Inheritances and work for pay — will the expected wave of bequests undermine active ageing policies?

Abstract

It is frequently assumed that the inheritance of wealth undermines economic activity. If such an assumption is valid, the expected wave of bequests may have a negative impact on labour market activity of heirs, what might further weaken the financing of state pension systems. This paper provides a detailed review of the empirical findings on the associations of inheritances with labour market activity, that is labour force participation status and working hours, and presents own analyses based on the survey of health, ageing, and retirement in Europe. We find that the receipt of an inheritance is not related to labour force participation in general. Inheritance expectations even have a small, but statistically significant positive effect on remaining in the labour force for men. Women who expect an inheritance tend to reduce working hours, but the effect of having received an inheritance is not significant, neither for men nor for women. We conclude that the receipt of an inheritance will not affect labour market decisions, so that the expected wave of bequests will not undermine active ageing policies.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10433-022-00706-1.

Introduction

The ageing of societies has given rise to concerns about future pension payments nearly worldwide. One very frequently taken policy measure is delaying retirement. At the same time, a wave of bequests is expected in many western countries, as large cohorts of “boomers” advance into old age, who had the chance to accumulate wealth in times of economic growth and longer periods of peace. It is widely taken for granted that inheriting wealth as well as the expectation to inherit wealth reduces the willingness to be economically active (e.g. Cox 2014). A well-known and frequently cited proverb is “the parent who leaves his son enormous wealth generally deadens the talents and energies of the son” (Carnegie 1891: p. 54). This idea made Carnegie praise the advantages of poverty as a “blessing” (1891: p. 55). As inheritances from parents are nowadays typically received when the heirs are close to retirement age (cf. OECD 2021), private transfers of wealth may undermine active ageing policies (e.g. Malo and Sciulli 2021). But are these assumptions justified? What is the empirical evidence on the association of inheritances and labour force participation?

Following an overview on the literature on behavioural consequences of inheritance receipt and inheritance expectation, we analyse data from the Survey of Health, Ageing and Retirement in Europe (SHARE). We find that the receipt of an inheritance is not related to labour force participation in general, neither for men nor for women. Our results for inheritance expectations even point to the opposite direction: a significant positive effect of inheritance expectations on labour force participation is observed at least for men, but not for women. Women who expect an inheritance reduce working hours, but the effect of having received an inheritance is not significant. We conclude that the receipt of an inheritance does not impact labour market decisions such as to quit the job or to reduce the number of hours worked substantially, and hence, private transfers of wealth will not undermine active ageing policies.

Literature review

Behavioural consequences of inheritances for labour market participation have been relevant in scientific debates, primarily in economics, for generations. Authors have mainly investigated three questions: the impact of inheritances on labour market participation in general, that is, if individuals stop working for pay, whether heirs adjust the number of working hours per week, and the impact of inheritances on the timing of the transition into retirement. However, only a few empirical studies have been conducted and the empirical evidence, so far, is somewhat mixed.

Analysing administrative data from the US, Holtz-Eakin et al. (1993) confirm the view that inheritances discourage labour market participation of heirs. Excluding people close to retirement age, they find a strong effect: 18% of the group with inheritances greater than $150,000 left the labour force in between 1982 and 1985, compared to 5% in the group with inheritances under $25,000. Furthermore, “the percentage of individuals who were out of the labour force in 1982 and entered in 1985 decreases with the size of the inheritance, falling from 53% in the low inheritance group to 16% in the high inheritance group” (Holtz-Eakin et al. 1993: p. 420). Although a large number of millionaires in the sample makes generalization questionable,1 it seems highly plausible that large inheritances of millions of dollars have an effect on economic activity. Imbens et al. (2001) report a similar effect in case of lottery wins in a non-representative sample of lottery players. Income from labour decreases by about 11% after lottery wins, and the effect is found to be stronger for winners close to retirement age. As expected, the effect is strongest in case of large wins (more than $100,000 per year, what is more than 2 million in total). However, are such effects relevant in case of average or typical amounts received?

Joulfaian and Wilhelm (1994) do not find such strong effects. Analysing both the Panel Study of Income Dynamics (PSID) and an estate-income tax match dataset which is representative for the top 2.5% of the wealth distribution, the authors conclude that “the large inheritances to be enjoyed by the baby boom are not expected to lead to a substantial reduction in their labor supply” (1994: p. 1232). Having calculated a large number of statistical models, most effects of expected and unexpected inheritances were statistically insignificant, some in the opposite direction (namely, inheritances reduced the probability to be retired). For married women, they report an inheritance effect of 0.2% reduction in labour force participation and a 1.4% reduction in working hours for every $10,000 received based on the PSID data, but for the tax match data inheritances of married women were “inversely related to the amount” (1994: p. 1231).

Brown et al. (2010), using data from the Halth and Retirement Study (HRS), find that inheritance receipt increases the probability of retirement by 4% points within an 8 year period of observation, and this effect increases with the size of the inheritance. Furthermore, they find that unexpected inheritances have a larger effect than expected ones, although the mechanism behind this phenomenon remains unclear. Blau and Goldstein (2016) also analyse data from the HRS to test a bargaining model on labour force participation of older married couples. They report that receiving an unexpected inheritance reduces labour force participation of the recipient by 4% points but has no effect on the labour force participation of the spouse. This finding is consistent with the assumption that inheritances do have an impact on the retirement decision of the recipient.2

Using non-representative Swedish data3 compiled from tax registers, Elinder et al. (2012) find a substantial negative correlation between the amount inherited and income from labour after the receipt. They furthermore suspect an anticipation effect, e.g. that expected inheritances may have induced a reduction in working hours already before the inheritance was actually received, what could have further reduced the observed difference in income before and after the transfer. However, they do not have data on expected inheritances.

Amedah and Fougère (2017) make use of a proportional hazard model to investigate the impact of inheritances larger than $20,000 on the timing of retirement, based on data from a Survey of Financial Security (SFS) in Canada. They find that the conditional probability to retire for inheritance recipients is about 20% higher compared to non-recipients, so that their data “supports the hypothesis that receiving an inheritance significantly increases the retirement hazard” (10f.). They also find that women have a 27% higher retirement hazard, but do not check for different effects of inheritances for men and women. As the amount seems rather small (approx. €15,000 in 2012), the effect seems surprisingly strong: even in case of successful investment, this amount seems too small to increase retirement income significantly or quit working in the long run.

Doorley and Pestel (2020) analyse the German Socio-Economic Panel (GSOEP) and compare behavioural responses to both expected and unexpected inheritances by men and women, namely the reduction of working hours and the desired hours of work. They find that both expected and unanticipated inheritances do not result in a decrease in the number of hours worked by male employees. Only in case of female employees, they find a decrease by about 1.5 h per week, on average, for both actual and desired working hours. They conclude that “any concerns that policy makers may have about inheritances discouraging work among the working age population are limited” (2020: p. 862). This finding from Germany is in line with findings from Niizeki and Hori (2018) from Japan. Analysing the family and lifestyle survey (FLS), they additionally considered income from spouses, but found no significant decline in the probability that women worked when their husband received an inheritance. For men, no significant effect was found, but “the probability that women worked did decrease when they received an inheritance” (2018: p. 12) by about 16% points 3 years after the parent’s death.

Malo and Sciulli (2021) employ data from the Household Finance and Consumption Survey (HFSC) for 14 European countries to analyse the impact of wealth transfers received in 2008–2013 on the employment status reported in 2014. This cross-sectional data have been collected in 20 countries, but six countries had to be dropped because of missing questions or too many missing values in the variables of interest. They report significant effects for women only, not for men. However, as changes in employment status remain unobserved, respondents may as well have been inactive before the receipt of an inheritance of gift. Furthermore, age groups from 25 to 59 years are analysed, so that other incidents like birth of a child might interfere. In a separate analysis, they report on findings for the age group 45 to 59. Here, they do not find significant effects for receiving inheritances or gifts from family members—what is by far the most frequent case (e.g. OECD 2021)—but significant effects for receiving inheritances or gifts from non-family donors. (This part of the analysis is not differentiated for men and women.) Nevertheless, the authors conclude that “we should expect that more people would leave the labour force when receiving an inheritance” in the future (Malo and Sciulli 2021: p. 19).

Some authors already have used the data from the Survey of Health, Ageing and Retirement in Europe (SHARE) to analyse consequences of inheritance receipt and inheritance expectations.4 Eder (2016) compared labour market participation in waves 1 and 2 (2004/2005) to wave 4 (2011/2012). He finds that the receipt of an inheritance significantly increases the probability of retirement and the effect increases with the size of the inheritance. He also compares retirement planning before the receipt of an inheritance to factual retirement, and reports “a significant increase in the probability of retiring earlier than previously planned” (2016: p. 302)—with no difference between expected and unexpected inheritances. Eder mainly reports findings on the labour force participation status indicated by the respondents answer “retired” to a question on employment status, what—from our point of view—is not perfectly on target. Individuals may work in retirement or make use of pre-retirement regulations even if these will result in lower incomes from work or stop working completely without being retired. Furthermore, the status “retired” is in many countries defined through age limits of national pension systems, which are applied independent of an inheritance. Eder has also used a second dependent variable (results available online) that includes house makers, but not unemployed, permanently sick, and disabled—these categories are considered indicating “not retired”, although some of these individuals may be participating in early exit schemes which may become more attractive when receiving an inheritance. In sum, we would favour analysing exit from work and factual working hours instead of transition into retirement, as the former is the interesting behavioural response to inheritance receipt. However, we think the general approach chosen by Eder (2016) is promising. As wave four of SHARE covers a special historic period—shortly after the global financial crisis of 2007/2008, it is well worth to extend the time frame of this approach, for example, because of unemployment episodes or loss of retirement savings in these years.

In contradiction to these findings, Suari-Andreu (2018) does not find large effects, neither on working hours nor on the timing of retirement. He analysed SHARE waves one to six in ten countries, using two dummy variables, one indicating the transition from any labour force status to “retired”, the other indicating any change from “employed” or “self-employed”. Nearly, all effects are “very close to zero” (2018: p. 23). The few significant effects are even negative—the receipt of an inheritance correlates with a biographically later transition into retirement. The author suspects that either the amounts inherited might be too small or social security and labour market regulations might not be flexible enough in these ten countries.

Alternatively, only some population groups might react on receipt of inheritances, for example, only women who are in paid employment, not men. Already Joulfaian and Wilheln (1994) argued that it might be more likely that women respond to received inheritances by quitting work or reducing working hours. On average, their incomes are lower and they work part-time more frequently so that their income might be replaced more easily by income from inheritances. The findings from Niizeki and Hori (2018), Doorley and Pestel (2020) and Malo and Sciulli (2021) provide further evidence that analyses on inheritances and paid employment have to be done separately for men and women.

Taken together, we will add to the existing literature by analysing longitudinal data from 14 European countries that allows for identifying the effects of (a) inheritance expectations and (b) the receipt of inheritances on labour market participation and reduction of hours worked per week for both men and women, controlling for crucial variables impacting labour market participation and hours worked in paid employment. As the empirical evidence is mixed, we start with the same assumptions which are hypothesised in the classical economic theory, that is:

H1a

Inheritance expectations cause labour market exits;

H1b

Receipt of inheritances cause labour market exits;

H2a

Inheritance expectations induce a substantial reduction in working hours;

H2b

Receipt of inheritances induces a substantial reduction in working hours.

Data and methods

To study the consequences of inheritances and inheritance expectations for labour market participation, we use data from the Survey of Health, Aging and Retirement in Europe (SHARE)—a well-known multidisciplinary, cross-national bank of microdata on health, psychological, and economic variables (Börsch-Supan et al. 2013). The data offer a unique means by which to compare the health, economic situation, and welfare of older people in different European countries over time (Lowenstein et al. 2019; Tur-Sinai and Lewin-Epstein, 2020). As our baseline data—Time 1 [T1]—we use data from waves 1 (2004) and 2 (2006) of the survey. If individuals were surveyed in both waves, the earlier information is used. This baseline data are cross-referenced with data collected roughly nine years later—our Time 2 [T2]. For individuals surveyed in wave 1, information gathered in wave 5 (2013) is used; for those examined for the first time in wave 2, the cross-referenced information was that gathered in wave 6 (2015). Our study includes all individuals employed (including self-employed) and aged 50–54 at T1, who will in any case be younger than 65 years at T2 to minimize the number of regular transitions to retirement (OECD 2019). Participants in the study came from fourteen European countries: Austria, Belgium, Czech Republic, Denmark, France, Germany, Greece, Israel, Italy, Poland, Spain, Switzerland, and the Netherlands.

Each participant was asked at T2 about the current state of employment. Following a common definition, our first explained variable, participating in the labour force at T2, takes the value 1 if the respondent defined the labour force status as being employed, self-employed or unemployed, 0 in case of being retired, homemaker, or permanently sick or disabled. Hence, we are comparing those still in the labour force to those who left the labour force since T1, as our study includes only individuals which were employed (including self-employed) at T1.5 In addition, each participant who was active in the labour market also at T2 was asked at both times (T1, T2) about the usual number of hours worked per week. For this subsample, we calculated the change in the weekly workload between T1 and T2 (the latter as against the former) to construct our second dependent variable as a substantial reduction in number of hours worked per week (1 in case of a reduction of 20% or more, 0 in case of no change, smaller reduction, or increase).

The two independent variables of interest are expecting an inheritance at T1 and having received an inheritance at T2. The first variable examines the individual’s estimation of receiving an inheritance in the next 10 years and may take 3 values: 0 = small chance (0–24% chance of receiving an inheritance in the next ten years), 1 = medium chance (25–74%), and 2 = strong chance (75–100%).6 The second variable examines whether the respondent or a spouse received an inheritance (in cash, goods, or assets) since T1. Answers have been compiled across waves as the question wording asks for inheritances since the last participation in the SHARE-survey and copied from us to each partner in the household as only one household member has received this question7 (we used this strategy in all situations where information for the couple or household has been provided by only one respondent, e.g. the financial respondent’s subjective assessment of the ability of the household to make ends meet). This allows to analyse the behavioural consequences of expectations and inheritances for labour force participation, which are individual-level outcomes, and to compare effects for men and women. Logistic regression is used to analyse the effects of the expectation to receive an inheritance (measured at T1) and the receipt of inheritances (measured at T2, including information from all waves in between T1 and T2) on labour force participation (measured at T2).8

In the vast literature on labour force participation and transitions into retirement, a wide range of important factors have been identified, beyond gender, for example, education, income, health and living arrangements as well as working conditions. Working conditions affect retirement timing decisions, although the relationship is complex. For example, analysing the English longitudinal study of aging (ELSA), Carr et al. (2016) report significant effects of job resources (decision authority, social support and recognition), but no systematic effect for physical and psychosocial job demands. Dal Bianco et al. (2015), using data from the Survey of Health, Aging and Retirement in Europe (SHARE), analysed measures of effort (stress, physical demands) and rewards (opportunity to develop new skills, social support, recognition, freedom in doing the job and job security). Poor quality of work was significantly related to both the intention to retire early and transitions to retirement. Career prospects, salary adequacy and perceived job security did not show significant correlations with the desire to retire as early as possible. Hence, we have to use measures of working conditions as controls when analysing the consequences of inheritances and inheritance expectations.

In order to control for such potential factors that may influence the decision to quit labour force participation or reduce working hours, we make use of variables available in SHARE (all from T1, that is before the receipt of an inheritance): Gender (1 = males, 0 = females), age (in years, continuous), living alone (1 = yes, 0 = no), education (in years, continuous), employment status (1 = private-sector or public-sector employee, 0 = self-employed), economic situation (financial respondent’s subjective assessment of the ability of the household to make ends meet) (1 = with great difficulty, 2 = with some difficulty, 3 = fairly easily, 4 = easily), working conditions (a summative index of ten questions on working conditions: satisfaction with the workplace, physical exertion that the job requires; stress and workload on the job; latitude on the job; opportunity to develop new skills; support received in difficult job situations; recognition received for the work; the extent to which the person’s pay is commensurate with his or her exertions and achievements; the likelihood of promotion; and job security—the higher the value of the index, the higher the perceived quality of the workplace), and health (self-rated health, 1 = excellent, 2 = very good, 3 = good, 4 = fair, and 5 = poor). Finally, we control for country-specific retirement regulations by including country dummies with Germany—the country with the largest number of inhabitants in our sample—as the reference group.

Results

Table 1 presents weighted descriptive statistics on the dependent, independent and control variables used in our models (total and by gender). The subsample of 3502 SHARE respondents aged 50–54 and active in the labour market at T1 consists about equally of men and women. At T2, 66% are still in the labour force. Of those who are still in paid employment at T2, 19% have reduced their working hours substantially. For inheritance expectations, at T1, the majority reports a small chance to receive an inheritance in the future. At T2, about 18% report that they have received an inheritance. Both the expectation to receive an inheritance and the receipt of an inheritance are not statistically different distributed for men and women.

Table 1.

Descriptives (percentages/means and standard deviations, weighted)

	Total (n = 3502)	Gender
		Male (n = 1681)	Female (n = 1821)	F/Χ2
Dependent variables
Labour force participation (Time 2)
In Labour force	66.13	66.98	65.35	0.046
Not in labour force	33.87	33.02	34.65
Reduction of working hours (Time 2)*
Yes	19.42	17.84	20.97	0.000
No	80.58	82.16	79.03
Independent variables of interest
Inheritance expectation (Time 1)
Small (< 25%)	54.18	55.61	52.79	0.512
Medium (25–74%)	22.83	21.97	23.67
Strong (> 74%)	22.99	22.42	23.54
Inheritance received (Time 2)
Yes	17.59	17.94	17.26	0.301
No	82.41	82.06	82.74
Control variables
Gender
Male	49.49
Female	50.51
Age (Time 1)	52.18 (1.32)	52.31 (1.30)	52.09 (1.34)	0.000
Living alone (Time 1)	19.01	16.44	21.54	0.000
Education (Time 1)	12.03 (4.05)	12.00 (3.93)	12.06 (4.17)	0.415
Self-employed (Time 1)	16.42	20.43	12.49	0.000
Self-rated health (Time 1)	2.51 (0.99)	2.47 (0.98)	2.55 (1.00)	0.063
Make ends meet (Time 1)	2.91 (0.91)	2.91 (0.91)	2.91 (0.92)	0.915
Workplace index (Time 1)	17.55 (4.38)	17.57 (4.14)	17.54 (4.61)	0.278

^*Based on those still active in the labour force at time 2 (2316 obs., male: 1126 obs., female: 1190 obs.)

Tables 2 and 3 show the results of our logistic regression models for participation in the labour force—for all respondents (Table 2) and male and female respondents separately (Table 4). Model 1 includes all the control variables, and models 2 and 3 add the expectation to receive an inheritance (measured at T1) and, respectively, the receipt of an inheritance (T2). It turns out that the expectation to receive an inheritance in the future has a significant positive effect on labour force participation at T2, controlling for gender, age, education, self-rated health, our index of working conditions, self-employment, the financial respondent’s ability of the household to make ends meet, and country of residence. However, participating in labour force at T2 is not statistically significant dependent on the receipt of an inheritance. Table 3 additionally shows that the expectation to receive an inheritance has a significant positive effect on labour force participation at T2 only for males, not for females. Receiving an inheritance has no significant impact on labour force participation, neither of men nor of women.9

Table 2.

Participation in the labour force at time 2 (logistic regressions, odds ratios and standard errors)

	Model 1	Model 2	Model 3
Gender: male	1.307*** (0.11)	1.326*** (0.11)	1.308*** (0.11)
Age (years)	0.651*** (0.02)	0.649*** (0.02)	0.651*** (0.02)
Living alone	1.236* (0.13)	1.201* (0.13)	1.237* (0.13)
Education (years)	1.043*** (0.01)	1.039*** (0.01)	1.043*** (0.01)
Self-employed	1.294** (0.15)	1.264** (0.15)	1.295** (0.15)
Self-rated health	1.271*** (0.06)	1.276*** (0.06)	1.271*** (0.07)
Make ends meet	0.873*** (0.04)	0.867*** (0.04)	0.873*** (0.04)
Workplace index	1.030*** (0.01)	1.030*** (0.01)	1.030*** (0.01)
Austria	0.070*** (0.02)	0.071*** (0.02)	0.070*** (0.02)
Belgium	0.244*** (0.05)	0.234*** (0.05)	0.244*** (0.05)
Czech Republic	0.325*** (0.08)	0.328*** (0.08)	0.325*** (0.08)
Denmark	1.086 (0.22)	1.055 (0.22)	1.076 (0.22)
France	0.299*** (0.06)	0.292*** (0.06)	0.298*** (0.06)
Greece	0.584** (0.14)	0.582** (0.14)	0.581** (0.14)
Israel	1.139 (0.29)	1.120 (0.29)	1.137 (0.29)
Italy	0.431*** (0.09)	0.431*** (0.09)	0.426*** (0.09)
The Netherlands	0.751 (0.18)	0.731 (0.17)	0.750 (0.18)
Poland	0.243*** (0.06)	0.246*** (0.06)	0.243*** (0.06)
Spain	0.625** (0.14)	0.646* (0.15)	0.616** (0.14)
Sweden	2.546*** (0.61)	2.390*** (0.58)	2.540*** (0.61)
Switzerland	1.068 (0.26)	1.024 (0.25)	1.070 (0.26)
Expectation: medium		0.954 (0.10)
Expectation: strong		1.292** (0.15)
Inheritance receipt			1.055 (0.12)
Constant	3.929e + 09*** (6.74e + 09)	4.881e + 09*** (8.51e + 09)	3.867e + 09*** (6.63e + 09)
Pseudo R2	0.156	0.158	0.156
N	3502	3427	3502

Reference categories: Germany, expectation: small (^*p < 0.1, ^**p < 0.05, ^***p < 0.01)

Table 3.

Participation in the labour force at time 2 (logistic regressions, odds ratios and standard errors)

	Male			Female
	Model 1	Model 2	Model 3	Model 1	Model 2	Model 3
Age	0.655*** (0.03)	0.651*** (0.03)	0.655*** (0.03)	0.636*** (0.03)	0.634*** (0.03)	0.636*** (0.03)
Living alone	0.854 (0.14)	0.819 (0.14)	0.854 (0.14)	1.749*** (0.26)	1.702*** (0.26)	1.751*** (0.26)
Education	1.064*** (0.02)	1.060*** (0.02)	1.063*** (0.02)	1.025 (0.02)	1.023 (0.02)	1.026 (0.02)
Self-employed	1.887*** (0.32)	1.837*** (0.32)	1.888*** (0.32)	0.855 (0.15)	0.846 (0.15)	0.855 (0.15)
Self-rated health	1.235*** (0.08)	1.232*** (0.08)	1.235*** (0.08)	1.310*** (0.08)	1.319*** (0.08)	1.310*** (0.08)
Make ends meet	0.796*** (0.06)	0.793*** (0.06)	0.796*** (0.06)	0.972 (0.07)	0.963 (0.07)	0.973 (0.07)
Workplace index	1.025* (0.02)	1.029* (0.02)	1.026* (0.02)	1.032** (0.01)	1.030** (0.01)	1.031** (0.01)
Austria	0.083*** (0.04)	0.082*** (0.04)	0.083*** (0.04)	0.057*** (0.02)	0.058*** (0.02)	0.057*** (0.02)
Belgium	0.249*** (0.07)	0.230*** (0.07)	0.249*** (0.07)	0.247*** (0.06)	0.244*** (0.06)	0.248*** (0.06)
Czech Republic	1.799 (0.93)	1.768 (0.91)	1.798 (0.93)	0.123*** (0.04)	0.115*** (0.04)	0.123*** (0.04)
Denmark	1.714 (0.56)	1.594 (0.53)	1.708 (0.56)	0.757 (0.20)	0.760 (0.20)	0.745 (0.20)
France	0.279*** (0.08)	0.266*** (0.08)	0.278*** (0.08)	0.333*** (0.09)	0.330*** (0.09)	0.331*** (0.09)
Greece	0.535* (0.19)	0.525* (0.19)	0.534* (0.19)	0.654 (0.21)	0.655 (0.21)	0.648 (0.21)
Israel	1.363 (0.57)	1.325 (0.55)	1.360 (0.57)	1.048 (0.34)	1.030 (0.34)	1.046 (0.34)
Italy	0.446** (0.14)	0.458** (0.15)	0.444** (0.14)	0.436*** (0.12)	0.427*** (0.12)	0.428*** (0.12)
The Netherlands	0.918 (0.32)	0.870 (0.31)	0.916 (0.32)	0.664 (0.21)	0.656 (0.21)	0.668 (0.21)
Poland	0.408** (0.15)	0.408** (0.15)	0.408** (0.15)	0.159*** (0.05)	0.160*** (0.05)	0.159*** (0.05)
Spain	0.868 (0.30)	0.979 (0.35)	0.863 (0.30)	0.470** (0.14)	0.447** (0.14)	0.459** (0.14)
Sweden	3.190*** (1.22)	2.907*** (1.12)	3.184*** (1.22)	2.157** (0.68)	2.055** (0.65)	2.152 (0.68)
Switzerland	1.361 (0.55)	1.288 (0.52)	1.361 (0.55)	0.876 (0.27)	0.835 (0.26)	0.879 (0.27)
Expectation: medium		0.907 (0.14)			1.004 (0.14)
Expectation: strong		1.431** (0.24)			1.173 (0.18)
Inheritance received			1.025 (0.17)			1.092 (0.17)
Constant	4.044e + 09*** (1.05e + 10)	5.618e + 09*** (1.49e + 10)	3.997e + 09*** (1.04e + 10)	1.197e + 10*** (2.83e + 10)	1.449e + 10*** (3.47e + 10)	1.174e + 10*** (2.77e + 10)
Pseudo R2	0.178	0.182	0.178	0.169	0.172	0.170
N	1681	1643	1681	1821	1784	1821

Reference categories: Germany, expectation: small (^*p < 0.1, ^**p < 0.05, ^***p < 0.01)

Table 4.

Reduction of working hours at time 2 (logistic regressions, odds ratios and standard errors)

	Model 1	Model 2	Model 3
Gender: male	0.770** (0.08)	0.778** (0.09)	0.770** (0.08)
Age (years)	1.171*** (0.05)	1.169*** (0.05)	1.172*** (0.05)
Living alone	0.817 (0.12)	0.795 (0.12)	0.819 (0.12)
Education (years)	0.984 (0.01)	0.982 (0.02)	0.984 (0.01)
Self-employed	2.377*** (0.32)	2.325*** (0.32)	2.376*** (0.32)
Self-rated health	0.908* (0.05)	0.908 (0.05)	0.907* (0.05)
Make ends meet	1.169** (0.08)	1.150** (0.08)	1.169** (0.08)
Workplace index	0.997 (0.01)	0.998 (0.01)	0.997 (0.01)
Austria	2.143* (0.87)	2.102* (0.87)	2.154* (0.88)
Belgium	1.255 (0.29)	1.279 (0.30)	1.256 (0.29)
Czech Republic	0.523* (0.20)	0.485* (0.19)	0.525* (0.20)
Denmark	0.636** (0.14)	0.645* (0.15)	0.631** (0.14)
France	0.900 (0.22)	0.833 (0.21)	0.897 (0.22)
Greece	0.381*** (0.13)	0.375*** (0.12)	0.380*** (0.12)
Israel	0.600* (0.18)	0.578* (0.17)	0.599* (0.18)
Italy	0.937 (0.25)	0.931 (0.25)	0.931 (0.25)
The Netherlands	0.952 (0.25)	0.977 (0.26)	0.951 (0.25)
Poland	1.068 (0.33)	1.058 (0.33)	1.070 (0.33)
Spain	0.488** (0.15)	0.500** (0.16)	0.483** (0.15)
Sweden	0.521*** (0.12)	0.536*** (0.13)	0.520** (0.12)
Switzerland	0.524** (0.15)	0.527** (0.15)	0.526** (0.15)
Expectation: medium		1.260* (0.17)
Expectation: strong		0.917 (0.13)
Inheritance receipt			1.048 (0.16)
Constant	0.000*** (0.00)	0.000*** (0.00)	0.000*** (0.00)
Pseudo-R2	0.148	0.148	0.148
N	2316	2274	2316

Reference categories: Germany, expectation: small (^*p < 0.1, ^**p < 0.05, ^***p < 0.01)

Tables 4 and 5 show the results of our logistic regression models for the reduction in working hours, only for all those respondents who have been active in the labour market both at T1 and T2 (Table 4) and separate for male and female respondents of this subsample, respectively (Table 5). No significant effects are found for substantially reducing working hours—neither in case of expecting an inheritance nor in case of receipt of an inheritance. A positive significant effect is found for females expecting an inheritance at T1—those with medium expectations are more likely to have reduced working hours by 20% or more at T2. But having received an inheritance has again no significant impact on working hours of men and women.

Table 5.

Reduction of working hours at time 2 (logistic regressions, odds ratios and standard errors)

	Male			Female
	Model 1	Model 2	Model 3	Model 1	Model 2	Model 3
Age	1.159** (0.07)	1.151** (0.07)	1.159** (0.07)	1.201*** (0.07)	1.205*** (0.07)	1.201*** (0.07)
Living alone	0.796 (0.19)	0.770 (0.18)	0.796 (0.19)	0.802 (0.15)	0.765 (0.15)	0.807 (0.15)
Education	1.009 (0.02)	1.013 (0.02)	1.009 (0.02)	0.967 (0.02)	0.960* (0.02)	0.967 (0.02)
Self-employed	2.768*** (0.51)	2.707*** (0.51)	2.768*** (0.51)	1.954*** (0.41)	1.868*** (0.40)	1.951*** (0.41)
Self-rated health	0.909 (0.08)	0.922 (0.08)	0.909 (0.08)	0.912 (0.07)	0.914 (0.08)	0.913 (0.07)
Make ends meet	1.180* (0.12)	1.158 (0.12)	1.180* (0.12)	1.159 (0.11)	1.137 (0.11)	1.161 (0.11)
Workplace index	0.975 (0.02)	0.973 (0.02)	0.975 (0.02)	1.016 (0.02)	1.020 (0.02)	1.016 (0.02)
Austria	3.414** (2.11)	3.159* (2.00)	3.412** (2.11)	1.606 (0.89)	1.621 (0.91)	1.626 (0.90)
Belgium	1.467 (0.58)	1.615 (0.64)	1.467 (0.58)	1.234 (0.37)	1.209 (0.36)	1.240 (0.37)
Czech Republic	0.419 (0.26)	0.407 (0.26)	0.419 (0.26)	0.700 (0.34)	0.676 (0.35)	0.708 (0.34)
Denmark	0.930 (0.35)	1.008 (0.38)	0.930 (0.35)	0.485** (0.15)	0.468** (0.14)	0.473** (0.14)
France	1.406 (0.56)	1.358 (0.55)	1.407 (0.56)	0.657 (0.21)	0.586 (0.19)	0.652 (0.21)
Greece	0.289** (0.16)	0.296** (0.17)	0.289** (0.16)	0.558 (0.23)	0.533 (0.22)	0.552 (0.23)
Israel	1.045 (0.50)	1.078 (0.52)	1.045 (0.50)	0.439** (0.17)	0.410** (0.17)	0.437** (0.17)
Italy	1.205 (0.52)	1.171 (0.50)	1.206 (0.52)	0.823 (0.29)	0.811 (0.29)	0.810 (0.29)
The Netherlands	1.169 (0.50)	1.269 (0.55)	1.170 (0.50)	0.917 (0.32)	0.940 (0.33)	0.920 (0.32)
Poland	1.047 (0.53)	1.013 (0.51)	1.047 (0.53)	1.241 (0.50)	1.225 (0.50)	1.251 (0.50)
Spain	0.650 (0.31)	0.688 (0.33)	0.650 (0.31)	0.423** (0.18)	0.416** (0.18)	0.415** (0.18)
Sweden	0.892 (0.36)	0.966 (0.39)	0.892 (0.36)	0.368*** (0.11)	0.360*** (0.11)	0.366*** (0.11)
Switzerland	0.348* (0.19)	0.362* (0.20)	0.348* (0.19)	0.636 (0.21)	0.606 (0.20)	0.640 (0.21)
Expectation: medium		0.921 (0.19)			1.633*** (0.29)
Expectation: strong		0.773 (0.16)			1.018 (0.20)
Inheritance received			0.996 (0.22)			1.111 (0.23)
Constant	0.000*** (0.00)	0.000*** (0.00)	0.000** (0.00)	0.000*** (0.00)	0.000*** (0.00)	0.000*** (0.00)
Pseudo-R2	0.1620	0.1583	0.1620	0.1513	0.1597	0.1515
N	1126	1105	1126	1190	1169	1190

Reference categories: Germany, expectation: small (^*p < 0.1, ^**p < 0.05, ^***p < 0.01)

Taken together, our result does not confirm the economic expectation of inheritances reducing economic activity—at least not for the age group 50–54, followed over 9 years of observation. To the contrary, the expectation of an inheritance increases the odds of staying in the labour force for males. As inheritances from parents and parents-in-law are received frequently in this age group, we may conclude that the consequences of inheritances—although these may take place in rare cases of exceptionally high amounts transferred—are not significant in the broader population.

Discussion

In this paper, we analyse if inheritance expectations as well as inheritances received discourage economic activity of men and women, both in terms of participation in the labour market and in terms of reducing the hours worked per week. For labour force participation, the empirical evidence shows a positive correlation between a strong expectation—meaning that the probability to inherit is rated between 75 and 100%—and participation in the labour market. That means, those who strongly expect to inherit in the near future are more likely to continue to do paid employment than the reference group of those who rate the expectation to inherit as small (0 to 24%). This finding contradicts the hypothesis H1a that expected inheritances reduce labour force participation, controlling for—among other crucial variables—self-rated health, working conditions, and country of residence. Contrary to what economic theory suggests, a strong inheritance expectation is positively correlated with participation in the labour force instead of causing labour market exits.

Furthermore, separate models for men and women show that a significant positive effect of strong inheritance expectation on labour force participation is observed for men, but not for women. This might be due to the selective group of women who are at all active in the labour market, as in most European countries the participation rates of women of this age group are way below the participation rates of men, often following still a traditional gender-specific division of labour pattern within couples. However, other than the strong expectation of a future inheritance, actual receipt of an inheritance is not correlated with labour force exits, neither for men nor for women, contradictory to hypothesis H1b.

Regarding the number of hours in paid work, inheritance expectations are correlated with a reduction of hours by 20% or more. However, it is the group rating a medium chance to inherit in the future that shows a positive effect on reduction of number of hours in paid work, in comparison with the reference group with a small expectation. For the group with a strong chance to inherit in the future, there is no correlation with a reduction of hours in paid work, contradicting the hypothesis H2a that inheritance expectations per se induce a substantial reduction in working hours. In separate models for men and women, medium expectations have a positive effect on the reduction of hours worked only for women (contradicting evidence on H2a for men). Again, actual receipt of an inheritance is not correlated with decrease in hours worked by 20% or more. (Empirical evidence is contradicting the hypothesis H2b that receipt of inheritances induces a substantial reduction in working hours.) However, as SHARE does not have continuous observation of working hours per week over time, we cannot rule out that individuals reduced working hours for a limited time in between T1 and T2.

Labour market decisions of older employees might be influenced by inheritance expectations, but the receipt of an inheritance does not impact labour market decisions such as to quit the job or at least to reduce the number of hours worked substantially. Therefore, although inheritances are nowadays, with long life expectancies among the testators, typically received when the heirs are close to retirement age, private transfers of wealth will not undermine active ageing policies. Explanations can be found in the inheritance phenomena itself, and this is the focus of our paper: Heirs in general receive rather small amounts, way too small to quit working for pay.10 And the transferred wealth might be illiquid, like real estate that has to be sold or rented out to gain additional income. Furthermore, because inheritances depend on the death of, e.g. a parent or other relatives, they hardly can be planned. Finally, if people are healthy and well-educated, and if they have good jobs, they tend to work longer on average—regardless of inheritance expectations or the receipt of inheritances. This is in line with findings that employees with higher private provisions for old age tend to work longer than those with lower provisions instead of entering into retirement earlier. And as Doorley and Pestel (2020) argue, inheritances may also have positive effects on labour in general because self-employed heirs invest into their businesses and create new jobs. Similar arguments came from Cox (2014), and already from Holtz-Eakin et al. (1993). Nonetheless, the majority of older employees does not receive an inheritance at all. So, the picture of a wave of bequests expected to come, that will flood everybody with large amounts inherited in the future, a picture which is highly loved by the media, might itself be a myth. At least, the phenomena of inheritances do not depend on the increasing numbers of elderly people per se, but on how (un-)equal wealth is distributed among the oldest age groups. As there is a tendency towards increasing wealth concentration in many European countries, especially in times of crisis, the expected wave of bequests will feel for the huge majority rather like a desiccated streamlet than a sea wave.

Taken together, we of course cannot reject the hypothesis that very large inheritances have an impact on economic activity, as Carnegie and others would predict. However, such effects are obviously not relevant in the population as a whole. We may instead follow a judgement of Cox and others, although from a slightly different context: “crowding out might loom larger in the minds of some economics than in the data” (Cox et al. 1999: p. 1).

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declarations

Conflict of interest

None.

Ethical approval

Not applicable.