Money also is sunny in a retiree’s world: financial incentives and work after retirement

Svenja Lorenz; Thomas Zwick

doi:10.1186/s12651-021-00304-1

Money also is sunny in a retiree’s world: financial incentives and work after retirement

Lorenz, Svenja; Zwick, Thomas 2021-12-01 00:00:00 This paper assesses the impact of financial incentives on working after retirement. The empirical analysis is based on a large administrative individual career data set that includes information about 2% of all German employees subject to social security or in marginal employment until age 67 and their employers in the period 1975–2014. We use the classical labor supply model and differentiate between the impact of (potential) labor and non-labor (pension entitle - ments) income. A Heckman-type two step selection model corrects for endogeneity. We show that labor income has a positive and non-labor income a negative impact on the decision to work after retirement. Especially individuals who can substantially increase their earnings in comparison to their pension entitlements accordingly have a higher probability to work. Men are more attracted by labor earnings incentives than women. Also individuals who work until retirement are easier attracted to work after retirement by higher labor income than those with gaps between employment exit and retirement. Our results allow the calculation of the impact of changes in taxes on labor and non-labor income and changes in earnings offers by employers on work after retirement for different demographic groups. Keywords: Work after retirement, Financial incentives, Labor and non-labor income JEL Classification: J14, J22, J26 1 Introduction scientific literature (Anger et al. 2018; Burkert and Hoch - Labor force participation of retirees increased strongly fellner 2017; Büsch et al. 2010; Cahill et al. 2015; Fas- in most European countries in recent years (Eurofound bender et al. 2016; Maxin and Deller 2010; Micheel et al. 2012; Hofäcker and Naumann 2015). On average, the 2010; Pleau and Shauman 2013; Wang et al. 2008; West- employment rate for individuals aged between 65 and ermeier 2019). 74 years has risen from 5.2% in 2002 to 8.7% (an increase This paper concentrates on the impact of financial of 67%) in 2014 in the EU-15 Countries (Rhein 2016). incentives of older individuals on their decision to work Individual, social and work-related push and pull fac- after retirement or not. More specifically, we estimate tors have been put forward as drivers of post-retirement the percentage change in labor force participation after employment in a dynamically developing strand of retirement in response to a percentage change in the financial situation. On the basis of a Heckman (1976, 1979) approach to correct for sample selection, we show that (potential) labor earnings are an important pull fac- tor for the decision to engage in employment beyond Also in Germany, the labor force participation rate of people aged 65 to 74 increased to 9.6% in 2014 from 4.2% in 2002, implying a rise of almost 140% (Rhein 2016). *Correspondence: thomas.zwick@uni-wuerzburg.de Faculty of Business Management and Economics, University of Würzburg, Sanderring 2, 97070 Würzburg, Germany Full list of author information is available at the end of the article © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. 21 Page 2 of 17 S. Lorenz , T. Zwick retirement. We separately assess the second main finan - retirement—Kanabar (2015) however does not find this cial incentive mechanism and find that there is a size - correlation. Micheel et al. (2010) stress a negative correla- able negative pension entitlement (non-labor income) tion between equivalent household income before retire- effect on the probability of working after retirement. As a ment and the interest in work after retirement—Engstler consequence, we show that the higher the ratio between and Romeu-Gordo (2014) however find a u-shaped pat - (potential) post-retirement labor earnings and pension tern for work after retirement and equivalent household entitlements, the larger the likelihood of post-retirement income. Maestas (2010) finds a weakly positive correla - employment. The second aim of the paper is the provi - tion between the decision to unretire and earnings before sion of empirical evidence of heterogeneous earnings retirement. Komp et al. (2010) and Kanabar (2015) find elasticities between retiree groups. We show that women no correlation between working after retirement and are less responsive to higher labor earnings than men. (non-pension) wealth. Engstler and Romeu-Gordo (2014) Especially men with low earnings before retirement have point to the fact that higher financial pressure such as the high earnings elasticities. We in addition find that indi - presence of a real estate loan also can explain work after viduals who use the bridge options partial retirement or retirement. unemployment before early retirement are less respon- When looking closer at the measurement of the finan - sive to higher labor earnings after retirement than their cial situation of retirees, it becomes evident that so far counterparts who work until retirement. no contribution on work after retirement made a dis- This paper makes the following contributions to the lit - tinction between labor and non-labor income. Some erature on empirical determinants of work beyond retire- papers only include non-labor income after retirement ment: first, we empirically assess the importance of labor such as pension entitlements (Hochfellner and Burkert earnings in comparison to non-labor income for work 2013; Kanabar 2015). Other papers bundle up non-labor beyond retirement for the first time. Second, large-scale and labor income after retirement by for example using and reliable administrative social security panel data equivalent household income after retirement (Anger covering all types of employees except civil and military et al. 2018; Engstler and Romeu-Gordo 2014; Micheel servants are used instead of small-scale and potentially et al. 2010; Pfarr and Maier 2015). A last group of papers specific sub-populations or survey-based data that were looks at wealth or the subjective economic status before the basis of most relevant studies so far. Third, a rich retirement (Fasbender et al. 2016; Kanabar 2015; Komp set of employer and individual characteristics allows us et al. 2010; Wang et al. 2008). to control for additional determinants of work beyond Empirical approaches that do not separate labor and retirement. Fourth, based on hypotheses, we show heter- non-labor earnings after retirement may report biased ogeneities in labor earnings elasticities between men and results on financial determinants of work participation women, employees who have their work and home in the after retirement. According to standard labor supply the- same region or not, employees with and without a gap ory, labor earnings are assumed to increase work partici- between employment exit and retirement, and employees pation and non-labor income is assumed to decrease it with high and low earnings before retirement. The analy - (Cahuc et al. 2004; Blundell et al. 2007). When work has sis of differences in elasticities between these sub-groups a higher reward, workers substitute leisure against work. allows us to derive more specific implications of changes An increase in non-labor income has a negative effect on in financial incentives for these sub-groups. work participation or working hours. The income effect induces employees to reduce work because they can 2 Background maintain a target standard of living with lower labor bur- 2.1 F inancial situation as driver of work after retirement den. This paper for the first time distinguishes between There are strong indications that the individual finan - income and substitution effects of work and non-work cial situation is a key driver of work after retirement. income after retirement on work participation and shows More specifically, Hochfellner and Burkert (2013) the empirical relevance of both drivers. and Pleau and Shauman (2013) find a negative cor - The extensive labor supply margin response to mar - relation between pension entitlements and work after ginal work income changes depends on the probability density function of the distribution of reservation wages around the economy’s equilibrium. The variance in labor supply for small changes in wages around the equilib- Examples are the German Ageing Survey (Deutscher Alterssurvey DEAS) or rium wage is low for the entire working-age population the Survey of Health, Ageing and Retirement in. Europe (SHARE). because relatively few people who do not work are will- We only briefly summarize the literature on the impact of financial situa- tion on working after retirement. There are several recent literature reviews ing or able to work for the current wage (Chetty et al. that describe the knowledge about other drivers, compare for example Eng- 2011). It is therefore hard to estimate reliable earnings stler and Romeu-Gordo (2014) or Sullivan and Al Ariss (2019). Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 3 of 17 21 elasticities on labor supply for the working-age popula- tion. Relatively many employees at the end of their career are closer to the margin of re-entering the labor force for available earnings opportunities, however. As the previ- ous labor contract ends when entering retirement, all employees close to retirement have to decide whether to work in a new labor contract or not besides receiving retirement benefits. Upon retirement, labor re-entry inci - dence and potential extensive margin responses to labor earnings options therefore are relatively high in compari- son to the earnings elasticity for the entire population (Rogerson and Wallenius 2009). Analogously to people close to retirement, all women with a small child have to decide whether to work again or continue to provide full time childcare. As a conse- quence, also the extensive margin response to labor earn- ings incentives is relatively high (and easier to measure) for women with small children. The empirical literature on mothers with young children shows that besides non- Fig. 1 Labor supply model of retirees (Source: own illustration) labor financial means such as income received by other household members, childcare cost subsidies or public transfers such as parental allowance, also (potential) labor and leisure. The shape of the indifference curves corre - earnings are an important driver of labor market re-entry sponds to consumption and leisure being defined as “nor - (Allègre et al. 2015; Borra 2010; Connelly 1992; Del Boca mal” goods with decreasing marginal returns. and Vuri 2007; Ribar 1992; Viitanen 2005). The empirical In this model, individual income consists of non-labor identification strategy to analyze financial incentives on income (in our case pension entitlements) V and labor labor supply of mothers after their parental leave period earnings W. The earnings derived from work after retire - can be applied to the labor supply decision of older work- ment correspond to the opportunity costs of leisure. The ers after their retirement. As a consequence, we use the point where one of the retiree’s indifference curves tan - Heckman selection model proposed for example by Con- gents the budget line is the combination of earnings and nelly (1992) or Borra (2010) and assume that for retirees, leisure with which the retiree maximizes utility under the (potential) labor earnings are an important driver for budget constraint. work after retirement besides non-labor earnings. Figure 1 shows that if potential (hourly) labor earn- ings after retirement are too low, the individual does not 2.2 Basic model work after retirement. The maximum indifference curve, According to the standard labor supply model, retir- U crosses the budget line at point A and the retiree con- ees maximize their utility by comparing consumption sumes only non-labor earnings V . If (hourly) labor earn- made possible by labor and non-labor income with lei- ings and accordingly the slope of the budget line increase, sure (Cahuc et al. 2004). In the case of work after retire- utility is maximized at point B with U > U . The retiree ment, individuals decide whether to work or not after B A therefore decides to work after retirement (substitution they receive an old-age pension. The observation period effect) and we can derive the following hypothesis: therefore starts at the earliest at the early retirement age (ERA) for those eligible for an early retirement option or at the normal retirement age (NRA) for the other employ- Hypothesis 1 The higher the (potential) labor earnings ees. Figure 1 shows different possible budget lines as well after retirement, the greater the likelihood to work after as the retirees’ corresponding indifference curves. The retirement. slope of the indifference curves (U, U, U, U ) defines A B C D the marginal rate of substitution between consumption An increase in non-labor income from V to V results 0 1 in a parallel shift of the budget line. With higher non- labor income, labor force participation decreases because the retiree can now consume the same quantity of goods Please note that “not working” when receiving an old-age pension means when working less (income effect). In Fig. 1, the income for many individual in our dataset to be engaged in unpaid care, household or effect is depicted by steeper indifference curves at a given volunteer work or being self-employed. 21 Page 4 of 17 S. Lorenz , T. Zwick leisure level with increasing V. Thus, an increase in the care of relatives, childcare of grandchildren) is higher pension entitlements decreases the opportunity cost of than the comparable share of men (Hank and Buber leisure. In Fig. 1 an increase in the pension entitlements 2009). In addition, there is a second spike in household is illustrated as a shift from point B to point C. Note that production when individuals approach retirement age. with an increase in V that is strong enough, at a given This spike is larger for women than for men, see Var - labor earnings level, pensioners may decide to stop work- gha et al. (2017). There is a decisive difference between ing. Therefore, the following hypothesis is derived: the average individual in working age and the decision for work after retirement when calculating the employ- Hypothesis 2 The higher the pension entitlements ment earnings elasticities, however. Until retirement, the (non-labor income), the lower the likelihood to work share of men who do not work is much smaller than the after retirement. share of women (Blau and Kahn 2007). The labor market attachment of men who potentially can join the work- force therefore is lower than that of women. Hence, the Given the higher pension entitlement, V , higher earnings elasticity of males during their working age on (potential) labor earnings (W instead of W ) are needed 2 1 average is lower than the earnings elasticity of females to incentivize the retiree to work as much as in point (Bastani et al. 2021). On day 1 after retirement, labor B (point D, for example). A higher substitution effect market participation of men and women is set to zero, (higher labor earnings) has to compensate for a higher however. As a consequence, there may be more males income effect (higher non-labor income) for labor mar - willing to re-enter work after retirement than females ket participation after retirement. Hypothesis 3 combines at available labor earnings given the higher home pro- the first two hypotheses: duction duties of women that make labor market par- ticipation less attractive or impossible for them. In other Hypothesis 3 The higher the ratio between the (poten - words, the density of male retirees to re-enter employ- tial) labor earnings after retirement and non-labor ment may be higher than the density of female retirees. income, the greater the likelihood that the retiree works Therefore, we assume that the labor earnings elasticity after retirement. for old women is lower than that for old men: Hypothesis 4 Women have lower post-retirement 2.3 Het erogeneous responses to financial incentives labor earnings elasticities than men. We now derive factors that explain differences in labor earnings elasticities for the labor force participation Some previous papers on the decision to work after between retiree groups. Arguments concerning differ - retirement look at potential differences between men and ent responses of individuals to financial incentives from women. As financial drivers however only individual and labor supply theory and existing empirical results on equivalent household income before retirement as well post-retirement employment are built upon to derive as wealth and pension entitlements have been included hypotheses that can be tested in the empirical analysis. but not our key variable (potential) labor earnings. In the labor supply literature, a robust finding is that Pleau (2010) and Pleau and Shauman (2013) stress that earnings elasticities of women exceed that of men (Bar- women with a high individual income before retirement gain et al. 2014; Evers et al. 2005; Meyer and Rosenbaum more often work after retirement. Equivalent household 2001). The difference between gender groups is explained income and wealth seem to play no role for women. For by the division of labor within the family (Blundell et al. men however work after retirement is positively corre- 2007). According to Blau and Kahn (2007) and Mincer lated with equivalent household income and negatively (1962), women substitute their time between market with wealth (but not individual income). Engstler and work, home production, and leisure, whereas men pri- Romeu-Gordo (2014) however find an impact of pension marily substitute their time between market work and entitlements for men only. leisure. As women have closer substitutes for market In Germany, there are various ways to transit from work than men, they are expected to have larger earn- employment to retirement (Rasner and Etgeton 2014). A ings substitution effects. The literature therefore points substantial portion of older workers have a gap between out that household production increases the elasticity of employment exit and retirement entry. This gap is called labor supply because home-made goods are substitutes “bridge option” if individuals do not withdraw from the for commercial goods (Cahuc et al. 2004; Mincer 1962). labor market between the exit from their last employ- At first sight, the labor supply mechanisms for all indi - ment and retirement (Brussig 2015). There are two bridge viduals in working age also apply for retirees. The share options, an unemployment spell between work exit and of older women engaged in household production (e.g., Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 5 of 17 21 (early) retirement and partial retirement. Partial retire- 2009). Employers therefore are more successful in nudg- ment may mean that employees work full time for half ing employees with a lower labor market attachment at the partial retirement period at reduced earnings and old age into partial retirement. The group of employees exit employment completely for the second half. The so- who bridge the time between employment and retire- called block model version of partial retirement has been ment by unemployment may mainly consist of workers chosen by more than 80% of participants (Brussig 2015). who have been dismissed against their wills and there- A much smaller share of retirees chose the so-called con- fore their value of leisure is unclear. The German labor tinuous version of partial retirement. Here the employee law especially protects employees with high tenure and/ reduced working-time to half during the entire partial or experience against dismissal, however (Ullmann and retirement period and no gap between labor market exit Bothfeld 2008). In addition, many judges in labor courts and pension entry occurs. Both bridge options can be are employee-friendly in dismissal disputes. As a con- used by employees after 15 contribution years to pension sequence, the appeal against an involuntary dismissal retirement and at least 8 years of contributions periods in before a labor court has a high success rate especially the last 10 years before retirement. Bridge options there- for older workers. Employers are exposed to high uncer- fore are not viable for employees with large labor market tainty with respect to the outcome of the lawsuit that eas- gaps and a low labor market attachment in general. ily takes more than 1 year (Ullmann and Bothfeld 2008) In both bridge options, employees usually have to and they frequently face a detrimental effect on their accept financial disadvantages in comparison to their reputation when they unilaterally dismiss older employ- other (early) retirement options, but they can exit ees. Therefore, only a very small share of older employees employment already before their ERA (Lorenz et al. is dismissed against their will and most dismissed older 2020). Indeed, labor market exit age on average is employees obtain substantial severance pay (Schmähl between one to two years earlier for those using par- 2003; Grund 2006). Severance and compensation pay- tial retirement or retirement after unemployment than ments for a dismissal or partial retirement however are for those choosing the early retirement options without subject to individual bargaining between employer and employment gaps before retirement that also would have employee. Therefore, employers may prefer to nudge been available for those who used bridge options (Geyer employees with a relatively strong preference for leisure et al. 2021). For those who use both bridge paths, we into unemployment before early retirement because dis- hardly observe a return to regular employment before missing them is cheaper. Our fifth hypothesis is: retirement. We therefore assume that employees who choose a bridge option put a relatively high value on Hypothesis 5 Individuals using bridge paths into leisure at old age and they therefore are less inclined to retirement have lower earnings elasticities than individu- work after retirement. A lower inclination to work after als who are employed until retirement. retirement also has been found for those who did not work full time until retirement (Burkert and Hochfell- Individuals with different places of work and resi - ner 2017; Maestas 2010; Smeaton and McKay 2003)— dence before retirement accept commuting costs. those who use the so-called continuous version of partial According to labor supply theory, at given earnings, retirement are a case in point for part-time work before commuting costs reduce the opportunity cost of lei- retirement. sure and therefore reduce labor supply (Cogan 1977, The argument that employees who select partial retire - 1981). Gutiérrez-i-Puigarnau and van Ommeren (2010) ment have a higher value of leisure seems to be obvious assume a strict complementarity between commuting because employees have to give their consent for trad- costs and labor supply. Their arguments for example ing earnings against an earlier employment exit (Wanger Erlinghagen (2005) and Ullmann and Bothfeld (2008) show that the average Please note that the concept of bridge employment in Germany differs from share of involuntary job terminations was between 20 and 30 percent after that used in the US (Beehr and Bennett 2015). In Germany, bridge employ- 1996 and that the share of involuntary job terminations was strongly nega- ment means that employees exit employment earlier then retirement or tively correlated with age. reduce their working time in their career job or they exit into unemployment This assumption of course applies under the ceteris paribus assump- from their career job before they enter retirement. In the US, bridge employ- tion. We know that partial retirement is mainly accepted by better edu- ment usually means taking another job after their career job. cated employees who work for large (industrial) employers (Engstler and Almost all employees in bridge options in the cohorts we look at also Romeu-Gordo 2017) and bridge options are mainly offered by employers are entitled to use other early retirement options such as the pension for with higher employee adaptation costs, i.e. large but shrinking employers in women and the pension for the long-term insured. Indeed, many employees regions with high unemployment (Lorenz et al. 2020). It is therefore impor- in bridge options used these alternative early retirement options after their tant to control for the mentioned individual and employer characteristics in bridge option. order to avoid biased results. 21 Page 6 of 17 S. Lorenz , T. Zwick on the negative effects of road pricing on labor sup - 1975 to 2014 (SIAB, 1975–2014). Out data therefore ply however are based on the average employee situa- only include employees subject to social security and tion in which males work full-time more often than employees in marginal employment —civil and mili- females, commute relatively long distances, and are lit- tary servants (around five percent of the workforce) and tle affected by economic incentives with respect to their self-employed (about ten percent of the workforce) are work participation (compare footnote 38 in Gutiérrez-i- excluded and unpaid care, household as well as volunteer Puigarnau and van Ommeren 2010). The insights from work is not unfortunately unknown. this paper on the impact of commuting costs and labor The data provide daily information about earnings and supply therefore may not apply for the decision to work employment as well as receipt of benefits according to after retirement. Given that older employees only work German Social Books II and III. Moreover, an extensive few hours and that they have a time perspective that is set of establishment information from the IAB Establish- too short to look for a new home near a new employer, ment History Panel is linked to the individual employ- the necessity to (continue to) commute to a job after ment history. We know the day at which employees exit retirement may have a negative effect on labor supply. from the last employment subject to social insurance We therefore assume that potential commuting costs contributions before retirement or the day at which the of individuals whose place of work is not the place of bridge options end. Individual and company-specific residence before retirement are higher. This group of characteristics are measured at this point in time. employees therefore may be less attracted by financial We know daily earnings of employees after retirement incentives to work after retirement: and whether an employee was in marginal employment or worked more. Earnings from marginal employment Hypothesis 6 Individuals whose place of work and and pension entitlements are income tax free and do not residence is in the same region before retirement have require social security contributions in contrast to labor higher earnings elasticities than individuals who have to earnings beyond marginal employment. We want to commute. have comparable earnings indicators for employees in marginal employment and employees with higher earn- ings. We argue that disposable income is more relevant Our last hypothesis concerns heterogeneity by income for the decision of whether to work or not than gross quartile before retirement. Bastani et al. (2021) show earnings. Therefore, we use net earnings after taxes and that earnings elasticity decreases with potential earn- social contributions for labor earnings above the mar- ings. They argue that those with low potential earnings ginal employment threshold. More specifically, we deduct have a lower labor market attachment and therefore earnings taxes for earnings tax class 1, the solidarity sur- their potential for entry into the labor market is higher. charge, church tax and social security contributions to For our group of old workers who have to decide health and nursing care insurance. Approximately 80% whether to re-enter employment, this argument is not of all employees are in marginal employment after retire- valid because their labor market attachment is zero ment. The share of retired women who have labor earn - when entering retirement. We however argue that the ings beyond the marginal employment level is with 18% risk of having pension entitlements below the individual somewhat smaller than the share of men (22%). income aspiration level after retirement (“old-age pov- The age at which pensions can be received in Ger - erty”) is higher for those with low labor earnings before many depends on the fact whether the individual is eli- retirement (Burkert and Hochfellner 2017). The finan - gible for early retirement or not. There are several early cial aspect of work after retirement therefore may play a larger role for those who earned less before retirement (Fasbender et al. 2016): A detailed description of the SIAB can be found in Antoni et al. (2016). Hypothesis 7 Individuals with lower earnings before Marginal employment is defined as dependent employment with a maxi- retirement have higher earnings elasticities. mum monthly salary of 400 EUR (raised to 450 EUR in 2013). The employment shares have been taken from Institut für Arbeit und Qualifikation (2020). In Germany, it is compulsory to report the end of employment to the 3 Data social insurance system after the statutory retirement age. In the case of Our analysis is based on a large and high-quality admin- employment after retirement, the employer must provide a new contract istrative dataset provided by the Federal Employment and submit a new declaration to the social insurance, even if the employee was already employed by the company. Thus, employment is interrupted Agency in Germany (Bundesagentur für Arbeit). The after pension entry by 1 day (Burkert and Hochfellner 2017). data comprise a sample of 2% of employees from the Tax and deduction exemptions on marginal employment were abolished Integrated Employment Biographies ranging from just in 2012. Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 7 of 17 21 retirement options that are tied to different requirements reduce pension claims (Westermeier 2019). Unfortu- such as pension contribution years, gender, or disabil- nately, there are no statistics on the incidence of this ity. The most important early retirement options are the group. pension for long-term insured employees, pension for women, pension for severely disabled, and the two bridge 3.1 Estimation sample options pension after unemployment and pension after Our estimation sample consists of women and men born partial retirement, see Lorenz et al. (2018) for an in- between 1935 and 1947. For each birth cohort, post- depth description of the early retirement rules and their retirement employment is observed for a maximum changes over time. Individuals who are eligible for one or of two years after the NRA. As work incidence quickly several early retirement options can exit the labor market declines with age after 65 in Germany (compare for and start to collect pensions at the ERA specified by the example Pfarr and Maier 2015), we are confident that we early retirement option (in many cases 60 or 63 years). do not miss any retirees who start to work in retirement All other individuals only can collect pensions at the after their 67th birthday. We restrict our sample to West- NRA (65 years). People born after 1937 in addition can German men and women with a strong labor market retire after the NRA. They receive a 6% increase in pen - attachment in old age. More specifically, all individuals in sion entitlements for each year of pension postponement. the sample must be employed subject to social security We cannot observe when individuals start to receive at least once after the age of 55 and they must be active pensions directly in our data and therefore define in the labor market (employed, in partial retirement, or two groups with employment after receiving a pen- unemployed) at age 59. The reason for these sample sion. The first group consists of individuals who have restrictions is to obtain a relatively homogeneous sample a labor market spell after their NRA. The second group of employees who in principle all have the option to work consists of individuals who already started a marginal after retirement. The restriction reduces our sample by employment before the age of 65 that extends beyond about 30%. the age of 65 without interruption and who are eligi- It is relatively hard in Germany to return to the labor ble for early retirement. We may incur three sources market after longer labor market gaps at higher age—in of measurement errors with this identification of indi - the group of employees excluded from our sample, the viduals who work after retirement. First, some indi- share or employees who work after retirement accord- viduals work after NRA without receiving a pension. ingly is less than one percent. Our sample restrictions The number of employees who postponed their entry also reduce the risk that we include individuals who also into retirement after NRA always was negligible. We were civil servants or self-employed besides or after their include these individuals because they are employed employment subject to social security. There are age after NRA but incorrectly assume that they also draw restrictions on starting a job with a civil servant status pension benefits after age 65. Second, there may be and it is hardly possible to work in a job subject to social individuals who receive pensions besides working but security besides a job as civil servant. Employers usually stop to work before age 65. We cannot observe this do not allow their employees to generate a substantial group of people and therefore our results only cover all income from self-employment besides their job subject individuals who work besides receiving retirement ben- to social security. efits after NRA. The third group works more than mar - After the sample restrictions we are left with 30,784 ginally before NRA in addition to receiving pensions. women and 44,887 men—we therefore include 75,671 This group again is negligibly small because earnings observations. In the sample, 6273 women and 9549 beyond the threshold of marginal employment strongly men continue to work after retirement, correspond- ing to an average of 20% for women and 21% for men, The SIAB data do not include relevant characteristics of the pension insur - ance, such as pension entitlements or eligibility. Therefore, individual pension entitlements are calculated and eligibility criteria for old age pensions and the corresponding statutory retirement dates NRA and ERA are identified according to Pfister et al. (2018) and Lorenz et al. (2018). We are grateful to Labor market careers are only available for East Germany after January 1, Philip vom Berge and Dana Müller from the FDZ at the IAB for merging the 1991 (Antoni et al. 2016). day of birth as part of the Custom Shaped Administrative Data for the Analy- sis of Labor Market (CADAL) project because the calculation of pension enti- This sample restriction was also made by Hanel (2010) and Geyer et al. tlements requires the exact birth date. (2021). 15 18 Bäcker et al. (2017) for example report that in 2014, from about 825,000 We also exclude seamen and miners (less than one percent of women retired individuals only about two percent or 22,000 individuals received a and of men) because they have special protection of legitimate expectation bonus. rules for retirement that cannot be identified in the dataset. 21 Page 8 of 17 S. Lorenz , T. Zwick Table 1 Summary statistics on work and non-work earnings Daily net labor earnings (EUR) Daily pension entitlements (EUR) Share of net labor earnings in pension entitlements (%) Women Men Women Men Women Men All employees who work after retirement Mean 14.013 19.770** 23.016 39.122** 65.94 52.97** SD 14.037 24.660 9.210 11.737 60.68 65.59 N 6273 9549 6273 9549 6273 9549 Employees with marginal employment after retirement Mean 9.406 9.514 22.637 38.510** 46.93 27.06** SD 3.416 3.453 8.643 10.584 24.70 15.28 N 5175 7350 5175 7350 5175 7350 SD standard deviation, N number of observations Data: SIAB 7514, own calculations Significance levels of differences between men and women: *p < 0.05, **p < 0.01 19 20 compare Appendix Table 7. Across birth cohorts, the retirement are 14 EUR for women and 20 EUR for men. share of individuals who work beyond retirement has For those who are marginally employed, daily earnings risen slightly. The number of observations and the share for both sexes are around 9 EUR. Pension entitlements of individuals who continue to work after retirement by of women who continue to work after retirement are birth cohort are shown in Appendix Table 7. on average 40% lower than those of men. Of particular Our data set has some drawbacks such as the absence of interest is the share of labor earnings after retirement information on pension entry and other earnings sources in pension entitlements. This share is quite substantial beyond labor earnings and pension entitlements. It how- with 66% for women and 53% for men. For employees in ever has the advantage of including reliable and large- marginal employment, the share is 47% for women and scale information on work behavior of older employees 27% for men. These findings are in line with evidence from many birth cohorts. Studies on the relationship from the UK where labor earnings after retirement were between financial incentives and work beyond retire - more than 90% of average pension entitlements (Kanabar ment so far typically relied on small-scale and potentially 2015). We therefore conclude that earnings after retire- specific populations (Kim and Feldmann 2000; Saba and ment are an important additional source of income for Guerin 2005; Torka et al. 2012) or on survey-based data securing the standard of living in old age. that may be affected by self-selection and measurement We also can show that those who continue to work error in key variables (Anger et al. 2018; Büsch et al. after retirement on average have a significantly lower 2010; Dittrich et al. 2011; Kanabar 2015; Pfarr and Maier pension entitlement than individuals who do not work 2015, Saba and Guerin 2005). after retirement (Appendix Table 8). This finding corre - sponds with the negative correlation between non-labor 3.2 Descriptive statistics earnings and re-entry into employment after retirement Our analysis focuses on labor and non-labor financial found in most of the previous literature (Burkert and incentives for employment beyond retirement. The fol - Hochfellner 2017; Büsch et al. 2010; Dittrich et al. 2011; lowing descriptive statistics accordingly show average Hochfellner and Burkert 2013; Kim and Feldmann 2000; daily labor earnings, average pension entitlements, and Micheel et al. 2010). Appendix Table 8 also summarizes the average share of labor earnings in pension entitle- ments by gender and type of employment after retire- ment (Table 1). The average daily net earnings after Women have no earnings after retirement above the social contribution earnings threshold and for men the proportion with earnings above the earn- ings threshold is 0.3%. We therefore do not have to correct for censored earn- ings. Please note that we cannot take into account other sources of income, The proportion of employees who continue to work after retirement is such as private and occupation pensions, capital and real estate rents, inde- slightly lower than that calculated by Anger et al. (2018). In their study, how- pendent work or savings. ever, employment after retirement is identified from survey answers. The share of older employees who indicate that they intend to work after retire- Two studies from Germany (Anger et al. 2018; Pfarr and Maier 2015) and ment in surveys even is between a third and a half (Micheel et al. 2010; one study from England (Kanabar 2015) however show that employees with Micheel 2021). higher financial means are more likely to work beyond retirement. Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 9 of 17 21 differences in individual and establishment characteris - accumulated pension entitlements. Pension entitlements tics prior to retirement between employees who continue indicate the negative income effect in the participation to work after retirement and those who do not. This table equation and they are an indicator of past labor market shows, for example, that there is a negative correlation performance. Given that all employees obtain a new labor between work after retirement and employer size as well contract, usually entailing less labor hours and new tasks, as the share of highly skilled workers. previous labor market performance is hardly related to (potential) earnings after retirement, compare Appendix 4 Estimation strategy Table 9. Pension entitlements therefore can be excluded We propose an empirical choice model to examine the from the earnings equation. decision on working after receiving a pension: LFP = β + X β + Z β + γpe + δwa ge + ε LFP = 1 if LFP > 0 (1) i 0 i x j i z i ( ) i i LFP = 0 if LFP = 0 The labor force participation variable (LFP) is a dichot - Besides the exclusion restriction and the Mills ratio in the omous variable with a value of zero for non-participation earnings equation, we include core individual characteris- (no labor earnings) and a value of one (positive labor tics that have been used in the literature as determinants earnings) for employment after receiving a pension. The of work after retirement (Anger et al. 2018; Brenke 2013; unobservable continuous variable “LFP*” reflects the util - Brussig 2010; Burkert and Hochfellner 2017; Engstler and ity difference between non-participation and participa - Romeu-Gordo 2017; Fasbender et al. 2016; Hofäcker and tion in the labor market beyond retirement. Our main Naumann 2015; Larsen and Pedersen 2013; Maestas 2010; variables of interest are the log disposable pension enti- Micheel et al. 2017; Rhein 2016; Westermeier 2019). More tlements “pe” (non-labor earnings) and the net (poten- specifically, we include education, occupation, accumu- 23 25 tial) labor earnings “wage” when receiving a pension. lated labor market gaps until age 65 in months, and the Labor earnings after retirement may be endogenous year of labor market exit. We keep the auxiliary regressions because they also reflect time allocation decisions, that parsimonious in order to avoid multicollinearity problems is, they are only observed for those who work after retire- in our main labor participation regression (Borra 2010). ment. In order to circumvent this endogeneity problem, In the main labor participation equation, we include predicted daily labor earnings after retirement wa ge all explanatory variables used in the auxiliary participa- are estimated for all individuals in an auxiliary estima- tion equations and (potential) labor earnings. In addition, tion. The standard Heckman sample selection approach we include dummies that equal one if an individual uses (Heckman 1976, 1979) is used to predict the earnings of the bridge option partial retirement or unemployment post-retirement employment. More specifically, we first because labor market gaps before retirement negatively estimate a reduced-form post-retirement employment affect the continuation of employment after retirement participation Probit model for all individuals in the sam- (Anger et al. 2018; Burkert and Hochfellner 2017; Schel- ple. We use the Probit parameters from this participation lenberg et al. 2005; Smeaton and McKay 2003; West- equation to generate the inverse Mills ratio. Including ermeier 2019). We also include several characteristics this ratio controls for possible sample-selection bias in of the last employer before retirement. Previous stud- the wage equation. We then take the results of the wage ies have shown that employees in small establishments equation to generate values of the predicted wage for all have a higher probability of working beyond retire- individuals in the sample (Borra 2010). ment (Anger et al. 2018; Burkert and Hochfellner 2017; Technically, this type of selectivity corrected model Micheel et al. 2010). In addition, Brussig and Bellmann can achieve identification by functional form assump - (2008) show that attitudes towards hiring older workers tions (Cameron and Trivedi 2002). Nonetheless, most differ among companies. For example, employees in com - researchers feel more comfortable if at least one regres- panies with a high proportion of older employees have a sor in the participation equation is excluded from the labor earnings equation (Connelly and Kimmel 2003; The Blossfeld classification of occupations is intended to form occupational Borra 2010). We accordingly include in the reduced form groups as homogeneously as possible with regard to average educational and labor participation equation as exclusion restriction the vocational background and professional fields of activity (Blossfeld 1985). An alternative for the aggregated gaps in the labor market before the age of 65 is the exit age from employment before the statutory retirement age. In a further estimation specification, instead of the expected earnings and The results are robust. It is not possible to include both variables in the esti- the pension entitlements, the ratio of the (potential) earnings in pension enti- mations because they are strongly correlated. Early retirement is associated tlements is included in the equation. with more gaps in the labor market before the age of 65. 21 Page 10 of 17 S. Lorenz , T. Zwick Table 2 Linear probability model of labor force participation Dependent variable: labor force participation Women Men Individual characteristics Estimated earnings (log) 0.056* (0.019) 0.246** (0.055) Pension entitlements (log) − 0.055** (0.006) − 0.165** (0.008) Bridge option: partial retirement − 0.078** (0.007) − 0.060** (0.005) Bridge option: unemployment − 0.323** (0.006) − 0.266** (0.005) No degree Ref. Ref. Vocational training 0.032** (0.006) 0.025 (0.013) University degree 0.005 (0.012) − 0.075* (0.031) Manufacturing occupations Ref. Ref. Service occupations 0.011 (0.008) − 0.021 (0.014) Administrative occupations − 0.010 (0.008) − 0.042** (0.012) Accumulated labor market gaps prior to age 65 in months − 0.006** (0.0007) − 0.001 (0.001) Employer characteristics Firm size: < 10 employees Ref. Ref. 10 to 100 employees − 0.027** (0.006) − 0.019** (0.007) > 100 employees − 0.044** (0.007) − 0.059** (0.008) Mean imputed earnings of all full-time employees − 0.0004** (0.00009) − 0.000006 (0.00006) Share of highly qualified employees − 0.027 (0.016) − 0.112** (0.016) Share of employees in partial retirement − 0.204** (0.061) − 0.128** (0.049) Share of employees aged 55–59 − 0.030 (0.017) − 0.013 (0.017) Share of employees aged 60–64 − 0.033 (0.023) − 0.065** (0.021) Share of employees aged 65 and older 0.514** (0.048) 0.527** (0.051) N 30,784 44,887 R 0.307 0.210 Additional controls: year of employment exit, and economic sector of the employer. Standard errors in parentheses are computed by bootstrapping with 200 repetitions and are clustered on the individual level. Data: SIAB 7514 *p < 0.05 **p < 0.01 higher chance of being hired after their retirement than in Appendix Table 10. We use potential earnings for those in companies with a high proportion of employees individuals who do not work after retirement from this in partial retirement and severance payments. Moreo- estimation. In the earnings equations, selectivity of ver, some studies also use the general economic environ- employees who decide to work after retirement is taken ment as a determinant of employment after retirement. in account by including the Mills’ ratio (“lambda”) derived from the parsimonious first-step labor mar We therefore include the mean imputed earnings of all - full-time employees, economic sector, establishment size, ket participation equation. The Hausman specification share of employees in partial retirement, share of highly test checks whether our exclusion restriction is valid to qualified employees, share of employees aged 55–59, control for endogeneity. The high χ values indicate that share of employees aged 60–64, and the share of employ - the pension entitlements variable is a valid exclusion ees aged 65 and older. restriction for the labor market participation equation. We take into account that the decision to work after retire- We present our estimation results of the second estima- ment is fundamentally different between men and women tion step, the decision to work beyond retirement, in Table 2, because the non-market production situation usually differs separately for women (column I) and men (column II). The between the genders. We therefore estimate all earnings and OLS estimates suggest a substantial positive effect of the participation equations separately for men and women. (potential) earnings on labor supply for women and men. According to Hypothesis 1, an increase in (potential) labor earnings by one percent significantly increases the prob 5 Estimation results ability of being active in the labor force by 0.056 percentage 5.1 Emplo yment decision to work beyond retirement points for women and by 0.246 percentage points for men, The first estimation step, the calculation of the earn - respectively. In other words, an increase in (potential) labor ings equation for all individuals in our sample, is shown Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 11 of 17 21 earnings by 10% has the potential to increase labor market Table 3 Linear probability model of labor force participation with the share of estimated earnings in disposable income participation of men from their average 20.5% (see Appendix Table 7) to about 23% and of women from 19.3% to almost Women Men 20%. According to Hypothesis 4, labor earnings elasticity Share of estimated earnings in 0.222** (0.054) 0.816** (0.057) is much higher for men than for women. Consistent with disposable income Hypothesis 2, an increase in pension entitlements by one N 30.784 44.887 percent significantly reduces the employment probability R 0.306 0.206 by 0.055 percentage points for women and 0.165 percentage Same list of covariates as in Table 2. Standard errors in parentheses are points for men. Hochfellner and Burkert (2013) find a simi - computed by bootstrapping with 200 repetitions and clustered on the lar pattern by gender, their income elasticities are however individual level. Data: SIAB 7514 somewhat lower. Pleau and Shauman (2013) find a smaller *p < 0.05 negative effect of retirement income on working after retire - **p < 0.01 ment for males than for females in their US sample, however. A higher pension entitlement leads to a stronger reduc- Table 4 Extensive margin of labor supply for employees with tion in labor market participation for men than for women, and without bridge paths accordingly. Bridge paths The individual characteristics show the expected signs on employment after retirement (compare for exam- Women Men ple Burkert and Hochfellner 2017; Larsen and Pedersen No Yes No Yes 2013; Maestas 2010): employees who use bridge paths have a significantly lower probability to work and there - Estimated 0.095** (0.024) 0.057* (0.029) 0.307** (0.095) 0.244** earnings (log) (0.064) fore labor market gaps before retirement have a negative N 19,887 10,897 19,223 25,664 correlation with labor market participation after retire- R 0.407 0.067 0.287 0.071 ment. Academics work less after retirement than skilled Same list of covariates as in Table 2. Standard errors in parentheses are employees. Occupations before retirement hardly have 27 computed by bootstrapping with 200 repetitions and clustered on the an influence on working after retirement, however. individual level. Data: SIAB 7514 These are the effects of the establishment character - *p < 0.05 istics of the last employer prior to retirement on the **p < 0.01 labor market decision of the retiree: workers who are employed in firms with more than ten employees prior to retirement show a lower labor force participation after The estimation equation in Table 3 replaces the sepa- retirement. Furthermore, there is a negative correla- rate labor earnings and pension entitlement variables by tion between mean earnings of all full-time employees, the share of earnings in disposable income. This variable the share of highly qualified employees, and the share has a substantial positive impact on the extensive margin of employees in partial retirement and work after retire- of labor supply. An increase in the share of (potential) ment. The higher the share of employees who work after labor earnings in disposable income after retirement by retirement at the last employer, the higher is the prob- one percentage point increases the probability of working ability of being employed after retirement. beyond retirement by 0.22 percentage points for women Note that the (potential) labor earnings regressor is and 0.82 percentage points for men. This result supports generated because it has been estimated in a separate Hypothesis 3. step with uncertainty (Borra 2010). Therefore, the stand - ard errors of this variable are bootstrapped. A bivariate 5.2 Heterogeneity in earnings elasticities Probit model is used to verify the results of the linear Table 4 shows differences in the estimated elasticity of probability model presented in Table 2. The results of labor force participation with respect to (potential) labor both estimation models are consistent with each other. earnings between employees who use bridge paths versus those who do not use bridge paths prior to retirement. In line with Hypothesis 5 and prior empirical evidence We obtain larger coefficients for the estimated earnings variables in the (Anger et al. 2018; Burkert and Hochfellner 2017; Hoch- labor participation equation if we drop the endogeneity correction (Mills fellner and Burkert 2013), the estimates suggest a higher ratio) in the earnings estimation step. (Potential) earnings however remain positive earnings effect on labor supply for employees significant and larger for men than for women without controlling for endo - geneity. without bridge paths than for employees with bridge All results remain robust when only individual characteristics are paths. The earnings elasticities for employees without a included in the participation equation and employer characteristics are not bridge path are twice as large for women and 50% larger controlled for. 21 Page 12 of 17 S. Lorenz , T. Zwick Table 5 Extensive margin of labor supply for employees whose Table 6 Extensive margin of labor supply for employees with region of residence is the same as or different from the place of low and high earnings before retirement work Earnings before retirement Region of residence is same as region of work Women Men Women Men Low High Low High earnings earnings earnings earnings No Yes No Yes Estimated 0.084 (0.045) 0.063 (0.039) 0.384** 0.097 (0.110) Estimated 0.062** 0.080 (0.083) 0.163** 0.234* (0.146) ernings (0.129) earnings (0.026) (0.084) (log) (log) N 7441 7441 10.706 10,706 N 19,080 2289 25,799 6769 R 0.365 0.273 0.283 0.171 R 0.313 0.315 0.239 0.233 High earnings are defined as earnings from the last employment subject to List of covariates is the same as in Table 2. Number of observations lower than in social insurance contributions with earnings above the 75th percentile of all last Table 2 because region of residence is not reported for all employees. Standard earnings of employment subject to social insurance contributions. Low earnings errors in parentheses are computed by bootstrapping with 200 repetitions and are defined as earnings from the last employment subject to social insurance clustered on the individual level. Data: SIAB 7514 contributions with earnings less than 25th percentile of all last earnings of *p < 0.05 employment subject to social insurance contributions. List of covariates **p < 0.01 is the same as in Table 2. Standard errors in parentheses are computed by bootstrapping with 200 repetitions and clustered on the individual level. Data: SIAB 7514 *p < 0.05 for men than those for employees with bridge paths, see **p < 0.01 Table 5. Employees without bridge paths therefore seem to derive lower utility from leisure and they are more responsive to financial incentives when deciding about quartiles both have lower elasticities than the second and employment after retirement. third earnings quartiles and therefore Hypothesis 7 is not When we compare earnings elasticities between supported for women. Our evidence only is in accord- employees who have to commute to work to a different ance with findings by Pleau (2010) and Pleau and Shau - region in their last employment prior to retirement and man (2013) for US men. They however find an increasing those who live and work in the same region, we use the labor market supply elasticity with pre-retirement earn- approach proposed by Kropp and Schwengler (2016) ings for US women. The low labor supply elasticity of to delineate 50 functional West German labor market women with very low earnings before retirement may be regions based on commuting flows. The dummy has a consequence of high alternative financial sources for value one if an employee commutes to a workplace out- example from other household members that we cannot side the own labor market region. Our results do not sup- observe in our data (Blundell et al. 2007; Micheel et al. port Hypothesis 6: there are practically no differences 2010). between commuters and non-commuters for females and the differences between both employee groups are small 6 Discussion and limitations for men, see Table 6. We hardly find changes in the Our paper uses the empirical identification approach commuting situation before and after retirement. Com- developed for the employment decision of women with muting costs therefore seem not to play a central role for small children (Allègre et al. 2015; Borra 2010; Ram- the labor supply decision of older employees for example mohan and Whelan 2007; Ribar 1992; Viitanen 2005). in comparison to financial needs of older employees. This literature mainly uses structural estimation mod - Our last sample split differentiates employees accord - els because there are hardly any quasi-natural experi- ing to their earnings situation before exiting the labor ments that can be exploited to obtain causal relationships market into retirement. According to Hypothesis 7, we between financial incentives and the employment deci - find a higher elasticity of men in the lowest earnings sion. Also the literature on determinants for work after quartile in comparison to men in the highest earnings quartile. For women, the lowest and highest earnings The labor market regions are defined based on the place of work and the residential regions are defined based on the place of residence at the last employment prior to retirement. 29 30 The results remain robust when the information about place of work and A notable exception is Bastani et al. (2021) who analyze the impact of a place of residence are directly used from the dataset instead of the approach change in labor market participation tax rates on labor supply of secondary proposed by Kropp and Schwengler (2016). earners. Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 13 of 17 21 retirement usually does not exploit quasi-natural experi- example were marked differences with respect to the ments (Chetty et al. 2011). The main reason for the dearth role of women on the labor market and in the household of quasi-experimental evidence is that there were hardly between East and West Germany before and shortly after any changes in the rules for labor earnings after retire- the reunification of both countries (Sprengholz et al. ment. In Germany, earnings beyond marginal employ- 2020). In addition, there seems to be a secular reduc- ment have to be taxed at the common rates, marginal tion in wage elasticities over the last decades in many employment is tax free and the maximum earnings level countries (Bargain and Peichl 2016). It therefore remains for marginal employment was practically unchanged. In unclear what the influence of the observation period and 2005, a pension supplement of 6% per year of retirement institutional differences is for estimations of labor sup - after NRA was introduced for all pension entitlements. ply elasticities in other countries or in the Eastern part of This option was used by a negligible share of retirees, Germany. however. Besides using a structural approach our study is limited by 7 Conclusions several gaps in relevant information. Our data set does not This paper provides a labor supply model for retirees and report hours worked and we therefore only could analyze empirical assessments based on a large administrative labor the extensive margin. Future work, as a consequence, should market history data set. We for the first time use detailed analyze retiree´s labor supply respect to the number of hours information about labor earnings and non-labor income of worked conditional on employment (intensive margin). In a large administrative 2% sample of all employees after their addition, our administrative data only include pension enti- retirement in Germany. We concentrate on the effect of tlements derived from labor earnings. Future work therefore labor earnings on the labor force participation beyond retire- should include additional available financial means during ment and take endogeneity of the labor decision and hetero- retirement such as occupation pensions, additional pen- geneities between different employee groups into account. sion entitlements for example from raising children or rent We first establish that labor earnings achieved by pension - from wealth, assets, and real estate. Also considerable finan - ers after retirement constitute a substantial part of disposable cial demands such as mortgage debt or financially depend - income. The share of post-retirement earnings in dispos - ent household have an important impact on the decision to able income is more than 60% for women and 50% for men. work that is unobservable for us. Moreover, decisions about We then show that men are more responsive to financial labor supply after retirement may result from considera- incentives than women. A one percent increase in (poten- tions involving other individuals such as pension redistribu- tial) post-retirement labor earnings increases the employ- tions after a divorce or widow/widower pensions. Especially ment probability of pensioners by 0.246 percentage points equivalent household income may be a more important for males and by 0.056 percentage points for females. Labor driver for work after retirement than the individual financial earnings therefore have a strong incentive effect for work situation. We know in addition that males and females react after retirement. Higher pension entitlements however lead differently to financial stimuli within the household context to a negative labor participation effect. We analogously find (Blundell et al. 2007; Micheel et al. 2010). The distribution that the higher the share of (potential) labor earnings in dis- of wage elasticities of single childless women is very similar posable income of the retiree, the higher is the labor force to that of men. It therefore is mainly married women and participation of the individual. We also show that individu- mothers who have higher wage elasticities (Bargain and Pei- als who work until retirement instead of using the bridge chl 2016). As a consequence, also the household and marital options unemployment or partial retirement as well as men situation and changes in behavior of household members not with low earnings prior to retirement are stronger attracted affected by financial incentive changes should be included in by (potential) earnings incentives to work after retirement. future research in order to get a full and unbiased picture of Having to commute before retirement in contrast to residing financial incentives for work after retirement. in the same region as the place of work does not influence Transferability of our results to other countries work participation after retirement, however. depends on the empirical question whether there are dif- Our analysis therefore shows that tax incentives or ferences in labor supply elasticities between countries. higher earnings offers by employers are most effective It seems that there are little international differences in to stimulate labor market participation after retire- work preference (Bargain et al. 2014), although there for ment for males who earned not much before retire- ment and did not use bridge options. The calculation of the separate elasticities for labor earnings and pen- sion eligibilities for example allows us the prediction The maximum earnings level for minimum employment was increased on 1 of changes in financial incentives on work supply after January 2013 from 400€ to 450€, but further unchanged during our observa- retirement for different demographic groups. More tion period. 21 Page 14 of 17 S. Lorenz , T. Zwick specifically, our analysis shows that the gradual reduc- Appendix tion of tax-free allowance to zero until 2040 that was See Tables 7, 8, 9, and 10. induced by the Retirement Income Act 2004 (Genser Table 7 Number and share of individuals who work beyond and Holzmann 2019) may increase incentives to work retirement by birth cohort after retirement. We also show that an increase in the tax-free allowance for work during pension also may Birth cohort Women Men have a positive effect for the older employees affected N Share N Share by the measure. Average 2056 0.193 3180 0.205 8 Methods 1935 1934 0.192 3075 0.204 On the basis of theoretical hypotheses derived from a 1936 1992 0.202 3142 0.215 labor supply model, we propose an empirical choice 1937 2384 0.206 3586 0.178 model to examine the employment decision after 1938 2675 0.197 3933 0.207 receiving a pension. The decision to participate in the 1939 2582 0.206 4091 0.193 labor market is estimated based on a linear probabil- 1940 2807 0.174 4306 0.200 ity model using the employment status after retirement 1941 2783 0.192 4052 0.196 as the outcome variable. In order to impute (potential) 1942 2336 0.194 3292 0.201 labor earnings for all individuals included in the sam- 1943 2330 0.204 3334 0.215 ple as explanatory variable, a two-step Heckman-type 1944 2418 0.208 3201 0.206 selection model is used. For the estimations, a sample 1945 1775 0.235 2551 0.254 of 2% of employees from the Integrated Employment 1946 2264 0.223 2934 0.261 Biographies from 1975 to 2014 (SIAB, 1975–2014) is 1947 2504 0.229 3390 0.265 used. Data: SIAB 7514 Table 8 Sample characteristics Work beyond retirement Women Men No Yes Diff. No Yes Diff. Employee characteristics prior to retirement Pension entitlements 790.119 (314.7) 696.246 (278.6) 93.872** (4.354) 1283.311 (414.2) 1183.442 (355.01) 99.869** (4.641) Estimated earnings (log) − 0.016 (0.982) 1.082 (0.593) 1.098** (0.013) 1.236 (0.618) 1.710 (0.498) 0.474** (0.007) Observed earnings (log) 2.352 (0.721) 2.520 (0.887) Share of unemployment bridge path 0.258 (0.438) 0.159 (0.365) 0.099** (0.006) 0.409 (0.492) 0.250 (0.433) 0.160** (0.006) Share of partial retirement bridge path 0.136 (0.342) 0.045 (0.207) 0.091** (0.005) 0.223 (0.416) 0.109 (0.312) 0.114** (0.005) Manufacturing occupations 0.175 (0.380) 0.143 (0.350) 0.033** (0.006) 0.550 (0.498) 0.468 (0.499) 0.082** (0.006) Service occupations 0.308 (0.462) 0.356 (0.479) 0.048** (0.007) 0.188 (0.390) 0.289 (0.453) 0.101** (0.005) Administrative occupations 0.517 (0.500) 0.502 (0.500) 0.015* (0.007) 0.262 (0.440) 0.244 (0.429) 0.019** (0.005) Accumulated labor market gaps prior to 32.777 (25.380) 4.309 (13.736) 28.468** (0.332) 19.575 (22.322) 3.438 (11.382) 16.137** (0.236) age 65 in months Characteristics of last employer prior to retirement Firm size 681.909 (2400.6) 366.367 (1549.8) 315.5* (31.885) 1788.5 (5732.3) 952.760 (4163.6) 835.7** (62.703) Mean imputed earnings of all full-time 89.652 (33.863) 80.465 (32.472) 9.188** (0.475) 108.250 (44.180) 96.706 (38.056) 12.54** (0.495) employees Share of highly qualified employees 0.116 (0.162) 0.088 (0.145) 0.029** (0.002) 0.131 (0.161) 0.098 (0.144) 0.033** (0.002) Share of employees in partial retirement 0.015 (0.038) 0.008 (0.028) 0.007** (0.0005) 0.022 (0.045) 0.015 (0.043) 0.007** (0.0005) Share of employees aged of 55 to 59 0.133 (0.135) 0.120 (0.134) 0.012** (0.002) 0.121 (0.110) 0.113 (0.110) 0.008** (0.001) Share of employees aged of 60 to 64 0.059 (0.105) 0.083 (0.137) 0.024** (0.002) 0.057 (0.098) 0.077 (0.123) 0.020** (0.001) Share of employees aged 65 and older 0.012 (0.039) 0.032 (0.081) 0.020 (0.0007) 0.009 (0.036) 0.028 (0.078) 0.020** (0.0005) N 24,511 6273 30,784 35,338 9549 44,887 Mean is given for continuous variables. Standard deviations of the continuous variables and standard errors for the mean value differences are given in parentheses. Data: SIAB 7514 Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 15 of 17 21 Table 9 Correlations between previous labor market experience and estimated as well as observed earnings Pension entitlements (log) Observed earnings before retirement (log) Women Estimated earnings (log) − 0.1117 − 0.0277 Observed earnings after retirement (log) 0.1582 0.2299 Men Estimated earnings (log) 0.0979 0.0592 Observed earnings after retirement (log) 0.1494 0.1962 Table 10 Auxiliary estimations Women Men Observations 30,784 44,887 Censored observations 24,511 35,338 Uncensored observations 6273 9549 Log likelihood − 11,791 − 19,581 χ (9) 7545.47 7300.83 Prob > χ : 0.2424 0.000 Labour force participation Log wage Labour force participation Log wage Coef. SE Coef. SE Constant 1.292** 0.176 − 0.637** 0.201 1.186** 0.0171 − 1.320** 0.211 No degree Ref. Ref. Ref. Ref. Ref. Ref. Ref. Ref. Vocational training 0.227** 0.027 0.231** 0.041 0.378** 0.028 0.231** 0.043 University degree 0.043 0.049 0.351** 0.062 0.149** 0.034 0.566** 0.046 Manufacturing occupations Ref. Ref. Ref. Ref. Ref. Ref. Ref. Ref. Service occupations 0.228** 0.029 0.261** 0.046 0.253** 0.018 0.223** 0.029 Administrative occupations 0.096** 0.028 0.239** 0.037 0.073** 0.018 0.198** 0.023 Acc. labor market gaps until 65 − 0.033** 0.0006 − 0.035** 0.004 − 0.031** 0.0005 − 0.022** 0.002 in months Pension entitlements (log) − 0.275** 0.025 − 0.494** 0.024 Lambda 1.205** 0.157 0.705** 0.084 Hausman test Chi (8) 57.73 69.00 Prob > chi : 0.0000 0.0000 In both equations, year of employment exit prior to the age of 65 is controlled for. SE: standard error. Data: SIAB 7514 *p < 0.05 **p < 0.01 Acknowledgements at the Institute for Employment Research (IAB) and subsequently via remote We are grateful to Mona Bruns for helpful discussions on this paper. data access. These data are publicly available. Authors’ contributions Declarations Both authors contributed in equal shares to the paper. Both authors read and approved the final manuscript. Ethics approval and consent to participate Not applicable. Funding Not applicable. Consent for publication Not applicable. Availability of data and materials The paper uses the SIAB7514. Data access was provided via on site use at the Research Data Centre (FDZ) of the German Federal Employment Agency (BA) 21 Page 16 of 17 S. Lorenz , T. Zwick Competing interests Cogan, J.F.: Labor supply with time and money costs of participation. Rand Not applicable. Report R-2044-HEW. Santa Monica CA (1977) Cogan, J.F.: Fixed costs and labor supply. Econometrica 49(4), 945–963 (1981) Author details Connelly, R.: The effect of child care costs on married women’s labor force Faculty of Business Management and Economics, University of Würzburg, participation. Rev. Econ. Stat. 74(1), 83–90 (1992) Sanderring 2, 97070 Würzburg, Germany. ZEW Leibniz Centre for European Connelly, R., Kimmel, J.: Marital status and full-time/part-time work status in Economic Research, Mannheim, Germany. Research Centre for Education child care choices. Appl. Econ. 35(7), 761–777 (2003) and the Labour Market, University of Maastricht, Maastricht, The Netherlands. Del Boca, D., Vuri, D.: The mismatch between employment and child care in Italy: the impact of rationing. J. Popul. Econ. 20(4), 805–832 (2007) Received: 12 April 2021 Accepted: 21 September 2021 Dittrich, D., Büsch, V., Micheel, F.: Working beyond retirement age in Germany: the employee’s perspective. In: Ennals, R., Salomon, R. (eds.) Older Work- ers in a Sustainable Society, pp. 189–202. Peter Lang, Frankfurt am Main (2011) Engstler, H., Romeu-Gordo, L.: Arbeiten im Ruhestand—Entwicklung, Faktoren und Motive der Erwerbstätigkeit von Altersrentenbeziehern. In: Kistler, References E., Trischler, F. (eds) Reformen auf dem Arbeitsmarkt und in der Alters- Allègre, G., Simonnet, V., Sofer, C.: Child care and labor market participation sicherung—Folgen für die Einkunftslage im Alter. edition der Hans-Böck- in France: do monetary incentives matter? Ann. Econ. Stat. 117(118), ler-Stiftung, Arbeit und Soziales 196, Düsseldorf, pp. 115–147 (2014) 115–139 (2015) Engstler, H., Romeu-Gordo, L.: Der Übergang in den Ruhestand: Alter, Pfade Anger, S., Trahms, A., Westermeier, C.: Erwerbstätigkeit nach dem Übergang in und Ausstiegspläne. In: Mahne, K., Wolff, C., Simonson, J., Tesch-Römer, C. Altersrente: Soziale Motive überwiegen, aber auch Geld ist wichtig. IAB- (eds.) Altern im Wandel, pp. 65–80. Springer, Wiesbaden (2017) Kurzbericht, 24/2018. Nuremberg (2018) Erlinghagen, M.: Entlassungen und Beschäftigungssicherheit im Zeitverlauf. Z. Antoni, M., Ganzer, A., vom Berge, P.: Stichprobe der Integrierten Arbeitsmark- Soziol. 34(2), 147–168 (2005) tbiografien (SIAB) 1975–2014. FDZ Datenreport, 04/2016. Nuremberg Eurofound: Income From Work After Retirement in the EU. Publications Office (2016) of the European Union, Luxembourg (2012) Bäcker, G., Jansen, A., Schmitz, J.: Rente erst ab 70? Probleme und Perspektiven Evers, M., De Mooij, R., van Vuuren, D.: What explains the variation in estimates des Altersübergangs. IAQ-Forschung 2017-02. University of Duisburg- of labor supply elasticities? Ifo Institute for Economic Research Working Essen (2017) Paper Series, 1633. Munich (2005) Bargain, O., Orsini, K., Peichl, A.: Comparing labor supply elasticities in Europe Fasbender, U., Wang, M., Voltmer, J.B., Deller, J.: The meaning of work for postre- and the United States—new results. J. Hum. Resour. 49(3), 723–838 tirement employment decisions. Work Aging Retire. 2(1), 12–23 (2016) (2014) Genser, B. Holzmann, R.: Taxing German old-age pensions fairly and efficiently, Bargain, O., Peichl, A.: Own-wage labor supply elasticities: variation across time ifo DICE report volume 17. Munich (2019) and estimation methods. IZA J. Labor Econ. 5(10), 1–31 (2016) Geyer, J., Haan, P., Lorenz, S., Zwick, T. Bruns, M.: The role of labor demand in the Bastani, S., Moberg, Y., Selin, H.: The anatomy of the extensive margin labor labor market effects of pension reforms. Industrial Relations, forthcoming supply response. Scand. J. Econ. 123(1), 33–59 (2021) (2021) Beehr, T., Bennett, M.: Working after retirement: features of bridge employment Grund, C.: Severance payments for dismissed employees in Germany. Eur. J. and research directions. Work Aging Retire. 1(1), 112–128 (2015) Law Econ. 22(1), 49–71 (2006) Blau, F., Kahn, L.: Changes in the labor supply behavior of married women: Gutiérrez-i-Puigarnau, E., van Ommeren, J.: Labour supply and commuting. J. 1980–2000. J. Law Econ. 25(3), 393–438 (2007) Urban Econ. 68(1), 82–89 (2010) Blossfeld, H.P.: Berufseintritt und Berufsverlauf. Mitt. Arbeitsmarkt Berufsforsch. Hanel, B.: Financial incentives to postpone retirement and further effects on 18(2), 177–197 (1985) employment—evidence from a natural experiment. Labour Econ. 17(3), Blundell, R., Chiappori, P., Magnac, T., Meghir, C.: Collective labour supply: het- 474–486 (2010) erogeneity and non-participation. Rev. Econ. Stud. 74(2), 417–445 (2007) Hank, K., Buber, I.: Grandparents caring for their grandchildren: findings from Borra, C.: Childcare costs and Spanish mothers’ labor force participation. Haci- the 2004 survey of health, ageing, and retirement in Europe. J. Fam. Issues enda Pública Española/revista De Economía Pública 194(3), 9–40 (2010) 30(1), 53–73 (2009) Brenke, K.: Immer mehr Menschen im Rentenalter sind berufstätig. DIW Heckman, J.: The common structure of statistical models of truncation, sample Wochenbericht 80(6), 3–12 (2013) selection and limited dependent variables and a simple estimator for Brussig, M.: Erwerbstätigkeit im Alter hängt vom Beruf ab. Altersübergangs- such models. Ann. Econ. Soc. Meas. 5(4), 475–492 (1976) Report, 2010-05. University of Duisburg-Essen (2010) Heckman, J.: Sample selection bias as a specification error. Econometrica 47(1), Brussig, M.: Alter beim Austritt aus sozialversicherungspflichtiger Beschäfti- 153–161 (1979) gung ist gestiegen. Altersübergangs-Report, 2015-01 (2015) Hochfellner, D., Burkert, C.: Berufliche Aktivität im Ruhestand. Z. Gerontol. Brussig, M., Bellmann, L.: Betriebliche Einflussfaktoren auf Rekrutierungen Geriatr. 46(3), 242–250 (2013) Älterer und deren Bewerbungen. German J. Hum. Resour. Manag. 22(1), Hofäcker, D., Naumann, E.: The emerging trend of work beyond retirement 35–57 (2008) age in Germany. Z. Gerontol. Geriatr. 48(5), 473–479 (2015) Burkert, C., Hochfellner, D.: Employment trajectories beyond retirement. J. Institut für Arbeit und Qualifikation.: Erwerbstätige nach Stellung im Beruf Aging Soc. Policy 29(2), 143–167 (2017) 1994–2019. Sozialpolitik aktuell in Deutschland, University of Duisburg- Büsch, V., Dorbritz, J., Heien, T., Micheel, F.: Weiterbeschäftigung im Rentenalter: Essen (2020) Wünsche, Bedingungen, Möglichkeiten. Materialien zur Bevölkerungswis- Kanabar, R.: Post-retirement labour supply in England. J. Econ. Ageing 6, senschaft, 129/2010 (2010) 123–132 (2015) Cahill, K., Giandrea, M., Quinn, J.: Retirement patterns and the macroeconomy, Kim, S., Feldmann, D.C.: Working in retirement: the antecedents of bridge 1992–2010: the prevalence and determinants of bridge jobs, phased employment and its consequences for quality of life in retirement. retirement, and reentry among three recent cohorts of older Americans. Acad. Manag. J. 43(6), 1195–1210 (2000) Gerontologist 55(3), 384–403 (2015) Komp, K., van Tilburg, T., Broese van Groenou, M.: Paid work between age 60 Cahuc, P., Carcillo, S., Zylberberg, A.: Labor Economics. MIT press, Cambridge and 70 years in Europe: a matter of socio-economic status? Int. J. Ageing (2004) Later Life 5(1), 45–75 (2010) Cameron, C., Trivedi, P.: Microeconometrics—Methods and Applications. Kropp, P., Schwengler, B.: Three-step method for delineating functional labor Cambridge University Press, Cambridge (2002) market regions. Reg. Stud. 50(3), 429–445 (2016) Chetty, R., Guren, A., Manoli, D., Weber A.: Does indivisible labor explain the Larsen, M., Pedersen, P.: To work, to retire- or both? Labor market activity after difference between micro and macro elasticities? A meta-analysis of 60. IZA J. Eur. Labor Stud. 2, 21 (2013) extensive margin elasticities. NBER Working Paper 16729. Cambridge MA (2011) Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 17 of 17 21 Lorenz, S., Pfister, M., Zwick T.: Identification of the statutory retirement dates in Ribar, D.C.: Child care and the labor supply of married women: reduced form the Sample of Integrated Labor Market Biographies (SIAB). FDZ Method- evidence. J. Hum. Resour. 27(1), 134–165 (1992) enreport, 08/2018. Nuremberg (2018) Rogerson, R., Wallenius, J.: Micro and macro elasticities in a life cycle model Lorenz, S., Pfister, M., Zwick, T.: Beware of the employer: financial incentives for with taxes. J. Econ. Theory 144(6), 2277–2292 (2009) employees may fail to prolong old-age employment. ZEW Discussion Saba, T., Guerin, G.: Extending employment beyond retirement age: the case of Paper, 20-007. Mannheim (2020) health care managers in Quebec. Public Pers. Manag. 34, 195–214 (2005) Maestas, N.: Back to work: expectations and realizations of work after retire- Schellenberg, G., Turcotte, M., Ram, B.: Post-retirement employment. Stat. Can. ment. J. Hum. Resour. 45(3), 718–748 (2010) Perspect. 75–001, 14–17 (2005) Maxin, L., Deller, J.: Activities in retirement: individual experience of silver work. Schmähl, W.: Ageing workforce: firm strategies and public policy in Germany. Comp. Popul. Stud. 35(4), 801–832 (2010) Geneva Pap. Risk Insur. 28(4), 575–595 (2003) Meyer, B., Rosenbaum, D.: Welfare, the earned income tax credit, and the labor Smeaton, D., McKay, S.: Working after state pension age: quantitative analysis. supply of single mothers. Q. J. Econ. 116(3), 1063–1114 (2001) UK Department for Work and Pensions Research Report 182. London Micheel, F.: The intention to paid and/or unpaid activities in retirement—a (2003) study of older workers in Germany. J. Aging Soc. Policy 33(2), 101–119 Sprengholz, M., Wieber, A., Holst, E.: Gender identity and wives’ labor market (2021) outcomes in West and East Germany between 1983 and 2016. Socio- Micheel, F., Cihlar, V., Riedl, S.: “Der Geist ist willig, aber...”—Was Menschen Econ. Rev. (2020). https:// doi. org/ 10. 1093/ ser/ mwaa0 48 daran hindert, im Ruhestand zu arbeiten. Bevölkerungsforschung Aktuell, Sullivan, S.E., Al Ariss, A.: Employment after retirement: a review and framework 6/2017, 2–9 (2017) for future research. J. Manag. 45(1), 262–284 (2019) Micheel, F., Roloff, J., Wickenheiser, I.: The impact of socioeconomic character - Torka, N., Goedegebure, I., Van Ewijk, I., Looise, J.: On the motives and needs for istics on older employees’ willingness to continue working in retirement work beyond age 65: comparing voluntary workers versus agency work- age. Comp. Popul. Stud. Zeitschrift Für Bevölkerungswissenschaft 35(4), ers. German J. Hum. Resour. Manag. 26(2), 167–188 (2012) 869–902 (2010) Ullmann, K., Bothfeld, S.: The German employment protection act—how does Mincer, J.: Labor force participation of married women. In: Lewis, H. (ed.) it work in company practice? WSI Discussion Paper 161, Duesseldorf Aspect of Labor Economics. Universities National Bureau of Economic (2008) Research Conference Series, vol. 14, pp. 63–105. Princeton University Vargha, L., Gál, R., Crosby-Nagy, M.: Household production and consumption Press, Princeton (1962) over the life cycle: national time transfer accounts in 14 European coun- Pfarr, C., Maier, C.: Arbeit trotz Rente: Warum bleiben Menschen in Ruhestand tries. Demogr. Res. 36(32), 905–944 (2017) erwerbstätig. Deutsches Institut für Altersvorsorge, Berlin (2015) Viitanen, T.K.: Cost of childcare and female employment in the UK. Labour 19, Pfister, M., Lorenz, S., Zwick, T.: Implementation of pension entitlements in the 149–170 (2005) sample of integrated labor market biographies (SIAB). FDZ Methodenre- Wang, M., Zhan, Y., Liu, S., Shultz, K.: Antecedents of bridge employment: a port, 01/2018. Nuremberg (2018) longitudinal investigation. J. Appl. Psychol. 93(4), 818–830 (2008) Pleau, R.: Gender differences in postretirement employment. Res. Aging 32(3), Wanger, S.: Altersteilzeit: Beliebt, aber nicht zukunftsgerecht. IAB-Kurzbericht, 267–303 (2010) 08/2009 (2009) Pleau, R., Shauman, K.: Trends and correlates of post-retirement employment, Westermeier, C.: Ältere am Arbeitsmarkt. Eine stabile Beschäftigung vor dem 1977–2009. Hum. Relat. 66(1), 113–141 (2013) Rentenalter begünstigt die Weiterarbeit. IAB-Kurzbericht, 15/2019 (2019) Rammohan, A., Whelan, S.: The Impact of childcare costs on the full-time/part- time employment decisions of Australian mothers. Australian Economic Publisher’s Note Papers 46(2), 152–169 (2007) Springer Nature remains neutral with regard to jurisdictional claims in pub- Rasner, A., Etgeton, S.: Rentenübergangspfade: Reformen haben großen lished maps and institutional affiliations. Einfluss. DIW-Wochenbericht 81(19), 431–441 (2014) Rhein, T.: Arbeiten im Rentenalter: Erwerbstätigkeit 65plus in Europa. Aktuelle Berichte, 25/2016. Institute for Employment Research, Nuremberg (2016) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal for Labour Market Research Springer Journals http://www.deepdyve.com/lp/springer-journals/money-also-is-sunny-in-a-retiree-s-world-financial-incentives-and-work-3l2rir0lq2

Loading next page...

Publisher: Springer Journals
Copyright: Copyright © The Author(s) 2021
ISSN: 1614-3485
eISSN: 2510-5027
DOI: 10.1186/s12651-021-00304-1
Publisher site: See Article on Publisher Site

Abstract

This paper assesses the impact of financial incentives on working after retirement. The empirical analysis is based on a large administrative individual career data set that includes information about 2% of all German employees subject to social security or in marginal employment until age 67 and their employers in the period 1975–2014. We use the classical labor supply model and differentiate between the impact of (potential) labor and non-labor (pension entitle - ments) income. A Heckman-type two step selection model corrects for endogeneity. We show that labor income has a positive and non-labor income a negative impact on the decision to work after retirement. Especially individuals who can substantially increase their earnings in comparison to their pension entitlements accordingly have a higher probability to work. Men are more attracted by labor earnings incentives than women. Also individuals who work until retirement are easier attracted to work after retirement by higher labor income than those with gaps between employment exit and retirement. Our results allow the calculation of the impact of changes in taxes on labor and non-labor income and changes in earnings offers by employers on work after retirement for different demographic groups. Keywords: Work after retirement, Financial incentives, Labor and non-labor income JEL Classification: J14, J22, J26 1 Introduction scientific literature (Anger et al. 2018; Burkert and Hoch - Labor force participation of retirees increased strongly fellner 2017; Büsch et al. 2010; Cahill et al. 2015; Fas- in most European countries in recent years (Eurofound bender et al. 2016; Maxin and Deller 2010; Micheel et al. 2012; Hofäcker and Naumann 2015). On average, the 2010; Pleau and Shauman 2013; Wang et al. 2008; West- employment rate for individuals aged between 65 and ermeier 2019). 74 years has risen from 5.2% in 2002 to 8.7% (an increase This paper concentrates on the impact of financial of 67%) in 2014 in the EU-15 Countries (Rhein 2016). incentives of older individuals on their decision to work Individual, social and work-related push and pull fac- after retirement or not. More specifically, we estimate tors have been put forward as drivers of post-retirement the percentage change in labor force participation after employment in a dynamically developing strand of retirement in response to a percentage change in the financial situation. On the basis of a Heckman (1976, 1979) approach to correct for sample selection, we show that (potential) labor earnings are an important pull fac- tor for the decision to engage in employment beyond Also in Germany, the labor force participation rate of people aged 65 to 74 increased to 9.6% in 2014 from 4.2% in 2002, implying a rise of almost 140% (Rhein 2016). *Correspondence: thomas.zwick@uni-wuerzburg.de Faculty of Business Management and Economics, University of Würzburg, Sanderring 2, 97070 Würzburg, Germany Full list of author information is available at the end of the article © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. 21 Page 2 of 17 S. Lorenz , T. Zwick retirement. We separately assess the second main finan - retirement—Kanabar (2015) however does not find this cial incentive mechanism and find that there is a size - correlation. Micheel et al. (2010) stress a negative correla- able negative pension entitlement (non-labor income) tion between equivalent household income before retire- effect on the probability of working after retirement. As a ment and the interest in work after retirement—Engstler consequence, we show that the higher the ratio between and Romeu-Gordo (2014) however find a u-shaped pat - (potential) post-retirement labor earnings and pension tern for work after retirement and equivalent household entitlements, the larger the likelihood of post-retirement income. Maestas (2010) finds a weakly positive correla - employment. The second aim of the paper is the provi - tion between the decision to unretire and earnings before sion of empirical evidence of heterogeneous earnings retirement. Komp et al. (2010) and Kanabar (2015) find elasticities between retiree groups. We show that women no correlation between working after retirement and are less responsive to higher labor earnings than men. (non-pension) wealth. Engstler and Romeu-Gordo (2014) Especially men with low earnings before retirement have point to the fact that higher financial pressure such as the high earnings elasticities. We in addition find that indi - presence of a real estate loan also can explain work after viduals who use the bridge options partial retirement or retirement. unemployment before early retirement are less respon- When looking closer at the measurement of the finan - sive to higher labor earnings after retirement than their cial situation of retirees, it becomes evident that so far counterparts who work until retirement. no contribution on work after retirement made a dis- This paper makes the following contributions to the lit - tinction between labor and non-labor income. Some erature on empirical determinants of work beyond retire- papers only include non-labor income after retirement ment: first, we empirically assess the importance of labor such as pension entitlements (Hochfellner and Burkert earnings in comparison to non-labor income for work 2013; Kanabar 2015). Other papers bundle up non-labor beyond retirement for the first time. Second, large-scale and labor income after retirement by for example using and reliable administrative social security panel data equivalent household income after retirement (Anger covering all types of employees except civil and military et al. 2018; Engstler and Romeu-Gordo 2014; Micheel servants are used instead of small-scale and potentially et al. 2010; Pfarr and Maier 2015). A last group of papers specific sub-populations or survey-based data that were looks at wealth or the subjective economic status before the basis of most relevant studies so far. Third, a rich retirement (Fasbender et al. 2016; Kanabar 2015; Komp set of employer and individual characteristics allows us et al. 2010; Wang et al. 2008). to control for additional determinants of work beyond Empirical approaches that do not separate labor and retirement. Fourth, based on hypotheses, we show heter- non-labor earnings after retirement may report biased ogeneities in labor earnings elasticities between men and results on financial determinants of work participation women, employees who have their work and home in the after retirement. According to standard labor supply the- same region or not, employees with and without a gap ory, labor earnings are assumed to increase work partici- between employment exit and retirement, and employees pation and non-labor income is assumed to decrease it with high and low earnings before retirement. The analy - (Cahuc et al. 2004; Blundell et al. 2007). When work has sis of differences in elasticities between these sub-groups a higher reward, workers substitute leisure against work. allows us to derive more specific implications of changes An increase in non-labor income has a negative effect on in financial incentives for these sub-groups. work participation or working hours. The income effect induces employees to reduce work because they can 2 Background maintain a target standard of living with lower labor bur- 2.1 F inancial situation as driver of work after retirement den. This paper for the first time distinguishes between There are strong indications that the individual finan - income and substitution effects of work and non-work cial situation is a key driver of work after retirement. income after retirement on work participation and shows More specifically, Hochfellner and Burkert (2013) the empirical relevance of both drivers. and Pleau and Shauman (2013) find a negative cor - The extensive labor supply margin response to mar - relation between pension entitlements and work after ginal work income changes depends on the probability density function of the distribution of reservation wages around the economy’s equilibrium. The variance in labor supply for small changes in wages around the equilib- Examples are the German Ageing Survey (Deutscher Alterssurvey DEAS) or rium wage is low for the entire working-age population the Survey of Health, Ageing and Retirement in. Europe (SHARE). because relatively few people who do not work are will- We only briefly summarize the literature on the impact of financial situa- tion on working after retirement. There are several recent literature reviews ing or able to work for the current wage (Chetty et al. that describe the knowledge about other drivers, compare for example Eng- 2011). It is therefore hard to estimate reliable earnings stler and Romeu-Gordo (2014) or Sullivan and Al Ariss (2019). Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 3 of 17 21 elasticities on labor supply for the working-age popula- tion. Relatively many employees at the end of their career are closer to the margin of re-entering the labor force for available earnings opportunities, however. As the previ- ous labor contract ends when entering retirement, all employees close to retirement have to decide whether to work in a new labor contract or not besides receiving retirement benefits. Upon retirement, labor re-entry inci - dence and potential extensive margin responses to labor earnings options therefore are relatively high in compari- son to the earnings elasticity for the entire population (Rogerson and Wallenius 2009). Analogously to people close to retirement, all women with a small child have to decide whether to work again or continue to provide full time childcare. As a conse- quence, also the extensive margin response to labor earn- ings incentives is relatively high (and easier to measure) for women with small children. The empirical literature on mothers with young children shows that besides non- Fig. 1 Labor supply model of retirees (Source: own illustration) labor financial means such as income received by other household members, childcare cost subsidies or public transfers such as parental allowance, also (potential) labor and leisure. The shape of the indifference curves corre - earnings are an important driver of labor market re-entry sponds to consumption and leisure being defined as “nor - (Allègre et al. 2015; Borra 2010; Connelly 1992; Del Boca mal” goods with decreasing marginal returns. and Vuri 2007; Ribar 1992; Viitanen 2005). The empirical In this model, individual income consists of non-labor identification strategy to analyze financial incentives on income (in our case pension entitlements) V and labor labor supply of mothers after their parental leave period earnings W. The earnings derived from work after retire - can be applied to the labor supply decision of older work- ment correspond to the opportunity costs of leisure. The ers after their retirement. As a consequence, we use the point where one of the retiree’s indifference curves tan - Heckman selection model proposed for example by Con- gents the budget line is the combination of earnings and nelly (1992) or Borra (2010) and assume that for retirees, leisure with which the retiree maximizes utility under the (potential) labor earnings are an important driver for budget constraint. work after retirement besides non-labor earnings. Figure 1 shows that if potential (hourly) labor earn- ings after retirement are too low, the individual does not 2.2 Basic model work after retirement. The maximum indifference curve, According to the standard labor supply model, retir- U crosses the budget line at point A and the retiree con- ees maximize their utility by comparing consumption sumes only non-labor earnings V . If (hourly) labor earn- made possible by labor and non-labor income with lei- ings and accordingly the slope of the budget line increase, sure (Cahuc et al. 2004). In the case of work after retire- utility is maximized at point B with U > U . The retiree ment, individuals decide whether to work or not after B A therefore decides to work after retirement (substitution they receive an old-age pension. The observation period effect) and we can derive the following hypothesis: therefore starts at the earliest at the early retirement age (ERA) for those eligible for an early retirement option or at the normal retirement age (NRA) for the other employ- Hypothesis 1 The higher the (potential) labor earnings ees. Figure 1 shows different possible budget lines as well after retirement, the greater the likelihood to work after as the retirees’ corresponding indifference curves. The retirement. slope of the indifference curves (U, U, U, U ) defines A B C D the marginal rate of substitution between consumption An increase in non-labor income from V to V results 0 1 in a parallel shift of the budget line. With higher non- labor income, labor force participation decreases because the retiree can now consume the same quantity of goods Please note that “not working” when receiving an old-age pension means when working less (income effect). In Fig. 1, the income for many individual in our dataset to be engaged in unpaid care, household or effect is depicted by steeper indifference curves at a given volunteer work or being self-employed. 21 Page 4 of 17 S. Lorenz , T. Zwick leisure level with increasing V. Thus, an increase in the care of relatives, childcare of grandchildren) is higher pension entitlements decreases the opportunity cost of than the comparable share of men (Hank and Buber leisure. In Fig. 1 an increase in the pension entitlements 2009). In addition, there is a second spike in household is illustrated as a shift from point B to point C. Note that production when individuals approach retirement age. with an increase in V that is strong enough, at a given This spike is larger for women than for men, see Var - labor earnings level, pensioners may decide to stop work- gha et al. (2017). There is a decisive difference between ing. Therefore, the following hypothesis is derived: the average individual in working age and the decision for work after retirement when calculating the employ- Hypothesis 2 The higher the pension entitlements ment earnings elasticities, however. Until retirement, the (non-labor income), the lower the likelihood to work share of men who do not work is much smaller than the after retirement. share of women (Blau and Kahn 2007). The labor market attachment of men who potentially can join the work- force therefore is lower than that of women. Hence, the Given the higher pension entitlement, V , higher earnings elasticity of males during their working age on (potential) labor earnings (W instead of W ) are needed 2 1 average is lower than the earnings elasticity of females to incentivize the retiree to work as much as in point (Bastani et al. 2021). On day 1 after retirement, labor B (point D, for example). A higher substitution effect market participation of men and women is set to zero, (higher labor earnings) has to compensate for a higher however. As a consequence, there may be more males income effect (higher non-labor income) for labor mar - willing to re-enter work after retirement than females ket participation after retirement. Hypothesis 3 combines at available labor earnings given the higher home pro- the first two hypotheses: duction duties of women that make labor market par- ticipation less attractive or impossible for them. In other Hypothesis 3 The higher the ratio between the (poten - words, the density of male retirees to re-enter employ- tial) labor earnings after retirement and non-labor ment may be higher than the density of female retirees. income, the greater the likelihood that the retiree works Therefore, we assume that the labor earnings elasticity after retirement. for old women is lower than that for old men: Hypothesis 4 Women have lower post-retirement 2.3 Het erogeneous responses to financial incentives labor earnings elasticities than men. We now derive factors that explain differences in labor earnings elasticities for the labor force participation Some previous papers on the decision to work after between retiree groups. Arguments concerning differ - retirement look at potential differences between men and ent responses of individuals to financial incentives from women. As financial drivers however only individual and labor supply theory and existing empirical results on equivalent household income before retirement as well post-retirement employment are built upon to derive as wealth and pension entitlements have been included hypotheses that can be tested in the empirical analysis. but not our key variable (potential) labor earnings. In the labor supply literature, a robust finding is that Pleau (2010) and Pleau and Shauman (2013) stress that earnings elasticities of women exceed that of men (Bar- women with a high individual income before retirement gain et al. 2014; Evers et al. 2005; Meyer and Rosenbaum more often work after retirement. Equivalent household 2001). The difference between gender groups is explained income and wealth seem to play no role for women. For by the division of labor within the family (Blundell et al. men however work after retirement is positively corre- 2007). According to Blau and Kahn (2007) and Mincer lated with equivalent household income and negatively (1962), women substitute their time between market with wealth (but not individual income). Engstler and work, home production, and leisure, whereas men pri- Romeu-Gordo (2014) however find an impact of pension marily substitute their time between market work and entitlements for men only. leisure. As women have closer substitutes for market In Germany, there are various ways to transit from work than men, they are expected to have larger earn- employment to retirement (Rasner and Etgeton 2014). A ings substitution effects. The literature therefore points substantial portion of older workers have a gap between out that household production increases the elasticity of employment exit and retirement entry. This gap is called labor supply because home-made goods are substitutes “bridge option” if individuals do not withdraw from the for commercial goods (Cahuc et al. 2004; Mincer 1962). labor market between the exit from their last employ- At first sight, the labor supply mechanisms for all indi - ment and retirement (Brussig 2015). There are two bridge viduals in working age also apply for retirees. The share options, an unemployment spell between work exit and of older women engaged in household production (e.g., Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 5 of 17 21 (early) retirement and partial retirement. Partial retire- 2009). Employers therefore are more successful in nudg- ment may mean that employees work full time for half ing employees with a lower labor market attachment at the partial retirement period at reduced earnings and old age into partial retirement. The group of employees exit employment completely for the second half. The so- who bridge the time between employment and retire- called block model version of partial retirement has been ment by unemployment may mainly consist of workers chosen by more than 80% of participants (Brussig 2015). who have been dismissed against their wills and there- A much smaller share of retirees chose the so-called con- fore their value of leisure is unclear. The German labor tinuous version of partial retirement. Here the employee law especially protects employees with high tenure and/ reduced working-time to half during the entire partial or experience against dismissal, however (Ullmann and retirement period and no gap between labor market exit Bothfeld 2008). In addition, many judges in labor courts and pension entry occurs. Both bridge options can be are employee-friendly in dismissal disputes. As a con- used by employees after 15 contribution years to pension sequence, the appeal against an involuntary dismissal retirement and at least 8 years of contributions periods in before a labor court has a high success rate especially the last 10 years before retirement. Bridge options there- for older workers. Employers are exposed to high uncer- fore are not viable for employees with large labor market tainty with respect to the outcome of the lawsuit that eas- gaps and a low labor market attachment in general. ily takes more than 1 year (Ullmann and Bothfeld 2008) In both bridge options, employees usually have to and they frequently face a detrimental effect on their accept financial disadvantages in comparison to their reputation when they unilaterally dismiss older employ- other (early) retirement options, but they can exit ees. Therefore, only a very small share of older employees employment already before their ERA (Lorenz et al. is dismissed against their will and most dismissed older 2020). Indeed, labor market exit age on average is employees obtain substantial severance pay (Schmähl between one to two years earlier for those using par- 2003; Grund 2006). Severance and compensation pay- tial retirement or retirement after unemployment than ments for a dismissal or partial retirement however are for those choosing the early retirement options without subject to individual bargaining between employer and employment gaps before retirement that also would have employee. Therefore, employers may prefer to nudge been available for those who used bridge options (Geyer employees with a relatively strong preference for leisure et al. 2021). For those who use both bridge paths, we into unemployment before early retirement because dis- hardly observe a return to regular employment before missing them is cheaper. Our fifth hypothesis is: retirement. We therefore assume that employees who choose a bridge option put a relatively high value on Hypothesis 5 Individuals using bridge paths into leisure at old age and they therefore are less inclined to retirement have lower earnings elasticities than individu- work after retirement. A lower inclination to work after als who are employed until retirement. retirement also has been found for those who did not work full time until retirement (Burkert and Hochfell- Individuals with different places of work and resi - ner 2017; Maestas 2010; Smeaton and McKay 2003)— dence before retirement accept commuting costs. those who use the so-called continuous version of partial According to labor supply theory, at given earnings, retirement are a case in point for part-time work before commuting costs reduce the opportunity cost of lei- retirement. sure and therefore reduce labor supply (Cogan 1977, The argument that employees who select partial retire - 1981). Gutiérrez-i-Puigarnau and van Ommeren (2010) ment have a higher value of leisure seems to be obvious assume a strict complementarity between commuting because employees have to give their consent for trad- costs and labor supply. Their arguments for example ing earnings against an earlier employment exit (Wanger Erlinghagen (2005) and Ullmann and Bothfeld (2008) show that the average Please note that the concept of bridge employment in Germany differs from share of involuntary job terminations was between 20 and 30 percent after that used in the US (Beehr and Bennett 2015). In Germany, bridge employ- 1996 and that the share of involuntary job terminations was strongly nega- ment means that employees exit employment earlier then retirement or tively correlated with age. reduce their working time in their career job or they exit into unemployment This assumption of course applies under the ceteris paribus assump- from their career job before they enter retirement. In the US, bridge employ- tion. We know that partial retirement is mainly accepted by better edu- ment usually means taking another job after their career job. cated employees who work for large (industrial) employers (Engstler and Almost all employees in bridge options in the cohorts we look at also Romeu-Gordo 2017) and bridge options are mainly offered by employers are entitled to use other early retirement options such as the pension for with higher employee adaptation costs, i.e. large but shrinking employers in women and the pension for the long-term insured. Indeed, many employees regions with high unemployment (Lorenz et al. 2020). It is therefore impor- in bridge options used these alternative early retirement options after their tant to control for the mentioned individual and employer characteristics in bridge option. order to avoid biased results. 21 Page 6 of 17 S. Lorenz , T. Zwick on the negative effects of road pricing on labor sup - 1975 to 2014 (SIAB, 1975–2014). Out data therefore ply however are based on the average employee situa- only include employees subject to social security and tion in which males work full-time more often than employees in marginal employment —civil and mili- females, commute relatively long distances, and are lit- tary servants (around five percent of the workforce) and tle affected by economic incentives with respect to their self-employed (about ten percent of the workforce) are work participation (compare footnote 38 in Gutiérrez-i- excluded and unpaid care, household as well as volunteer Puigarnau and van Ommeren 2010). The insights from work is not unfortunately unknown. this paper on the impact of commuting costs and labor The data provide daily information about earnings and supply therefore may not apply for the decision to work employment as well as receipt of benefits according to after retirement. Given that older employees only work German Social Books II and III. Moreover, an extensive few hours and that they have a time perspective that is set of establishment information from the IAB Establish- too short to look for a new home near a new employer, ment History Panel is linked to the individual employ- the necessity to (continue to) commute to a job after ment history. We know the day at which employees exit retirement may have a negative effect on labor supply. from the last employment subject to social insurance We therefore assume that potential commuting costs contributions before retirement or the day at which the of individuals whose place of work is not the place of bridge options end. Individual and company-specific residence before retirement are higher. This group of characteristics are measured at this point in time. employees therefore may be less attracted by financial We know daily earnings of employees after retirement incentives to work after retirement: and whether an employee was in marginal employment or worked more. Earnings from marginal employment Hypothesis 6 Individuals whose place of work and and pension entitlements are income tax free and do not residence is in the same region before retirement have require social security contributions in contrast to labor higher earnings elasticities than individuals who have to earnings beyond marginal employment. We want to commute. have comparable earnings indicators for employees in marginal employment and employees with higher earn- ings. We argue that disposable income is more relevant Our last hypothesis concerns heterogeneity by income for the decision of whether to work or not than gross quartile before retirement. Bastani et al. (2021) show earnings. Therefore, we use net earnings after taxes and that earnings elasticity decreases with potential earn- social contributions for labor earnings above the mar- ings. They argue that those with low potential earnings ginal employment threshold. More specifically, we deduct have a lower labor market attachment and therefore earnings taxes for earnings tax class 1, the solidarity sur- their potential for entry into the labor market is higher. charge, church tax and social security contributions to For our group of old workers who have to decide health and nursing care insurance. Approximately 80% whether to re-enter employment, this argument is not of all employees are in marginal employment after retire- valid because their labor market attachment is zero ment. The share of retired women who have labor earn - when entering retirement. We however argue that the ings beyond the marginal employment level is with 18% risk of having pension entitlements below the individual somewhat smaller than the share of men (22%). income aspiration level after retirement (“old-age pov- The age at which pensions can be received in Ger - erty”) is higher for those with low labor earnings before many depends on the fact whether the individual is eli- retirement (Burkert and Hochfellner 2017). The finan - gible for early retirement or not. There are several early cial aspect of work after retirement therefore may play a larger role for those who earned less before retirement (Fasbender et al. 2016): A detailed description of the SIAB can be found in Antoni et al. (2016). Hypothesis 7 Individuals with lower earnings before Marginal employment is defined as dependent employment with a maxi- retirement have higher earnings elasticities. mum monthly salary of 400 EUR (raised to 450 EUR in 2013). The employment shares have been taken from Institut für Arbeit und Qualifikation (2020). In Germany, it is compulsory to report the end of employment to the 3 Data social insurance system after the statutory retirement age. In the case of Our analysis is based on a large and high-quality admin- employment after retirement, the employer must provide a new contract istrative dataset provided by the Federal Employment and submit a new declaration to the social insurance, even if the employee was already employed by the company. Thus, employment is interrupted Agency in Germany (Bundesagentur für Arbeit). The after pension entry by 1 day (Burkert and Hochfellner 2017). data comprise a sample of 2% of employees from the Tax and deduction exemptions on marginal employment were abolished Integrated Employment Biographies ranging from just in 2012. Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 7 of 17 21 retirement options that are tied to different requirements reduce pension claims (Westermeier 2019). Unfortu- such as pension contribution years, gender, or disabil- nately, there are no statistics on the incidence of this ity. The most important early retirement options are the group. pension for long-term insured employees, pension for women, pension for severely disabled, and the two bridge 3.1 Estimation sample options pension after unemployment and pension after Our estimation sample consists of women and men born partial retirement, see Lorenz et al. (2018) for an in- between 1935 and 1947. For each birth cohort, post- depth description of the early retirement rules and their retirement employment is observed for a maximum changes over time. Individuals who are eligible for one or of two years after the NRA. As work incidence quickly several early retirement options can exit the labor market declines with age after 65 in Germany (compare for and start to collect pensions at the ERA specified by the example Pfarr and Maier 2015), we are confident that we early retirement option (in many cases 60 or 63 years). do not miss any retirees who start to work in retirement All other individuals only can collect pensions at the after their 67th birthday. We restrict our sample to West- NRA (65 years). People born after 1937 in addition can German men and women with a strong labor market retire after the NRA. They receive a 6% increase in pen - attachment in old age. More specifically, all individuals in sion entitlements for each year of pension postponement. the sample must be employed subject to social security We cannot observe when individuals start to receive at least once after the age of 55 and they must be active pensions directly in our data and therefore define in the labor market (employed, in partial retirement, or two groups with employment after receiving a pen- unemployed) at age 59. The reason for these sample sion. The first group consists of individuals who have restrictions is to obtain a relatively homogeneous sample a labor market spell after their NRA. The second group of employees who in principle all have the option to work consists of individuals who already started a marginal after retirement. The restriction reduces our sample by employment before the age of 65 that extends beyond about 30%. the age of 65 without interruption and who are eligi- It is relatively hard in Germany to return to the labor ble for early retirement. We may incur three sources market after longer labor market gaps at higher age—in of measurement errors with this identification of indi - the group of employees excluded from our sample, the viduals who work after retirement. First, some indi- share or employees who work after retirement accord- viduals work after NRA without receiving a pension. ingly is less than one percent. Our sample restrictions The number of employees who postponed their entry also reduce the risk that we include individuals who also into retirement after NRA always was negligible. We were civil servants or self-employed besides or after their include these individuals because they are employed employment subject to social security. There are age after NRA but incorrectly assume that they also draw restrictions on starting a job with a civil servant status pension benefits after age 65. Second, there may be and it is hardly possible to work in a job subject to social individuals who receive pensions besides working but security besides a job as civil servant. Employers usually stop to work before age 65. We cannot observe this do not allow their employees to generate a substantial group of people and therefore our results only cover all income from self-employment besides their job subject individuals who work besides receiving retirement ben- to social security. efits after NRA. The third group works more than mar - After the sample restrictions we are left with 30,784 ginally before NRA in addition to receiving pensions. women and 44,887 men—we therefore include 75,671 This group again is negligibly small because earnings observations. In the sample, 6273 women and 9549 beyond the threshold of marginal employment strongly men continue to work after retirement, correspond- ing to an average of 20% for women and 21% for men, The SIAB data do not include relevant characteristics of the pension insur - ance, such as pension entitlements or eligibility. Therefore, individual pension entitlements are calculated and eligibility criteria for old age pensions and the corresponding statutory retirement dates NRA and ERA are identified according to Pfister et al. (2018) and Lorenz et al. (2018). We are grateful to Labor market careers are only available for East Germany after January 1, Philip vom Berge and Dana Müller from the FDZ at the IAB for merging the 1991 (Antoni et al. 2016). day of birth as part of the Custom Shaped Administrative Data for the Analy- sis of Labor Market (CADAL) project because the calculation of pension enti- This sample restriction was also made by Hanel (2010) and Geyer et al. tlements requires the exact birth date. (2021). 15 18 Bäcker et al. (2017) for example report that in 2014, from about 825,000 We also exclude seamen and miners (less than one percent of women retired individuals only about two percent or 22,000 individuals received a and of men) because they have special protection of legitimate expectation bonus. rules for retirement that cannot be identified in the dataset. 21 Page 8 of 17 S. Lorenz , T. Zwick Table 1 Summary statistics on work and non-work earnings Daily net labor earnings (EUR) Daily pension entitlements (EUR) Share of net labor earnings in pension entitlements (%) Women Men Women Men Women Men All employees who work after retirement Mean 14.013 19.770** 23.016 39.122** 65.94 52.97** SD 14.037 24.660 9.210 11.737 60.68 65.59 N 6273 9549 6273 9549 6273 9549 Employees with marginal employment after retirement Mean 9.406 9.514 22.637 38.510** 46.93 27.06** SD 3.416 3.453 8.643 10.584 24.70 15.28 N 5175 7350 5175 7350 5175 7350 SD standard deviation, N number of observations Data: SIAB 7514, own calculations Significance levels of differences between men and women: *p < 0.05, **p < 0.01 19 20 compare Appendix Table 7. Across birth cohorts, the retirement are 14 EUR for women and 20 EUR for men. share of individuals who work beyond retirement has For those who are marginally employed, daily earnings risen slightly. The number of observations and the share for both sexes are around 9 EUR. Pension entitlements of individuals who continue to work after retirement by of women who continue to work after retirement are birth cohort are shown in Appendix Table 7. on average 40% lower than those of men. Of particular Our data set has some drawbacks such as the absence of interest is the share of labor earnings after retirement information on pension entry and other earnings sources in pension entitlements. This share is quite substantial beyond labor earnings and pension entitlements. It how- with 66% for women and 53% for men. For employees in ever has the advantage of including reliable and large- marginal employment, the share is 47% for women and scale information on work behavior of older employees 27% for men. These findings are in line with evidence from many birth cohorts. Studies on the relationship from the UK where labor earnings after retirement were between financial incentives and work beyond retire - more than 90% of average pension entitlements (Kanabar ment so far typically relied on small-scale and potentially 2015). We therefore conclude that earnings after retire- specific populations (Kim and Feldmann 2000; Saba and ment are an important additional source of income for Guerin 2005; Torka et al. 2012) or on survey-based data securing the standard of living in old age. that may be affected by self-selection and measurement We also can show that those who continue to work error in key variables (Anger et al. 2018; Büsch et al. after retirement on average have a significantly lower 2010; Dittrich et al. 2011; Kanabar 2015; Pfarr and Maier pension entitlement than individuals who do not work 2015, Saba and Guerin 2005). after retirement (Appendix Table 8). This finding corre - sponds with the negative correlation between non-labor 3.2 Descriptive statistics earnings and re-entry into employment after retirement Our analysis focuses on labor and non-labor financial found in most of the previous literature (Burkert and incentives for employment beyond retirement. The fol - Hochfellner 2017; Büsch et al. 2010; Dittrich et al. 2011; lowing descriptive statistics accordingly show average Hochfellner and Burkert 2013; Kim and Feldmann 2000; daily labor earnings, average pension entitlements, and Micheel et al. 2010). Appendix Table 8 also summarizes the average share of labor earnings in pension entitle- ments by gender and type of employment after retire- ment (Table 1). The average daily net earnings after Women have no earnings after retirement above the social contribution earnings threshold and for men the proportion with earnings above the earn- ings threshold is 0.3%. We therefore do not have to correct for censored earn- ings. Please note that we cannot take into account other sources of income, The proportion of employees who continue to work after retirement is such as private and occupation pensions, capital and real estate rents, inde- slightly lower than that calculated by Anger et al. (2018). In their study, how- pendent work or savings. ever, employment after retirement is identified from survey answers. The share of older employees who indicate that they intend to work after retire- Two studies from Germany (Anger et al. 2018; Pfarr and Maier 2015) and ment in surveys even is between a third and a half (Micheel et al. 2010; one study from England (Kanabar 2015) however show that employees with Micheel 2021). higher financial means are more likely to work beyond retirement. Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 9 of 17 21 differences in individual and establishment characteris - accumulated pension entitlements. Pension entitlements tics prior to retirement between employees who continue indicate the negative income effect in the participation to work after retirement and those who do not. This table equation and they are an indicator of past labor market shows, for example, that there is a negative correlation performance. Given that all employees obtain a new labor between work after retirement and employer size as well contract, usually entailing less labor hours and new tasks, as the share of highly skilled workers. previous labor market performance is hardly related to (potential) earnings after retirement, compare Appendix 4 Estimation strategy Table 9. Pension entitlements therefore can be excluded We propose an empirical choice model to examine the from the earnings equation. decision on working after receiving a pension: LFP = β + X β + Z β + γpe + δwa ge + ε LFP = 1 if LFP > 0 (1) i 0 i x j i z i ( ) i i LFP = 0 if LFP = 0 The labor force participation variable (LFP) is a dichot - Besides the exclusion restriction and the Mills ratio in the omous variable with a value of zero for non-participation earnings equation, we include core individual characteris- (no labor earnings) and a value of one (positive labor tics that have been used in the literature as determinants earnings) for employment after receiving a pension. The of work after retirement (Anger et al. 2018; Brenke 2013; unobservable continuous variable “LFP*” reflects the util - Brussig 2010; Burkert and Hochfellner 2017; Engstler and ity difference between non-participation and participa - Romeu-Gordo 2017; Fasbender et al. 2016; Hofäcker and tion in the labor market beyond retirement. Our main Naumann 2015; Larsen and Pedersen 2013; Maestas 2010; variables of interest are the log disposable pension enti- Micheel et al. 2017; Rhein 2016; Westermeier 2019). More tlements “pe” (non-labor earnings) and the net (poten- specifically, we include education, occupation, accumu- 23 25 tial) labor earnings “wage” when receiving a pension. lated labor market gaps until age 65 in months, and the Labor earnings after retirement may be endogenous year of labor market exit. We keep the auxiliary regressions because they also reflect time allocation decisions, that parsimonious in order to avoid multicollinearity problems is, they are only observed for those who work after retire- in our main labor participation regression (Borra 2010). ment. In order to circumvent this endogeneity problem, In the main labor participation equation, we include predicted daily labor earnings after retirement wa ge all explanatory variables used in the auxiliary participa- are estimated for all individuals in an auxiliary estima- tion equations and (potential) labor earnings. In addition, tion. The standard Heckman sample selection approach we include dummies that equal one if an individual uses (Heckman 1976, 1979) is used to predict the earnings of the bridge option partial retirement or unemployment post-retirement employment. More specifically, we first because labor market gaps before retirement negatively estimate a reduced-form post-retirement employment affect the continuation of employment after retirement participation Probit model for all individuals in the sam- (Anger et al. 2018; Burkert and Hochfellner 2017; Schel- ple. We use the Probit parameters from this participation lenberg et al. 2005; Smeaton and McKay 2003; West- equation to generate the inverse Mills ratio. Including ermeier 2019). We also include several characteristics this ratio controls for possible sample-selection bias in of the last employer before retirement. Previous stud- the wage equation. We then take the results of the wage ies have shown that employees in small establishments equation to generate values of the predicted wage for all have a higher probability of working beyond retire- individuals in the sample (Borra 2010). ment (Anger et al. 2018; Burkert and Hochfellner 2017; Technically, this type of selectivity corrected model Micheel et al. 2010). In addition, Brussig and Bellmann can achieve identification by functional form assump - (2008) show that attitudes towards hiring older workers tions (Cameron and Trivedi 2002). Nonetheless, most differ among companies. For example, employees in com - researchers feel more comfortable if at least one regres- panies with a high proportion of older employees have a sor in the participation equation is excluded from the labor earnings equation (Connelly and Kimmel 2003; The Blossfeld classification of occupations is intended to form occupational Borra 2010). We accordingly include in the reduced form groups as homogeneously as possible with regard to average educational and labor participation equation as exclusion restriction the vocational background and professional fields of activity (Blossfeld 1985). An alternative for the aggregated gaps in the labor market before the age of 65 is the exit age from employment before the statutory retirement age. In a further estimation specification, instead of the expected earnings and The results are robust. It is not possible to include both variables in the esti- the pension entitlements, the ratio of the (potential) earnings in pension enti- mations because they are strongly correlated. Early retirement is associated tlements is included in the equation. with more gaps in the labor market before the age of 65. 21 Page 10 of 17 S. Lorenz , T. Zwick Table 2 Linear probability model of labor force participation Dependent variable: labor force participation Women Men Individual characteristics Estimated earnings (log) 0.056* (0.019) 0.246** (0.055) Pension entitlements (log) − 0.055** (0.006) − 0.165** (0.008) Bridge option: partial retirement − 0.078** (0.007) − 0.060** (0.005) Bridge option: unemployment − 0.323** (0.006) − 0.266** (0.005) No degree Ref. Ref. Vocational training 0.032** (0.006) 0.025 (0.013) University degree 0.005 (0.012) − 0.075* (0.031) Manufacturing occupations Ref. Ref. Service occupations 0.011 (0.008) − 0.021 (0.014) Administrative occupations − 0.010 (0.008) − 0.042** (0.012) Accumulated labor market gaps prior to age 65 in months − 0.006** (0.0007) − 0.001 (0.001) Employer characteristics Firm size: < 10 employees Ref. Ref. 10 to 100 employees − 0.027** (0.006) − 0.019** (0.007) > 100 employees − 0.044** (0.007) − 0.059** (0.008) Mean imputed earnings of all full-time employees − 0.0004** (0.00009) − 0.000006 (0.00006) Share of highly qualified employees − 0.027 (0.016) − 0.112** (0.016) Share of employees in partial retirement − 0.204** (0.061) − 0.128** (0.049) Share of employees aged 55–59 − 0.030 (0.017) − 0.013 (0.017) Share of employees aged 60–64 − 0.033 (0.023) − 0.065** (0.021) Share of employees aged 65 and older 0.514** (0.048) 0.527** (0.051) N 30,784 44,887 R 0.307 0.210 Additional controls: year of employment exit, and economic sector of the employer. Standard errors in parentheses are computed by bootstrapping with 200 repetitions and are clustered on the individual level. Data: SIAB 7514 *p < 0.05 **p < 0.01 higher chance of being hired after their retirement than in Appendix Table 10. We use potential earnings for those in companies with a high proportion of employees individuals who do not work after retirement from this in partial retirement and severance payments. Moreo- estimation. In the earnings equations, selectivity of ver, some studies also use the general economic environ- employees who decide to work after retirement is taken ment as a determinant of employment after retirement. in account by including the Mills’ ratio (“lambda”) derived from the parsimonious first-step labor mar We therefore include the mean imputed earnings of all - full-time employees, economic sector, establishment size, ket participation equation. The Hausman specification share of employees in partial retirement, share of highly test checks whether our exclusion restriction is valid to qualified employees, share of employees aged 55–59, control for endogeneity. The high χ values indicate that share of employees aged 60–64, and the share of employ - the pension entitlements variable is a valid exclusion ees aged 65 and older. restriction for the labor market participation equation. We take into account that the decision to work after retire- We present our estimation results of the second estima- ment is fundamentally different between men and women tion step, the decision to work beyond retirement, in Table 2, because the non-market production situation usually differs separately for women (column I) and men (column II). The between the genders. We therefore estimate all earnings and OLS estimates suggest a substantial positive effect of the participation equations separately for men and women. (potential) earnings on labor supply for women and men. According to Hypothesis 1, an increase in (potential) labor earnings by one percent significantly increases the prob 5 Estimation results ability of being active in the labor force by 0.056 percentage 5.1 Emplo yment decision to work beyond retirement points for women and by 0.246 percentage points for men, The first estimation step, the calculation of the earn - respectively. In other words, an increase in (potential) labor ings equation for all individuals in our sample, is shown Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 11 of 17 21 earnings by 10% has the potential to increase labor market Table 3 Linear probability model of labor force participation with the share of estimated earnings in disposable income participation of men from their average 20.5% (see Appendix Table 7) to about 23% and of women from 19.3% to almost Women Men 20%. According to Hypothesis 4, labor earnings elasticity Share of estimated earnings in 0.222** (0.054) 0.816** (0.057) is much higher for men than for women. Consistent with disposable income Hypothesis 2, an increase in pension entitlements by one N 30.784 44.887 percent significantly reduces the employment probability R 0.306 0.206 by 0.055 percentage points for women and 0.165 percentage Same list of covariates as in Table 2. Standard errors in parentheses are points for men. Hochfellner and Burkert (2013) find a simi - computed by bootstrapping with 200 repetitions and clustered on the lar pattern by gender, their income elasticities are however individual level. Data: SIAB 7514 somewhat lower. Pleau and Shauman (2013) find a smaller *p < 0.05 negative effect of retirement income on working after retire - **p < 0.01 ment for males than for females in their US sample, however. A higher pension entitlement leads to a stronger reduc- Table 4 Extensive margin of labor supply for employees with tion in labor market participation for men than for women, and without bridge paths accordingly. Bridge paths The individual characteristics show the expected signs on employment after retirement (compare for exam- Women Men ple Burkert and Hochfellner 2017; Larsen and Pedersen No Yes No Yes 2013; Maestas 2010): employees who use bridge paths have a significantly lower probability to work and there - Estimated 0.095** (0.024) 0.057* (0.029) 0.307** (0.095) 0.244** earnings (log) (0.064) fore labor market gaps before retirement have a negative N 19,887 10,897 19,223 25,664 correlation with labor market participation after retire- R 0.407 0.067 0.287 0.071 ment. Academics work less after retirement than skilled Same list of covariates as in Table 2. Standard errors in parentheses are employees. Occupations before retirement hardly have 27 computed by bootstrapping with 200 repetitions and clustered on the an influence on working after retirement, however. individual level. Data: SIAB 7514 These are the effects of the establishment character - *p < 0.05 istics of the last employer prior to retirement on the **p < 0.01 labor market decision of the retiree: workers who are employed in firms with more than ten employees prior to retirement show a lower labor force participation after The estimation equation in Table 3 replaces the sepa- retirement. Furthermore, there is a negative correla- rate labor earnings and pension entitlement variables by tion between mean earnings of all full-time employees, the share of earnings in disposable income. This variable the share of highly qualified employees, and the share has a substantial positive impact on the extensive margin of employees in partial retirement and work after retire- of labor supply. An increase in the share of (potential) ment. The higher the share of employees who work after labor earnings in disposable income after retirement by retirement at the last employer, the higher is the prob- one percentage point increases the probability of working ability of being employed after retirement. beyond retirement by 0.22 percentage points for women Note that the (potential) labor earnings regressor is and 0.82 percentage points for men. This result supports generated because it has been estimated in a separate Hypothesis 3. step with uncertainty (Borra 2010). Therefore, the stand - ard errors of this variable are bootstrapped. A bivariate 5.2 Heterogeneity in earnings elasticities Probit model is used to verify the results of the linear Table 4 shows differences in the estimated elasticity of probability model presented in Table 2. The results of labor force participation with respect to (potential) labor both estimation models are consistent with each other. earnings between employees who use bridge paths versus those who do not use bridge paths prior to retirement. In line with Hypothesis 5 and prior empirical evidence We obtain larger coefficients for the estimated earnings variables in the (Anger et al. 2018; Burkert and Hochfellner 2017; Hoch- labor participation equation if we drop the endogeneity correction (Mills fellner and Burkert 2013), the estimates suggest a higher ratio) in the earnings estimation step. (Potential) earnings however remain positive earnings effect on labor supply for employees significant and larger for men than for women without controlling for endo - geneity. without bridge paths than for employees with bridge All results remain robust when only individual characteristics are paths. The earnings elasticities for employees without a included in the participation equation and employer characteristics are not bridge path are twice as large for women and 50% larger controlled for. 21 Page 12 of 17 S. Lorenz , T. Zwick Table 5 Extensive margin of labor supply for employees whose Table 6 Extensive margin of labor supply for employees with region of residence is the same as or different from the place of low and high earnings before retirement work Earnings before retirement Region of residence is same as region of work Women Men Women Men Low High Low High earnings earnings earnings earnings No Yes No Yes Estimated 0.084 (0.045) 0.063 (0.039) 0.384** 0.097 (0.110) Estimated 0.062** 0.080 (0.083) 0.163** 0.234* (0.146) ernings (0.129) earnings (0.026) (0.084) (log) (log) N 7441 7441 10.706 10,706 N 19,080 2289 25,799 6769 R 0.365 0.273 0.283 0.171 R 0.313 0.315 0.239 0.233 High earnings are defined as earnings from the last employment subject to List of covariates is the same as in Table 2. Number of observations lower than in social insurance contributions with earnings above the 75th percentile of all last Table 2 because region of residence is not reported for all employees. Standard earnings of employment subject to social insurance contributions. Low earnings errors in parentheses are computed by bootstrapping with 200 repetitions and are defined as earnings from the last employment subject to social insurance clustered on the individual level. Data: SIAB 7514 contributions with earnings less than 25th percentile of all last earnings of *p < 0.05 employment subject to social insurance contributions. List of covariates **p < 0.01 is the same as in Table 2. Standard errors in parentheses are computed by bootstrapping with 200 repetitions and clustered on the individual level. Data: SIAB 7514 *p < 0.05 for men than those for employees with bridge paths, see **p < 0.01 Table 5. Employees without bridge paths therefore seem to derive lower utility from leisure and they are more responsive to financial incentives when deciding about quartiles both have lower elasticities than the second and employment after retirement. third earnings quartiles and therefore Hypothesis 7 is not When we compare earnings elasticities between supported for women. Our evidence only is in accord- employees who have to commute to work to a different ance with findings by Pleau (2010) and Pleau and Shau - region in their last employment prior to retirement and man (2013) for US men. They however find an increasing those who live and work in the same region, we use the labor market supply elasticity with pre-retirement earn- approach proposed by Kropp and Schwengler (2016) ings for US women. The low labor supply elasticity of to delineate 50 functional West German labor market women with very low earnings before retirement may be regions based on commuting flows. The dummy has a consequence of high alternative financial sources for value one if an employee commutes to a workplace out- example from other household members that we cannot side the own labor market region. Our results do not sup- observe in our data (Blundell et al. 2007; Micheel et al. port Hypothesis 6: there are practically no differences 2010). between commuters and non-commuters for females and the differences between both employee groups are small 6 Discussion and limitations for men, see Table 6. We hardly find changes in the Our paper uses the empirical identification approach commuting situation before and after retirement. Com- developed for the employment decision of women with muting costs therefore seem not to play a central role for small children (Allègre et al. 2015; Borra 2010; Ram- the labor supply decision of older employees for example mohan and Whelan 2007; Ribar 1992; Viitanen 2005). in comparison to financial needs of older employees. This literature mainly uses structural estimation mod - Our last sample split differentiates employees accord - els because there are hardly any quasi-natural experi- ing to their earnings situation before exiting the labor ments that can be exploited to obtain causal relationships market into retirement. According to Hypothesis 7, we between financial incentives and the employment deci - find a higher elasticity of men in the lowest earnings sion. Also the literature on determinants for work after quartile in comparison to men in the highest earnings quartile. For women, the lowest and highest earnings The labor market regions are defined based on the place of work and the residential regions are defined based on the place of residence at the last employment prior to retirement. 29 30 The results remain robust when the information about place of work and A notable exception is Bastani et al. (2021) who analyze the impact of a place of residence are directly used from the dataset instead of the approach change in labor market participation tax rates on labor supply of secondary proposed by Kropp and Schwengler (2016). earners. Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 13 of 17 21 retirement usually does not exploit quasi-natural experi- example were marked differences with respect to the ments (Chetty et al. 2011). The main reason for the dearth role of women on the labor market and in the household of quasi-experimental evidence is that there were hardly between East and West Germany before and shortly after any changes in the rules for labor earnings after retire- the reunification of both countries (Sprengholz et al. ment. In Germany, earnings beyond marginal employ- 2020). In addition, there seems to be a secular reduc- ment have to be taxed at the common rates, marginal tion in wage elasticities over the last decades in many employment is tax free and the maximum earnings level countries (Bargain and Peichl 2016). It therefore remains for marginal employment was practically unchanged. In unclear what the influence of the observation period and 2005, a pension supplement of 6% per year of retirement institutional differences is for estimations of labor sup - after NRA was introduced for all pension entitlements. ply elasticities in other countries or in the Eastern part of This option was used by a negligible share of retirees, Germany. however. Besides using a structural approach our study is limited by 7 Conclusions several gaps in relevant information. Our data set does not This paper provides a labor supply model for retirees and report hours worked and we therefore only could analyze empirical assessments based on a large administrative labor the extensive margin. Future work, as a consequence, should market history data set. We for the first time use detailed analyze retiree´s labor supply respect to the number of hours information about labor earnings and non-labor income of worked conditional on employment (intensive margin). In a large administrative 2% sample of all employees after their addition, our administrative data only include pension enti- retirement in Germany. We concentrate on the effect of tlements derived from labor earnings. Future work therefore labor earnings on the labor force participation beyond retire- should include additional available financial means during ment and take endogeneity of the labor decision and hetero- retirement such as occupation pensions, additional pen- geneities between different employee groups into account. sion entitlements for example from raising children or rent We first establish that labor earnings achieved by pension - from wealth, assets, and real estate. Also considerable finan - ers after retirement constitute a substantial part of disposable cial demands such as mortgage debt or financially depend - income. The share of post-retirement earnings in dispos - ent household have an important impact on the decision to able income is more than 60% for women and 50% for men. work that is unobservable for us. Moreover, decisions about We then show that men are more responsive to financial labor supply after retirement may result from considera- incentives than women. A one percent increase in (poten- tions involving other individuals such as pension redistribu- tial) post-retirement labor earnings increases the employ- tions after a divorce or widow/widower pensions. Especially ment probability of pensioners by 0.246 percentage points equivalent household income may be a more important for males and by 0.056 percentage points for females. Labor driver for work after retirement than the individual financial earnings therefore have a strong incentive effect for work situation. We know in addition that males and females react after retirement. Higher pension entitlements however lead differently to financial stimuli within the household context to a negative labor participation effect. We analogously find (Blundell et al. 2007; Micheel et al. 2010). The distribution that the higher the share of (potential) labor earnings in dis- of wage elasticities of single childless women is very similar posable income of the retiree, the higher is the labor force to that of men. It therefore is mainly married women and participation of the individual. We also show that individu- mothers who have higher wage elasticities (Bargain and Pei- als who work until retirement instead of using the bridge chl 2016). As a consequence, also the household and marital options unemployment or partial retirement as well as men situation and changes in behavior of household members not with low earnings prior to retirement are stronger attracted affected by financial incentive changes should be included in by (potential) earnings incentives to work after retirement. future research in order to get a full and unbiased picture of Having to commute before retirement in contrast to residing financial incentives for work after retirement. in the same region as the place of work does not influence Transferability of our results to other countries work participation after retirement, however. depends on the empirical question whether there are dif- Our analysis therefore shows that tax incentives or ferences in labor supply elasticities between countries. higher earnings offers by employers are most effective It seems that there are little international differences in to stimulate labor market participation after retire- work preference (Bargain et al. 2014), although there for ment for males who earned not much before retire- ment and did not use bridge options. The calculation of the separate elasticities for labor earnings and pen- sion eligibilities for example allows us the prediction The maximum earnings level for minimum employment was increased on 1 of changes in financial incentives on work supply after January 2013 from 400€ to 450€, but further unchanged during our observa- retirement for different demographic groups. More tion period. 21 Page 14 of 17 S. Lorenz , T. Zwick specifically, our analysis shows that the gradual reduc- Appendix tion of tax-free allowance to zero until 2040 that was See Tables 7, 8, 9, and 10. induced by the Retirement Income Act 2004 (Genser Table 7 Number and share of individuals who work beyond and Holzmann 2019) may increase incentives to work retirement by birth cohort after retirement. We also show that an increase in the tax-free allowance for work during pension also may Birth cohort Women Men have a positive effect for the older employees affected N Share N Share by the measure. Average 2056 0.193 3180 0.205 8 Methods 1935 1934 0.192 3075 0.204 On the basis of theoretical hypotheses derived from a 1936 1992 0.202 3142 0.215 labor supply model, we propose an empirical choice 1937 2384 0.206 3586 0.178 model to examine the employment decision after 1938 2675 0.197 3933 0.207 receiving a pension. The decision to participate in the 1939 2582 0.206 4091 0.193 labor market is estimated based on a linear probabil- 1940 2807 0.174 4306 0.200 ity model using the employment status after retirement 1941 2783 0.192 4052 0.196 as the outcome variable. In order to impute (potential) 1942 2336 0.194 3292 0.201 labor earnings for all individuals included in the sam- 1943 2330 0.204 3334 0.215 ple as explanatory variable, a two-step Heckman-type 1944 2418 0.208 3201 0.206 selection model is used. For the estimations, a sample 1945 1775 0.235 2551 0.254 of 2% of employees from the Integrated Employment 1946 2264 0.223 2934 0.261 Biographies from 1975 to 2014 (SIAB, 1975–2014) is 1947 2504 0.229 3390 0.265 used. Data: SIAB 7514 Table 8 Sample characteristics Work beyond retirement Women Men No Yes Diff. No Yes Diff. Employee characteristics prior to retirement Pension entitlements 790.119 (314.7) 696.246 (278.6) 93.872** (4.354) 1283.311 (414.2) 1183.442 (355.01) 99.869** (4.641) Estimated earnings (log) − 0.016 (0.982) 1.082 (0.593) 1.098** (0.013) 1.236 (0.618) 1.710 (0.498) 0.474** (0.007) Observed earnings (log) 2.352 (0.721) 2.520 (0.887) Share of unemployment bridge path 0.258 (0.438) 0.159 (0.365) 0.099** (0.006) 0.409 (0.492) 0.250 (0.433) 0.160** (0.006) Share of partial retirement bridge path 0.136 (0.342) 0.045 (0.207) 0.091** (0.005) 0.223 (0.416) 0.109 (0.312) 0.114** (0.005) Manufacturing occupations 0.175 (0.380) 0.143 (0.350) 0.033** (0.006) 0.550 (0.498) 0.468 (0.499) 0.082** (0.006) Service occupations 0.308 (0.462) 0.356 (0.479) 0.048** (0.007) 0.188 (0.390) 0.289 (0.453) 0.101** (0.005) Administrative occupations 0.517 (0.500) 0.502 (0.500) 0.015* (0.007) 0.262 (0.440) 0.244 (0.429) 0.019** (0.005) Accumulated labor market gaps prior to 32.777 (25.380) 4.309 (13.736) 28.468** (0.332) 19.575 (22.322) 3.438 (11.382) 16.137** (0.236) age 65 in months Characteristics of last employer prior to retirement Firm size 681.909 (2400.6) 366.367 (1549.8) 315.5* (31.885) 1788.5 (5732.3) 952.760 (4163.6) 835.7** (62.703) Mean imputed earnings of all full-time 89.652 (33.863) 80.465 (32.472) 9.188** (0.475) 108.250 (44.180) 96.706 (38.056) 12.54** (0.495) employees Share of highly qualified employees 0.116 (0.162) 0.088 (0.145) 0.029** (0.002) 0.131 (0.161) 0.098 (0.144) 0.033** (0.002) Share of employees in partial retirement 0.015 (0.038) 0.008 (0.028) 0.007** (0.0005) 0.022 (0.045) 0.015 (0.043) 0.007** (0.0005) Share of employees aged of 55 to 59 0.133 (0.135) 0.120 (0.134) 0.012** (0.002) 0.121 (0.110) 0.113 (0.110) 0.008** (0.001) Share of employees aged of 60 to 64 0.059 (0.105) 0.083 (0.137) 0.024** (0.002) 0.057 (0.098) 0.077 (0.123) 0.020** (0.001) Share of employees aged 65 and older 0.012 (0.039) 0.032 (0.081) 0.020 (0.0007) 0.009 (0.036) 0.028 (0.078) 0.020** (0.0005) N 24,511 6273 30,784 35,338 9549 44,887 Mean is given for continuous variables. Standard deviations of the continuous variables and standard errors for the mean value differences are given in parentheses. Data: SIAB 7514 Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 15 of 17 21 Table 9 Correlations between previous labor market experience and estimated as well as observed earnings Pension entitlements (log) Observed earnings before retirement (log) Women Estimated earnings (log) − 0.1117 − 0.0277 Observed earnings after retirement (log) 0.1582 0.2299 Men Estimated earnings (log) 0.0979 0.0592 Observed earnings after retirement (log) 0.1494 0.1962 Table 10 Auxiliary estimations Women Men Observations 30,784 44,887 Censored observations 24,511 35,338 Uncensored observations 6273 9549 Log likelihood − 11,791 − 19,581 χ (9) 7545.47 7300.83 Prob > χ : 0.2424 0.000 Labour force participation Log wage Labour force participation Log wage Coef. SE Coef. SE Constant 1.292** 0.176 − 0.637** 0.201 1.186** 0.0171 − 1.320** 0.211 No degree Ref. Ref. Ref. Ref. Ref. Ref. Ref. Ref. Vocational training 0.227** 0.027 0.231** 0.041 0.378** 0.028 0.231** 0.043 University degree 0.043 0.049 0.351** 0.062 0.149** 0.034 0.566** 0.046 Manufacturing occupations Ref. Ref. Ref. Ref. Ref. Ref. Ref. Ref. Service occupations 0.228** 0.029 0.261** 0.046 0.253** 0.018 0.223** 0.029 Administrative occupations 0.096** 0.028 0.239** 0.037 0.073** 0.018 0.198** 0.023 Acc. labor market gaps until 65 − 0.033** 0.0006 − 0.035** 0.004 − 0.031** 0.0005 − 0.022** 0.002 in months Pension entitlements (log) − 0.275** 0.025 − 0.494** 0.024 Lambda 1.205** 0.157 0.705** 0.084 Hausman test Chi (8) 57.73 69.00 Prob > chi : 0.0000 0.0000 In both equations, year of employment exit prior to the age of 65 is controlled for. SE: standard error. Data: SIAB 7514 *p < 0.05 **p < 0.01 Acknowledgements at the Institute for Employment Research (IAB) and subsequently via remote We are grateful to Mona Bruns for helpful discussions on this paper. data access. These data are publicly available. Authors’ contributions Declarations Both authors contributed in equal shares to the paper. Both authors read and approved the final manuscript. Ethics approval and consent to participate Not applicable. Funding Not applicable. Consent for publication Not applicable. Availability of data and materials The paper uses the SIAB7514. Data access was provided via on site use at the Research Data Centre (FDZ) of the German Federal Employment Agency (BA) 21 Page 16 of 17 S. Lorenz , T. Zwick Competing interests Cogan, J.F.: Labor supply with time and money costs of participation. Rand Not applicable. Report R-2044-HEW. Santa Monica CA (1977) Cogan, J.F.: Fixed costs and labor supply. Econometrica 49(4), 945–963 (1981) Author details Connelly, R.: The effect of child care costs on married women’s labor force Faculty of Business Management and Economics, University of Würzburg, participation. Rev. Econ. Stat. 74(1), 83–90 (1992) Sanderring 2, 97070 Würzburg, Germany. ZEW Leibniz Centre for European Connelly, R., Kimmel, J.: Marital status and full-time/part-time work status in Economic Research, Mannheim, Germany. Research Centre for Education child care choices. Appl. Econ. 35(7), 761–777 (2003) and the Labour Market, University of Maastricht, Maastricht, The Netherlands. Del Boca, D., Vuri, D.: The mismatch between employment and child care in Italy: the impact of rationing. J. Popul. Econ. 20(4), 805–832 (2007) Received: 12 April 2021 Accepted: 21 September 2021 Dittrich, D., Büsch, V., Micheel, F.: Working beyond retirement age in Germany: the employee’s perspective. In: Ennals, R., Salomon, R. (eds.) Older Work- ers in a Sustainable Society, pp. 189–202. Peter Lang, Frankfurt am Main (2011) Engstler, H., Romeu-Gordo, L.: Arbeiten im Ruhestand—Entwicklung, Faktoren und Motive der Erwerbstätigkeit von Altersrentenbeziehern. In: Kistler, References E., Trischler, F. (eds) Reformen auf dem Arbeitsmarkt und in der Alters- Allègre, G., Simonnet, V., Sofer, C.: Child care and labor market participation sicherung—Folgen für die Einkunftslage im Alter. edition der Hans-Böck- in France: do monetary incentives matter? Ann. Econ. Stat. 117(118), ler-Stiftung, Arbeit und Soziales 196, Düsseldorf, pp. 115–147 (2014) 115–139 (2015) Engstler, H., Romeu-Gordo, L.: Der Übergang in den Ruhestand: Alter, Pfade Anger, S., Trahms, A., Westermeier, C.: Erwerbstätigkeit nach dem Übergang in und Ausstiegspläne. In: Mahne, K., Wolff, C., Simonson, J., Tesch-Römer, C. Altersrente: Soziale Motive überwiegen, aber auch Geld ist wichtig. IAB- (eds.) Altern im Wandel, pp. 65–80. Springer, Wiesbaden (2017) Kurzbericht, 24/2018. Nuremberg (2018) Erlinghagen, M.: Entlassungen und Beschäftigungssicherheit im Zeitverlauf. Z. Antoni, M., Ganzer, A., vom Berge, P.: Stichprobe der Integrierten Arbeitsmark- Soziol. 34(2), 147–168 (2005) tbiografien (SIAB) 1975–2014. FDZ Datenreport, 04/2016. Nuremberg Eurofound: Income From Work After Retirement in the EU. Publications Office (2016) of the European Union, Luxembourg (2012) Bäcker, G., Jansen, A., Schmitz, J.: Rente erst ab 70? Probleme und Perspektiven Evers, M., De Mooij, R., van Vuuren, D.: What explains the variation in estimates des Altersübergangs. IAQ-Forschung 2017-02. University of Duisburg- of labor supply elasticities? Ifo Institute for Economic Research Working Essen (2017) Paper Series, 1633. Munich (2005) Bargain, O., Orsini, K., Peichl, A.: Comparing labor supply elasticities in Europe Fasbender, U., Wang, M., Voltmer, J.B., Deller, J.: The meaning of work for postre- and the United States—new results. J. Hum. Resour. 49(3), 723–838 tirement employment decisions. Work Aging Retire. 2(1), 12–23 (2016) (2014) Genser, B. Holzmann, R.: Taxing German old-age pensions fairly and efficiently, Bargain, O., Peichl, A.: Own-wage labor supply elasticities: variation across time ifo DICE report volume 17. Munich (2019) and estimation methods. IZA J. Labor Econ. 5(10), 1–31 (2016) Geyer, J., Haan, P., Lorenz, S., Zwick, T. Bruns, M.: The role of labor demand in the Bastani, S., Moberg, Y., Selin, H.: The anatomy of the extensive margin labor labor market effects of pension reforms. Industrial Relations, forthcoming supply response. Scand. J. Econ. 123(1), 33–59 (2021) (2021) Beehr, T., Bennett, M.: Working after retirement: features of bridge employment Grund, C.: Severance payments for dismissed employees in Germany. Eur. J. and research directions. Work Aging Retire. 1(1), 112–128 (2015) Law Econ. 22(1), 49–71 (2006) Blau, F., Kahn, L.: Changes in the labor supply behavior of married women: Gutiérrez-i-Puigarnau, E., van Ommeren, J.: Labour supply and commuting. J. 1980–2000. J. Law Econ. 25(3), 393–438 (2007) Urban Econ. 68(1), 82–89 (2010) Blossfeld, H.P.: Berufseintritt und Berufsverlauf. Mitt. Arbeitsmarkt Berufsforsch. Hanel, B.: Financial incentives to postpone retirement and further effects on 18(2), 177–197 (1985) employment—evidence from a natural experiment. Labour Econ. 17(3), Blundell, R., Chiappori, P., Magnac, T., Meghir, C.: Collective labour supply: het- 474–486 (2010) erogeneity and non-participation. Rev. Econ. Stud. 74(2), 417–445 (2007) Hank, K., Buber, I.: Grandparents caring for their grandchildren: findings from Borra, C.: Childcare costs and Spanish mothers’ labor force participation. Haci- the 2004 survey of health, ageing, and retirement in Europe. J. Fam. Issues enda Pública Española/revista De Economía Pública 194(3), 9–40 (2010) 30(1), 53–73 (2009) Brenke, K.: Immer mehr Menschen im Rentenalter sind berufstätig. DIW Heckman, J.: The common structure of statistical models of truncation, sample Wochenbericht 80(6), 3–12 (2013) selection and limited dependent variables and a simple estimator for Brussig, M.: Erwerbstätigkeit im Alter hängt vom Beruf ab. Altersübergangs- such models. Ann. Econ. Soc. Meas. 5(4), 475–492 (1976) Report, 2010-05. University of Duisburg-Essen (2010) Heckman, J.: Sample selection bias as a specification error. Econometrica 47(1), Brussig, M.: Alter beim Austritt aus sozialversicherungspflichtiger Beschäfti- 153–161 (1979) gung ist gestiegen. Altersübergangs-Report, 2015-01 (2015) Hochfellner, D., Burkert, C.: Berufliche Aktivität im Ruhestand. Z. Gerontol. Brussig, M., Bellmann, L.: Betriebliche Einflussfaktoren auf Rekrutierungen Geriatr. 46(3), 242–250 (2013) Älterer und deren Bewerbungen. German J. Hum. Resour. Manag. 22(1), Hofäcker, D., Naumann, E.: The emerging trend of work beyond retirement 35–57 (2008) age in Germany. Z. Gerontol. Geriatr. 48(5), 473–479 (2015) Burkert, C., Hochfellner, D.: Employment trajectories beyond retirement. J. Institut für Arbeit und Qualifikation.: Erwerbstätige nach Stellung im Beruf Aging Soc. Policy 29(2), 143–167 (2017) 1994–2019. Sozialpolitik aktuell in Deutschland, University of Duisburg- Büsch, V., Dorbritz, J., Heien, T., Micheel, F.: Weiterbeschäftigung im Rentenalter: Essen (2020) Wünsche, Bedingungen, Möglichkeiten. Materialien zur Bevölkerungswis- Kanabar, R.: Post-retirement labour supply in England. J. Econ. Ageing 6, senschaft, 129/2010 (2010) 123–132 (2015) Cahill, K., Giandrea, M., Quinn, J.: Retirement patterns and the macroeconomy, Kim, S., Feldmann, D.C.: Working in retirement: the antecedents of bridge 1992–2010: the prevalence and determinants of bridge jobs, phased employment and its consequences for quality of life in retirement. retirement, and reentry among three recent cohorts of older Americans. Acad. Manag. J. 43(6), 1195–1210 (2000) Gerontologist 55(3), 384–403 (2015) Komp, K., van Tilburg, T., Broese van Groenou, M.: Paid work between age 60 Cahuc, P., Carcillo, S., Zylberberg, A.: Labor Economics. MIT press, Cambridge and 70 years in Europe: a matter of socio-economic status? Int. J. Ageing (2004) Later Life 5(1), 45–75 (2010) Cameron, C., Trivedi, P.: Microeconometrics—Methods and Applications. Kropp, P., Schwengler, B.: Three-step method for delineating functional labor Cambridge University Press, Cambridge (2002) market regions. Reg. Stud. 50(3), 429–445 (2016) Chetty, R., Guren, A., Manoli, D., Weber A.: Does indivisible labor explain the Larsen, M., Pedersen, P.: To work, to retire- or both? Labor market activity after difference between micro and macro elasticities? A meta-analysis of 60. IZA J. Eur. Labor Stud. 2, 21 (2013) extensive margin elasticities. NBER Working Paper 16729. Cambridge MA (2011) Money also is sunny in a retiree’s world: financial incentives and work after retirement Page 17 of 17 21 Lorenz, S., Pfister, M., Zwick T.: Identification of the statutory retirement dates in Ribar, D.C.: Child care and the labor supply of married women: reduced form the Sample of Integrated Labor Market Biographies (SIAB). FDZ Method- evidence. J. Hum. Resour. 27(1), 134–165 (1992) enreport, 08/2018. Nuremberg (2018) Rogerson, R., Wallenius, J.: Micro and macro elasticities in a life cycle model Lorenz, S., Pfister, M., Zwick, T.: Beware of the employer: financial incentives for with taxes. J. Econ. Theory 144(6), 2277–2292 (2009) employees may fail to prolong old-age employment. ZEW Discussion Saba, T., Guerin, G.: Extending employment beyond retirement age: the case of Paper, 20-007. Mannheim (2020) health care managers in Quebec. Public Pers. Manag. 34, 195–214 (2005) Maestas, N.: Back to work: expectations and realizations of work after retire- Schellenberg, G., Turcotte, M., Ram, B.: Post-retirement employment. Stat. Can. ment. J. Hum. Resour. 45(3), 718–748 (2010) Perspect. 75–001, 14–17 (2005) Maxin, L., Deller, J.: Activities in retirement: individual experience of silver work. Schmähl, W.: Ageing workforce: firm strategies and public policy in Germany. Comp. Popul. Stud. 35(4), 801–832 (2010) Geneva Pap. Risk Insur. 28(4), 575–595 (2003) Meyer, B., Rosenbaum, D.: Welfare, the earned income tax credit, and the labor Smeaton, D., McKay, S.: Working after state pension age: quantitative analysis. supply of single mothers. Q. J. Econ. 116(3), 1063–1114 (2001) UK Department for Work and Pensions Research Report 182. London Micheel, F.: The intention to paid and/or unpaid activities in retirement—a (2003) study of older workers in Germany. J. Aging Soc. Policy 33(2), 101–119 Sprengholz, M., Wieber, A., Holst, E.: Gender identity and wives’ labor market (2021) outcomes in West and East Germany between 1983 and 2016. Socio- Micheel, F., Cihlar, V., Riedl, S.: “Der Geist ist willig, aber...”—Was Menschen Econ. Rev. (2020). https:// doi. org/ 10. 1093/ ser/ mwaa0 48 daran hindert, im Ruhestand zu arbeiten. Bevölkerungsforschung Aktuell, Sullivan, S.E., Al Ariss, A.: Employment after retirement: a review and framework 6/2017, 2–9 (2017) for future research. J. Manag. 45(1), 262–284 (2019) Micheel, F., Roloff, J., Wickenheiser, I.: The impact of socioeconomic character - Torka, N., Goedegebure, I., Van Ewijk, I., Looise, J.: On the motives and needs for istics on older employees’ willingness to continue working in retirement work beyond age 65: comparing voluntary workers versus agency work- age. Comp. Popul. Stud. Zeitschrift Für Bevölkerungswissenschaft 35(4), ers. German J. Hum. Resour. Manag. 26(2), 167–188 (2012) 869–902 (2010) Ullmann, K., Bothfeld, S.: The German employment protection act—how does Mincer, J.: Labor force participation of married women. In: Lewis, H. (ed.) it work in company practice? WSI Discussion Paper 161, Duesseldorf Aspect of Labor Economics. Universities National Bureau of Economic (2008) Research Conference Series, vol. 14, pp. 63–105. Princeton University Vargha, L., Gál, R., Crosby-Nagy, M.: Household production and consumption Press, Princeton (1962) over the life cycle: national time transfer accounts in 14 European coun- Pfarr, C., Maier, C.: Arbeit trotz Rente: Warum bleiben Menschen in Ruhestand tries. Demogr. Res. 36(32), 905–944 (2017) erwerbstätig. Deutsches Institut für Altersvorsorge, Berlin (2015) Viitanen, T.K.: Cost of childcare and female employment in the UK. Labour 19, Pfister, M., Lorenz, S., Zwick, T.: Implementation of pension entitlements in the 149–170 (2005) sample of integrated labor market biographies (SIAB). FDZ Methodenre- Wang, M., Zhan, Y., Liu, S., Shultz, K.: Antecedents of bridge employment: a port, 01/2018. Nuremberg (2018) longitudinal investigation. J. Appl. Psychol. 93(4), 818–830 (2008) Pleau, R.: Gender differences in postretirement employment. Res. Aging 32(3), Wanger, S.: Altersteilzeit: Beliebt, aber nicht zukunftsgerecht. IAB-Kurzbericht, 267–303 (2010) 08/2009 (2009) Pleau, R., Shauman, K.: Trends and correlates of post-retirement employment, Westermeier, C.: Ältere am Arbeitsmarkt. Eine stabile Beschäftigung vor dem 1977–2009. Hum. Relat. 66(1), 113–141 (2013) Rentenalter begünstigt die Weiterarbeit. IAB-Kurzbericht, 15/2019 (2019) Rammohan, A., Whelan, S.: The Impact of childcare costs on the full-time/part- time employment decisions of Australian mothers. Australian Economic Publisher’s Note Papers 46(2), 152–169 (2007) Springer Nature remains neutral with regard to jurisdictional claims in pub- Rasner, A., Etgeton, S.: Rentenübergangspfade: Reformen haben großen lished maps and institutional affiliations. Einfluss. DIW-Wochenbericht 81(19), 431–441 (2014) Rhein, T.: Arbeiten im Rentenalter: Erwerbstätigkeit 65plus in Europa. Aktuelle Berichte, 25/2016. Institute for Employment Research, Nuremberg (2016)

Journal

Journal for Labour Market Research – Springer Journals

Published: Dec 1, 2021

Keywords: Work after retirement; Financial incentives; Labor and non-labor income; J14; J22; J26

References

Child care and labor market participation in France: do monetary incentives matter?

Allègre, G; Simonnet, V; Sofer, C
Comparing labor supply elasticities in Europe and the United States—new results

Bargain, O; Orsini, K; Peichl, A
Own-wage labor supply elasticities: variation across time and estimation methods

Bargain, O; Peichl, A
The anatomy of the extensive margin labor supply response

Bastani, S; Moberg, Y; Selin, H
Working after retirement: features of bridge employment and research directions

Beehr, T; Bennett, M
Changes in the labor supply behavior of married women: 1980–2000

Blau, F; Kahn, L
Berufseintritt und Berufsverlauf

Blossfeld, HP
Collective labour supply: heterogeneity and non-participation

Blundell, R; Chiappori, P; Magnac, T; Meghir, C
Childcare costs and Spanish mothers’ labor force participation

Borra, C
Immer mehr Menschen im Rentenalter sind berufstätig

Brenke, K
Betriebliche Einflussfaktoren auf Rekrutierungen Älterer und deren Bewerbungen

Brussig, M; Bellmann, L
Employment trajectories beyond retirement

Burkert, C; Hochfellner, D
Retirement patterns and the macroeconomy, 1992–2010: the prevalence and determinants of bridge jobs, phased retirement, and reentry among three recent cohorts of older Americans

Cahill, K; Giandrea, M; Quinn, J
Fixed costs and labor supply

Cogan, JF
The effect of child care costs on married women’s labor force participation

Connelly, R
Marital status and full-time/part-time work status in child care choices

Connelly, R; Kimmel, J
The mismatch between employment and child care in Italy: the impact of rationing

Del Boca, D; Vuri, D
Working beyond retirement age in Germany: the employee’s perspective

Dittrich, D; Büsch, V; Micheel, F; Ennals, R; Salomon, R
Der Übergang in den Ruhestand: Alter, Pfade und Ausstiegspläne

Engstler, H; Romeu-Gordo, L; Mahne, K; Wolff, C; Simonson, J; Tesch-Römer, C
Entlassungen und Beschäftigungssicherheit im Zeitverlauf

Erlinghagen, M
The meaning of work for postretirement employment decisions

Fasbender, U; Wang, M; Voltmer, JB; Deller, J
Severance payments for dismissed employees in Germany

Grund, C
Labour supply and commuting

Gutiérrez-i-Puigarnau, E; van Ommeren, J
Financial incentives to postpone retirement and further effects on employment—evidence from a natural experiment

Hanel, B
Grandparents caring for their grandchildren: findings from the 2004 survey of health, ageing, and retirement in Europe

Hank, K; Buber, I
The common structure of statistical models of truncation, sample selection and limited dependent variables and a simple estimator for such models

Heckman, J
Sample selection bias as a specification error

Heckman, J
Berufliche Aktivität im Ruhestand

Hochfellner, D; Burkert, C
The emerging trend of work beyond retirement age in Germany

Hofäcker, D; Naumann, E
Post-retirement labour supply in England

Kanabar, R
Working in retirement: the antecedents of bridge employment and its consequences for quality of life in retirement

Kim, S; Feldmann, DC
Paid work between age 60 and 70 years in Europe: a matter of socio-economic status?

Komp, K; van Tilburg, T; Broese van Groenou, M
Three-step method for delineating functional labor market regions

Kropp, P; Schwengler, B
To work, to retire- or both? Labor market activity after 60

Larsen, M; Pedersen, P
Back to work: expectations and realizations of work after retirement

Maestas, N
Activities in retirement: individual experience of silver work

Maxin, L; Deller, J
Welfare, the earned income tax credit, and the labor supply of single mothers

Meyer, B; Rosenbaum, D
The intention to paid and/or unpaid activities in retirement—a study of older workers in Germany

Micheel, F
The impact of socioeconomic characteristics on older employees’ willingness to continue working in retirement age

Micheel, F; Roloff, J; Wickenheiser, I
Labor force participation of married women

Mincer, J; Lewis, H
Gender differences in postretirement employment

Pleau, R
Trends and correlates of post-retirement employment, 1977–2009

Pleau, R; Shauman, K
The Impact of childcare costs on the full-time/part-time employment decisions of Australian mothers

Rammohan, A; Whelan, S
Rentenübergangspfade: Reformen haben großen Einfluss

Rasner, A; Etgeton, S
Child care and the labor supply of married women: reduced form evidence

Ribar, DC
Micro and macro elasticities in a life cycle model with taxes

Rogerson, R; Wallenius, J
Extending employment beyond retirement age: the case of health care managers in Quebec

Saba, T; Guerin, G
Post-retirement employment

Schellenberg, G; Turcotte, M; Ram, B
Ageing workforce: firm strategies and public policy in Germany

Schmähl, W
Gender identity and wives’ labor market outcomes in West and East Germany between 1983 and 2016

Sprengholz, M; Wieber, A; Holst, E
Employment after retirement: a review and framework for future research

Sullivan, SE; Al Ariss, A
On the motives and needs for work beyond age 65: comparing voluntary workers versus agency workers

Torka, N; Goedegebure, I; Van Ewijk, I; Looise, J
Household production and consumption over the life cycle: national time transfer accounts in 14 European countries

Vargha, L; Gál, R; Crosby-Nagy, M
Cost of childcare and female employment in the UK

Viitanen, TK
Antecedents of bridge employment: a longitudinal investigation

Wang, M; Zhan, Y; Liu, S; Shultz, K

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Money also is sunny in a retiree’s world: financial incentives and work after retirement

Money also is sunny in a retiree’s world: financial incentives and work after retirement

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Money also is sunny in a retiree’s world: financial incentives and work after retirement

Money also is sunny in a retiree’s world: financial incentives and work after retirement

Abstract

Journal

Recommended Articles

References

Our policy towards the use of cookies