Moderating Effect of Gender, Income and Age towards an Islamic Health Protection Retirement Plan (i-HPRP)

This study examines the moderating effect of demographic variables like gender, income, and age on the proposed Islamic Health Protection Retirement Plan Model (i-HPRP). The study adopts a quantitative method by surveying 498 respondents located in different states in Malaysia. Structural Equation Modelling (SEM) is adopted in this study to measure the moderating effect of demographic variables. The analysis results showed that saving behavior has significantly influenced the plan's preferences while the spending behavior did not. Besides, demographic variables play a role as a moderator in the relationship between saving and the retirement plan. Overall, this research findings reveal that gender, age, and income level were important in determining individual saving behavior and preparation for retirement. An effort to increase awareness of retirement preparation should be enforced on various age groups among citizens to reduce the problem of insufficient funds among retirees.  


INTRODUCTION
Retirement means that a person permanently ends his or her paid employment. According to Denton and Spencer (2009), retirement usually happens to employees aged between 50 to 60 years. However, the age of retirement is different from country to country. The issue of the awareness of retirement plans is still unclear among society in Malaysia. In Malaysia, current government policy makes it compulsory for all its working population to set a minimum amount of their salary into the Employee Provident Fund (EPF). This policy is significant to ensure that employees have enough savings to cover living maintenance after retirement. Since most think that companies and government retirement plans should be enough to cover golden years expenses. Nevertheless, many studies have shown that money saved in companies and government pension plans may not be adequate for future consumption. Based on consumption and saving decisions, a rational and forethought individual exploit an expected lifetime utility subject to the constraint of the flow of income and spending (Lusardi, 2008). In fact, household spending on recreation and transport significantly reduced while food consumption slightly increased during the Covid-19 crisis (ECB Economic Bulletin, 2020). Perpetually, Covid-19 restrictions have influenced the responsibility of individuals' saving to secure retirement financially and for medical purposes in Malaysia (Durgahyen, 2020). Linnet Lee (CEO of Financial Planning Association Malaysia) suggested that individuals should start planning for retirement, developing saving behaviour, and investing for their retirement from the perspective of income, health, security and long-term health services (Durgahyen,

Keywords
Saving behavior, spending behavior, income, gender, age, Islamic retirement plan 2020). Although current EPF (employees provident fund) withdrawals conceivably settle the current predicament, it will cause larger issues in the future, especially postretirement (Durgahyen, 2020). Idiosyncratically, individuals who spend more than they should, have less disposable income for savingless contribute towards a retirement product (The RinggitPlus, 2020). According to a survey conducted by the Asian Institute of Finance (2019), young Malaysian individuals (between 20-33 years) are stressing with financial constraints, many of them are having high costs, credit cards and borrowings (The Edge Financial Daily, 2015). Similarly, the RinggitPlus survey (2018) reported that 47% of Malaysians (aged 35 years and above) did not start saving for retirement, and an increasing percentage was reported among the younger generation (The RinggitPlus, 2020). In other words, young individuals are hindered with debt due to unhealthy spending behaviour. Furthermore, saving behaviour improved with age growth as the highest proportion was recorded between 27 to 33 years aged individuals (57%) (The Edge Financial Daily, 2015). Meanwhile, gender differences in retirement wealth ascend due to the different personal and job characteristics of men and women (Eleanor, 2021). It continues to emerge the patterns of retirement planning (Topa et al., 2018). Women devote fewer years in employment, are likely to work as part-time employees, their income lesser than men counterparts, and have greater longevity -have accumulated lesser assets and face greater retirement costs than men (Eleanor, 2021). The Malaysia Economic Monitor Report (2019) has reported that the younger generation's income growth was obstinately slothful than older employees (The World Bank, 2019). Further, most Malaysian argued that their income was not rising with the growing living cost; many young employees sought informal employment where mandatory savings plan unavailable (The World Bank, 2019). Yet, the young generation was found better in saving than expected as 64% of them save a minimum of 20% of their income monthly (The Edge Financial Daily, 2015). Instead, these findings exhibit a lower tendency for savings among young employees who ignore future prospects due to impulse spending (The RinggitPlus, 2020). All as above, it is crucial to identify an adequate and proven high-impact policy to form a shared wealth and develop individuals' access to investments, services and long-term growth opportunities (The World Bank, 2019). A mix of short-term evaluations and long term policies is demanded in order to alleviate the rising living cost (The World Bank, 2019). Whereby, short term evaluations should focus on strengthening social protection nets and, in the long term, grander cooperation or partnership among agencies and enactment of structural reforms to substitute better market competition and hasten productivity to boost real income (The World Bank, 2019). In these circumstances, this study designed a protracted Islamic health protection retirement plan (i-HPRP) to investigate the impact of employees' saving and spending behaviour on their preferences toward a retirement plan. Further, it examined the moderating effect of employees' demographic factors (gender, age, and income) on the demonstrated retirement product. The findings may shed light on the literature on retirement planning and the influential economic behaviour of individuals. Individuals would explore their preferences while practitioners would able to identify the consumers' preferences. Policymakers may find this model and findings useful during the reformation of existing retirement products for both public and private sector.

THEORETICAL BACKGROUND
Life-Cycle Theory (LCT) is an economic theory that pertains to people's spending and saving habits throughout a lifetime. The life-cycle approach offers a specific account of consumption and saving. Still, it is derived from fundamental underlying principles that could be used to extend the model to deal with a wide range of issues about consumption and saving, many of which had not been thought about in 1950 (Deaton, 2005;Modigliani & Brumberg, 1954). Accordingly, Modigliani and Brumberg's original formulation recognised that lifecycle planning requires people to look into an uncertain future. Consequently, it is challenging to formulate theoretically satisfactory and tractable models of how people behave in the face of uncertainty. Thus, this research concerned how individuals plan their spending and saving to prepare for retirement prospects.

Saving and Spending behaviour
People save too small for retirement (Lusardi, 1999), spend more (Sotiropoulos & d'Astous, 2013), purchase things and regret later. Accordingly, they do not make bills-payment timely (Abendroth & Diehl, 2006), thus make terrific financial decisions (Strömbäck et al., 2017). Nonetheless, individuals do not always make terrific financial decisions as some of them are more or less inclined to make bad financial decisions (Strömbäck et al., 2017). Saving is an instant expense with possible advantages in the future, while consumption allows the instant satisfaction of an individual's needs (Barboza, 2018). Therefore, positive saving behaviour forecasts better individual wellbeing (Shim et al., 2012). Saving attitude for retirement influence to prefer social models (Croy et al., 2010), and money management determine the perceived retirement model (Cude et al., 2006). However, unforeseen household expenses incurred against liquid savings can be held as insurance (Gathergood & Weber, 2014). Meanwhile, spending behaviour is mainly adopted from individuals' current-biased preferences, and they behave differently than future-oriented individuals who have a healthy saving attitude related to retirement planning (Brown & Previtero, 2015). Idiosyncratically, impulsive spending behaviour grounds various risks related to income shocks, unanticipated expenditures on durables (Gathergood, 2012). Followingly, such spending behaviour pledge to save later, less likely to hold a fixed amount every month, thus less likely to participate in retirement plans (Brown & Previtero, 2015). Indeed, individuals with impulsive spending behaviour suffer from financial problems over time (Gamst-Klaussen et al., 2019). More so, individuals who have higher tendencies in spending and impulse buying behaviour suffer from financial worries (Gardarsdóttir & Dittmar, 2012). An individual can associate spending behaviour with shops or online shopping using debit/credit card or other online payment accounts; in the meantime, saving or investment products not available during impulse spending (Gathergood & Weber, 2014). In other words, individuals do not care about any saving or investment retirement products during their impulsiveness spending, thus less likely to participate in a retirement plan. Researchers identified the negative relationship between spending behaviour (self-control problems) and financial products (accumulation of wealth) among individuals (Ameriks et al., 2007). Lack of self-control or impatience treated as behavioural biases among individuals (Gathergood, 2012), which was denoted as "intrapersonal decision time-inconsistency problem", a skirmish between manifold selves (Thaler & Shefrin, 1981). This study treated this uncontrolled or lack of self-control as individual spending behaviour. Past studies evidenced both significant and insignificant influence of saving behaviour towards retirement planning in the Malaysian context. For instances, Afthanorhan et al. (2020), Hassan et al. (2016), and Kadir et al. (2020) acknowledged a significant positive impact, while Moorthy et al. (2012) and Ibrahim et al. (2012) showed insignificantly poor https://journals.e-palli.com/home/index.php/ajebi Am. J. Econ. Bus. Innov. 1(1) 19-29, 2022 saving behaviour among Malaysians. On the other side, Haron et al. (2019) and Kimiyagahlam et al. (2019) stated saving behaviour towards retirement planning as a new concept for Malaysian. These findings exhibit inconsistent and unstable prediction of the influence of saving behaviour. Simultaneously, a scarce study was conducted on spending behaviour on a retirement plan and retirement planning behaviour preference.

Proposed Islamic Health Protection Retirement Plan (i-HPRP)
The life span theory enables humans to think over the future by expecting a better life without stress and financial disturbance. Notwithstanding, individuals prefer to plan for retirement through employers' contribution (Butt et al., 2018) due to loyalty and dedication towards the corporation. Nonetheless, several things are demanded by the employee in their retirement life, such as; food, medical services and housing (Worsley et al., 2011). Besides, cultural norms, rituals and living standards (Alaudin et al., 2016), values and perceptions (Gupta & Pradhan, 2017;Hershey et al., 2007), perspectives, race/ethnic, religious aspects, investment (Angel et al., 2014), and individuals' characteristics and preferences of employees (Butt et al., 2018; significantly affect the retirement plan. Indeed, Malaysia has demonstrated a significant part of the global Islamic finance market (ICD-Refinitiv, 2020). So, an effective retirement plan that considers cultural practices, economic behaviour and retirement demands and necessities . Meanwhile, providing the information and direction on saving for long-term investment may not be enough (Hardcastle, 2012) as individuals' saving behaviour relies on income and exchange effects in response to a new policy or scheme (Hardcastle, 2012). Further, lifecycle consumption entails a "consumption smoothing", whereby an individual should continue the consumption route in time (Hurd & Rohwedder, 2013). A rearrangement of consumption will increase lifetime utility if the path of consumption is not continuous (Hurd & Rohwedder, 2013). Meanwhile, Islamic banking and finance were unable to fulfil the needs for investment products, retirement plan products, and travelling needs products (Chowdhury et Yumna, 2019). Constantly, a reliable, executable and beneficial model should be available to attract consumers for retirement. As these consequences, this study demonstrated an Islamic health protection retirement plan (i-HPRP) by considering the social practices and retirement features. This study followed a similar approach to Lusardi et al. (2009), who designed planning support to simplify the retirement decision and encourage employees to make an active involvement. The designed planning support included several features; it breaks down the participation steps and required steps to participate online. Further, it provided other information such as savings, minimum amount of income as contribution, life cycle fund, and designed messages and pictures to encourage participants to save. During the working period, both employees and employers' contribution is demonstrated the accumulated savings and investment income separated to manage both retirement income and medical expenses upon designated retirement period. A special feature of this plan is that it is an Islamic-based retirement plan where its operation and framework would comply with the Islamic rule (Shariah). It is basically a retirement plan attach with the existing Takaful scheme offered by Takaful operators where all the operations are Shariah compliant. As Takaful based on Tabarru' (mutual donation) and Taawun (mutual assistance) principles, the proposed plan will be done using the same spirit. As an Islamic retirement product, the aim is not only to attain the worldy benefits, but the most important objective is to achieve rewards in the hereafter by take care of life, family, and property (financial) as encouraged by Islam. These three elements basically were the objective of Shariah (Islamic Law).

Mechanism of i-HPRP
An Islamic retirement product is designed to offer a stream of regular retirement income and medical benefits in exchange for a regular contribution during the working period, named a paid contribution. The contribution here is considered donation contributed by participants given an intention to help others who are in need during retirement age including themselves. This is a spirit of an Islamic insurance that is practiced currently via the principle of Tabarru' which means mutual donation. In order to show support from employer, the contribution might be subsidized by the employer through out the working tenor (by percentage) for the employees to enjoy the benefits once they retire. Among the responsibility of employer is to reduce burden and assist the employees even after they stop working with them. This actually shows the gratitude or appreciation of employer towards the employees. In addition, it offers peace of mind to the participant especially Muslim because the donated fund will be invested to only on Shariah compliant investment instrument. All investment pracrices will be closely monitored by the Central Bank to ensure its comply with the Shariah. Other than that, through the contract of Wakala, Takaful operator will manage the fund on behalf of the participants as well as responsible to distribute the fund for retirement income and medical benefits.

Demographic effect on retirement plan
Demographic factors form a binding effect on behavioural, psychological, and financial aspects. Demographic variables such as age, gender, and income interact with retirement planning and influential factors (Afthanorhan et al., 2020). Meanwhile, these factors determine the lifecycle dimensions (e.g., total income and saving prospects) (Eriksson & Hermansson, 2014). A study conducted by Loke (2017) confirmed the demographic (gender, income, age, etc.) effect on individuals' financial management behaviour in Malaysia. In details, consistent with a study by Falahati and Sabri (2015), the author found that gender differences would determine the individual's overall financial management level. On the other hand, gender https://journals.e-palli.com/home/index.php/ajebi Am. J. Econ. Bus. Innov. 1(1) 19-29, 2022 significantly revealed the paths difference from financial knowledge to public security and self-insurance, further from goals clarity to public security (Topa et al., 2018). Sabri and Zakaria (2015) found that females with better income (more than RM3,500) are financially stable than males. However, Burke et al. (2017) found that male with a higher income is prone to participate in different investment portfolios than the default employer's plan. Based on market observation and findings from past studies on gender effect on retirement or financial planning, it believed that gender would still play a significant role in retirement planning in this century. In general, women are more particular and concerned about wealth planning, especially for married women, where they would think about family financial planning (child education, saving for emergency, holiday, and others), including financial planning for retirement. Accordingly, the individuals' age group moderates the process to prepare a retirement plan (Mansor et al., 2015). Unlikely, Mokhtar et al. (2015) unveiled the insignificant effect of demographic factors on public sector employees' financial wellbeing (financial commitment, retirement savings, tendency). Nevertheless, income is a significant factor for retirement preparation among individuals (Lee et al., 2018). An individual usually does not start saving and plan for retirement at an early age of working life (Jiménez et al., 2019). Overwhelmingly, there is a selfmotivated character of protecting tendencies across adulthood; thus, age differences drive the retirement saving behaviour in the psychological biases (Rolison et al., 2017). Individuals with an increase in age and income likely to borrow at a young age to smooth consumption (Lusardi, 2008). Likewise, individuals with rich pensions do not need to accumulate extra savings to provide for retirement over the life cycle (Lusardi, 2008). Accordingly, lifetime resources (income, wealth), their distribution, and age play an important role in retirement saving decisions (Lusardi, 2008). Under these circumstances, this study argues that demographic factors, such as gender, income and age moderate individuals' economic behaviour to perceived Islamic health protection plan. Figure 1 exhibits the theoretical model for this study.

METHODOLOGY Data collection and analysis
The developed plan (i-HPRP) provided information regarding contribution, potential investment accounts and benefits based on a life cycle fund. The researchers formulated a survey instrument based on the designed plan, questioning the designed plan's preferences, saving, and spending behaviour overtly. These data collection methods are standard in marketing and appropriate for capturing the extensive heterogeneity in retirement saving decisions (Lusardi, 2008). This study developed the survey instruments by referring to past literature (H. Patton et al., 1995;Harun-Or-Rashid et al., 2011;MacFarland et al., 2003) and contextual modification in a focus group discussion of relevant academic researchers. The variables were modified based on the concept of proposed retirement model. The convenience sampling adopted, and the data collected online due to restrictions on movement caused by the pandemic crisis. Precisely, the data collected from all states in Malaysia by employing enumerators in different states. They shared the google form through social networks among their friends, relatives and mutual connections who are currently working between the age range 18-59. Upon the normality test and removal of outliers, a total sample of 498 selected for final data analysis. The sample size fulfilled the minimum sample requirement for the multivariate analysis . Structural equation modelling (SEM) was applied for data analysis upon undertaking the exploratory factor analysis (EFA) to confirm the grouping items for studied constructs. Further, confirmatory factor analysis (CFA) was used to estimate the model fitness, calculate the reliability and validity of measured items, followed by estimation and analysis of the theoretical model developed by this study. SEM is a multivariate technique; combines the factor analysis and multiple regression method to estimate a sequence of interconnected dependence relationships concurrently and governs the level to which a theorised model is reliable with the sample data (Byrne, 2010;Hair et al., 2006). Meanwhile, the CFA fit indices (RMSEA 1 < 0.08; GFI, CFI, TLI > 0.9. see Table 1) confirmed the well fit of the studied model.
Based on three reasons, SEM was adopted as the most   496 . 930 . 972 . 965 .07

Required
Upper is better <5 >.9 >.9 >.9 <.08 appropriate technique for this study. Firstly, it allows to estimate and specify models with the moderation effect among independent and dependent variables, enable to discourse a single relationship at a time by overcoming the limitation of several multivariate methods (Byrne, 2010). Secondly, it can decrease the measurement error and allow for multi-item scales, enabling the researcher to measure the relationship's strength among any two variables (Hair et al., 2006). Lastly, it allows defining and examining theoretical models, specifically, those that comprehend latent construct variables (Anderson & Gerbing, 1988).
The parameter critical ratio values and estimates in the CFA loaded higher than 0.5 at a significant level (p<0.05). Cronbach's alpha results (all values exceeded more than 0.7 suggested requirements by Garver & Mentzer's 1999) portrayed the measurement scales' internal consistency and reliability. Furthermore, the importance of the average variance explained (AVE) exhibited a standard value of ≥ 0.5 (Fornell & Larcker, 1981;Garver & Mentzer, 1999). Meanwhile, all variables were supported in terms of discriminant validity ( DV & AVE) for each construct surpassed the squared correlation amongst latent variables (Fornell & Larcker, 1981)). Table 2 exhibited the results of the reliability and validity test. Several techniques were introduced in SEM to perform the multigroup analysis, for instance, the heterogeneity test, user-defined estimand, critical ratios for difference tests, and pairwise deletion (Afthanorhan et al., 2020). This study used the heterogeneity test to examine the moderation effect of categorical variables (gender, income, and age) as it is imperious to identify the significant outcome on each group. The heterogeneity test procedure (Byrne, 2010), namely: chi-square difference, was used to examine the measurement invariance before conducting a moderating effect analysis. In other words, the metric invariance can be implied by measuring the chi-square values from constrained and unconstrained models (baseline). The difference predicts the further analysis for moderation effect. This study confirmed the configural invariance by assessing a large difference of chi-square (102.394) between the two models. The fulfilment of recommended fit indices (X2=chi-square, RMSEA, TLI, CFI, see Table 3) thus concluded the existence of moderation effect. Based on , when the difference of chi-square is higher than 3.84, the moderation effect is statistically significant.

Direct Path Analysis
According to the conceptual model, five hypotheses were developed; two paths indicated the direct path relationship while three with moderation effects. Based on the path estimates (Figure 2 &Table 4), H1 was supported. In other words, the relationship between saving behaviour and perceived Islamic retirement plan was positive as hypothesised. The path estimates reported the values of β 3 = 0.377, C. R. = 6.899, p-value = 0.001. Thus, it indicated the relationship statistically significant at a 99.99% confidence level. On the other side, H2 was not supported even the direction showed as hypothesised.
https://journals.e-palli.com/home/index.php/ajebi Am. J. Econ. Bus. Innov. 1(1) 19-29, 2022 In other words, the relationship between spending behaviour and perceived Islamic retirement plan was negative but unable to achieve the significance level. The path estimates displayed the β = -0.018, C. R. = -0.664, p-value = 0.507, lead to statistically insignificant. The next section discusses the effect of moderating variable on the relationship between independent and dependent variables. Among the five paths depicted in the conceptual model, three hypotheses are developed based on the moderation effect of categorical variables, namely, gender, income, and age, on the relationship of H1 (saving and perceived Islamic retirement plan) and H2 (spending and perceived Islamic retirement plan). These two relationships were examined for moderation effect of different gender (men and women), income (Low = ≤ RM 4,000 & High = > RM 4,000), and age (18-29, 30-39, and 40 & above). Table  5 displayed the results of moderation effects of each variable and hypothesis paths. The moderating role of gender exhibited a positive effect on saving behaviour and perceived Islamic retirement plan while negative for spending and perceived Islamic retirement plan. Based on the gender perspective, men are more intend to save for a perceived Islamic retirement plan compared to women, as the coefficients (β) were shown (Male>Female = 0.494 > 0.303). Thus, gender significantly moderated the relationship at a p-value of 0.001. However, gender did not manage to moderate the spending behaviour and perceived Islamic retirement plan.

Moderation Effects
On other hand, income group denoted the partial moderation effect on the relationship between saving and retirement plan (High>Low = 0.432 > 0.334), while full moderation effect on the spending and retirement plan (β = -0.096, p-value = 0.03). In other words, both income groups are interested in saving for the retirement plan, and the high-income group exhibited greater behaviour than the low income. In contrast, the lower-income group spends less on the retirement plan, and the relationship is statistically significant. Meanwhile, age differences exposed no moderation effect on the relationship between spending behaviour and perceived Islamic retirement plan. Hence, it showed the moderating effect on the relationship between saving behaviour and perceived Islamic retirement plan. Specifically, age group 18-29 and 30-39; both groups exhibited the interests in saving for the retirement plan, while younger group (18-29) shown the higher compared to young adults (30-39). The difference in coefficient (β=0.48 >β=0.442) assumed the hypotheses. On the other side, the age group 40 and above did not show significant (p-value = 0.09) saving behaviour towards the retirement plan.

DISCUSSION
According to the life-cycle consumption, saving is the disposal income that subtracted from income after consumption. While, human needs theory underpins that individuals spend for the basic needs for a comfortable life (Max-neef et al., 1987). Thereby, individuals may enhance the saving by reducing spending on needs to enjoy a comfortable and happy life. Based on Keynes (1936) and (Durgahyen, 2020), individuals intend to save more during the recession (like Covid-19) because they are concerned about surviving their future life. The finding of this study supports this assumption where there is a significant influence of current saving behaviour towards perceived Islamic retirement plan while at the same spending behaviour showed a negative direction.
In other words, individuals were more concerned about saving for future retirement and reduced spending on comfortable needs, especially during the difficult time of economic downturn. Based on a recent study conducted by Clair et al. (2021), Zhang (2021) and ECB Economic Bulletin (2020), individuals' behaviour have currently changed towards basic needs, and there were decreased in travelling and other expenses (like recreation and transportation) due to the Covid-19 crisis.
In addition, it is found that male employees unveiled higher saving behaviour than female employees to the preference for an Islamic health protection plan. This is in line with the past study conducted by Loke (2017). Am. J. Econ. Bus. Innov. 1(1) 19-29, 2022 On the other hand, spending behaviour remained a statistically insignificant factor in the retirement plan's preference. The rationale of these findings might be that male become more concerned about family financial planning due to the economic situation and restricted movement order. Hence, it is undeniable that employees with higher income have a higher tendency to save for a retirement plan for their future financial protection. In this context, they put more precautions to take care of future needs towards health and income preparation as they have a surplus to save. Unfortunately, those who lost the job and have income reduction have no option to increase their saving for retirement preparation. The low-income group might be more concerned to sustain their lives during the pandemic crisis. There were efforts taken by the government to face the challenging situation caused by Covid-19, which including Economic Stimulus Package, PRIHATIN package, and National Recovery Plan in order assist all sectors, especially the low-income group (B40), either in terms of direct cash assistance or deferment of monthly financing instalment (moratorium). The financial institutions have also taken the initiative to help the small and medium businesses to sustain their businesses with small financing with low-profit rate. It is hoped that with all assistance from the government and other parties may support the Malaysian to continue their saving behaviour even during a hard time. Surprisingly, the young generation has shown a higher saving behaviour compare to the older group of employees. This is consistent with the Asian Institute of Finance survey report (2015), which found a higher saving behaviour among Malaysian young individuals compared to the senior generation. However, this finding has contradicted a study conducted by Jiménez et al. (2019) from a different context. Nevertheless, the effect of different age groups was significant, in line with past studies, including Barnhoorn et al. (2016) and Ekici and Koydemir (2016). In details, mid-aged employees showed an expected significant tendency of saving towards retirement planning. Hence, it is assumed that the older groups might have prepared their retirement age as they were closed to retire, and the proposed plan seems not relevant to them even interesting. They might think it's too late to save in a retirement plan. In another finding, the age group did not show any significant differences over spending behaviour on the Islamic health protection plan's preference. Although the young and mid-aged generation unveiled a positive spending behaviour means they were not concerned about overspending. On the other hand, older employees started to spend less and more concern over future and financial behaviour. Overall. The findings have supported previous studies on the significant effect of demographic variables (gender, income and age) on retirement planning (Afthanorhan et al., 2020;Loke, 2017). Based on the above discussions and findings, it can be concluded that the proposed Islamic health retirement protection plan has the potential viability in order to  Econ. Bus. Innov. 1(1) 19-29, 2022 commercialise in the market. However, with government or employers' interventions, the plan will manage to get acceptability among employees within a short period.

CONCLUSION
The main purpose of this study is to investigate the moderating effect of selected demographic factors (gender, income, and age) on an illustrated Islamic health protection plan's preference, mainly financial behaviour (saving and spending). Based on the findings, saving behaviour has significantly influenced the employees' preference towards the i-HPRP, while spending behaviour had not significantly affected their preferences. Interestingly, demographic factors are found to have a moderating effect on the saving and retirement plan relationship effect. At the same time, income moderated both relationships, saving and spending behaviour on the i-HPRP preferences. Several limitations of this study are the proposed retirement plan, i-HPRP still not available in the market where the respondents have to give their responses based on the illustrated plan in the survey form. Secondly, the survey data was collected during the Covid-19 crisis which responses are submitted online without researchers' verbal explanation. Thirdly, only economic behaviour, mainly saving and spending, was included, while future studies may adopt psychological and psychometric factors toward a retirement plan's preference. Finally, future research may adopt different data analysis technique such as PLS-SEM etc. Nevertheless, this study has several implications. Firstly, the findings contribute to the knowledge related to Islamic retirement plan and planning. In practice, individuals will be concerned about retirement planning, and Muslim individuals would be encouraged to seek for Islamic retirement plan. Besides, individuals would be concerned to invest in shariah linked products to gain spiritual and physical benefits. Practitioners, both conventional and Islamic insurance companies, maybe encouraged to develop new retirement products with the inclusion of health benefits. Further, the findings will help takaful operators reshape marketing strategies and improve existing annuity products to cater to better market share. Hence, policymakers may find the idea of designing an Islamic health retirement plan considering the greater social practices (religious norms). Besides, they might enable to include the health services into a single pension or retirement product. Upon the implementation of the proposed plan, takaful operators will manage to enhance their market capital while the government will manage to reduce retirement expenditures, thus contribute to economic growth over the long term. Finally, the findings demonstrated Malaysian employees' economic behaviour based on demographic factors, which will help designated policymakers to build an effective policy toward a sustainable retirement plan.