Available via license: CC BY
Content may be subject to copyright.
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
160
THE IMPACT OF IN-HOUSE TRAINING ON THE DIVERSITY
DIMENSIONS OF EMPLOYEE PRODUCTIVITY IN THE
SOUTH AFRICAN WORKPLACE
Gerhardus van Zyl*
University of Johannesburg
Received: May 2016
Accepted: December 2016
Abstract
The aim of this article is to determine the impact of in-house training (defined as any training provided
by firms in the workplace) on employee productivity, employee remuneration and net employee
productivity gains when diversity attributes of the workplace are taken into consideration. The
manufacturing industry of Gauteng Province of South Africa is used as a case study. Fixed-effect panel
data estimations were performed in order to determine the diversity-based employee productivity,
remuneration and net productivity differentials of in-house training. The results accentuate the
important positive productivity, remuneration and net productivity spill-over effects created by in-
house training opportunities. The outcomes of the study also confirm the importance of a workplace
that is more gender diverse, racial diverse and in which skilled and older experienced employees are
retained if the productivity spill-over effects generated by in-house training opportunities are to be
enhanced.
Keywords
In-house trained employees, diversity dimensions, fixed -effect panel data estimations, employee productivity,
employee remuneration, net productivity gains
_______________________________
*Prof G van Zyl is a professor in the Department of Economics and Econometrics, University of Johannesburg, South Africa.
[hardusvz@uj.ac.za]
THE IMPACT OF IN-HOUSE TRAINING ON THE DIVERSITY DIMENSIONS OF EMPLOYEE PRODUCTIVITY IN SA
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
161
1. INTRODUCTION
The aim of this article is to determine the impact of in-house training (defined as any training
provided by firms in the workplace) on employee productivity, employee remuneration and net
employee productivity gains when diversity attributes of the workplace are taken into
consideration. The manufacturing industry of Gauteng Province of South Africa is used as a case
study.
The article forms part of an extensive research project on different aspects of employee
productivity in the South African workplace. This particular study covers both the employee
productivity and real employee remuneration impacts of in-house training on various employee
diversity dimensions. The diversity dimensions that are taken into consideration are differences
in skill levels, gender differences, different age groups and the racial composition of the
workforce. Literature on the employee productivity impacts of in-house training in developing
countries is very limited, while published literature for developed economies in this regard
generally indicates positive impacts of in-house training on both employee productivity and real
wages (Konings & Vanormelingen, 2010; Acemoglu & Pischke, 1998; Black & Lynch, 2001; Dearden,
Reed & Reenen, 2006; Moretti, 2004; Van Biesbroeck, 2010; Bauernschuster, Falck & Heblick, 2009,
Arumlampalam & Booth, 2001, Gavrel & Lebon, 2009). In order to expand the debate on employee
productivity in South Africa it is deemed important to determine what the magnitude of the
positive employee productivity and real employee remuneration impacts are if in-house training
practices are enhanced. In the estimation process the study encompasses important diversity
attributes, in-house training intensities, marginal efficiencies of in-house trained employees, in-
house trained employee remuneration costs and the net employee productivity gains for the firm.
2. LITERATURE STUDY
The majority of published research articles indicate the constant pressure to upgrade employee
skill levels in the workplace in order to keep pace with rapid technological advancements and the
contribution of in-house training practices fulfilling an integral part in this regard (Konings, et
al., 2010; Acemoglu et al., 1998; Black et al, 2001; Dearden et al., 2006; Moretti, 2004; Van
Biesbroeck, 2010 and Gavrel et al., 2009). These studies also conclude that in-house training
practices have a positive impact on employee productivity levels, that the marginal employee
productivity efficiencies of in-house trained employees is substantially higher when compared to
employees who have not undergone any in-house training, that employee remuneration levels are
enhanced by in-house training practices and that there are industry differences in the magnitude
of the in-house training practices–employee productivity link.
Konings et al. (2010) specifically argue in favour of in-house training that is specific (not general)
in nature due to the efficient nature of such training in terms of the enhancement of employee
productivity effects and the spill-over effects that it creates in the workplace. Moretti (2004) is
of the opinion that a lack of firm-based in-house training data is the single biggest reason for
relatively limited research on the employee productivity effect of specific in-house training
practices when compared to the vast literature on employee productivity effects of general
training (such as more efficient education levels). The study by Ackerberg, Caves and Frazer
(2015) concludes that the greater availability of firm-based data on in-house training practices
creates very important benefits for research in this regard. These benefits are multiple
applications of firm-based in-house training data sets, the elimination of aggregation biases and
Van Zyl
162
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
control for any endogeneity issues. Bauernschuster et al. (2009) conclude that firms’ decisions to
invest in in-house training are based purely on the real employee productivity benefits that may
flow from in-house training.
The studies by Koning et al. (2010), Bauernschuster et al. (2009) and Gavrel et al. (2009) indicate
that in nearly all of the cases the marginal employee productivity efficiencies generated by in-
house training practices are higher than real increases in employee remuneration levels that flow
from in-house training. Net employee productivity benefits are thus created for firms that invest
in in-house training practices. Acemoglu et al. (1998) further indicate a lower employee turnover
as a result of effective in-house training practices.
The studies of Jones (2001), Hellerstein, Newmark and Troske (1999), Van Biesbroeck (2010),
Konings et al. (2010) and Dearden et al. (2006) employ both production function and employee
remuneration estimation models in order to compare the real employee remuneration levels with
the real employee productivity levels that resulted from in-house training practices. These
studies use a Hicks-neutral technical efficient Cobb-Douglas format in which value added,
capital input and labour input are included. In order to differentiate between specific firm-based
and general components and to cater for time-specific deviations (from an average employee
productivity level) a natural logarithm format is used. Added advantages are that these models
also cater for the amount of in-house training practices and the intensity thereof. The estimation
models are also driven by the implicit assumptions that i) in-house trained and untrained
employees are perfect substitutes, ii) that in-house trained employees have similar employee
productivity gains, and iii) that in-house training is not treated as a binary variable. The basic
premises of these particular assumptions are that the level of the marginal efficiencies of in-
house trained employees can be determined relative to untrained employees, the in-house
trained productivity estimates are treated as average employee productivity impacts and
employees differ only in the amount of in-house training that they receive. The training intensity
estimation allows researchers to estimate average training costs and the determination of the
spill-over effect on employee remuneration levels should the intensity of in-house training
change. The real in-house training employee productivity gains are derived from the linear
estimation of value added on the employee input, capital input and the share of in-house trained
employees. A positive estimate of the share of in-house trained employees would indicate that
the marginal efficiency levels of in-house trained employees are higher than the marginal
efficiency levels of employees who have not undergone any form of in-house training.
The estimation models also consider the employee remuneration differentials between employees
who have undergone in-house training and those employees who have not undergone any in-house
training. The aim is to determine the impact of in-house training on employee remuneration
levels. Basically, the relative employee remuneration benefit for in-house trained employees is
the difference between the average remuneration levels of in-house trained employees and the
untrained employees divided by the average employee remuneration level of untrained
employees.
3. RESEARCH DESIGN
3.1 Research approach and method
The research design comprises the
THE IMPACT OF IN-HOUSE TRAINING ON THE DIVERSITY DIMENSIONS OF EMPLOYEE PRODUCTIVITY IN SA
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
163
identification of the various diversity dimensions to be included in the in-house training–
employee productivity estimation model,
specification of the adapted in-house training–employee productivity estimation model,
compilation of firm-based data sets of in-house training, diversity dimensions and employee
remuneration levels for the proxy firms in the manufacturing industry of Gauteng Province,
estimation process and
interpretation of the estimation results.
3.2 Model specification
The study employs an adapted simplified version of the Konings et al. (2010) model. In the
adapted version of the estimation model employee diversity dimensions (gender, race, skill levels
and different ages) are included. The adapted model is explained in the following few paragraphs.
Yit = βo + βLitgras + βLitgrasøt
+ βmitgras + βKKit + wit + ƞitgras
(1)
(where Yit = real sales for firm i in period t; βo = fixed component that is common to all the firms;
βLitgras = the quality of the employee aggregate for firm i for period t and for each gender grouping,
race grouping, age grouping and skills level; βLitgrasøt = the marginal productivity differential
between in-house trained employees (per gender grouping, race grouping, age and skills
groupings) and untrained employees, where
is the ratio of trained employees per gender
grouping, race grouping, age and skills groupings for firm i in period t; βmitgras = the production
material input for firm t in period i per gender grouping, race grouping, age and skills groupings;
βKKit = capital outlay for firm i in period t; wit = unobserved employee productivity effects such as
employee remuneration disputes, technological advancement and production run stoppages for
firm i in period t; ƞitgras = mean zero error term)
The Konings et al. (2010) model explicitly caters for the timing of capital accumulation in order
to circumvent correlation of the capital outlay input and employee productivity in the same year
(t). The decision to investment in production capital outlay is taken in the previous period (t-1)
and the aim is to identify the capital outlay coefficient in the estimation. The model is also based
on the assumption that a firm’s expectations of future employee productivity (per diversity
dimension) depend in the main on current employee productivity levels. The model clearly
distinguishes between the production material input and the capital outlay input based on the
assumption that the production material input is chosen after the employee and in-house training
inputs. This is especially true for rigid labour markets such as the South African labour market.
The demand for the production material input is a function of the production capital outlay,
employee productivity and in-house training. It is also important to note that the production
material input is based on a monotonic conditionality between employee productivity levels and
the efficient usage of the production material input (the higher the employee productivity levels
the higher the efficient usage of the production material input). Employee productivity is the only
unobservable element in mitgras based on the assumption that all other input prices are treated as
constant. The production material input (mitgras) is inverted to obtain an expression for employee
productivity, as the production material input is directly dependent on the quality of the
employee input (per gender grouping, race grouping, age and skills groupings) of firm i in period
t, the production capital outlay of firm i in period t and the fraction of in-house trained employees
Van Zyl
164
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
for firm i in period t (per gender grouping, race grouping, age and skills groupings). The production
material input function (mitgras) can be presented as ft-1(mitgras, Litgras,
, Kit).
It is also deemed important to determine the employee productivity differential between in-
house trained employees and non in-house trained employees. The coefficient on the share of in-
house trained employees is divided by the labour coefficient (øt
).
In order to compare the employee productivity benefits of in-house trained employees with the
remuneration benefits of in-house trained employees, employee remuneration estimates are
derived.
rem = remu + λT
+ Xγ + ε
(2)
(where rem = employee remuneration, remu = average remuneration, λT
= relative remuneration
premium for in-house trained employees, Xγ = diversity dimensions of employee remuneration)
In the Konings et al. (2010) model the intensity of the employee productivity in-house training
effect is defined as βT =
and the aim of
estimating βT is to determine the impact of the in-house training intensity on employee
productivity and to compare the calculated coefficients to the employee remuneration in-house
training intensity estimates. The estimations are controlling for possible endogeneity of in-house
training in both the employee productivity and employee remuneration equations.
The study employs fixed-effect panel data estimations for the in-house trained employee
productivity–gender dimension, the in-house trained employee productivity–race dimension, the
in-house trained–age dimension, the in-house trained employee productivity–skills dimension,
the in-house trained employee remuneration–gender dimension, the in-house trained employee
remuneration–race dimension, the in-house trained employee remuneration–skills dimension
and the in-house trained employee remuneration–age dimension. In order to facilitate
comparisons with other research and studies that deal with the diversity dimensions of employee
productivity the following gender, race, age and skills level categories are used in this study:
For gender attributes two categories are used, namely a gender distribution of less than 25%
female participation and a gender distribution of more than 25% female participation in the
workplace (Van Zyl, 2013 and Van Zyl, 2014).
For race attributes a category in which one specific race group has more than a 60% share
and a category where no particular race group has more than a 60% share in the workplace
are used (Van Zyl, 2013 and Van Zyl, 2014).
In terms of the age attribute three groups are used, namely employees 35 years of age and
younger, between 35 and 55 years of age and 55 years of age and older (Van Zyl, 2013 and
Van Zyl, 2014).
For the skills attribute the International Standard Classification of Occupations (ISCO-88) is
used in order to distinguish between more skilled occupations (Category A) and less skilled
occupations (Category B) in the workplace (Van Zyl, 2013).
Fixed-effect panel data estimations are done for each of the in-house trained employees–
diversity attributes.
THE IMPACT OF IN-HOUSE TRAINING ON THE DIVERSITY DIMENSIONS OF EMPLOYEE PRODUCTIVITY IN SA
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
165
3.3 Data collection process
In order to capture the employee productivity impacts of the different in-house trained-diversity
attribute dimensions, the manufacturing industry of Gauteng Province is used as a case study,
given the importance of the manufacturing industry in the gross geographical product (GGP) of
Gauteng Province and also given the availability of firm-based data.
Contact information was supplied by the Manufacturing Sector Education and Training Authority
(MERSETA), Department of Labour and the Chamber of Business. Data was supplied by the
individual firm in the sample group. Statistical validation requires a representative sub-sector
spread of firms in the manufacturing industry of Gauteng Province. The sample response of 107
firms, which covers a variety of sub-sectors in the Gauteng manufacturing industry, is confirmed
to be statistically significant.
The sample period was for the calendar years 2010-2013. For each firm in the sample group data
was collected on the number of employees, the number of employees per gender group, race, age
and skills groupings, average firm real sales turnover, average employee remuneration per gender,
race, age and skills level groupings, the proportion of in-house trained employees per gender,
race, age and skills groupings, the average training cost per employee per gender, race, age and
skills level groupings, time spent on in-house training per gender, race, age and skills groupings,
the size of the production capital outlay and the size of the production material outlay. A
summary of the sample statistics is provided in Appendix 1 and the log format of the data set is
provided in Appendix 2.
It should be noted that in-house training encompasses the total number of in-house training
opportunities. Individual employees could have attended more than one in-house training
opportunity. The second column of Appendix 1 indicates the average number of employees for the
sample of firms who have undergone some form of in-house training (per diversity category).
There are individual employees who have not undergone any form of in-house training. The third
column of Appendix 1 indicates the average number of employees for the sample of firms (for all
the diversity categories) who have not undergone any form of in-house training.
For the sample of firms in the study the following can be deduced from Appendix 1:
The average number of employees is 97 employees. An average number of 65.7 employees
have undergone some form of in-house training while an average of 24.9 employees have not
undergone any form of in-house training. The majority of employees on average have
undergone some form of in-house training.
An average number of 33 employees were employed in workplaces where the female
participation rate was higher than 25%. For this category of female representation an
average of 19.5 employees have received some form of in-house training while an average of
8.6 employees have received no in-house training. In terms of female participation of less
than 25% representation in the workplace an average of 64 employees were employed. An
average of 34.2 employees have received some form of in-house training while an average of
17.9 employees have received no form of in-house training. It is clear that for workplaces
where female representation was less than 25%, employment levels on average were nearly
double that of employment levels for workplaces in which female representation was more
than 25%. On average, in-house training employee levels were much higher for workplaces
where female participation levels were less than 25%.
Van Zyl
166
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
For a workplace in which one race group has a more than 60% representation of the workforce
an average of 63 employees were employed, while for a workplace employee structure in
which no individual race group has a greater than a 60% representation an average of 27
employees were employed. For a greater than 60% representation of one race group an
average of 29.5 employees have received some form of in-house training, while an average
of 14.2 employees have received no form of in-house training. For the less than 60%
representation of one race group category an average of 14.3 employees have received some
form of in-house training, while an average of 8.3 employees have received no form of in-
house training.
In terms of the different age groups an average number of 29 employees in the 35 years and
younger bracket, 46 employees in the 35–55 year bracket and 22 employees in the 55 and
older age bracket are employed per firm in the sample of firms. For the sample of firms an
average of 22.4 employees in the 35 years and younger age bracket, 35.8 employees in the
35–55 age bracket and 19.7 employees in the 55 years and older age bracket have undergone
some form of in-house training. For the sample of firms an average of 15.4 employees in the
35 years and younger age bracket, 16.7 employees in the 35–55 age bracket and 3.6
employees in the 55 years and older age bracket have undergone no form of in-house
training. In terms of skill levels, an average of 33 employees are classified as category A
employees and 64 employees per firm are classified as category B employees (for the sample
of firms). For category A employees an average of 29.2 employees have received some form
of in-house training, while an average of 9.4 employees have received no form of in-house
training. For category A employees on average 9.4 employees per firm have received no form
of in-house training while an average of 17.5 category B employees per firm have received no
formal in-house training.
The average real sales turnover for the sample of firms was R17 052 000. Average real sales
turnover attributed to in-house trained employees was R13 403 000 and for employees who
have received no in-house training the average sales turnover was R4 283 000.
The average employee remuneration for the sample of firms was R85 400. Employees who have
undergone some form of in-house training received an average remuneration of R115 100,
whereas employees who have received no form of in-house training received an average
remuneration of R38 400.
Average employee productivity for the sample of firms is calculated as the average real sales
turnover divided by the average number of employees. For the sample of firms the average
employee productivity is R175 790 and for employees who have undergone some form of in-
house training average employee productivity is R204 000.
For the sample of firms the average capital outlay is R19 063 000, the average
capital/employee ratio is R196 520, the average material output is R2 720 000, the average
material/employee ratio is R28 040, the portion of in-house trained employee opportunities
is 0.65, the average cost of in-house trained employees is R1 423 and the average hours of
in-house training per employee is 18 hours.
In order to determine the magnitude of in-house training of the dataset (when compared to non
in-house training) log regressions are done for each of the key variables (including the sub-
diversity components) on a dummy variable that is equal to 1 when in-house training is provided
and 0 otherwise. The coefficients of the in-house training dummy are then interpreted as either
absolute size differences (average number of employees and real sales added) or a percentage
THE IMPACT OF IN-HOUSE TRAINING ON THE DIVERSITY DIMENSIONS OF EMPLOYEE PRODUCTIVITY IN SA
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
167
difference (employee remuneration cost, employee productivity and the capital/employee ratio)
(Konings et al., 2010).
From Appendix 2 the following can be deduced:
A firm that offers in-house training is typically about three times the size of non in-house
training firms.
Employee remuneration is 26% higher, employee productivity is 33% higher and the
capital/employee ratio is 22% higher for firms that offer in-house training.
4. ESTIMATION RESULTS
The panel data estimates for the impact of in-house training on employee productivity, employee
remuneration, in-house training intensity and the net productivity benefits for the sample of
firms (for the different diversity dimensions) are presented in this section.
TABLE 1: Panel data estimates for the impact of in-house training on employee productivity (for
the different diversity dimensions)
Full sample
estimates
(excluding material
cost)
Full sample
estimates (including
material cost)
Full sample
estimates
(controlling for
endogeneity of
inputs)
Labour:
Total
0.913
0.864
0.814
Less than 25% female
presentation
0.902
0.812
0.758
More than 25% female
presentation
0.970
0.902
0.834
More than 60% single race
presentation
0.754
0.711
0.671
Less than 60% single race
presentation
0.808
0.724
0.693
Category A employees
0.974
0.910
0.879
Category B employees
0.825
0.723
0.650
Age 35 years and less
0.651
0.591
0.561
Age between 35 and 55
years
0.756
0.712
0.591
Age 55 years and older
0.711
0.658
0.581
Captial
0.241
0.215
0.192
Van Zyl
168
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
Full sample
estimates
(excluding material
cost)
Full sample
estimates (including
material cost)
Full sample
estimates
(controlling for
endogeneity of
inputs)
In-house training:
Total
0.571
0.504
0.482
Less than 25% female
presentation
0.511
0.484
0.431
More than 25% female
presentation
0.668
0.524
0.514
More than 60% single race
presentation
0.473
0.413
0.375
Less than 60% single race
presentation
0.513
0.461
0.399
Category A employees
0.643
0.543
0.509
Category B employees
0.462
0.418
0.368
Age 35 years and younger
0.477
0.425
0.386
Age between 35 and 55
years
0.539
0.471
0.426
Age 55 years and older
0.501
0.437
0.413
Source: Author’s analysis
*The estimates are significant at a 5% confidence level
All the panel data estimates are positive, thus indicating a positive relationship between in-house
trained employees and employee productivity. Employee productivity is represented by the
positive impact of in-house training opportunities on real sales. The panel data estimates become
smaller when material costs are included in the estimations and even smaller when the
estimations are controlled for endogeneity of inputs. The reason why the estimations are
controlled for the endogeneity of inputs is the observation that larger firms are more likely to
provide in-house training opportunities for employees. The effect is that the positive relationship
between in-house training and employee productivity can result in an upward bias of the in-house
training coefficients. The estimates are expressed as the impact of a 10% increase in the number
of in-house trainined employees on real sales. The discussion of the panel data estimates will
focus only on the last column of TABLE 1 (panel data estimates that controls for endogeneity of
inputs), and the following can be deduced in terms of the positive impact of in-house training on
real sales:
A 10% increase in in-house training opportunities has a 4.82% positive impact on real sales.
In terms of the gender dimension the positive impact on real sales is greater when the
workplace is more gender diverse (5.14% versus 4.31% when in-house trained employees are
increased by 10%).
THE IMPACT OF IN-HOUSE TRAINING ON THE DIVERSITY DIMENSIONS OF EMPLOYEE PRODUCTIVITY IN SA
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
169
A more racially diverse workplace has a stronger impact on employee productivity when in-
house training opportunities are increased by 10% (3.99% versus 3.75% increase in real
sales).
The employee productivity impacts of in-house training for the higher skilled employee
segment are more profound when compared to the less skilled employee segment (5.09%
versus 3.68% increase in real sales).
The greatest employee productivity benefit generated by a 10% increase in in-house training
opportunities is for the 35–55 years of age grouping of employees, followed by the 55 years
and older grouping (4.26% versus 4.13%).
In order to determine the employee productivity differential between in-house trained employees
and non in-house trained employees the marginal efficiency of each diversity segment is
calculated. The employee marginal efficiency of in-house training opportunities is calculated as:
× %. The employee marginal efficiencies for the different diversity
dimensions are listed in TABLE 2.
TABLE 2: Marginal efficiency differential for in-house trained employees
Diversity dimensions
Marginal efficiency differential (%)
Total
59.2
Less than 25% female presentation
56.8
More than 25% female presentation
61.6
More than 60% single race presentation
55.9
Less than 60% single race presentation
57.6
Category A employees
57.9
Category B employees
56.6
Age 35 years and younger
68.8
Age between 35 and 55 years
72.1
Age 55 years and older
71.1
Source: Author’s analysis
The positive differentials between in-house trained employees and non in-house trained
employees (for all the diversity dimensions) are in excess of 50%. For the sample of firms the total
employee productivity differential (marginal efficiency) for in-house trained employees is 59.2%.
The following in-house trained employee marginal efficiency differential observations for the
different diversity dimensions can be deduced from TABLE 2.
The same diversity employee productivity patterns of in-house trained employees are
indicated when TABLE 2 is compared to TABLE 1.
In terms of the gender dimension a more diverse workplace creates a greater employee
marginal efficiency differential (61.6% versus 56.8%).
A more race diverse workplace greates higher employee marginal efficiency differentials
(57.6% versus 55.9%).
Van Zyl
170
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
Higher skilled employees create a higher employee marginal efficiency differential (57.9%
versus 56.6%).
The greatest employee marginal efficiency differential is created by the 35–55 year age
grouping followed by the 55 year and older grouping.
TABLE 3 represents the panel data estimates for the impact of in-house training on employee
remuneration. The aim of the estimates is to determine the remuneration differential of
employees who have undergone some form of in-house training. The estimates are interpreted as
percentage changes.
TABLE 3: Panel data estimates for the employee remuneration impacts of in-house training
opportunities (for the different diversity dimensions)
Full sample estimates
(log remuneration
regressed on share of
in-house trained
employees)
Full sample estimates
(log remuneration
controlled for the
endogeneity of
inputs)
Full sample estimates
(log remuneration,
capital/labour ratio
& total factor
productivity included
as control variable)
Total
0.524
0.347
0.284
Less than 25% female
presentation
0.501
0.308
0.264
More than 25% female
presentation
0.608
0.389
0.298
More than 60% single
race presentation
0.576
0.432
0.212
Less than 60% single
race presentation
0.599
0.451
0.232
Category A employees
0.534
0.352
0.217
Category B employees
0.502
0.320
0.117
Age 35 years and
younger
0.429
0.301
0.134
Age between 35 and 55
years
0.539
0.386
0.224
Age 55 years and older
0.492
0.362
0.193
Source: Author’s analysis
*The estimates are significant at a 5% confidence level
For the sample of firms all the employee remuneration estimates are positive, indicating that
employees who have undergone some form of in-house training are receiving remuneration levels
that are higher than for those employees who have not undergone any form of in-house training.
The last column of TABLE 3 indicates that for the sample of firms the remuneration levels of
employees who have undergone some form of in-house training is 28.4% higher than the
remuneration levels of employees who have not undergone any form of in-house training. In terms
THE IMPACT OF IN-HOUSE TRAINING ON THE DIVERSITY DIMENSIONS OF EMPLOYEE PRODUCTIVITY IN SA
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
171
of all the diversity dimensions, the employee remuneration estimates indicate that the positive
employee remuneration differentials are greater for a more diverse gender composition, a more
race diverse workplace, higher skilled employees and for the 35-55 year age grouping. The positive
employee remuneration differentials of lower skilled and the younger age grouping are the lowest
for the diversity dimensions of the sample group.
Net employee productivity gains for the sample of firms are possible only if the employee
remuneration differentials are less than the employee productivity differentials. The last column
of TABLE 2 (marginal efficiency differentials (%)) is compared with the last column of TABLE 3
(% differential of employee remuneration differential). It is clear that for all the diversity
dimensions the marginal efficiency differentials are greater than the employee remuneration
differentials. It can thus be concluded that in-house trained employees create positive net
employee productivity benefits for the sample of firms.
It is important for the purposes of this particular study to compare the in-house training
intensities of employee productivity and employee remuneration levels. The in-house training
variable is expressed as the average training hours per employee. The aim is to determine the
impact of the in-house training intensity on employee productivity and employee remuneration
levels respectively. The average in-house training intensity estimates are presented in TABLE 4.
TABLE 4: In-house training intensity estimates (for the various diversity dimensions)
Employee
productivity
β
T
Employee
remuneration
Capital outlay
0.104
Labour
0.829
Average in-house training intensity:
Total
0.0064
0.0077
0.0048
Less than 25% female presentation
0.0092
0.0111
0.0064
More than 25% female presentation
0.0050
0.0060
0.0040
More than 60% single race presentation
0.0046
0.0055
0.0038
Less than 60% single race presentation
0.0051
0.0062
0.0043
Category A employees
0.0078
0.0094
0.0062
Category B employees
0.0048
0.0058
0.0038
Age 35 years and younger
0.0035
0.0042
0.0027
Age between 35 and 55 years
0.0047
0.0057
0.0034
Age 55 years and older
0.0045
0.0054
0.0032
Source: Author’s analysis
*The estimates are significant at a 5% confidence level
For all the diversity dimensions of the sample of firms the employee productivity intensities (βT)
are greater than the employee remuneration intensities. This is a further confirmation of the
Van Zyl
172
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
existence of net employee productivity gains for the sample of firms due to in-house training
opportunities.
5. CONCLUSION
The aim of the article was to determine the impact of in-house training (defined as any training
provided by firms in the workplace) on employee productivity, employee remuneration and net
employee productivity gains when diversity attributes of the workplace are taken into
consideration.
Three aspects of the diversity dimensions of the in-house training–employee productivity
relationship were estimated, namely the productivity differentials of in-house trained employees,
the differentials of in-house trained employee remuneration levels and the net productivity gains
for the sample of firms. The results of this particular study accentuate the important positive
spill-over effects generated by in-house trained employees such as higher employee productivity
levels, higher employee remuneration levels and net productivity gains for the sample of firms. In
terms of the diversity dimensions, the results of the study confirm the importance of a more
gender diverse workplace, a more racially diverse workplace, a more skilled workforce and the
retaining of older more experienced employees if the net productivity benefits of in-house
training are to be enhanced.
A possible further area of research is a comparable diversity-based multiple industry and
geographical analysis of the in-house training-employee productivity relationship.
LIST OF REFERENCES
Acemoglu, D. & Pischke, JS. (1998). Why do firms train? Theory and evidence.
Quarterly Journal of
Economics
, 113(1), pp. 79-119.
Ackerberg, D., Caves, K. & Frazer, G. (2015). Identification properties of recent production function
estimators.
Econometrica
, 83(6), pp. 2411-2451.
Arumlampalam, W. & Booth, A. (2001). Learning and earning: Do multiple training events pay? A
decade of evidence from a cohort of young British men.
Economica
, 68(271), pp. 379-400.
Bauernschuster, S., Falck, O. & Heblick, S. (2009). Training and innovation.
Journal of Human Capital
,
3(4), pp. 323-353.
Black, SE. & Lynch, LM. (2001). How to compete: The impact of workplace practices and information
technology on productivity.
The review of economics and statistics,
83(3), pp. 434-445.
Dearden, L., Reed, H. & Reenen, J.V. (2006). The impact of training on productivity and wages:
Evidence from British panel data.
Oxford Bulletin of Economics and Statistics
68(4), pp. 397-421.
Gavrel, F. & Lebon, I. (2009). Minimum wage, unemployment benefits and labour market efficiency.
Louvain Economic Review
, 74(1). pp. 53-75.
Hellerstein, J.K., Neuwmark, D. & Troske, K.R. (1999). Wages, productivity and worker characteristics:
Evidence from plant level production functions and wage equations.
Journal of Labor Economics
,
17(3). pp. 409-446.
THE IMPACT OF IN-HOUSE TRAINING ON THE DIVERSITY DIMENSIONS OF EMPLOYEE PRODUCTIVITY IN SA
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
173
Jones, P. (2001). Are educated workers really more productive?
Journal of Development Economics,
64(1). pp. 57-79.
Konings, J. & Vanormelingen, S. (2010). The impact of training on productivity and wages: Firm level
evidence,
IZA DP
, No. 4731. pp. 1-53
Moretti, E. (2004). Workers education, spillovers and productivity: Evidence from plant-level
production functions.
American Economic Review
, 94(3). pp. 656-690.
Van Biesebroeck, J. (2010). Wages equal productivity. Fact or Fiction?: Evidence from Sub Saharan
Africa.
World Development,
39(8). pp. 1333-1346.
Van Zyl, G. (2013). The relative labour productivity contribution of different age-skill categories for a
developing economy: The Gauteng province of South Africa as a case study.
The South African Journal
of Human Resource Management
, 11(1), pp. 1-8.
Van Zyl, G. (2014). Labour productivity and employee diversity in the South African workplace.
Journal of Economic & Financial Sciences
, 7(2), pp. 451-466.
Van Zyl
174
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
APPENDIX 1: SUMMARY STATISTICS
Total
In-house
training
No in-house
training
Average number of employees
97
65.7
24.9
Gender distribution of a more than 25% female
employee participation rate
33
19.5
8.6
Gender distribution of a less than 25% female
employee participation rate
64
34.2
17.9
Average number of specific race group greater
than 60% representation
63
29.5
14.2
Average number of specific race group less
than 60% representation
27
14.3
8.3
Average number of age 35 and younger
29
22.4
15.4
Average number of age between 35 and 55
years of age
46
35.8
16.7
Average number of 55 years and older
22
19.7
3.6
Average category A employees
33
29.2
9.4
Average category B employees
64
34.1
17.5
Average firm sales turnover (x R1 000)
17 052
13 403
4 283
Average employee remuneration (x R1 000)
85.4
115.1
38.4
Average employee productivity (x R1 000)
175.79
204
143
Average production capital outlay (x R1 000)
19 063
Average capital/employee ratio (x R1 000)
196.52
Average production material outlay (x R1 000)
2720
Average material/employee ratio (x R1 000)
28.04
Portion of in-house trained employees
0.65
Average cost of in-house training per employee
R1 423
Average hours of in-house training per
employee
18 hours
Source: Author’s analysis
THE IMPACT OF IN-HOUSE TRAINING ON THE DIVERSITY DIMENSIONS OF EMPLOYEE PRODUCTIVITY IN SA
Journal of Economic and Financial Sciences |
JEF
| April 2017, 10(1), pp. 160-175
175
APPENDIX 2: THE LOG REGRESSION COEFFICIENTS OF THE DIVERSITY
ATTRIBUTES OF IN-HOUSE TRAINING
In-house training log
coefficients
Average number of employees (Ln):
Total
3.14
Less than 25% female presentation
3.67
More than 25% female presentation
2.98
More than 60% single race presentation
4.18
Less than 60% single race presentation
3.22
Category A employees
3.78
Category B employees
2.65
Age 35 years and less
3.13
Age between 35-55 years
5.42
Age 55 years and older
4.78
Average real sales turnover per employee (Ln):
Total
3.24
Less than 25 % female presentation
3.05
More than 25% female presentation
3.25
More than 60% single race presentation
2.97
Less than 60% single race presentation
3.26
Category A employees
3.55
Category B employees
3.09
Age 35 years and less
2.54
Age between 35-55 years
2.03
Age 55 years and older
3.76
Employee remuneration cost (Ln)
0.26
Employee productivity (Ln)
0.33
Capital/employee ratio (Ln)
0.22
Source: Author’s analysis
