Do optimists plan for retirement? A behavioural explanation for non-participation in pension schemes

Do optimists plan for retirement? A behavioural explanation for non-participation in pension schemes

Economics Letters 125 (2014) 396–399 Contents lists available at ScienceDirect Economics Letters journal homepage: www.elsevier.com/locate/ecolet D...

329KB Sizes 0 Downloads 42 Views

Economics Letters 125 (2014) 396–399

Contents lists available at ScienceDirect

Economics Letters journal homepage: www.elsevier.com/locate/ecolet

Do optimists plan for retirement? A behavioural explanation for non-participation in pension schemes Jiayi Balasuriya ∗ , Orla Gough, Kristina Vasileva Westminster Business School, University of Westminster, London NW1 5LS, UK

article

info

Article history: Received 4 May 2014 Received in revised form 14 October 2014 Accepted 16 October 2014 Available online 31 October 2014

abstract This paper investigates the relationship between financial optimism and non-participation in pension schemes in the UK. We show that financial optimism reduces the probability of employees joining employer run pension schemes and also the probability of the self-employed subscribing to private pension plans. © 2014 Elsevier B.V. All rights reserved.

JEL classification: E2 D1 Keywords: Optimism Pension decision Psychology and economics

‘‘The man who is a pessimist before 48 knows too much; if he is an optimist after it he knows too little’’. —Mark Twain’s Notebook, 1902–1903 1. Introduction and background Over 11 million people in the United Kingdom do not save sufficiently for their retirement (Department for Work and Pensions, 2012). A recent survey found that British retirement saving shortfall to be by far the worst among all 15 countries surveyed, with retirement savings only covering 37% of the required retirement provision (Twigg, 2013). By not contributing into pension schemes taxpayers lose tax relief available on pension payments and additional employer contributions. As individuals are also undersaving to secure themselves enough retirement income (Banks et al., 1998), inadequate pension contribution has become an increasingly critical factor as governments and individuals struggle to eliminate the problem of old-age poverty. The decision on retirement savings is a function of a complex set of factors (Agnew, 2010). In addition to rational explanations, behavioural biases can have effects on non-participation. The effect of behavioural issues, such as optimism, has attracted researchers’

∗ Correspondence to: Westminster Business School, University of Westminster, 35 Marylebone Road, London NW1 5LS, UK. Tel.: +44 0 207911 5000x67103. E-mail address: [email protected] (J. Balasuriya). http://dx.doi.org/10.1016/j.econlet.2014.10.012 0165-1765/© 2014 Elsevier B.V. All rights reserved.

attention since the seventies. Prior research has shown that optimism has a significant impact on decision making in various social domains including finance and business related fields (Weinstein, 1980; Lee et al., 1991). While optimism helps us sustain motivation and encourages us to overcome difficulties, it leads us to neglect risks (Weinstein and Lyon, 1999). Having a positive outlook about the future could result in a lower tendency to worry about the potentially negative consequences of risky decisions. Optimists may not be able to overcome their optimistic views and limit risk-friendly decisions, even though risky investments may lead to loss of wealth (Brunnermeier and Parker, 2005). Nofsinger (2005) suggests that optimism could be a strong influence on decisionmaking processes and that it encourages investors to hold risky portfolios. This study extends previous literature on optimism by exploring the correlation between financial optimism and decisions on pension participation. Although optimism has been found to affect financial decisions and the tendency to ignore risks, prior studies have not investigated whether optimism plays a role in pension decisions. Our research questions are: (a) are optimistic employees more likely to not join employer pension schemes and, (b) when employer pensions are not available for the selfemployed, are they less likely to pay into private pension schemes if they are optimistic. The British Household Panel Survey (BHPS) is used as the longitudinal data source in this study. We provide evidence that participation in employer pension schemes significantly declines

J. Balasuriya et al. / Economics Letters 125 (2014) 396–399

397

Table 1 Pension participation. Sample 1 (employed):

Sample 2 (self-employed): Frequency (%)

Employer runs a pension scheme Employer does not run a pension scheme Total valid

12,302 (61.5) 7,711 (38.5) 20,013

Where employer runs a pension scheme: Member of employers pension scheme Not member of employers pension scheme Total valid

7,556 (61.4) 4,688 (38.1) 12,244

when employee optimism level increases. We also find that the probability of participating in private pension plans is lower for optimistic self-employed respondents in the survey. 2. Data and methodology The BHPS surveyed 10,300 individuals from 5500 households since 1991 by following the same households over consecutive years of the survey. We focus on the working population in the survey and extracted two samples from the full panel to investigate the relationship between optimism and pension participation. Sample 1 contains respondents who started new employment within the same year before the survey took place. This gives us the opportunity to observe an employee’s financial optimism level when they join a new firm and decide whether to ‘‘opt-in’’ to their employer’s pension scheme. Table 1 suggests that under 40% of employees join pension schemes when they start new jobs as less than two-thirds of employers run pension schemes. Sample 2 contains respondents from the full panel who stated that they are ‘‘self-employed’’. This enables us to examine whether optimists are less likely to pay into private pension schemes when employer-run pensions are not available. Over half the self-employed surveyed do not have a private personal pension as summarised in Table 1. We define optimism as the overestimation of the positive outcome in an individual’s future financial situation relative to a rational expectation. This is consistent with definitions for optimism found in psychology and social sciences (Weinstein, 1980). We use answers to the following two BHPS questions on individuals’ opinion on their financial situation for our measures. Values for financial expectation (E ) and realisation (R) were directly obtained from these questions. Question 1: Looking ahead, how do you think you will be financially a year from now, will you be Better off, worse off than you are now, Or about the same? (denoted as financial expectation, represented by E) and Question 2: Would you say that you yourself are Better off, or worse off financially than you were a year ago, Or about the same? (denoted as realisation, represented by R). Our first financial optimism measure Optimism(−) for year t is calculated as the difference between Et and Rt −1 , where Rt −1 stands for the reported change in financial situation during the past year. With Optimism(−) , the historically realised return Rt −1 is used as a benchmark and we assume that Rt −1 captures unbiased expected individual financial return characteristics and information at time t. Optimism(−) indicates that a respondent is either irrationally optimistic since she ignores her historical return, or she is rationally optimistic if she has private information that is not disclosed in the survey, or a combination of both scenarios. Our second measure Optimism(+) is calculated as the difference between Et and Rt , where Rt represents the reported change in financial situation during the current year t (collected in yeart +1 ). Optimism(+) suggests that a respondent is either irrationally optimistic since her actual realisation is less than her expected return, or she was rationally

Frequency (%) Paid into private personal pension No payment into private personal pension Total valid

5,547 (36.7) 8,755 (58.0) 14,302

optimistic if the forecasting error is due to the effect of unforeseeable information exposed during year t. Optimism(+) can be understood as a forecasting error and is similar to the definition of ‘‘unrealistic optimism’’ in some previous research (Arabsheibani et al., 2000; Coelho, 2010). However, the limitation of Optimism(+) is that what transpires in reality after a financial decision is often not ‘‘rationally expected’’. Individuals might make perfectly rational expectations based on the information they have at the time of forecasting (as with Optimism(−) ), but new information exposed during year t may be completely unforeseeable at time t and therefore should not affect our classification of the optimistic bias of the forecast (Balasuriya, 2012). Using both Optimism(−) and Optimism(+) could offset each other’s theoretical advantages and disadvantages in capturing optimism. Table 2 contains aggregate data of optimism and variables on pension participation from both employed and self-employed samples. While around half of the sampled population are neutral (Optimism = 0), the number of extreme optimists (Optimism = 2) are more than three times that of extreme pessimists (Optimism = −2). This indicates that in a representative sample of the British population, people tend to be exceedingly more optimistic than pessimistic. Pessimists tend to work for firms that run pension schemes more than optimists. Extreme optimists report a pension subscription rate over 10% less than extreme pessimists. These initial statistics indicate a tendency that with an increase in optimism, pension participation decreases for both the employed and self-employed. 3. Regression results and discussion We further investigated the relationship between financial optimism and pension participation using logistic regression, with participation in pension schemes as the dependent variable. A number of control variables are used as independent variables to capture demographic and employment-related influences on pension participation. Table 3—Sample 1 reports logistic regression results on pension subscription on the year an individual started working with a new employer when employer pension schemes were available. Results show that apart from income, optimism is the most important factor contributing to non-participation in pensions. Extreme pessimists are almost twice as likely to join pension schemes as extreme optimists (reference variable). Even moderately optimistic (Optimism = 1) employees are over 20% more likely than extreme optimists to take part in pension schemes. Individuals whose realised financial situation turns out to be much worse than expected (Optimism(+) = 2) are the least likely to participate in pensions. Although Table 3—Sample 1 reveals that optimists are significantly less likely to join employer pensions in the year they commence employment, we do not rule out the possibility that individuals may join employer pension schemes later in their

398

J. Balasuriya et al. / Economics Letters 125 (2014) 396–399

Table 2 Optimism scores and pension subscription rate. Sample 1 (employed):

Sample 2 (self-employed):

Frequency (%)

Employer runs a pension scheme%

Member of employers pension scheme %

Frequency (%)

Paid into private personal pension %

Optimism(−) = E(t ) − R(t −1) Optimism = −2 Optimism = −1 Optimism = 0 Optimism = 1 Optimism = 2 Total valid

621 (3.1) 4,761 (23.8) 9,231 (46.1) 3,368 (16.8) 2,032 (10.2) 20,013

62.6 64.7 62.0 58.9 55.4

66.8 63.5 61.6 61.2 53.2

186 (1.2) 2,416 (16.0) 8,077 (53.5) 3,183 (21.1) 1,237 (8.2) 15,099

34.4 38.3 38.0 35.3 29.3

Optimism(+) = E(t ) − R(t ) Optimism = −2 Optimism = −1 Optimism = 0 Optimism = 1 Optimism = 2 Total valid

303 (1.5) 2,965 (14.8) 8,495(42.4) 4,015 (20.1) 1,494 (7.5) 17,272

64.4 64.8 62.5 60.2 60.2

71.3 66.2 63.7 60.7 60.7

167 (1.1) 1,958 (13.0) 6,886 (45.6) 3,301 (21.9) 819 (5.4) 13,131

39.5 41.1 38.9 36.7 29.2

Table 3 Logistic regression results for optimism and pension participation.

Optimism = −2 Optimism = −1 Optimism = 0 Optimism = 1 Optimism = 2 (reference) Age No. of children Annual Income (ln) Year of interview Male (reference) Female Married (reference) Unmarried White (reference) Non-white Healthy (reference) Unhealthy Smoker (reference) Non-smoker University:first degree or above (reference) University:no first degree Firm size Finance-related occupation (reference) Non-finance-related occupation Permanent contract (reference) Non-permanent contract N Chi2 Pseudo R2 * ** ***

Sample 1 (employed):

Sample 2 (self-employed):

Marginal effect: member of employers pension scheme

Marginal effect: paid into private personal pension

Optimism(−) = E(t ) − R(t −1)

Optimism(−) = E(t ) − R(t −1)

Optimism(+) = E(t ) − R(t )

***

Optimism(+) = E(t ) − R(t ) ***

0.850 0.483*** 0.470*** 0.312*** 0.000 0.020*** 0.095*** 0.908*** −0.052*** 0.000 0.407*** 0.000 −0.185*** 0.000 −0.144 0.000 −0.008 0.000 0.818*** 0.000 −0.451*** 0.135*** 0.000 −0.139*** 0.000 −0.545***

0.939 0.386*** 0.338*** 0.219*** 0.000 0.019*** 0.098*** 0.979*** −0.048*** 0.000 0.383*** 0.000 −0.199*** 0.000 −0.127 0.000 −0.048 0.000 0.824*** 0.000 −0.432*** 0.134*** 0.000 −0.140*** 0.000 −0.555***

0.241 0.269*** 0.288*** 0.192** 0.000 0.002 0.011 0.150*** −0.063*** 0.000 −0.614*** 0.000 −0.448*** 0.000 −0.135 0.000 −0.113 0.000 0.724*** 0.000 −0.203***

0.321 0.452*** 0.353*** 0.274** 0.000 −0.001 0.003 0.164*** −0.058*** 0.000 −0.624*** 0.000 −0.456*** 0.000 −0.129 0.000 −0.165* 0.000 0.692*** 0.000 −0.202***

20,013 0.000 0.212

20,013 0.000 0.214

15,099 0.000 0.076

15,099 0.000 0.075

Stands for significance at the 10% levels or better. Stands for significance at the 5% levels or better. Stands for significance at the 1% levels or better.

career.1 However, by not joining or delaying joining pension schemes, optimists might lose out on the potential benefits of these schemes, such as tax efficiency and company paid contributions, and consequently optimists may not save enough for retirement. Results in Table 3—Sample 2 indicate that among the selfemployed, extreme optimists are at least 29% less likely to have

1 We selected individuals interviewed every year and constructed the mean optimism score for each respondent. We found that optimism is negatively correlated with pension participation regardless of which year respondents started employment or joined schemes.

paid into private pensions compared to neutral respondents. When the level of optimism increases from neutral, the probability of investing in private pension plans decreases. This finding reiterates the potential harmful effects of being extremely optimistic and not saving adequately for retirement. Participating in pension schemes is also associated with other variables. Higher income, being female, being a non-smoker, working for larger firms and having a permanent employment contract were significantly positively correlated with being a member of employer pensions. Not having a university degree and not having a finance-related occupation were inversely related to joining employer pension schemes. On each stratified optimism score, we

J. Balasuriya et al. / Economics Letters 125 (2014) 396–399

found that income and education play significant roles in joining employer pensions across all optimism levels.2 Individuals with lower income might suffer from poor financial well-being when they retire due to insufficient pension participation as well as have lower savings during their working life. It may therefore be necessary to enact new regulations on pensions to help protect optimistic low earners. Many countries are embarking on a journey of substantive changes in their pension systems to keep up with the changing landscape of increased life expectancy and the varying financial sustainability of existing pension systems. Our findings from Sample 1 are in favour of the British government’s recent Automatic Enrolment (Miscellaneous Amendments) Regulations 2013 No. 2556 legislation. Employees who do not have or have not joined employers’ pensions in Sample 1 should benefit from these new directives. However, auto-enrolment is a gradual process from 2012 to 2018, with smaller firms not starting this process until mid 2015 at the earliest. Optimistic employees could still continue to ignore saving for retirement for up to another six years. We found evidence in the full panel that optimism is negatively correlated with respondents’ pension income.3 This suggests that individuals who have already retired or are near retirement without an occupational pension will not benefit from the new regulations and the UK government may need to intervene and help with their retirement income. One concern with these new regulations is that as employees could opt-out of employer pensions after auto-enrolment, optimists may take this option since they are prone to undertake risky financial strategies. As optimists are more likely to have other financial obligations such as debt (Dawson and Henley, 2012), they may be more likely to opt-out of saving for retirement. The correlation between optimism and pension withdrawal should be investigated with future BHPS data gathered after auto-enrolment. Another significant issue is that the self-employed would not be protected by these new regulations. The optimistic self-employed might still face the reality of not saving sufficiently for retirement as they tend not to join private pension schemes.

2 The role of income and education is less prominent among the self-employed controlled for optimism. Correlations among demographic variables are available upon request. 3 Full estimating results are available upon request.

399

4. Conclusions Prior research has suggested that optimism leads to more risk taking behaviour within the financial decision-making domain. This paper explored the relationship between optimism and planning for retirement which is perhaps one of the most important financial decisions in an individual’s life. We presented evidence that optimism significantly contributes to non-participation in both employer run pension schemes and private pension plans. Our research suggests that both employed and self-employed individuals who are financially optimistic could face the very negative consequences of pension shortfall and low pensions income when they retire. References Agnew, J.R., 2010. Pension participant behavior. In: Baker, H.K., Nofsinger, J.R. (Eds.), Behavioral Finance Investors, Corporations, and Markets. Wiley, p. 577. Arabsheibani, G., de Meza, D., Maloney, J., Pearson, B., 2000. And a vision appeared unto them of a great profit: Evidence of self-deception among the selfemployed. Econom. Lett. 67, 35–41. Balasuriya, J.W., 2012. An Empirical Analysis of Financial Optimism and Portfolio Choice (Ph.D. thesis), Cass Business School, City University London. Banks, J., Blundell, R., Tanner, S., 1998. Is there a retirement-savings puzzle? Amer. Econ. Rev. 88 (4), 769–788. Brunnermeier, M., Parker, J., 2005. Optimal expectations. Amer. Econ. Rev. 95 (4), 1092–1118. Coelho, M.P., 2010. Unrealistic optimism: still a neglected trait. J. Bus. Psychol. 25 (3), 397–408. Dawson, C., Henley, A., 2012. Something will turn up? Financial over-optimism and mortgage arrears. Econom. Lett. 117 (1), 49–52. Department for Work and Pensions 2012. Estimates of the number of people facing inadequate retirement incomes. Ad Hoc Statistical Analysis Quarter 3. Lee, C.C., Shleifer, A., Thaler, R.H., 1991. Investor sentiment and the closed-end fund puzz. J. Finance 46 (1). Nofsinger, J.R., 2005. Social mood and financial economics. J. Behav. Finance 6 (3), 144–160. Twigg, M., 2013. The Future of Retirement. HSBC Insurance Holding Limited. Weinstein, N.D., 1980. Unrealistic optimism about future life events. J. Personality Soc. Psychol. 39 (5), 806–820. Weinstein, N.D., Lyon, J.E., 1999. Mindset, optimistic bias about personal risk and health-protective behaviour. Br. J. Health Psychol. 4, 289–300.