnaire. Cooper et al. (2018) posit that descriptive statistics entail the summary of the collected data according to a specific level of measurement. It describes the population or sample’s characteristics, therefore predicting outputs that are based on the inputs. To present the data, descriptive analysis (frequencies) was used. Inferential statistics draw conclusions about a population through the sampled data (Cooper et al., 2018). The Chi-square test was employed to determine the correlation between variables. The main purpose of the statistical analysis was to investigate the individuals’ perspectives towards participating in the retirement funds in the mining sector.

According to Easterby-Smith et al. (2021), an underlying factor that is 0.5 or greater has significant evidence of validity. The study employed the Cronbach alpha tests to determine the reliability of the retirement planning construct; according to Kennedy (2022), a value of 0.7 and above is considered reliable.

Table 2. Reliability statistics.

Reliability Statistics
Cronbach’s Alpha	Cronbach’s Alpha Based on Standardised Items	N of Items
0.871	0.868	7

Source: own composition.

presents the reliability statistics. It shows the item reliability and Cronbach’s alpha coefficient when evaluating the individuals’ perspectives towards participating in retirement funds. On the questionnaire (Appendix A), this section comprised seven items. The reliability of the construct to establish the participants’ perspectives towards participating in the retirement funds has an overall value of 0.871 for Cronbach’s alpha.

Table 3. Hosmer–Lemeshow test.

Hosmer–Lemeshow Test
Step	Chi-Square	df	Sig
1	7.325	8	0.502
2	5.320	8	0.723
3	5.049	8	0.752
4	3.479	8	0.901

Source: own composition.

demonstrates that the Hosmer–Lemeshow test is the alternative model to the Chi-square test. The Hosmer–Lemeshow test divides subjects into 10 groups and compares the number in each group (observed) to the number predicted by the model. The groups are created based on their estimated probability; those below the probability of 0.1 form one group, and so on, until groups with the probability of 0.9 to 1.0 are formed. These categories are further divided into two groups based on their actual observed outcome variables. If the Hosmer–Lemeshow goodness-of-fit test statistic is greater than 0.05, as we need for well-fitting models, we cannot reject the null hypothesis. This will imply that the model in question fits the data at an acceptable level. Our Hosmer–Lemeshow test statistic has a significance of 0.901, which means it is not statistically significant and, as such, our model is a good fit.

From

Table 4. Model summary.

Step	−2 Log Likelihood	Cox and Snell R-Square	Nagelkerke R-Square
1	90.356 a	0.463	0.633
2	90.367 a	0.463	0.633
3	91.076 a	0.460	0.629
4	94.251 a	0.447	0.610

Source: own composition. ^a Estimation terminated at iteration number 6 because parameter estimates changed by less than 0.001.

, Cox and Snell’s R-Square tries to simulate multiple R-Squares based on ‘likelihood’, but its maximum is less than 1.0, making it difficult to interpret. The results show that the logistic model explains 44.7% of the change in the dependent variable. The Negelkerke R-Square, which ranges from zero to one, is a more reliable measure of the relationship. Cox and Snell measures will normally be lower than Nagelkerke’s R². Nagelkerke’s R² forms part of the SPSS output in the model summary table. In this study, Nagelkerke’s R² is 0.610, showing a moderately strong relationship of 61% between the predictors and the prediction in the study.

The Equation

Logistic regression predicts the probability of an individual belonging to a particular group. In our case, it is the group where participants contribute to retirement funds.

$$p={\displaystyle \frac{{\mathrm{e}\mathrm{x}\mathrm{p} (\beta }_0{+\beta }_1{x}_1{+\beta }_2{x}_2+\dots +{\beta }_q{x}_q)}{{1+\mathrm{e}\mathrm{x}\mathrm{p} (\beta }_0{+\beta }_1{x}_1{+\beta }_2{x}_2+\dots +{\beta }_q{x}_q}}$$

$${\beta }_0=Constant/intercept$$

$${\beta }_1\to {\beta }_q=Coefficientsforpredictors$$

Here,

• p = the probability that a case is in a particular category;
• e = the base of natural logarithms;
• a = the constant of the equation;
• b = the coefficient of the predictor variables.

In

Table 5. Variables in the equation.

	N	B (Coeff)	S. E	Wald (1 Statistics)	df	Sig	Exp (B)	95% C.I. for EXP(B) Lower Upper
Age	133			5.055	2	0.080
Age (1)	133	0.084	0.688	0.015	1	0.903	1.087	0.282	4.189
Age (2)	133	−1.240	0.653	3.602	1	0.058	0.289	0.080	1.041
ES (1)	133	3.953	0.679	33.847	1	0.000	52.080	13.751	197.242
MS	133			8.776	3	0.032
MS (1)	133	−0.962	−0.959	1.007	1	0.316	0.382	0.058	2.502
MS (2)	133	0.947	0.938	1.020	1	0.313	2.578	0.410	16.21
MS (3)	133	0.733	1.036	0.500	1	0.479	2.081	0.273	15.85
Constant	133	-0.639	0.969	0.434	1	0.510	0.528

ES = employment status; MS = marital status.

, only three variables are significant in step 4. In step 1, gender was eliminated with a significance of 0.916. In step 2, financial education was eliminated with a 0.404 significance level. Education was eliminated in step 3 with a significance of 0.265. In this case, we note that age is significant at 10% and contributed significantly to the prediction (p = 0.08). Employment contributes at the significance level of 0.000 with a significance of 5% and, marital status contributes at a significance of 0.032.

4. Results and Discussion of Findings

4.1. Response Rate

Cooper et al. (2018), describe the response rate as the extent of the population or sample’s respondent representation. If a high response rate is achieved on the distributed questionnaire, there will be a lesser chance of significant response bias than compared to a low response rate. Out of the targeted population of 310, the respondents returned only 172 questionnaires, out of which only 133 questionnaires (which constitute 43% of the target population) were completed correctly, fully, and therefore usable for analysis. However, Bell et al. (2022) assert that heterogeneity and homogeneity of the population should be considered to the extent that a homogeneous population requires a smaller sample size than a heterogeneous population. This contention, therefore, supports the 43% response rate in this study, as the population was homogeneous, and the sample size is considered credible and representative of the entire population.

4.2. Chi-Square Test of Association

As shown in

Table 6. The significance level of the relationship between age, marital status, and employment and contributing to retirement funds.

		Contribution						Pearson Chi-Square Tests
		No		Yes		Total
		N	%	N	%	N	%	Chi-Square	df	p-Value
Age	18–25 years	14	43.8	18	56.3	32	100	7.932	2	0.019 *
	26–35 years	22	48.9	23	51.1	45	100
	36+ years	13	23.2	43	76.8	56	100
Marital status	Other	38	45.2	46	54.8	84	100	6.907	1	0.009 *
	Married	11	22.4	38	77.6	49	100
Employment status	Permanent	4	5.8	65	94.2	69	100	59.393	1	0.000 *
	Other	45	70.3	19	29.7	64	100

Source: own composition. * The Chi-square statistic is significant at the 0.05 level.

, the Pearson Chi-square was performed to determine if there is a relationship between age, marital status, and employment status and contributing to retirement funds. The Chi-square test reveals that the age group of 36+ years is 76.8% more likely to contribute to the retirement funds, followed by participants in the age group of 18–25 years, who are 56.3% more likely to contribute as compared to the age group of 26–35 years, which is 51.1% more likely to contribute to the retirement funds. As people or employees get older, their retirement planning behaviours increase significantly. Regarding the relationship between age groups and contributions to the retirement funds, we reject the null hypothesis at the 0.05 level which indicates that there is no association between age and contributions to the retirement funds. The p-value is 0.019, indicating that the significance level is < 0.05. These findings imply that there is a significant positive association between age and contribution to retirement funds. These results are consistent with Saber (2023) and Zom (2024) who found a positive significant impact of age on retirement planning. However, this is inconsistent with Kofarmata and Adhama (2024) who found an insignificant link between age and retirement planning.

In the Chi-square test, the results reveal that married respondents are 77.6% more likely to contribute to the retirement funds while the other groups, consisting of divorced, widowed, and single respondents, are 54.8% more likely to contribute to the retirement funds. Regarding, the association between marital status and contribution to retirement funds, we reject the null hypothesis at a p-value of 0.009, indicating that the significance level is <0.05 and that there is no association between marital status and contribution to retirement funds. The result implies that there is a positive and significant relationship between marital status and contribution to retirement funds. The result is consistent with Afthanorhan et al. (2020) and Saber (2023) who found that marital status has a positive influence on retirement planning. However, this is inconsistent with the findings of Gutura and Chisasa (2024) who found a negative relationship between marital status and retirement planning. Furthermore, this is inconsistent with Sharpe (2021) and Zom (2024) who found no significant correlation between marital status and retirement planning. As far as employment status is concerned, we reject the null hypothesis that there is no relationship between employment status and contributions to retirement funds owing to the p-value of 0.000, indicating a significance level of <0.05. Permanently employed participants are 94.2% more likely to contribute to the retirement funds than those who are fixed-term and casually employed, represented by “other” showing 29.7%. The findings suggest there is a significantly positive relationship between employment status and contribution to retirement funds. The finding is consistent with Van Rooij et al. (2011), who found a positive relationship between employment status and retirement planning.

4.3. Descriptive Analysis of Respondents’ Biographical Information

Of the total number of 133 mineworkers who responded, from the age of 56 to 65 years, the frequency is at equilibrium, with 50% of participants contributing and 50% not contributing. However, at 66 years and older, the frequency showed 100% contribution to the retirement funds with a single participant. This finding is consistent with the observation by Kalmi and Ruuskanen (2017) that older and middle-aged people contribute more to their retirement planning. These results are also corroborated by earlier theoretical models that are based on human capital accumulation, which shows that age is directly linked to retirement planning (Jappelli & Padula, 2013). However, this is inconsistent with Afthanorhan et al. (2020), who found that younger employees are sceptical about retirement planning as it is perceived as a burden since it is long-term planning.

The category with high frequency was the married and divorced participants. This implies a positive relationship between the marital status of a participant and their contribution to the retirement funds. In the single category, the contributors come third when compared to the married and divorced participants. Njoroge et al. (2024) found a positive correlation between marital satisfaction and retirement planning aspects, such as financial preparedness. Le Blanc et al. (2014) assert that the “linked lives” of individuals postulate an influence on their household circumstances. This is the reason the married and divorced category in the marital status has shown to perform much higher than the other categories. Similar results were found by Radl and Himmelreicher (2015) who found that family circumstances may influence and shape the individual’s retirement decision due to changed income prospects. It should be further noted that family circumstances influence household finances and their economic retirement incentives. Kaur and Hassan (2018) support the demographical factors (age, income, and gender) influencing participants’ retirement planning.

Permanent employees were at an overwhelming majority of 94.2% and showed dominance when compared to the non-contributors of 5.8%. Fixed-term contract employees were dominated by non-contributors with 60% and 40% for contributors. Part-time or casual employees show high dominance of non-contributors with 94.7%; meanwhile, contributors had 5.3%. Solem et al. (2014) found similar results that permanent employees are obliged to take compulsory retirement funds in their workplace, therefore allowing them to enjoy the benefits at a later stage. Fixed-term and casual employees may take up the optional retirement planning. The majority of non-contributors are in casual employment followed by 60% of fixed-term contract participants.

I am satisfied with my involvement in the retirement funds.

From the results in

Table 7. Participants’ perspectives towards participating in the retirement funds.

	Agree		Neutral		Disagree		Total
	N	%	N	%	N	%	N	%
I am satisfied with my involvement in the retirement funds.	69	51.9	15	11.3	49	36.8	133	100
I will be financially independent when I retire.	65	48.9	17	12.8	51	38.3	133	100
I can make a sound financial decision.	64	48.1	37	27.8	32	24.1	133	100
My age influences my involvement in retirement funds.	64	48.1	28	21.1	41	30.8	133	100
My marital status influences why I participate in retirement funds.	54	40.6	28	21.1	51	38.3	133	100
My employment status influences why I participate in retirement funds.	77	57.9	25	18.8	31	23.3	133	100

Source: own composition.

, only 51.9% of the respondents are satisfied with their involvement. The results are supported by Topa et al. (2018), who found that prior studies concur that an individual’s involvement in retirement planning is significant to all people. Rita et al. (2024) found that most respondents were satisfied with their current financial planning using Indonesian sandwich-generation employees between the ages of 25 to 55 years. Of the total number of 133 participants in this study, only 11.3% could not decide whether they were satisfied with their involvement in the retirement funds. However, 36.8% of the respondents are not satisfied with their involvement in retirement funds. There could be factors affecting their satisfaction with their contribution. Shariff and Ishak (2021) found retirement planning, financial literacy, liquidity preference of individuals, and level of income as some factors affecting the contributions of participants. However, the majority of the respondents agree that they are satisfied with their involvement. A study conducted by Talib and Manaf (2017) found that a majority of their respondents had a positive attitude towards retirement and were satisfied with their retirement planning. Igbozuruike and Eboh (2023) recommend that consulting experts on retirement planning is a measure that can enhance involvement in retirement planning.

I will be financially independent when I retire.

In Table 7, only 48.9% of the respondents feel that they will be financially independent when they retire. Consistent with the results, Snyman et al. (2017) found that respondents feel that they will be financially secure after retirement. Janetius and Singh (2023) argued that to become financially independent, financial planning is a requirement for individuals. Some participants who were not decisive of the question and selected neutral amounted to 12.8%. Those respondents are not sure whether they will be financially independent after retirement when they consider their current contribution to the retirement fund. Dhlembeu et al. (2022) reported that only 10% of South Africans will be financially independent after retirement; however, retired people’s lack of financial independence becomes a problem in the country. Furthermore, the lack of financial preparedness increases the fear of retirement. Only 38.3% of the respondents disagree that they will be financially independent after retirement. Inkinen (2024) found that the majority of their respondents were not confident that they would be financially independent after retirement. Similar results are found by Janetius and Singh (2023), who found that some people continue to work after retirement age due to financial commitments and not having sufficient funds to retire comfortably.

I can make a sound financial decision.

Table 7 above shows that it was clear from the majority of the responses that 48.1% showed that they can manage or make sound financial decisions. Taj and Shah (2022) found that sound financial decisions are associated with financial security, growth, and success, while poor financial decisions can have disastrous repercussions. On the contrary, 27.8% of participants were not certain whether they could make sound financial decisions and the remaining 24.1% disagreed with the question. Therefore, they cannot make sound decisions when it comes to their financial planning. According to Hammond et al. (2017), sound financial management is significant for individuals at all stages of life; nevertheless, sound financial decision-making can be challenging as people age.

My age influences my involvement in retirement funds.

The results in Table 7 show that 48.1% of the respondents showed that age influences their participation in retirement funds. The results imply that the age of the respondents has a significant impact on their involvement in retirement funds. As people grow older, they start to prepare for their retirement date financially. These results are consistent with Muthia et al. (2021) and Gutura and Chisasa (2024), who found that the ability to plan for retirement was influenced by the respondent’s age. However, 21.1% were not certain whether their contribution to retirement funds was influenced by their age. The results are inconsistent with Ye et al. (2022) who found that the age of participants leads to a higher probability of financial planning. A total of 30.8% of the respondents disagreed with the question. Respondents noted that age had no impact on their participation in the retirement funds. The reason could be that it is mandatory to contribute to the retirement funds with the employer if you are permanently employed regardless of their age.

My marital status influences why I participate in the retirement funds.

The results in Table 7 show that 40.6% of the respondents agreed that their participation level in retirement funds is influenced by their marital status, while 21.1% were not certain of their choice and selected neutral. In addition, 38.3% disagreed that their marital status has an impact on their participation in retirement funds. It might be that they only take part in the retirement funds that are compulsory at work and therefore do not have any other form of savings or retirement funds managed outside of work. These results imply that marital status influences respondents to consider retirement planning. The results revealed that married couples were more prepared for retirement than their unmarried counterparts. The results suggest that married couples may be forced to save due to children who will need to be cared for or when an unfortunate event happens. The results are consistent with Afthanorhan et al. (2020) and Gutura and Chisasa (2024) who found that spousal influence was a stronger component of retirement decision-making.

Furthermore, 21.1% of the respondents were not certain whether they agree or disagree that their marital status influences their retirement planning. These results imply that respondents are not sure whether their marital status has an impact on their retirement fund contribution. The remaining 38.3% disagree that marital status has an impact on their participation in retirement funds. The results are supported by Sarpong-Kumankoma (2023) who found that marital status was irrelevant for retirement planning. This may be because marriage cannot affect eligibility for employer-sponsored retirement plans; however, employees could make beneficiary arrangements for their spouses.

My employment status influences why I participate in the retirement funds.

In Table 7, the majority of the respondents (57.9%) showed that they agree that their employment status influences their participation in retirement funds. The results imply that the employment status of the respondents had a significant impact on their involvement in retirement fund contributions. When respondents are employed, their retirement planning behaviour increases; however, this was influenced by the status of employment. These findings are consistent with Moure (2016) and Hlabati (2020). However, 18.8% of the respondents were not certain whether their employment status influenced their participation in the retirement funds. A total of 23.3% of the respondents disagree that their employment status had an impact on their participation in the retirement funds. Contradictory to the results, McDonald (2017) found that an employee’s retirement planning is beneficial to individuals who plan accordingly to maintain their standard of life after retirement.

4.4. Logistic Regression Analysis

For robustness, this study employed a logistic regression analysis to determine the relationship between the independent variables (age, marital status, and employment) and the dependent variable (retirement planning) presented in

Table 8. Regression statistics.

R	R2	Adj R2	Std. Error of the Estimate	Obs
0.684	0.467	0.455	0.846	133

Source: own composition.

. The R-Square (coefficient of determination) shows that independent variables contribute 46.73% towards retirement planning. The table shows the standard error of the estimated regression, the adjusted regression square, the regression square, and the regression level.

Table 9. ANOVA.

	df	SS	Ms	F	Sig.
Regression	3	81.057	27.019	37.727	0.000
Residual	129	92.386	0.716
Total	133	173.444

SS = sum of squares; Ms = mean square. Source: own composition.

provides the ANOVA results to establish whether the means of all the groups are the same. If the results show any differences in the means, it implies that retirement planning depends on the independent variables. The F-statistics of 37.727 (p = 0.000) indicate that the model is fit to explain the correlation between variables.

In

Table 10. Coefficients.

	Coefficients	Standard Error	t Stat	p-Value	Lower 95%	Upper 95%
Intercept	0.626	0.185	3.318	0.000	0.992	0.259
Age	0.301	0.082	3.678	0.000	0.139	0.463
MS	0.194	0.088	2.204	0.029	0.0199	0.369
ES	0.225	0.078	2.881	0.005	0.0705	0.380

Collinearity statistics: tolerance = 1.000; variance inflation factor (VIF) = 1.000. MS = marital status; ES = employment status. Source: own composition.

, the coefficient of age (0.301), with a t-value of 3.678, indicates that the independent variable (age) has a statistically significant and positive (p = 0.000) correlation with retirement planning. These results imply that age contributes about 30% to determining retirement planning. Marital status has a coefficient of 0.194, with a t-value of 2.204. These results imply that marital status contributes around 19% to determining retirement planning. Marital status has a statistically significant and positive (p = 0.029) relationship with retirement planning. Employment status has a coefficient of 0.225, with a t-value of 2.881. These results imply that employment status contributes about 23% to determining retirement planning. Employment status has a statistically significant and positive (p = 0.005) relationship with retirement planning. These results presented an acceptable value of 1; therefore, there were no problems with VIF and tolerance. The data were evenly distributed, with no issues of skewness and Kurtosis.

5. Conclusions and Recommendations

This study was restricted to the selected mining company in South Africa, Gauteng province, to investigate the individual’s perspectives towards participating in retirement planning. SPSS was used in performing the descriptive and inferential statistics. The study employed the logistic regression model and the Hosmer–Lemeshow test was performed to evaluate the fit of the logistic regression and the Chi-square data to determine the significance of the results. The key findings are that retirement planning is underpinned by age, marital status, and employment status of the mineworkers in the selected company. Therefore, it implies that age, marital status, and employment status have a positive relationship with retirement planning. These three aspects are derived from the demographical factors of mineworkers. The practical implication of this paper informs the companies, policymakers, and government to prioritise awareness of retirement planning based on demographical factors such as age, marital status, and employment status to appropriately prepare mineworkers for retirement. Retirement planning programmes are necessary for employees based on their age group, marital status, and employment status to increase their participation in retirement funds. Furthermore, targeted programmes should be developed for younger mineworkers to enhance their understanding of retirement planning. Continued advocacy for increased understanding and awareness not only by employers but also by retirement investment companies and policymakers could increase mineworkers’ confidence in retirement planning. Furthermore, awareness of external retirement schemes offered by private companies to supplement their employer-sponsored retirement savings should be increased. Policymakers should establish retirement planning awareness guaranteeing that everyone can pre-commit to a regular payroll savings plan with help through tax or monetary incentives.

For further research, researchers may consider the role of family background to better understand the retirement planning of mining workers. This is necessary due to the fact that the crisis also affects the family members; hence, a study on this would enhance the literature on the retirement fund participation of mineworkers. Researchers could compare the pre- and post-retirement financial standards of living of mineworkers.