Application of decomposed theory of planned behaviour on intention to save in voluntary private retirement fund in Malaysia

The objectives of this research are four-pronged. Firstly, is to investigate the determinant of behavioural intention and to examine the relationship of its decomposed salient behavioural beliefs toward these determinants. Next, is to gauge if the perceived behavioural control mediates the relationship between incentives and the intention to save before finally inspecting the moderating effect of age on the attitude, subjective norms (SN) and the perceived behavioural control (PBC) towards the intention to save in a voluntary retirement fund. Findings of this study were based on the data collected from a sample of 484 workers both in the public and private organisations Malaysia. The structural equational modelling shows attitude, SN, and PBC has a positive effect in influencing an individual’s intention to save in a voluntary retirement fund. These three constructs explain 53.7 per cent of the coefficient of determination variance on the intention to save in a voluntary retirement fund. Among the three determinants of intention, through the path relationship reveals PBC (β = 0.361, t = 5.862) as the most important predictor, followed by attitude (β = 0.280, t = 4.763) and subjective norms (β = 0.209, t = 4.056) in determining the intention to save in a voluntary retirement fund. Decomposing attitude into specific behavioural belief namely relative advantage, compatibility, complexity, trust and awareness explains 39% of the variance of attitude towards savings in a voluntary retirement fund. Relative advantage (β = 0.482, t = 8.398) influences the attitude to save in a voluntary fund, while compatibility (β = - 0.192, t = 2.697) has a negative relationship. On the other hand, the complexity (β = 0.028, t = 0.401) in saving in the voluntary retirement fund is not significant. This study also extends the formation of attitude to include the construct of trust and awareness. In this study, trust (β = 0.133, t = 2.634) is positively influencing the attitude to save in the voluntary retirement fund. However, the finding shows that awareness (β = 0.044, t = 0.658) is not supported as a factor that influences attitude. Decomposing the subjective norms into internal (β = 0.219, t = 2.798) and external norms (β = 0.459, t = 6.057), shows both normative beliefs are positively significant in influencing subjective norms with an explained variance of 41%. Decomposing the perceived behavioural control into specific control beliefs shows that while both are positively influencing the PBC, the facilitating condition (β = 0.362, t = 5.862) variable has primacy over self-efficacy (β = 0.273, t = 3.643). The mediation analysis using Preacher and Hayes (2008) technique shows that tax incentives (β = 0.063, t = 2.905) are significantly positive, while cash incentives (β = 0.013, t = 0.634) are not in influencing the intention to save in a voluntary retirement fund. This show that the respondents are not moved with the immediate tangible benefit instead is more interested in saving because of tax incentives. The permutationbased PLS-MGA analysis confirms that there is no significant difference in attitude and perceived behavioural control between the younger and older group’s behavioural intention to save in a voluntary retirement fund. However, the same analysis shows a significant difference in term of subjective norms (permutation p-value = 0.976) of the younger and the older group. These findings were supported by relevant theories and are consistent with findings of previous studies. Decomposition of the primary variables into a specific belief provides a better explanation of the factors that influence saving intention. Based on the findings, implications for practice and academic research were offered. Limitations of the study and recommendations for future research were also discussed