عوامل تعیین کننده روانی اجتماعی از برنامه ریزی مالی برای دوران بازنشستگی در میان مهاجران در اروپا

The aim of this paper is twofold. First, to extend Hodges’ model of relationships between financial planning for retirement with psychosocial variables to predict both objective and subjective measures of financial planning. Second, to apply that model to a representative sample of immigrants in Europe provided by the first wave of SHARE study. Data from 1272 immigrants in Europe were obtained through structured interviews and questionnaires. The Structural Equation Modeling (SEM) analyses reveal that the total expected amount of pension was predicted by poor health, migration seniority, and job demands, among other variables. Total years of pension contribution were predicted by salary, job tenure, and migration seniority. These results validate and expand the previous research on bridge employment activities and partial retirement.

During the last decades, the European Union countries have undergone a notable increase in immigration, from the continent and from more distant countries (Österle, 2007). In general, immigrants have a lower socioeconomic level and poorer health than the populations from the host countries and these characteristics place them among the groups with the worst economic preparation for retirement (Ringbäck, Gullberg, Hjern, & Rosén, 1999). Financial Planning for Retirement (hereafter, FPR) is at the core of the current concerns of governments and citizens due to its great impact on well-being during retirement ( Ettner & Grzywacz, 2003) and its economical implications. However research on FPR among immigrants is still sparse and its determinants deserve further attention. An important contribution of this study is the extension of previous models of FPR to include psychosocial predictors. These results are based on a representative sample of immigrants in Europe proceeding from the first phase of the SHARE study. 1 1.1. The importance of FPR Europe is aging faster in this century than the rest of the world, with 15% of the population currently over 65 years of age. At the same time, many Europeans retire relatively early from the labor market and expect more generous benefits and social services than the inhabitants of other developed countries. Despite the positive trend that European economies showed during the past decades, a large segment of retirees still have incomes that fall below the poverty threshold. Lusardi and Mitchell (2007) revealed that in 2004, those baby boomers that were in the lowest income quartile were worse off than their predecessors (Hershey, Jacobs-Lawson, McArdle, & Hamagami, 2007). Several studies stated that retirement can be seen as a decision-making process and, in this sense, the person should plan a course of action to cope with retirement (Beehr, 1986). Moreover, a recent meta-analysis stated that consequences of retirement are more directly affected by retirement planning than for other facets of the process. Specifically, retirement planning showed a strong impact on illness and life satisfaction after retirement (Topa, Moriano, Depolo, Alcover, & Morales, 2009). Hence, due to the economical implications of retirement, financial planning has been considered one of the most relevant parts of general retirement preparation (Hershey et al., 2002 and Taylor and Geldhauser, 2007). This influence of financial facets on retirement is easy to understand if we acknowledge that retirement involves planning and making decisions about the assignment of limited resources to different ends that compete with each other (Hatcher, 2003). Specifically, retirement planning requires a person to decide when and how much to save, when to stop working and when to begin to spend the resources those have been saved up for that moment. To sum up, one’s financial planning can strongly affect the quality of one’s retirement savings decisions (Hershey & Mowen, 2000). Related to measures, FPR has been conceptualized in diverse ways, using objective (amount of dollars accumulated) or subjective indicators such as perceptions of financial engagement (Stoller & Stoller, 2003) or estimations of the percentage of salary saved (Hershey et al., 2002 and Hershey et al., 2007). Whereas the objective measures characterize studies using an economic approach, the latter measures are more frequent in studies of a psychological or sociological orientation. Nevertheless, there seem to be essential differences in the predictive validity of the various financial measures, which leads to recommending complementarity of measures (Taylor & Geldhauser, 2007). Specifically, objective measures may be the best predictors of more objective results, whereas subjective measures are used to predict variables with more emotional impact. In the present study, we used two measures of FPR, number of years of contribution to the pension system (objective measure) and total expected amount of pension (subjective measure), relating them to different types of predictors: sociodemographic, economic, and psychosocial. Furthermore, antecedents of FPR have been studied by researchers from an economic perspective (Helman et al., 2005 and Lusardi and Mitchell, 2007). The economic literature states that people plan for and decide about retirement like decision-makers who try to maximize their benefits, although they are subject to situational constraints. In this sense, sociodemographic characteristics (e.g., age, gender and income) have frequently been used as antecedents (Beedon and Wu, 2004 and Wu, 2001). Another perspective has been proposed by Hershey (2004), based on a general model of planning advanced by Friedman and Scholnick (1997), including psychological predispositions that influences financial planning behavior (e.g., personality factors, perceptions of task relevance and motivational factors). Recently, some theoretical revisions emphasized the unique contribution of psychosocial variables to retirement research. Shultz and Wang (2011) stated that the psychological perspective allows psychologists to explore both inter-individual and intra-individual differences on retirement processes, while other social scientists are not able to provide insights at the micro-individual level. Moreover, these authors recognized that the psychological conceptualization of retirement offers the possibility to take into account the interaction between retirees and their environment. According with this point of view, the Theory of Planned Behavior has been applied to these topic. Croy, Gerrans, and Speelman (2010) have founded that self reported attitudes, subjective norm and perceived behavioral control accounted for a high proportion of the variance in saving intentions. Moreover, other psychosocial antecedents of FPR have been explored by empirical studies. van Rooij, Lusardi, and Alessie (2011) proved that sophisticated financial literacy may be an important driver of FPR above and beyond the effects of demographic characteristics as age, income and education. In a similar vein, some authors have considered the influences of other psychological antecedents, as self-esteem, on FPR (Neymotin, 2010). However, several authors note that in order to further our comprehension of FPR, we need broader explanatory models of decisions and behaviors of saving for retirement (Taylor and Doverspike, 2003 and Taylor and Geldhauser, 2007). 1.2. Financial planning for retirement and immigration The increase in the total population of EU Member States in recent years was mainly due to high net migration. From 2004 to 2008 the population of EU Member States increased, on average, by 1.7 million per year, solely because inflows outweighed outflows (Oblak Flander, 2011). Such Lelkes (2007) stated, the situation of migrants is disadvantageous both in absolute and relative terms, characterized by both high poverty rates and relatively higher poverty rates than the “indigenous” population, with the group of non-EU immigrants exposed to the higher poverty risk. Related to this point, researchers have shown that those individuals most in need of financial and social planning for retirement may actually avoid this process (Kim and Moen, 2001, Long, 1987 and Turner et al., 1994). Summing up, the prevalence of poverty and low income among immigrants, the impact of discontinuous and shorter working careers and the fact that immigrants struggle with adaptation problems would affect retirement planning among this group in a long term. The relation between specific FPR and immigration can be analyzed from many perspectives, although most of the studies have emphasized the economic aspects (Beedon and Wu, 2004 and Wu, 2001) while ignoring the psychosocial implications. From the economic viewpoint, the relation arouses great interest due to the financial consequences on the medium- and long-term pension systems, among other aspects. From the social viewpoint, the relation is worrisome because it seems that – like other collectives with low incomes – immigrants have more difficulties to plan economically for their retirement (Kim & Moen, 2002). The exaggerated optimism about the social security system, the lack of financial knowledge, and the discontinuity of their professional careers have been pointed out as some antecedents of this scarce planning (Taylor & Geldhauser, 2007). The economic theories of migration state that people seek work in a free and competitive market in which individual characteristics lead to differential earnings (Borjas, 1989 and Stark, 1995). Immigrants take advantage of these financial benefits by means of their transfer from low-income areas to other areas with high income. The theory of human capital applies a broader approach to immigration, defining everything that can be used to generate benefits as capital, and including as benefits the competences possessed and developed by people. Thus, costs and benefits are not merely monetary, but instead include cultural and ecological considerations and long-term life goals (Borjas, 1989). The model of relations between FPR and immigration proposed by Hodges (2004) defends that economic considerations play a role in immigrants’ process of financial preparation for retirement, but so does the influence of life cycle, as well as cultural and personal preferences. Our contention is that psychosocial variables have not been adequately outlined by the model and this is one of the purposes of this study. According to this model, age is a very important factor in FPR, determining when one can work, when to retire, and when to begin to perceive pension benefits. This influence of age on FPR is coherent with studies that found an increase with age of concerns about pension and the sufficiency of the accumulated savings (Hershey et al., 2002 and Higgs et al., 2003). In the same sense, age would be increase FPR because older people have less time to recoup the benefits accumulated by financial preparation, so a positive effect of age both on number of years of contribution to the pension system and on total expected amount of pension is derived (H1a). In regard to health, several studies have shown the influence of illness on preparation for retirement, the decision to quit work at a certain time or to continue with bridge employments (Shultz & Wang, 2007). Frequently, illness would transform working in a difficult assignment and, in this sense, it would improve FPR. According to this, it is unanimously accepted that disease forces withdrawal from work (Hershey and Mowen, 2000 and Shultz and Wang, 2007), so it is reasonable to expect a positive influence both on number of years of contribution to the pension system and on total expected amount of pension (H1b). Regarding sociodemographic factors, migration seniority must be considered in two senses. First, researchers have shown that migrants suffered a wage disadvantage when compared with nonmigrants, but with a longer duration of residences, this wage differential disappeared (Borjas, Bronars, & Trejo, 1990). Taking into account this data, migration seniority would be considered as a proxy of availability of resources which the person could invest in FPR. Second, Bartel (1979) founded that most migrations involved an employment separation, meaning that the mover changed from one employment for another. This is significant in regard to FPR, “because of the implications changing employments has for interruptions and loses in vesting, tenure, and financial accumulations in employer – sponsored pension plans” (Hodges, 2004, 28). In this sense, in a previous study (Hodges, 2004), migration seniority seemed to be a positive antecedent of higher engagement in FPR and a negative antecedent of the economic sum accumulated for pension (H1c). Finally, salary is a decisive economic factor that will exert a positive influence both on number of years of contribution to the pension system and on total expected amount of pension (H1d). Early researchers reported that those who are lower paying are less likely to plan for retirement and less likely to have access to retirement planning seminars (Beck, 1984 and Kasshau, 1974) and many studies show that people do not engage in FPR because they lack money to save or because they think they do not earn enough to be able to set aside a part of their income (Kim and Moen, 2002 and Turner et al., 1994). The human capital theory underline those factors that can be used to generate income and wealth, including those talents that immigrants can use to increase their productivity and those skills that they can develop to improve their employability. Investors in human capital compare alternative costs and profits, which are not solely monetary, and decide taking into account their anticipated future gains. In this group of factors Hodges (2004) included education and job tenure. Years of formal education increase qualified job opportunities, salary levels, and job tenure in the most desired positions. Some authors have posited that education interacts with migration, because highly skilled workers have greater gains from migration. These employees are difficult to replace and employers frequently viewed more cost effective to move them that replace them. Moreover, recent research on FPR has shown that economic knowledge acquired during school may be an important antecedent of saving intentions ( van Rooij et al., 2011). In this sense, it is expected to exert a positive influence of education both on number of years of contribution to the pension system and on total expected amount of pension (H2a). Job tenure, as an indicator of stability, usually correlates with higher salaries and better organizational positions. Moreover, tenure proved that a worker has invested a substantial portion of time and effort in the job, and perhaps because is an important factor reducing labor turnover, job tenure is also closely linked to the possibility of saving and planning for retirement. In a psychological perspective, job tenure may be a proxy variable of risk aversion, which would be proved as an antecedent of FPR (van Rooij et al., 2011). Since these evidences, it is reasonable to expect a positive influence of job tenure both on number of years of contribution to the pension system and on total expected amount of pension (H2b). A broader understanding of the mutual influence between FPR and immigration requires the addition of psychosocial variables – both specific of the job and generic to older persons – to Hodges’ (2004) model. Firstly, full retirement from work can be conceptualized as a way of withdrawing from one’s position or from the organization as a whole (Barnes-Farrell, 2003 and Hanisch and Hulin, 1990) and, in this sense; retirement would be related to job quality. There has been a long tradition of research regarding employment quality (Siegrist, 1996, Siegrist et al., 2004 and Marmot et al., 2006), which applies Blau’s exchange theory. Under this view, employment quality is understood as the balance between the effort put out and the perceived rewards. Empirical studies have shown that situations with low perceived reward generate stress and increase quitting intentions. In the active population over 50 years of age, the predictive power of high pressure on personal well-being and health has been revealed (Debrand & Lengagne, 2007; Blanchet & Debrand, 2007), while employment quality has been identified as an antecedent of implication in bridge employment (Topa, Depolo, Moriano, & Morales, 2009), a form of partial withdrawal from job. Hence, FPR can be favored by conditions similar to those that led to quitting the job (Siegrist, Wahrendorf, von dem Knesebeck, Jürges, & Börsch-Supan, 2006): low quality of work, excess of physical or temporal demands and pressures, among others. In this sense, we expect a positive influence of job demands both on number of years of contribution to the pension system and on total expected amount of pension (H3a). Secondly, previous studies have shown that preferences to continue to work or to retire should be considered in the light of people’s desire to maintain personal control over their lives and positive self-images (Barnes-Farrell, 2003). Aged persons’ experiences of high autonomy and control over their lives – named older persons’ quality of life – have been conceptualized as the extent to which their needs are satisfied in the active situation. Quality of life thus defined has allowed us to distinguish various facets, such as autonomy, control, pleasure and self-realization (Higgs et al., 2003). Furthermore, lower levels of quality of life would be considered as a source of stress, and in the sense that individuals avoid them; it would improve planning for retirement or, in other words, quality of life will be a negative antecedent both of number of years of contribution to the pension system and on total expected amount of pension (H3b). Lastly, well-being in old age seems to be associated with a positive assessment of one’s life (Higgs et al., 2003). In this sense, high levels of life satisfaction may negatively affect people’s desire to quit work. Related to this point, empirical studies have showed that people more satisfied with their retirement and with life in general were more actively engaged in bridge employment (Wang, Zhan, Liu, & Shultz, 2008). According to this point, self-esteem has been proved as a positive antecedent of FPR (Neymotin, 2010). Hence, we propose that life satisfaction would have a negative impact both on number of years of contribution to the pension system and on total expected amount of pension, considered as a behavioral preparation to quit job (H3c). Summing up, we propose a broad model of antecedents of FPR among immigrants that includes sociodemographic factors (i.e. age, poor health, salary and migration seniority), factors of human capital (i.e. years of formal education, and job tenure) and psychosocial factors (i.e. physical and temporal job demands, life satisfaction, and quality of life) to predict economic preparation for retirement (i.e. years of pension contribution and total expected amount of pension). The specific hypotheses are represented graphically in Fig. 1.

The correlation matrix analysis provides preliminary support for the hypotheses presented in the model. Among the antecedents, age and poor health have significant and positive relations with each other and negative relations with formal education. At the same time, age is positively related to migration seniority and job tenure, and negatively to salary. Migration seniority is negatively related to education and to salary, and it has positive relations with poor health, but statistical significance was not reached in any case. Among the psychosocial factors, life satisfaction is positively related to quality of life and inversely to physical and temporal job demands. In regard to consequents, age, migration seniority, salary, and job tenure have positive relations with years of pension contribution; whereas the years of formal education, work demands, and job tenure have a positive relation with the total expected pension at retirement. Contrary to our hypotheses, age have not showed a positive relationship with expected yearly amount of pension and poor physical health have showed negative relationships with both objective and subjective measures of FRP. In the same sense, education has a relationship with years of pension contribution near zero, contrary to our expectations (see Table 1).We tested the model with Structural Equation Modeling (SEM), using the maximum likelihood procedure and the matrix of original data as input with AMOS 19.0. The model included nine latent exogenous variables and two latent endogenous variables. To test the fit of the model, several indexes are recommended, such as χ2 and its level of probability, although, due to its sensitivity to sample size and to deviations from data normality, other fit indexes are proposed, such as the Comparative Fit Index (CFI), the Incremental Fit Index (IFI), and the Root Mean Square Error of Approximation (RMSEA). Three more indexes are also used to compare models, the Expected Cross-Validation Index (ECVI, similar to the Akaike Information Criterion – AIC) and the Browne–Cudek Criterion (BCC). In both of them, the simpler models with better fit achieve low values, whereas the more complex models with poorer fit reach high scores because both indexes penalize complex models with a poor fit to the data. The fit indicators of the original model were improvable (χ2 = 177.4, p = .00, CMIN/DF = 8.5, IFI = .85, CFI = .85, RMSEA = .08, ECVI = .228, AIC = 289.4, BCC = 290.5), so the model was respecified. The relations that were not statistically significant based on their critical ratios (CRs) were eliminated, obtaining a progressively better fit. Specifically, the relations between age and quality of life, on the one hand, and the total amount of expected pension, on the other hand, were eliminated, as were the relations of education, physical health, life satisfaction, and work demands with the number of years of pension contribution. The respecified model showed better fit (χ2 = 192.7, p = .00, CMIN/DF = 6.2, IFI = .85, CFI = .85, RMSEA = .06, ECVI = .224, AIC = 290.4, BCC = 285) and its standardized estimations can be observed in Fig. 2. The percentage of variance explained by the model seems adequate for each indicator of FPR, because it ranges between 24% for the number of years of pension contribution and 57% for the total amount of expected pension.Among the socio-demographic antecedents, poor health and migration seniority have a decisive influence on the amount of expected pension, being the first positive and the second negative, whereas life satisfaction and job demands are its best psychosocial predictors. Migration seniority, salary, and job tenure are better predictors of the number of years of pension contribution.