International Journal of Manpower

Training older employees: what is effective?

Thomas Zwick
Article information:
To cite this document:
Thomas Zwick , (2015),"Training older employees: what is effective?", International Journal of
Manpower, Vol. 36 Iss 2 pp. 136 - 150
Permanent link to this document:
http://dx.doi.org/10.1108/IJM-09-2012-0138
Downloaded on: 03 May 2015, At: 12:11 (PT)
References: this document contains references to 35 other documents.
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

To copy this document: [email protected]

The fulltext of this document has been downloaded 47 times since 2015*
Users who downloaded this article also downloaded:
Wei-Tao Tai, (2006),"Effects of training framing, general self-efficacy and training motivation
on trainees' training effectiveness", Personnel Review, Vol. 35 Iss 1 pp. 51-65 http://
dx.doi.org/10.1108/00483480610636786
Susan Linz, Linda K Good, Michael Busch, (2015),"Promoting worker loyalty: an empirical
analysis", International Journal of Manpower, Vol. 36 Iss 2 pp. 169-191 http://dx.doi.org/10.1108/
IJM-06-2013-0129
T. Brad Harris, Wonjoon Chung, Holly M. Hutchins, Dan S. Chiaburu, (2014),"Do trainer style and
learner orientation predict training outcomes?", Journal of Workplace Learning, Vol. 26 Iss 5 pp.
331-344 http://dx.doi.org/10.1108/JWL-05-2013-0031

Access to this document was granted through an Emerald subscription provided by

Token:JournalAuthor:03563360-5CB0-460C-9C81-AB5BE35D4DF9:
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald
for Authors service information about how to choose which publication to write for and submission
guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as
well as providing an extensive range of online products and additional customer resources and
services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the
Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for
digital archive preservation.

*Related content and download information correct at time of

download.
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0143-7720.htm

IJM
36,2
Training older employees:
what is effective?
Thomas Zwick
136 Faculty of Business Management and Economics,
University of Würzburg, Würzburg, Germany
Received 14 September 2012
Revised 31 January 2013
5 May 2013
Abstract
17 October 2013 Purpose – Employees older than 55 years of age have a much lower share in training than other
Accepted 30 October 2013 employees. The purpose of this paper is to propose that one of the reasons for this phenomenon that
has not been taken into account so far is that their training is less effective.
Design/methodology/approach – This paper shows that training of older employees indeed is less
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

effective in the self-assessment of training participants. Training effectiveness is measured with

respect to key dimensions such as career development, earnings, adoption of new skills, flexibility or
job security. Besides age a broad range of explanatory variables is included as covariates in a large
linked employer-employee data set.
Findings – The paper finds that main reason for the differences in training effectiveness during the
life cycle is that firms do not take into account differences in training motivation. Older employees
get higher returns from informal and directly relevant training and from training contents that can be
mainly tackled by crystallised abilities. Training incidence in the more effective training forms is,
however, not higher for older employees. Given that other decisive variables on self-assessed
effectiveness such as training duration, financing and initiative are not sensitive to age, the wrong
allocation of training contents and training forms therefore is the critical explanation for the lower
effectiveness of training.
Originality/value – This paper therefore shows to human resource managers why old employees
rate training effectiveness lower and indicates what can be done in order to improve training
effectiveness of old employees. It uses a large and detailed data set entailing more than 6,000
employees from about 150 establishments.
Keywords Older workers, Training, Human resource management, Data analysis
Paper type Research paper

1. Introduction
Most papers on continuing training of older employees concentrate on their lower
training incidence (Taylor and Urwin, 2001; Addio et al., 2010). Obviously, it is a
problem in a greying economy when older employees get less than optimal training
because this might negatively affect their productivity and employability. The main
reasons for the lower training incidence of older employees proposed in the literature
are the shorter amortisation period of investments (Cunha et al., 2006), their lower
motivation to invest in training (Warr and Fay, 2001) and a perceived lower
adaptability of older employees (Warr, 1993).

JEL Classification — J10, J24, J28

The author thanked David Bills, Ruth Kanfer, Christian Stamov-Rossnagel, Ursula Staudinger
and Sven Voelpel for helpful comments. The author also thanked the Research Data Centre (FDZ)
International Journal of Manpower of the Federal Employment Agency at the Institute for Employment Research for the provision of
Vol. 36 No. 2, 2015
pp. 136-150 the data set and its useful documentation. The data basis of this publication is the “Berufliche
© Emerald Group Publishing Limited
0143-7720
Weiterbildung als Bestandteil Lebenslangen Lernens” panel data set (WeLL) (first two waves).
DOI 10.1108/IJM-09-2012-0138 Data access was via the scientific user file provided by FDZ.
Less attention has been given to the question whether and why training of older Training
employees might be less effective than training of younger employees. Some older
contributions show that training of older employees does not increase the relative
productivity of older employees (Göbel and Zwick, 2013) or that the performance of
employees
older workers in training programmes is lower (Ng and Feldman, 2008). Relatively well
researched are reasons for the lower effectiveness from the training supply side.
Personnel managers frequently seem to think that older employees are less able or 137
willing to learn (Warr and Birdi, 1998). Mainly caused by a lack of data, we do not
know very much about the training demand side – the opinion of older training
participants, however. This paper therefore uses recently published German linked
employer-employee data (WeLL). The data include detailed training information of
more than 6,000 employees from about 150 firms and allow us to analyse age patterns
in training characteristics and self-reported effectiveness for those employees who
participate in training with a focus on differences between age groups.
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

2. Background
Work motivation does not necessarily decline with age – motivation for some tasks
such as training may, however, be negatively affected by age. Warr and Fay (2001), for
example argue in a theoretical model that work motivation is influenced by incentives,
habits, comparisons with (younger) peers and social pressure. Older workers might be
less motivated to participate in training because (financial) incentives are lower than for
younger employees or comparable incentives are less attractive. Training might
demand time flexibility and be perceived as unwelcome break of routines that are more
entrenched for older employees (especially when they did not have training for a long
period of time). A comparison of training effectiveness with younger peers might be
unfavourable for older employees because the capacity to learn declines in some
skill dimensions. Warr and Fay (2001) also argue that the social pressure to participate
in training might be lower than for younger employees. In two meta-analyses, Ng and
Feldman (2008, 2012) indeed find that age is negatively related to performance in
training programmes, career development motivation, motivation to learn, training
motivation and learning self-efficacy. Age is, however, not related to training
participation. They conclude that their findings suggest that the stereotype of older
workers as being less willing to engage in further career development and training
activities is consistent with the cumulated research evidence, although the stereotype
may over-state the degree to which age negatively influences the willingness to engage
in further self-development.
Stamov-Roßnagel and Hertel (2010) stress that older employees mainly want to
match their resources to external demands. Younger people primarily strive for gains,
older people, however, more often focus on maintenance, harvesting of prior investment
returns and the prevention of losses. The authors argue that interest in tasks that
involve acquiring new skills, knowledge or career opportunities should decrease with
age. Motives such as autonomy, positive relationships with colleagues and supervisors
and self-realisation gain in importance during the life cycle. Callahan et al. (2003)
accordingly find in their meta-analysis that a clear motivation why training measures
are necessary with respect to relevant and topical work problems increases older
employee training performance. This might mean that training forms that support
the motivation of older employees such as training directly targeted at relevant
problems at the work place or an improvement of the work relationships are more
attractive and more efficient for older employees. A prominent and well-researched
IJM topic is information and communication technology (ICT) skills and ICT training for
36,2 older workers. Several papers show that older employees use less ICT and that this
may have a negative impact on their performance. Nevertheless, ICT training (that
usually implies abstract and new information that has to be processed by fluid
intelligence, see next part) has no measurable impact on ICT skills of older employees
(Bertschek and Meyer, 2009).
138 Kanfer and Ackerman (2004) argue that the motivation for certain tasks changes
with age on the basis of the distinction between crystallised[1] and fluid[2] intelligence.
They stress that motivation for training declines with age because a reduction in fluid
cognitive intelligence slows learning and the timeframe for the development of
crystallised expertise in which performance may be sustained with less effort
decreases. Callahan et al. (2003) also argue (but do not find) that the lecture method
that places a relatively heavy demand on cognitive ability is less effective than more
active learning methods. Efficacy in training of skills that do not place heavy demands
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

on fluid intellectual abilities such as conflict management might be higher for older
employees, however.
Older employees in addition have a higher interest in working time flexibility than
younger employees because they frequently have to care for their elderly parents or
need time for regeneration (OECD, 2006). In contrast to childcare, the time needs and
the time dimension of informal elderly care is uncertain. The (mainly older and female)
family members therefore only can combine informal care for elderly family members
with employment when they have sufficient working time flexibility (Ettner, 1995).
Informal and unplanned learning should therefore play a greater role for older
employees than formal and “normal” learning (Weiss, 2009). There are very few
empirical analyses on differences in training characteristics and effectiveness during
the life cycle. We also do not know whether employers try to adopt the training design
and methods that suit the preferences of older employees (Armstrong-Stassen and
Templer, 2005). Baethge and Baethge-Kinsky (2004) mention that self-assessed training
competence, self-managing disposition and competence development activities do not
differ between age groups. Only the anticipation of training needs declines with age in
their study. Warr and Birdi (1998) stress that voluntary learning activities and
training motivation decline with age. This goes hand in hand with the assessment of
personnel managers who say that the strongest disadvantage of older employees is
their low trainability and interest in training (Boockmann and Zwick, 2004; Loretto and
White, 2006). A more negative self-assessment and perception of managers of older
employees both may create a reduced interest in training. Peers do not expect regular
training participation of older workers and therefore social pressure is lower (Warr and
Birdi, 1998). Beicht et al. (2006) show that there are hardly any differences in the kind
of training and the financing of training, people attend during their life cycle in
Germany. The older the training participants the more sceptical are the assessments of
the benefits of training. In contrast to the present study, this paper includes employees
and people outside the labour force, does not take into account differences between
these groups, however.
The present paper concentrates on differences between training characteristics and
training efficiency during the life cycle. On the basis of the theoretical and empirical
evidence discussed above, the following four main hypotheses are proposed:

H1. Training characteristics (contents, financing, extent) do not change during

the life cycle.
It is a problem that employers do not offer age-specific training because training Training
motivation and efficiency changes over the life cycle. More specifically: older
H2. Older employees assess time flexible training forms to be more effective in employees
comparison to rigid courses.
H3. Older employees assess training contents they can immediately apply to topical
problems on the workplace to be more effective than training contents that 139
are more abstract.
H4. Older employees assess training contents in which they do not have a learning
disadvantage (such as knowledge mainly acquired by crystallised skills) to be
more effective than training contents that are easier to cope with for younger
employees (such as knowledge mainly acquired by fluid intelligence).
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

3. Data and estimation strategy

The “Berufliche Weiterbildung als Bestandteil Lebenslangen Lernens – Continuing
Training as Part of Lifelong Learning” (WeLL) data set combines individual answers
on training behaviour with socio-demographic information and some establishment
characteristics. So far, there are two waves available from the years 2007 and 2008. The
first wave entails answers by 6,404 employees in 149 enterprises[3]. The second wave
comprises repeated interviews with 4,259 employees from the first wave and interviews
with 636 newly hired employees from Autumn 2008 in the same enterprises. The
sample is not representative for the workforce but tailored towards analysing
intra-firm processes with respect to continuing training. It therefore allows the analysis
of differences between employee groups (Bender et al., 2009).
The WeLL data set covers a remarkable breadth of training forms and training
contents in the employer and employee part of the data set. The aim of these extensive
lists was to entail formal and informal training activities and make the answers
comparable to previous training questionnaires (Beicht et al., 2006; Bender et al., 2009).
This paper mainly looks at determinants of individual training participation and
training characteristics with a focus on employee age. These items are more or less time
invariant within less than one year. It therefore does not make sense to include the
panel dimension of the data set. In order to avoid biased estimations by including some
employees once and other employees twice, all employees from the second wave and
only those 4,084 employees from the first wave are included who do not have an
observation in the second wave. The final sample therefore consists of 6,349 employees.
In addition, some employees report more than one training episode. In order to avoid
that those employees with more than one observation (who probably differ from
the other training participants) dominate the results, only one training episode per
employee (the first one reported) is taken.
In order to guarantee anonymity, the data do not entail the precise age of the
employees but only report whether employees have been born in the year 1951 or
before, between 1952 and 1961, between 1962 and 1971 and in 1972 or after. In the
year 2007, the employees in the oldest age group therefore were at least 56 years old
and in 2008, they were at least 57 years old. Unfortunately, most establishment
information is reported only in aggregated form for anonymity reasons. We therefore
only know whether an establishment is in the size bracket 100-199, 200-499 or 500-1,999
employees. In addition, a division between manufacturing and services firms is
possible.
IJM The complete attribution of training differences to age in bivariate descriptive
36,2 statistics might create artefacts (Gallenberger, 2002). Lower training participation of
older employees, for example may be a consequence of differences between older and
younger employees with respect to qualification levels, gender or other characteristics
that are relevant for training. Tippelt et al. (2009) show for example that female, lower
educated, older or sick employees participate significantly less in continuing training.
140 We also know that employment is a crucial pre-requisite for training – the relative
training participation of the unemployed compared with the employed increases
with age (Alferoff, 1999; Von Rosenbladt and Bilger, 2008). Therefore selection into
training differs by employment status for different age groups. Even more problematic
for the measurement of differences in training effectiveness between age groups
are third factors that influence self-assessed training effectiveness and differ by age.
Adding controls for characteristics that are correlated with training assessment
and age therefore reduces the measurement bias of the correlations between age and
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

training effectiveness.
We therefore need a multivariate approach that controls as many of those third
factors as possible. On the basis of the WeLL data set, it is possible to integrate the
most decisive determinants for training participation of (older) employees (Bannwitz,
2008): gender, qualification, professional position and motivation on the individual
level as well as size and sector on the establishment level. In addition, tenure is added in
order not to confound the age and tenure effect on training (Göbel and Zwick, 2013).
Finally, two individual characteristics that are closely related to training – and easily
might be confounded with age – are included: self-assessed health (Wegge et al., 2008;
Friedrichs et al., 2011) and the prospect to leave the labour force within the next year
(Boockmann et al., 2012), see Table AI for detailed descriptions of the explanatory
variables. We also compare simple bivariate statistics with the multivariate regressions
in order to show the additional effect of adding the covariates.
Even when several decisive characteristics that vary with age and influence training
participation are controlled for in this analysis, we have to assume that – especially
given the lower employment participation of employees older than 60 years –
unobservable characteristics such as learning motivation, work ethics or general ability
might systematically differ by age and training participation. As older employees are
on average a more positively selected group than younger employees, selectivity based
on unobservables should, however, positively bias the correlations between age and
training effectiveness. The low effectiveness of training measures for older employees
calculated therefore might even be higher than in reality.
Although the dependent variables in the regressions are dummy variables, ordinary
least square regressions are performed because this is the most robust estimation
technique and the least demanding preconditions are necessary. Binary response
models such as Probit/Logit make strong (and convenient) modelling assumptions on
the behaviour of the error term in the underlying structural model. These assumptions
are not testable, however, and if the assumptions are wrong, presumably the bias
can also be large. The two main drawbacks of OLS with respect to Probit/Logit are that
the coefficients are biased when predicted values lie outside the interval [0,1] and
that they are heteroskedastic. The last problem is solved in this paper by using a
heteroskedasticity-consistent robust standard error estimator. The first problem does
not seem to be large. In our case, the maximal share of predicted observations is below
3 per cent, in most regressions all predicted values are inside the interval [0,1]. The
potential estimation biases incurred by predicted values outside the zero to one interval
therefore should be small. All regressions have been repeated also with Probit and Training
Logit regressions, however, as a robustness check. The influence of the age groups older
was the same in all regressions and therefore the additional evidence is not reported[4].
Our main variable of interest is self-assessed efficiency of training. The concrete
employees
question is: “Which effect did the courses, seminars or workshops you participated in
since [date about one year earlier] on the following aspects of your work from today’s
point of view?” Then follows a list of six efficiency dimensions: Professional 141
productivity, adaptation to new professional challenges, promotion chances, earnings,
job security, professional new orientation (employer change, new profession/
occupation, self-employment).

4. Training differences between age groups

Table I shows that the training extent, financing source and the party that took the
initiative for training are remarkably similar for different age groups. No indicator of
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

training intensity differs significantly between the oldest age group and the younger
training participants (see last column of Table I). These findings are analogous to
those reported for Germany by Beicht et al. (2006) and for the UK by Taylor and Urwin
(2001)[5] and confirm H1.
The descriptive analysis in Table I also reveals that there are highly significant
bi-variate differences between the self-assessed effects of training between the oldest
group of employees (these employees are at least 56 years old) and the younger
employees in all training effect dimensions. Training participants have been asked,
whether they experienced an increase in productivity, earnings, easier adapted to
a new job, have been promoted, increased job security or gained a new professional
orientation by their last training spell. These dimensions are considered to be the most
prominent training effects (compare, e.g. Dieckhoff, 2007).
The next estimation step shows that the differences between the age groups with
respect to the self-assessed training effectiveness are still significant when a broad
range of individual and employer characteristics are taken into account, see Table II.
This corresponds to the finding by Ng and Feldman (2008) that age is negatively
correlated with performance in training programmes measured by post-training
scores. The differences between the oldest age group and younger training participants
in addition are in the same size range with and without controlling for individual
and employer characteristics. Self-assessed training effectiveness is higher for
employees with higher tenure and higher qualification, employees in good health and
employees who do not intend to quit the labour market in the near future. Training in
larger enterprises also is more effective.
Obviously, the lower effectiveness of training of older employees is not a
consequence of lower or less costly training inputs, see Table I[6]. Another reason for
the lower training effectiveness might be that the oldest training participants are
generally more sceptical with respect to their work. Indeed, the oldest employees in the
sample are somewhat less satisfied with training, their overall satisfaction with
work is, however, not significantly lower than the satisfaction of their younger peers
(compare Table I).
Therefore, neither the training intensity, overall training satisfaction nor the general
attitude towards work can explain the differences in self-assessed training
effectiveness between training participants older than 56 years and younger training
participants. Based on H2-H4, differences in the assessment of training forms and
training contents may be a key for the explanation of this phenomenon. The WeLL
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

IJM
36,2

142

Table I.
Descriptive

and age groups

differences between
training dimensions
Birth year 1951 Birth years Birth years Birth year 1972
Entire sample or older 1952-1961 1962-1971 or younger Significance

Training duration in hours 44.57 41.71 47.77 43.30 42.36

(104.32) (87.87) (115.28) (101.07) (97.79)
Training duration in months 2.32 2.29 2.31 2.34 2.34
(1.89) (1.87) (1.87) (1.91) (1.93)
Number of trainings during last 12 months 1.77 1.77 1.74 1.78 1.80
(1.12) (1.39) (1.13) (1.26) (1.25)
Costs borne by participant (yes/no) 0.16 0.15 0.15 0.16 0.17
Initiative by participant (yes/no) 0.41 0.41 0.42 0.41 0.42
Initiative by employer (yes/no) 0.23 0.23 0.23 0.23 0.22
Training necessary by law (yes/no) 0.17 0.16 0.18 0.18 0.17
Training satisfaction 5.74 5.58 5.74 5.73 5.92 *
(2.64) (2.77) (2.69) (2.60) (2.50)
Work satisfaction 7.39 7.48 7.38 7.49 7.33
(1.69) (1.71) (1.71) (1.67) (1.71)
Self-assessed training effects
Higher productivity 0.38 0.32 0.38 0.38 0.41 ***
Adoption 0.34 0.27 0.34 0.40 0.39 ***
Promotion 0.08 0.04 0.06 0.08 0.13 ***
Higher earnings 0.03 0.02 0.03 0.03 0.06 ***
Job security 0.23 0.21 0.23 0.22 0.26 *
New orientation 0.09 0.05 0.06 0.10 0.16 ***
Notes: Standard deviation in brackets for non-dummy variables. Answer options for training and work satisfaction: 0 completely unsatisfied, 10 completely
satisfied, dummies coded as no: 0 and yes: 1. Difference between training participants who have been born in 1951 or older and all younger employees – all
significant differences are negative for the oldest group of training participants, rows without stars have significance levels above 10 per cent;
***,**,*significant at o 0.01, o0.05 and o0.1 levels, respectively
Higher Pro- Higher Job New
Training
productivity Adoption motion earnings security orientation older
employees
Birth years 1951 or older −0.04* −0.06** −0.04*** −0.02*** −0.02 −0.04***
Intermediate secondary
schooling 0.11*** 0.08*** 0.02*** 0.01* 0.05*** 0.02***
Higher secondary
schooling 0.18*** 0.18*** 0.05*** 0.01 0.02 0.08*** 143
Female 0.01 0.01 −0.02** −0.01*** −0.03*** 0.00
Tenure 2-5 years 0.06** 0.03 0.00 0.01 0.04* 0.01
Tenure 6-15 years 0.07*** 0.05*** 0.01 0.00 0.04** 0.01
Tenure more than
15 years 0.05*** 0.04*** 0.01* 0.00 0.02 0.00
Good health 0.06*** 0.04** 0.02*** 0.01** 0.04*** 0.01
High probability to quit
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

working −0.08** −0.07** −0.02 −0.01 −0.06** −0.00

Eastern Germany 0.01 0.00 −0.01 −0.00 0.01 −0.01
200-499 employees 0.03 0.02 0.01 0.01 0.01 −0.01
500-1,999 employees 0.06** 0.05** 0.03*** 0.01 0.01 −0.00
Services sector 0.05*** 0.05*** 0.00 −0.01 0.02 0.00
R2 0.14 0.13 0.11 0.11 0.10 0.12 Table II.
Notes: Heteroskedasticity-consistent robust OLS regressions, clustering adjusted for 149 enterprises; Determinants of
number of observations: 5,590; reference categories: secondary general school, birth year 1952 or older, self-assessed
employer with less than 200 and more than 50 employers, tenure less than two years training effects

questionnaire covers a broad range of training topics and training forms. In the next
step, the self-assessed training efficiency therefore is compared between the oldest
training group and the younger training participants for selected training topics and
forms. In order to make sure that relative differences between age groups in the
effectiveness of training forms and contents are not related to absolute differences in
the effectiveness, in Tables III and IV also absolute training effectiveness is reported.
Self-assessed absolute effectiveness is quite comparable between training forms and
contents ( probably with the exception of training-on-the job that is rate slightly less
effective than the other training forms).
This paper proposes a new explanation why training effectiveness of older
employees is lower: Employers do not take the changes in training preferences by age
into account. According to our hypotheses, older employees should assess the
effectiveness of more abstract and formal training forms (e.g. formal seminars and
training circles[7]) worse than the effectiveness of more applied and training forms
that are directly relevant for daily work problems (e.g. training on the job[8] or
self-motivated training[9]). Indeed, the effectiveness of seminars and training circles is
assessed significantly more negative in a descriptive comparison between the oldest
and younger training participants[10]. The training effectiveness of training on the
job and self-managed learning is, however, assessed the same by older and younger
employees. These results are also obtained in a multi-variate approach (see Table III),
and therefore the bivariate results are not reported here. An additional interpretation of
the relatively high effectiveness of self-motivated training for older employees is the
higher time flexibility of this training form (Callahan et al., 2003).
In accordance to the arguments proposed by Kanfer and Ackerman (2004), training
effectiveness of contents that mainly demand fluid cognitive intelligence (such as
IJM Higher Higher Job New
36,2 Effects of training productivity Adoption Promotion earnings security orientation

Seminar
Absolute effectiveness 0.53 0.47 0.12 0.05 0.27 0.14
Birth year 1951 or older −0.08* −0.10*** −0.08*** −0.09 *** −0.09*** −0.12***
144 Training circle
Absolute effectiveness 0.53 0.49 0.13 0.04 0.30 0.14
Birth years 1951 or older −0.07* −0.10** −0.08*** −0.04*** −0.06* 0.08***
Training on the job
Absolute effectiveness 0.44 0.40 0.10 0.04 0.29 0.10
Birth years 1951 or older −0.02 −0.04 −0.05* −0.02 0.00 0.01
Self-managed learning
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

Absolute effectiveness 0.54 0.50 0.13 0.05 0.31 0.13

Table III. Birth years 1951 or older −0.01 −0.08* −0.06* −0.01 −0.03 −0.03
Self-assessed Notes: First rows: share of answers indicating a positive effectiveness. Second rows:
effectiveness Heteroskedasticity-consistent robust OLS regressions explaining effectiveness, Number of
of different observations (enterprises): seminar: 1,401 (142), training on the job: 2,104 (146), self-managed learning:
training forms 950 (134); same covariates as in Table II

Higher Higher Job New

Effects of training productivity Adoption Promotion earnings security orientation

Information and communication technology

Absolute effectiveness 0.37 0.33 0.07 0.04 0.22 0.07
Birth year 1951 or older −0.03* −0.09** −0.04** −0.03** −0.02* −0.02*
Technical training
Absolute effectiveness 0.39 0.35 0.08 0.03 0.24 0.10
Birth year 1951 or older −0.08* −0.09** −0.07*** −0.02* −0.03* −0.05**
Communication training
Absolute effectiveness 0.36 0.36 0.08 0.02 0.22 0.08
Birth year 1951 or older 0.09 −0.01 −0.06* 0.02 0.00 −0.03
Management training
Absolute effectiveness 0.38 0.36 0.08 0.03 0.24 0.09
Table IV. Birth year 1951 or older −0.02 −0.03 0.00 0.01 −0.01 −0.04
Self-assessed Notes: First rows: share of answers indicating a positive effectiveness. Second rows:
effectiveness of Heteroskedasticity-consistent robust OLS regressions explaining effectiveness, number of observations
different training (enterprises): information and communication technology: 937 (140), technical contents: 1,009 (143),
contents communication training: 554 (126), management training: 423 (127); same covariates as in Table II

ICT[11] or technical contents) is negatively assessed by older employees. Ng and

Feldman (2008) also suggest that the lower performance of older employees in training
programmes mainly is driven by technical training. Training in communication and
management skills that mainly demand crystallised cognitive intelligence[12] seems to
have a higher effectiveness, compare Table IV. These findings support the fourth
hypothesis that older employees are not keen on comparing themselves with younger
training participants in ability areas in which many of them are disadvantaged. An Training
additional interpretation of the higher self-assessed effectiveness of communication older
and management training is that both training contents can be easily applied to
concrete and topical workplace problems (H3) and older employees might use their
employees
relatively high experience for easy knowledge acquisition. Finally, these topics are
aimed at improving the quality of the workplace instead of opening new career chances
or improving the workplace security. 145
In contrast to the literature (self-assessed), training efficiency for older employees
does not seem to be a consequence of a genuinely lower ability and willingness to learn
(Warr and Fay, 2001) or of differences in the perception by personnel managers
(Boockmann and Zwick, 2004). Overall training satisfaction indeed is the same for old
and young employees no matter for which kind of training content or training form
(not shown here). The main problem instead is that older employees do not get those
training contents and forms they assess to be as equally efficient as their younger
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

peers. Table V accordingly shows that the employers do not offer training forms
and contents associated with higher training efficiency for older employees. There are
no age-specific differences in the participation in training forms and contents. Actually,
training on the job which is assessed relatively favourably by older employees is
offered significantly less to this group of training participants. This evidence
corresponds to earlier findings that only around one-fourth of those enterprises that
include older employees in training measures also offers specific training measures for
them (Göbel and Zwick, 2013).
In a series of robustness checks, the differences in training effectiveness by age
group are also split by gender, health situation and the intention to quit employment.
These sample splits demonstrate that the age effects are different for females,
employees with bad health and employees who intend to leave the labour market.
These differences might be important because starting with age 60, the selectivity of
employees strongly increases in Germany (Boockmann et al., 2012). Mainly the
healthier employees tend to remain in the labour market (Friedrichs et al., 2011) and the

Birth year Birth Birth Birth year

Entire 1951 or years years 1972 or
sample older 1952-1961 1962-1971 younger Significance

Training forms
Seminar 0.25 0.25 0.25 0.26 0.24
Training circle 0.17 0.16 0.17 0.17 0.16
Training on the job 0.38 0.31 0.36 0.40 0.44 ***
Self-motivated training 0.17 0.17 0.18 0.16 0.17
Training contents
Information and
communication technology 0.17 0.17 0.17 0.17 0.16
Technical training 0.18 0.19 0.18 0.18 0.17 Table V.
Communication training 0.10 0.11 0.09 0.08 0.11 Descriptive
Management training 0.08 0.08 0.08 0.08 0.07 differences in
Notes: Significance column: difference between training participants who are 56 years or older and all training forms and
younger employees – all significant differences are negative; rows without stars have significance contents between
levels above 10 per cent; ***significant at o0.01 level age groups
IJM decision of older women to remain in the labour market may differ from that of older
36,2 men and younger women (Altonji and Blank, 1999; Zwick, 2012).
The youngest and oldest employee groups more often intend to quit employment
than the prime-age group[13]. Interestingly, there are no age effects in training
effectiveness for those who intend to quit employment – for this group, the age dummy
in an estimation analogously to Table II is insignificant for all self-assessed training
146 efficiency dimensions. The age effects therefore completely stem from those employees
who intend to stay in the labour market for more than one year. It is not surprising that
the share of employees who state that they are healthy declines from 85 per cent in the
youngest group to 69 per cent in the oldest group. The reductions in training
effectiveness by age are somewhat smaller for those who state that they are sick, but
they do not disappear completely and remain significant for most efficiency dimension.
Finally, the correlation between age and training effectiveness is somewhat
stronger for males than for women. Again, the age effects are not completely
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

absent for women nevertheless and most coefficients remain significant (these results
are not shown here).

5. Conclusions
This paper shows that old training participants assess the effectiveness of a series
of relevant training goals significantly worse than younger training participants.
This difference is stronger for male and healthy training participants and only applies
for those who intend to stay in the labour market. It does not seem to be the result
of lower training intensity of older training participants, a general decline in work
satisfaction, or a lower general effectiveness of training for older workers in dimensions
such as promotions, job security or earnings increases. Instead, this paper argues that
lower training efficiency mainly is caused by offering older employees training
contents and training forms that are not in accordance with their training preferences.
The theoretical literature on training motivation over the life cycle stresses that
older employees prefer time flexible training forms that deliver practical and
immediately relevant knowledge and training contents that mainly can be mastered by
crystallised intelligence. Indeed, this paper shows that the effectiveness of
communication and management training is higher for older employees than
training featuring abstract technical contents or information technology. Self-induced
training and training-on-the-job accordingly also is more effective for older employees
than participation in formal seminars and training circles. Unfortunately, firms
do not seem to take these preferences of older training participants into account. The
oldest group of training participants gets more or less the same training forms and
training contents as younger training participants.
The management implication of this paper is that the large gap between employers
that offer training for older employees and those that offer specific training measures
for older employees (Göbel and Zwick, 2013) should decrease. Management has to take
into consideration the specific training needs and interests of older employees in order
to increase training efficiency and the motivation to participate in training.
This paper only reports self-assessed answers of training participants. Therefore,
assessments of ( personnel) managers on training effectiveness would be valuable in
order to get the complete picture on differences in training efficiency over the life
cycle. In addition, differences in tasks and their complexity might have an influence on
the effectiveness of training during the life cycle (Wegge et al., 2008). Unfortunately,
we cannot observe tasks in this data set. Also only few establishment characteristics
can be included here. Probably the inclusion of further establishment characteristics Training
that potentially are correlated with training effectiveness and the age of the training older
participant (such as industrial relations, the qualification structure of the establishment
or profitability) provide additional explanations for the lower training effectiveness
employees
for older employees. In addition, changes in selectivity of employees and training
participation cannot be taken into account in this cross-section analysis. We might
assume that more motivated and more productive employees “survive” in the labour 147
market in the highest age group. This might imply that the true correlation between
age and training effectiveness is even under-estimated here.

Notes
1. Crystallised intelligence is general knowledge, extent of vocabulary and verbal
comprehension.
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

2. Fluid intelligence is associated with working memory, abstract reasoning, attention and
processing of novel information.
3. The individual employee telephone interviews have been conducted between October 2007
and January 2008.
4. These estimations are available on request.
5. Warr (1993) reports a reduction in time spent in training with age, however.
6. This result is also obtained when we include additional individual and employer
characteristics for the explanation of training intensity (not shown here).
7. Training circles are usually meetings at fixed dates at which employees get training inputs
or discuss ways to improve the productivity of their working group. Examples are quality
or team circle, workshop or temporary working group.
8. Examples for this training form are organised introductory training or teaching by
colleagues or managers.
9. Examples for this training form are computer-aided learning, training on the intranet or
distance learning.
10. This result also is in accordance to earlier findings by Beicht et al. (2006).
11. This result is also in accordance to the findings by Bertschek and Meyer (2009).
12. The only exception is the promotion effectiveness of communication training.
13. The shares are seven, respectively, 9 per cent – the middle age groups have a share of
around one per cent.

References
Addio, A.D′., Keese, M. and Whitehouse, E. (2010), “Population ageing and labour markets”,
Oxford Review of Economic Policy, Vol. 26 No. 4, pp. 613-635.
Alferoff, C. (1999), “Older workers: flexibility, trust and the training relationship”, Education and
Ageing, Vol. 14 No. 1, pp. 51-60.
Altonji, J. and Blank, R. (1999), “Race and gender in the labor market”, in Ashenfelter, O. and
Card, D. (Eds), Handbook of Labor Economics, Vol. 3, Elsevier, Amsterdam,
pp. 3143-3259.
Armstrong-Stassen, M. and Templer, A. (2005), “Adapting training for older employees”, Journal
of Management Development, Vol. 24 No. 1, pp. 57-67.
IJM Baethge, M. and Baethge-Kinsky, V. (2004), Der Ungleiche Kampf Um Das Lebenslange LERNEN
(The Unequal Fight For Lifelong Training), Waxmann, Münster.
36,2
Bannwitz, J. (2008), Ältere Beschäftigte und betriebliche Weiterbildung. Ergebnisse des CVTS 3
(Older Employees and Continuing Training – Results of CVTS 3), Wissenschaftliches
Diskussionspapier des BIBB 104, Bonn.
Beicht, U., Krekel, E. and Walden, G. (2006), Berufliche Weiterbildung: Welche Kosten und welchen
148 Nutzen haben die Teilnehmenden? (Continuing Training: Which Costs and Benefits Have
Participants?), Bertelsmann, Bielefeld.
Bender, S., Fertig, M, Görlitz, K., Huber, M. and Schmucker, A. (2009), “WeLL – unique linked
employer-employee data on further training in Germany”, Schmollers Jahrbuch, Vol. 129
No. 4, pp. 637-643.
Bertschek, I. and Meyer, J. (2009), “Do older workers lower IT-enabled productivity? Firm-level
evidence from Germany”, Jahrbücher für Nationalökonomie und Statistik, Vol. 229 Nos 2/3,
pp. 327-342.
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

Boockmann, B. and Zwick, T. (2004), “Betriebliche determinanten der Beschäftigung älterer

Arbeitnehmer”, Zeitschrift für ArbeitsmarktForschung, Vol. 37 No. 1, pp. 53-63.
Boockmann, B., Fries, J. and Göbel, C. (2012), “Specific measures for older employees and late
career employment”, ZEW Discussion Paper 12-059, Mannheim
Callahan, J., Kiker, D. and Cross, T. (2003), “Does method matter? A meta-analysis of the effects of
training method on older learner training performance”, Journal of Management, Vol. 29
No. 5, pp. 663-380.
Cunha, F., Heckman, J., Lochner, L. and Masterov, D. (2006), “Interpreting the evidence on life
cycle skill formation”, in Hanushek, E. and Welch, F. (Eds), Handbook of the Economics of
Education, North Holland, pp. 697-812.
Dieckhoff, M. (2007), “Does it work? The effect of continuing training on labour market outcomes:
a comparative study on Germany, Denmark, and the United Kingdom”, European
Sociological Review, Vol. 23 No. 3, pp. 295-308.
Ettner, S. (1995), “The impact of ‘parent care’ on female labor supply decisions”, Demography,
Vol. 32 No. 1, pp. 63-80.
Friedrichs, M., Jungmann, F., Liebermann, S., Lück, P., Schmidt, K., Wegge, J. and Wolters, J.
(2011), “Einschätzung der Erwerbsbevölkerung zum Stellenwert der Arbeit, zum
Gesundheitsverhalten, zum subjektiven Gesundheitszustand und zu der Zusammenarbeit
in altersgemischten Teams (Assessment of the workforce on the importance of work,
health behaviour, subjective health assessment and co-operation in age-mixed teams)”,
Report 21, iga, Berlin.
Gallenberger, W. (2002), Weiterbildungsabstinenz älterer Beschäftigter in einer alternden
Erwerbsbevölkerung? (No Training for Older Employees in an Ageing Workforce?),
Leske + Budrich, Opladen.
Göbel, C. and Zwick, T. (2013), “Are personnel measures effective in increasing productivity of
old workers?”, Labour Economics, Vol. 22, pp. 80-93.
Kanfer, R. and Ackerman, P. (2004), “Aging, adult development and work motivation”, Academy
of Management Review, Vol. 29 No. 3, pp. 440-458.
Loretto, W. and White, P. (2006), “Employers’ attitudes, practices and policies towards older
workers”, Human Resource Management Journal, Vol. 16 No. 3, pp. 313-330.
Ng, T. and Feldman, D. (2008), “The relationship of age to ten dimensions of job performance”,
Journal of Applied Psychology, Vol. 93 No. 2, pp. 392-423.
Ng, T. and Feldman, D. (2012), “Evaluating six common stereotypes about older workers with
meta-analytical data”, Personnel Psychology, Vol. 65 No. 4, pp. 821-858.
OECD (2006), Ageing and Employment Policies – Live Longer, Work Longer, OECD, Paris. Training
Stamov-Roßnagel, C. and Hertel, G. (2010), “Older worker’s motivation: against the myth of older
general decline”, Management Decision, Vol. 48 No. 6, pp. 894-906. employees
Taylor, P. and Urwin, P. (2001), “Age and participation in vocation education and training”,
Work, Employment and Society, Vol. 15 No. 4, pp. 763-779.
Tippelt, R., Schmidt, B., Schnurr, S., Sinner, S. and Theisen, C. (2009), Bildung Älterer – Chancen
im Demographischen Wandel (Education of Older People – Chances in Demographic 149
Change), Bertelsmann, Bielefeld.
von Rosenbladt, B. and Bilger, F. (2008), Weiterbildungsbeteiligung in Deutschland – Eckdaten zum
BSW-AES 2007 (Training participation in Germany – Key Data on BSW-AES 2007),
BMBF, Berlin.
Warr, P. (1993), “Training for older managers”, Human Resource Management Journal, Vol. 4
No. 2, pp. 22-38.
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

Warr, P. and Birdi, K. (1998), “Employee age and voluntary development activity”, International
Journal of Training and Development, Vol. 2 No. 3, pp. 190-204.
Warr, P. and Fay, D. (2001), “Short report: age and personal initiative at work”, European Journal
of Work and Organizational Psychology, Vol. 10 No. 3, pp. 343-353.
Wegge, J., Roth, C., Neubach, B., Schmidt, K. and Kanfer, R. (2008), “Age and gender diversity as
determinants of performance and health in a public organization: the role of task
complexity and group size”, Journal of Applied Psychology, Vol. 93 No. 2, pp. 1301-1313.
Weiss, R. (2009), “Ausgelernt? Befunde, interpretation und Empfehlungen zum
lebensbegleitenden Lernen älterer Menschen (no training any more? Findings,
interpretation and recommendations on lifelong training for older employees)”,
in Staudinger, U. and Heidemeier, H. (Eds), Altern Bildung und lebenslanges Lernen,
Altern in Deutschland 2, Nova Acta Leopoldina Neue Folge Band 100, Wissenschaftliche
Verlagsgesellschaft, Stuttgart, pp. 43-64.
Zwick, T. (2012), “Consequences of seniority wages on the employment structure”, Industrial and
Labor Relations Review, Vol. 65 No. 1, pp. 108-125.

Further reading
Kooning, J.de and Gelderblom, A. (2006), “ICT and older workers: no unwrinkled relationship”,
International Journal of Manpower, Vol. 27 No. 5, pp. 467-490.
Lois, D. (2007), “Determinanten der Weiterbildungsbeteiligung älterer Erwerbstätiger
(determinants of training participation of older employees)”, Zeitschrift für
Arbeitsmarktforschung, Arbeitsgestaltung und Arbeitspolitik, Vol. 16 No. 1, pp. 5-22.

(The Appendix follows overleaf.)

IJM Appendix
36,2

Variable name Mean Description

Main secondary 0.22 Employees with highest schooling degree main secondary
150 schooling schooling (Hauptschule)
Intermediate secondary 0.43 Employees with highest schooling degree intermediate secondary
schooling education (Realschule)
Higher secondary 0.34 Employees with highest schooling degree higher secondary
schooling schooling (Gymnasium)
Female 0.38 Female yes/no
Birth years 1951 or older 0.14 Employees born in year 1951 or before (aged 56/57 or older)
Birth years 1952-1961 0.37 Employees born in years 1952-1961 (aged 46/47-55/56)
Birth years 1962-71 0.33 Employees born in years 1962-1971 (aged 36/37-45/46)
Downloaded by Professor Thomas Zwick At 12:11 03 May 2015 (PT)

Birth years 1972 and 0.16 Employees born in year 1972 or after (aged 35 or younger)
younger
Tenure o2 years 0.12 Tenure o2 years
Tenure 2-5 years 0.10 Tenure between 2 and 5 years
Tenure 6-15 years 0.26 Tenure between 6 and 15 years
Tenure more than 15 0.42 Tenure more than 15 years
years
Good health 0.78 Topical health situation good or very good (dummy)
High probability to quit 0.03 High self-assessed probability to quit employment within next
12 months (dummy)
Eastern Germany 0.39 Workplace located in eastern Germany
Table AI. 100-199 employees 0.14 Establishment has between 100 and 199 employees
Descriptive statistics 200-499 employees 0.24 Establishment has between 200 and 499 employees
of explanatory 500-1,999 employees 0.61 Establishment has between 500 and 1,999 employees
variables Services sector 0.49 Establishment in services sector

Corresponding author
Professor Thomas Zwick can be contacted at: [email protected]

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]