You are on page 1of 19

Electronic copy available at: http://ssrn.

com/abstract=2174205
1

STAFFING AND ORGANISATION PERFORMANCE
SUPPORTS FOR A SLIGHT OVERSTAFFING LEVEL
Trong B. Tran
1
and Steven R. Davis
2

School of Civil and Environmental Engineering
The University of New South Wales, Sydney, Australia

Abstract
The present paper presents a theoretical approach to measure the relationship between staffing
levels and organisation performance. Results from the approach are used to find the optimum staffing
level that maximises the profit of the firm. This study assumes that incoming work to a firm follows a
random walk model, while the staffing level is managed by the firm in order to maximise the profit. The
model includes effects of variability of external demand to the firms profit such as overtime work, late
penalties, hiring and firing costs, and so on. The model was implemented using MATLAB and used to
compute the profit for different staffing levels. Results from the present study indicate that a firm
should be slightly overstaffed to maximise its profit. However, this may be related to the particular
parameters used in the investigation.
Keywords Overstaffing, Organisation performance, Random walk
model.
Fields of Research: Human resource management, Strategic human
resource management, Managing people and organisation, Operations
management

1
PhD Candidate, School of Civil and Environmental Engineering, The University of New South
Wales, Sydney, NSW 2052, Australia. Corresponding author. Tel: +61 2 9385 4290; fax: +61 2 9385
6139; email: z3213671@unsw.edu.au.
2
Lecture, School of Civil and Environmental Engineering, The University of New South Wales,
Sydney, Australia.
Electronic copy available at: http://ssrn.com/abstract=2174205
2

I. Introduction
An important question for an organisation is what staff level to adopt.
Overstaffing provides an organisation with the flexibility to resource
unanticipated tasks and absorbs the effects of absenteeism, as well as
maximising unexpected opportunities as they arrive. However, having
extra staff means more costs to the organisation. Hence too many staff
may reduce the profitability of the organisation. The question becomes
will the benefits that extra staff brings to the organisation exceed the
costs to maintain this extra staff.
The present study examines the effects of staffing levels on the profit
performance of a firm. Different levels of staff, from a slight understaffing
to a moderate overstaffing, are modelled to see how the profit of a firm
changes when external demands fluctuate. The staffing size is controlled
by the firm, whereas the incoming work is assumed to follow a random
walk model. Results from the present study prove that there is a point of
overstaffing level at that the firms profit reaches the maximum value.
Understaffing and too much overstaffing conditions will shrink the profit
of the organisation.
This paper is organised as follows. Section 2 presents a review of the
literature on the effects of staffing levels on organisation performance.
The modelling is explained in Section 3. Section 4 gives details in data
Electronic copy available at: http://ssrn.com/abstract=2174205
3

collection, while Section 5 shows the results of the research.
Conclusions and suggestions for future research are in the last section of
the paper.
II. The Relationship Between Staffing Levels and Performance in
Organisations
Different staffing levels have different effects on organisational
performance. However, the literature has not reached an agreement on
the relationship between workforce size and performance of an
organisation. On one hand, staffing theorists conclude that an
organisation should be at a moderate understaffing level as this state
improves organisation performance. On the other hand, organisation
researchers have found that a slight overstaffing condition has a positive
relationship with performance in an organisation. There is, however, a
common agreement that both too much overstaffing and extreme
understaffing conditions have negative effects on organisational
performance.
An understaffing condition is defined as lack of enough people to
carry out smoothly the essential program and maintenance tasks in a
setting (Wicker, 1984, p. 71). This condition puts pressure on the staff in
the organisation, as the same jobs have to be done by fewer people.
This means that staff must assume a greater share of jobs; they must do
4

more things. Consequently staff in an understaffing condition are busier
and more dependent upon another than those in the other staffing
condition (Wicker, 1984).
A moderately understaffed organisation may gain better performance
compared with other conditions since employees in this state are likely to
work more efficiently and to experience higher motivation. In a slightly
understaffed state, each participant is involved in a wider variety of tasks;
hence they need to use varied skills to complete those tasks.
Furthermore, workers also need to find ways to combine similar tasks
together in order to reduce wasted time from redundant activities. As a
consequence, their efficiency is improved (Vecchio and Sussmann,
1981; Treville and Antonakis, 2006).
Workers in understaffed conditions are also claimed to have more
freedom on deciding the way to organise and complete their tasks. This,
in return, increases their working motivation (Ganster and Dwyer, 1995).
These positive experiences lead to improvement in the outputs of
individuals and hence the performance of the whole organisation is
improved.
However, negative effects on workers have been reported in
understaffed organisations. Studies show that workers in understaffed
conditions normally suffer from higher burnout and higher emotional
5

exhaustion compared with those in an overstaffed situation (Rochefort
and Clarke, 2010). This leads to reduced motivation, lower productivity
and poor performance among individuals (Rafferty et al., 2007). Other
disadvantages of understaffing claimed by researchers include lower
levels of aggregate organisation outputs, losses in business
opportunities, and increases in error rates of staff (Ahmed, 2007).
Organisation researchers, on the other hand, do not support
understaffing in organisations. They argue that a slight overstaffing leads
to improved organisation performance compared to a moderate
understaffing condition. Numerous studies have shown that a slight
overstaffing condition is associated with better outcomes at both staff
and organisation levels. At the staff level, employees in overstaffed
groups are reported to suffer less from burnout, to have higher job
satisfaction, and to have better worklife balance; thus to work more
productive and to produce a higher quality of service (Rafferty et al.,
2007). At the organisation level, the existence of excess staff stabilises
operations and enables competitive advantages of the organisation.
Excess staff are considered as a resource cushion that buffers the
organisation from environmental shocks, aids the organisation to remain
flexible and to prepare better for unpredicted external threats (Welbourne
et al., 1999; George, 2005). Organisations are also encouraged to
6

deploy extra staff in order to create a pool of creative and innovative
ideas, which strengthens the organisation and enables the introduction of
new products or entering of new markets (Cheng and Kesner, 1997).
Furthermore, these staff may provide organisations with a competitive
advantage because competitors cannot obtain the same resource
configurations and copy the organisations strategies (Mishina et al.,
2004). The literature acknowledges that too many staff may inhibit
organisation performance. This is because of increasing levels of
management, times for making decisions, and costs. Hence researchers
suggest that an organisation should not be greatly overstaffed (Tan,
2003; George, 2005).
The relationship between staffing levels and organisation performance
has been investigated previously by the authors (Trong B. Tran and
Davis, 2011, 2012b; Trong Binh Tran and Davis, 2012a). In previous
works, the authors used mathematical models such as a queuing model
and Markov chain models to quantitatively measure the effects of
different staffing levels on performance of an organisation. These studies
found that an organisation should be at a slight overstaffing condition to
get better performance in profit. However, these previous works just
focus on the group level. Higher levels have not been investigated by the
authors. The present research extends the authors work to the
7

organisation level. In this study, a random walk model is used to predict
the external demand of a firm, while a profit model is proposed to
estimate the profit of the firm over time. Results from this work show that
a slight overstaffing leads to improved organisation performance
compared to understaffing.
III. Modelling
1. Random Walk Model
Consider a movement of a walker on an infinite one-dimensional
uniform lattice. Suppose that the motion is started at the origin (x=0), and
the walker moves a short distance ! either left or right in a short time ".
The movement is assumed to be completely random, so the probabilities
of moving both left and right are 0.5. After the first time step, the walker
can either be at a distance ! to the left or right of the origin, with
probability 0.5 each. After the next time step, the walker is either at a
distance 2! to the left or right of the origin (with probability 0.25 each) or
has returned to the origin (with probability 0.5). At the end of n steps the
walker can be any one of 2n+1 sites and the probability that the walker is
at any site can be calculated. The probability of being at a distance m! to
the origin after n time steps of the walker is given by (Codling et al.,
2008, p. 815)
8

! !! ! !
!!
!
!
! ! !
!
!
! ! !
!
!

(1)
2. Applying the Model to External Demands of an Organisation
The present study assumes that incoming work of a firm follows a
random walk model. That is, the amount of incoming work to an
organisation changes from time to time because of its volatility. Suppose
that a firm has D demands at the beginning, and the demands can go
either up to u*D or down to d*D (where u#1, d=1/u) after a time step ".
The changes are assumed to be completely random, so the probabilities
of going up and down are both 0.5. After the next time step, the demand
is either at u
2
*D up or d
2
*D down (with probability 0.25), or has been D
(with probability 0.5). After n time steps, the amount of work for the firm
can be at state m of u
m
*D up or d
m
*D down to the origin D. The
probability of being at state m after n time steps is calculated using
Equation 1. An example of potential movement of demand in three time
steps is illustrated as in Figure 1 below.
9



3. Profit Model
When the demand varies from month to month, the firms workload is
fluctuated. Hence overtime work and/or late work may occur as the
external demand exceeds the capacity; or workers may have less work
to do when the incoming work is less than the firms capacity. All these
scenarios have effects on the financial performance of the firm. That is,
overtime work leads to overtime paid, or late work has results on
penalties, and so on. All these costs are considered when calculating the
profit of the firm.
The monthly profit of a firm is given by
Profit = Work done * Benefit Salaries Overtime
paid Late penalties Hiring costs Firing
costs
(2)
!
#!
#
$
!
#
%
!
#!
!
&!
!
#!
!
&
$
!
&!
&
%
!
Figure 1: The range of potential movement of demand over time
10


Where:
Work done: the amount of work has been done after a
month.

Benefit: the amount of money that the firm receives
when completing an amount of work.

Salaries = Number of workers * Workers average
salary
(3)
!"#$%&'# !"#$ !
!"#$%&'# !"#$% ! !"#$%!
!
! !"#$!%# !"#"$%
!"#$%& !" !"#$%#&% !"#$%&' !"#$% !"# !"#$%
(4)
Late penalties = Number of late days * Penalties rate (5)
Hiring costs = Hiring rate * Workers average salary *
Number of hired workers
(6)
Firing costs = Firing rate * Workers average salary *
Number of fired workers
(7)
Different rules to control the staffing size when the workload changes
are investigated. The rules used in the present study are as follows:
1. The number of staff is constant over time.
2. The number of staff is changed immediately to maintain the
workload in a given range. When the workload is outside that
range, the number of staff is adjusted. That is, when the workload
excesses the hiring boundary, a number of workers are hired to
reduce the workload; and when the workload goes below the firing
boundary, a number of staff is fired to increase the workload. Hiring
11

and firing times as well as hiring and firing costs are exclusive from
this case.
3. The number of staff is changed when the workload changes. This
is the same with the previous scenario except rules for hiring and
firing staff are calculated in the model.
4. The number of staff is controlled by the number of backlogs. When
the number of backlogs after a month goes beyond a given value,
the workload in the following month increase hence more workers
are needed to cover the changes. Hiring times and hiring costs of
new employees are not included in the calculation.
5. The number of staff is driven by delayed work. Workers are hired
when the amount of delayed work excesses a particular point.
However, times and costs for hiring new staff are not calculated in
the present research.
4. Algorithm in Pseudo Code to Compute the Firm Profit
The algorithm to compute the profit of the firm is as follows
12



5. Modeling Processes
The above algorithm was coded using the MATLAB language to
compute profits in different rules of controlling the staffing size (ie values
and meanings of large for backlog size and values of low for demand /
staffing level). Different initial numbers of staff, ranging from an
understaffing to an overstaffing, were examined to see which staffing
level would maximize the firms profit. The exact staffing level is
identified as the number employees being such that their capacity to do
Determine demand volatility from previous data.
Determine the up and down changes in demand to be
used in the binomial lattice.
For each month:
For both up and down changes in demand:
For staff level staying the same or reducing
due to resignations:
Calculate current staff level allowing
for previous staffing changes
If Backlog is large
Start the process of hiring a new
staff person
Else if demand / staffing level is low
Start the process of firing one of the
staff
Calculate profit
Calculate expected value of profit for the whole
binomial lattice


Figure 2:Algorithm to compute the profit
13

work (without overtime) is exactly equal to the mean amount of work
available.
IV. Data Collection
Data was collected from a Vietnamese manufacturing firm over five
years, from 2000 to 2004. External demands of the firm in each month,
the number of workers, the average salary of workers, benefits, and the
cost structure were collected.
This was a small size company with approximately 20 workers.
Incoming work to the firm was measured by using a standard unit of
work. The standard unit of work was also used to estimate the benefit
that the firm would receive for new incoming work and to determine the
breakdown of cost for each project. The standard unit of work was sized
to be the amount of work that a worker could do in a month based on
firms experience. The firm had policies to change its staffing size in
order to maintain the workload at 1 unit per staff per month. Failure to do
this would otherwise cause various problems such as overtime costs,
late penalties, or other potential costs.
V. Results Analysis
Data collected from the firm was used to estimate the volatility of
external demands. The model was used to analyse a time period of one
14

year using time steps of one month. This resulted in an upmovement
value of 1.1232 and a downmovement of 0.8903. The random walk
model presented in section III then was used to compute the external
demands of the firm for each month of the twelve month time period.
Equation 2 was used to calculate the firms monthly profit. Benefits of
work done, costs for salaries, overtime costs, late penalties, hiring costs
and firing costs were calculated when applicable. Demand for the first
month was set to be 20 units of work. Different numbers of initial staff,
ranging from 15 to 25, were examined to see what initial staffing level
would maximise the profit of the firm. Different rules were used for
modifying staff levels during the year as described in Section 3.
Figure 3 shows
profit vs initial staffing
level for the case
where hiring and firing
was not permitted.
Other hiring and firing
rules gave similar
results. The results in all
examined rules indicate that initial staffing levels have little effects in the
earlier months but the cumulative effects in the later months when the
Figure 3: Monthly profits of the firm when the
numbers of staff are constant over time
15

range of workloads varied greatest became quite large, particularly for
the understaffing condition. In the constant staffing size case, for
example, looking at staffing levels ranging from 15 to 25 gave the firm
profits between $2200 and $3200 (range = $1000) after the first month
but this range expanded to approximately $17,000 to $25,800 (range
=42,800) by the end of the 12 month modelling period (Figure 3). The
difference is mainly due to the variation in incoming work.
Results for the
different hiring and
firing rules show that
the optimum staffing
level is 20 workers
when instant hiring
and firing is possible.
In the cases where
hiring experienced delays (eg advertising and interviewing times) the
optimum staffing level is 22 for each of the four hiring and firing rule sets
investigated. Figure 4 shows the results for each of the individual hiring
and firing rules.
A staffing level of 22 in this case corresponds to slight overstaffing.
The models were set up with parameters such that firing costs were 3
-$20
-$15
-$10
-$5
$-
$5
$10
$15
$20
$25
$30
15 16 17 18 19 20 21 22 23 24 25
P
r
o
f
i
t
s

(
T
h
o
u
s
a
n
d
s
)

Initial staff
Constant staff
Staff changed by workload
Staff changed by backlogs
Staff changed by delayed work
Staff changed by workload with hiring/firing rules
Figure 4: Cumulative profits of the firm after 12
months.
16

times higher than hiring costs, but hiring took time, while firing was
instant. The effect of the hiring delay was more influential on the
optimum staffing level than the need to payout workers who were fired.
VI. Conclusions and Future Research
Results from the present study have theoretically shown that
overstaffing leads to improved organisation performance compared to
understaffing. The present research shows that the investigated firm
should be at a slight overstaffing to get the maximum profit. Other states,
insufficient staff and too many staff, may inhibit the firms performance.
Further work planned in this area includes investigation of other
scenarios such as effects of time and costs for hiring and firing staff, or
different workload boundaries when changing staffing sizes to provide a
larger picture on the effects of staffing levels on profit of an organisation.
References
Ahmed, H. (2007). Improved Operations through Manpower
Management in the Oil Sector. Journal of Petroleum Science and
Engineering, Vol. 55, No. 1-2, pp. 187-199.
Cheng, J. L. C., and Kesner, I. F. (1997). Organizational Slack and
Response to Environmental Shifts: The Impact of Resource
Allocation Patterns. Journal of Management, Vol. 23, No. 1, pp. 1-
18. doi: 10.1177/014920639702300101
17

Codling, E. A., Plank, M. J., and Benhamou, S. (2008). Random Walk
Models in Biology. Journal of the Royal society interface, Vol. 5,
pp. 813-834.
Ganster, D. C., and Dwyer, D. J. (1995). The Effects of Understaffing on
Individual and Group Performance in Professional and Trade
Occupations. Journal of Management, Vol. 21, No. 2, pp. 175-190.
George, G. (2005). Slack Resources and the Performance of Privately
Held Firms. Academy of Management Journal, Vol. 48, No. 4, pp.
661-676.
Mishina, Y., Pollock, T. G., and Porac, J. F. (2004). Are More Resources
Always Better for Growth? Resource Stickiness in Market and
Product Expansion. Strategic Management Journal, Vol. 25, No.
12, pp. 1179-1197.
Rafferty, A. M., Clarke, S. P., Ball, J., James, P., McKee, M., and Aiken,
L. H. (2007). Outcomes of Variation in Hospital Nurse Staffing in
English Hospitals: Cross-Sectional Analysis of Survey Data and
Discharge Records. International journal of nursing studies, Vol.
44, No. 2, pp. 175-182.
Rochefort, C. M., and Clarke, S. P. (2010). Nurses' Work Environment,
Care Rationing, Job Outcomes, and Quality of Care on Neonatal
Units. Journal of advanced nursing, Vol. 66, pp. 2213-2224.
Tan, J. (2003). Curvilinear Relationship between Organizational Slack
and Firm Performance: Evidence from Chinese State Enterprises.
European Management journal, Vol. 21, No. 6, pp. 740-749.
18

Tran, T. B., and Davis, S. R. (2011). A Quantitative Approach to Measure
Organisation Performance. International journal of social science
and humanity, Vol. 1, No. 4, pp. 289-293.
Tran, T. B., and Davis, S. R. (2012a). Effects of Sizes on Performance in
Organisations - a Quantitative Approach. [Accepted]. World Review
of Business Research, pp.
Tran, T. B., and Davis, S. R. (2012b). Examining Group Performance
Using a Markov Chain Model. [Accepted]. The International Journal
of Knowledge, Culture, and Change Management, pp.
Treville, S. d., and Antonakis, J. (2006). Could Lean Production Job
Design Be Intrinsically Motivating? Contextual, Configurational, and
Levels-of-Analysis Issues. Journal of Operations Management, Vol.
24, No. 2, pp. 99-123.
Vecchio, R. P., and Sussmann, M. (1981). Staffing Sufficiency and Job
Enrichment: Support for an Optimal Level Theory. Journal of
Occupational Behaviour, Vol. 2, No. 3, pp. 177-187.
Welbourne, T. M., Neck, H. M., and Meyer, G. D. (1999). Human
Resource Slack and Venture Growth: An Exploratory Analysis of
Growing Employees at a Faster Rate Than Sales. In P. D.
Reynolds, W. D. Bygrave, S. Manigart, C. M. Mason, G. D. Meyer,
H. J. Sapienza and K. G. Shaver (Eds.), Frontiers of
Entrepreneurship Research 1999 (pp. 480-490). Babson Park, MA:
Babson College.
19

Wicker, A. W. (1984). An Introduction to Ecological Psychology.
Cambridge: Cambridge University Press.

You might also like