203

WHAT IS PRODUCTIVITY?

Mats BJORKMAN

Linkbping University, Dept of Mechanical Engineering

Div. Of Assembly Technology, Linkbping, SWEDEN

Kevwords: Productivity, definition, evaluation, comparison .

Abstract: There exists a widespread awarene ~ s of the importance of high productivity :wd tb a t
productivity must be one of the most important design criteria of the production systems It is
important to realize that proouctivity is not a uniform universal concept; there exist man y di ffe-
rent definitions that are suitable under different conditions . Productivity is n0n11ally defined as
the ratio between the output of production and the input of production factors/mean s. It i ~i
necessary to use general measurements of the input of production factors/m e ans and of the Oll t -
put of the production system as variables in order to achieve a more general definition of pro -
d uctivity. The onl~J truly objective way to measure the production value of a product is its price
o n an open free ideal m arket. To produce something that there is no demand for is not pI.:>d uc -
tive, no matter how rational and efficient the production system seems to be.

INTRODUCTION

There exists a widespread awareness of the importance of high productivity :lI1d that prodl! c ti -
vity must be one of the most important design criteria of the production systems . Produ c tivit y
can also be used as a meas-ure \vhen the performance of a total company is evaluated and as a
tool when comparing different companies or production systcms [Craig, I brrisl lDc(~arIll() ,
Black, Kohser].

But what is productivity? How is it defined ancllww is it best JlICa~lIlcd? Tllis p;\pcr f ll\..· \ iSl' S (lIi
the importance of us ing a correct definition of the productivity concepi.

Certain definitions are suitable under certain conditions and it is important to us :: a defill itioll
that is suitable for the specific situation. Even if YOll llse the best possible avaibble COIl'jlu [c l"i -
zed information, control, accounting and decision support systC'ms, the best production syQ e l11
can be misjudged as non -productive if i1l1 unsuitable definition of productivity is use d . It is
especially important to use a suitable de finition \"hen comparing a new highly autom a ted flc .'\i -
ble produ c tion system with production systems of a different nature and concept. Thi s is the
ca~e when comparing flexible 3utom:ltic assembl y systems with traditional manual assemb ly
sy stems or ded icated as sembly maciline s/systems. The use of a traditional, but uns ui t:1 blc
definition may unjustly favour a traditional production system.

It is of utmost importance that the des igners and users of computerized systems for autc>m atic
control for productivity have a sufficient knowledge of the theoretic al background o f th e
productivity concept. To use the concept in Cl wrong manner can do much harrn. It is o ften
belter not to use it at all than to use it in a wrong manner. It is better to realize that you d on 't
really know the prod uc tivity status of your production system, than to believe that y ou kn ow it
when you really don't! There is a risk of being misled by inconect information that you t.rust to
be correct.
 204

PRoDucrrVITY IS NOT A UNIFORM UNIVERSAL CONCEPT

It is important to realize that productivity is r.ot a uniform universal concept; there exist .nany
different definitions that are suitable under different conditions. This leads to a lot of confu sion
concerning the concept of productivity and it also leads to difficulties and problems when
productivity is used as a practical tool or measurement.

The difficulties and problems have their origin in the following :

• Prod uctivity is often not defined by the users; the users take it for granted that the ir
definition is the "correct uniform universal definition". They take it for granted that
all the other users use the same definition.

• Different definitions are used in p,lfallel without the users being avv'are of it. The
users are furthennore not aware of the consequence of this di screpancy.

• Defil\itions are used in an unsuitable or improper manner. The users use cen,lin
specified definitions as a routine, without analysing the suitability of the definitions
in the specific situation.

• Bad quality of the input data limits the usefulness of theoretically correct
definitions. Deficiencies in the accounting, information and control systcms of the
companies limit the possibility to utilize theoretically corrcct definitions,

DIFFERENT TYPES OF DEFINITIONS

Productivity can be defined verbally or mathematically . The verbal definitions can play all im -
portant role as strategies or goals for the companies. This has been the case in many succes \fu]
Japanese companies that have used the definition from the 1clpan Productivity Centre (JPC),
lHelling].
''Above all else , productivity is considered as an attitude of mind" in this definition;
• "Produ ctivity is the m en/ahty of progress, of the constant improvement of that
which exists,
• Productivity is the certainty of being able to do beaer today tlUln yesterday, and less
well than tomorrow .
• Productivity is the will to improve on the present situation, no matter how good it
may seem, no matter how good it may really be.
• Productiviry is the constant adoption of L'Conomic and social life (0 changing
conditions.
• Productivity is the continual efforts to apply new techniques and new methods.
• Productivity is thefailh in human progress".

This definition has become a guide-line for the Japanese indu stry. l11e definiti on empha sizes
the importance of continuous improvement (kaizen) and the faith in the capability of ma n,
l\'othing is fixed and constant; it is always possible to improve on your own perform ance,
Furthermore, the environment of the company will cenainly change so the company has to
adap t accordingly in order to survive. 'The survival of [he finest" does not mean the SUfviv3.1 of
the strongest or largest, but the survival of the one who is able to adapt most effectively.

An interesting point is th at this definition derives from a conference arranged by the Eu ropean
Productivity Agency held in Rome in 1958. This definition has not had the same massi ve
impact in Europe as in Jap~n. The definition has mainly become a "paper product" in Europe
while it has become a way of living in J~lr a ncse indu stry.
 205

MATHEMATICAL DEFINITIONS

There is a need for more concrete definitions in order to be able to measure prcxiuctivity; there is
a need for mathematical definitions. These definitions should relate to the verbal definition and
consider the issues that are emphasized by this definition.

Productivity is norrnally defined as the ratio between the output of production and the input of
factors/means of production [Hallgren] [Son, Park]. Common definitions are, for example:

• Productivity
= ~g r::ady-ma~e products
kg raw material (1)

• Productivity
= ~umb_er
of ready-made produ~ts
Number of workers/operators (2)

• PrOductivity
= Num~er ~.f
ready-made products
Number of mall-hours (3)

• Productivity = __Numb~I__Q.Lma!J- h..ill!I..L__

Number of ready-made products (4)

• Productivity = Investment cost

Production value (5)

The problem is to choose relevant measures of the output of the proJuctiOll anJ the ill put of
factors/means of prcxiuction. There exists no definition that is best in all situations, the specific
situation has to be analysed in order to choose the most suitable definition.

SUITABILITY FOR GENERAL EVALUATIONS AND COMPARISONS

None of the above mentioned definitions are suitable [or general evaluations <1nd comparisoils.
They are normally specific to certain companies, production systems or products. These de fi 11 j .
lions are suitable to evaluate and control the development of an activity; its evolution. Tile
definition (3) number of ready-made products per man-hour is, for example, suitable for mell-
suring the development of the productivity of a specific manual production system. The same
definition is on the other hand unsuitable when a manual system is going to be compared with
an automatic production system. If the definition is llsed as a guide-line it can lead to an invest-
ment in an automatic system that increases the output per man-hour but at the same time also
increases the production cost per product unit. The definition (3) is obviously unsuitable to use
as an evaluation measurement in this situation because of the different nature of the main
production factors/means of the two alternative types of system.

The productivity definition (3) is very common and very useful, but it must be used with
caution. When comparing companies from different countries is it essential to realize that the
main production means may differ. Some countries may be abundant in relatively cheap and
capable workers. It makes no sense to invest heavily in machines just in order to increase the
number of ready-made products per man-hour in this type of country. In other countries the
work force cost is very high which makes it necessary to invest in automatization. It is essential
that the most effective means of production is used in every situation. Of course there may be
other valid reasons to invest in new technology in countries with low work force costs. It is for
example often imponant to increase the level of technology know-how in the industry.

Another example of an unsuitable use ef a productivity definition was publicised in an article in

a large Swedish newspaper. It was an important article about the productivity in Swedish
industry. The article contained the following tlble and the articie was to a large extent based on
this;
 206

NUMMI SAAB
in Fremonl, USA in TrollMltan, Sweden

Number of 3 000 blue collar 4 000 blue collar

employees 500 white collar 600 white collar

Production 890 cars/day 390 cars/day

(Toyota Corolla, (Saab 900,
Chevrolet Geo) Sa ab 9000)

Productivity 19 man-hours/car 60 man-hours/car

-
Table I: Two car factories - a comparison of productivity, [Dagens Nyheter].

Before trying to interpret the above table it is important to first analyze if the outputs of produc-
tion and inputs of factors/means of production of the two factories are comparable. There exist
a lot questions that have to be asked and answered: Is the degree of automatization
approximately the same? Are the produced cars of the approximate same type and degree of
complexity? Is the degree of pre-production and pre-assembly of the cars approximately the
same? Are the cars competing on the same market segment; the same type of customers? Etc.

None of these questions were asked in the article and the author just assumed that the produc-
tivity figures were comparable. This is not an uncommon mistake, not only among economical
journalists, and it can lead to \\-Tong decisions being made.

A MORE GENERAL DEFINITION

There is obviously a need for a more general definition suitable even if the activity is 110t jllt't
undergoing an evolution but a revolution, for example from a marlllal to an automatic pwduc-
tion system. It is necessary lO use more gcneral measurements of the input or the productioll
factors/means and of the production output of the system as variables in order to achieve Ihis
more general defInition of productivity.

The production output can be measured in many different wnys. The prncessing value is lll1C
way to try to create a more objective measurcment of the outJ1ut. The definition Illay get tlle
following fonnula;

.. Processing value
• Prod uetlVlty = .
Material or Time or Capital cost (6)
The processing value can be defIned as;

• Processing value = Market value/revenues - Direct material (7)

The factors/means of production must also be measured in a unit that is more universal than in
tbe definitions (1) - (5). It should for example be possible to compare manual assembly systems
with flexible automatic systems in a just manner. No type of system should be favoured in a
unjust manner. The total cost is often suitable to use for this purpose. It often makes it possible
to compare production systems of different nature;
 207

. ~ Processing value
• Prod UCUvlty =- T al
ot cost (8)

The use of the total cost is not always suitable. This is the case if the productivity of a subsys-
tem and not the product's total production system is going to be evaluated. In this situation it is
better to use the processing cost of the subsystem.

. .
• P rod ucn vlty = Processing-'='---
.--~.
value
Processmg cost (9)
The processing COSt can be defined as;
Processing cost = Total cost - Direct material
This results in the following definition;

• Productivity = ;\1arkeLYaLu.ro.~).les - Dir~c.t material.

. Total cost - Direct material (IO)

In this definition (10) the difference in nature of the main production factors/means of manual
and automatic production systems is considered in a better manner. The definition measures the
utilization of given resources in the production system. The possible difference in revenues
generated from the systems is also considered. A technologically more advanced system will
sometimes make it possible to produce products with a superior quality. If this superior quality
leads to an increase in revenues this increase must be taken into account when calculating the
productivity of the system.

The total cost includes all the costs that are tied to the production of the product, not only the
direct production costs but also indirect costs that emanate from the product. It is important that
the differences in useful lifetime and/or remaining value of a flexible production system such as
a robot installation and a non-flexible dedicated machine are considered when the total cos t is
calculated. The relatively high investment cost of a flexible robot installation must be distributed
over the total useful lifetime of the installation . This can be done llsing the net present value and
the annuity rate methods.

If the purpose of the evaluation is to measure the total productivity of the production of a
product, the total cost may be used instead of the processing cost as in definition (R ). The
choice of direct material can be considered as a part of the total productivity. It is morc produc-
tive to use a less expensive material under the condition that the other parameters are not
affected by the choice of material.

Perhaps the most controversial variable in the definitions (R) - (10) is the market
value/revenues. The only truly objective way to measure the pnxluctillll value uf a proclu ct i~ il s
price on an open free ideal market. The value of a product is the price the market is prepared to
pay! To produce something that there is no demand for is not productive. no matter how ratio-
nal and efficient the production seems to be. The purpose of the producti(.ln is to sa tisfy the
market demands. Mentally it is vital not to dc-couple the proouction from the market.

Henry Ford was almost obsessed with eliminatillg waste antI "llIl1WCCS.WIf)!" cxpemiitlllCS . lie
was concentrated on the production costs, which led to a very high productivity ill the Ford
factories during most of the period that the Ford model T was produced. The problem was that
the long term product development was neglected. The production of the T -ford was extremel y
efficient even during the last years of production . But it was not productive enough, in spi le o l
this efficiency. The market preferences changed and the customers demanded more modern
cars. A low price was no longer the dominating competitive mean on the market. Henry Ford
did not realize this in time, he concentrated on the denominator in the definitions (8) and Cl 0)
when he tried to increase the productivity. General Motors on the other hand tried to improve
both the denominator and the numerator by offering modern cars to reasonable prices . The
result of these two different policies was that Ford was passed by GM as a car manufacturer.
 208

DON'T REGARD PRODUCTIVITY FROM A TOO NARROW PERSPECDVE

There is a risk that productivity is regarded from a too narrow perspective. Cost reduction is
often a faster way of achieving higher productivity than efforts that increase the market compe-
titivness. Investments in a new production technology will for example often lead to higher
initial costs that decrease productivity from a short term point of view, but increase it in the long
term. When calculating productivity of a flexible production system such as a flexible robot
installation, the high investment costs have to be distributed over a longer period of time.

Market dominating companies or companies with a monopoly will often concentrate on the
revenues. If the customers don't have any choice but to buy the company's products there is a
great risk that the company regards it easier to increase the revenues instead of decreasing th e
costs. The result of this is an apparently high productivity . nut history shows that su c h a
company is extremely vulnerable, if competitors suddenly appear on the market.

There exists no significant contradiction between productivity and quality since it is not produc-
tive to produce non-competitive products with a low quality. These kinds of products norl11 i1 ll y
generate low revenues on a long term basis, even if it is possible to increase short tel1n produc-
tivity by decreasing short term costs through a decrease in quality specifications. Furthermore.
high quality is often a way to reduce the true costs . High quality may be a way to achieve both
an increase in revenues and a decrease in costs, which both contribute to a higher produclivity
[Deming] [Sandholm) . This is an important aspect of productivity as new technology often
contributes to a more consistent quality.

A FREE MARKET - A CONDmON FOR HIGH PRODUCTIVHY

An open stiff competition is normally a condition for achieving high productivity . If this com-
petition is restricted the productivity will suffer. The pressure from the market and the competi-
tors is a condition for the right "attitude of mind" that is described in the verbal productivity
definition from JPc.

Without an open free market the productivity will ultimately decrease. An example of this is
what was experienced by the former communistic countries in Europe. The "attitude of mind"
suffer that much that the productivity ultimately decreases even if the definitio ns (1) - (5) are
used to measure the productivity.

There can be political and sc>ciological reasons for restricting the free market. When introducing
such restrictions you have to realize and analyze the long term effects on the productivity. What
will the effects of the loss of productivity lead to, are we prepared to pay the price for this loss?
Countries that protect their domestic industry from international competitors wiJi have grea t
difficulties in achieving an industry that is tnlly competitive on the world market.

There exist a few examples in the world where extremely competent domestic competitors m ake
it very difficult for foreign competitors to compete on the domestic market. This is for example
the case on the Japanese car market and on the Swedish market for heavy trucks. It is important
to realize that this domination by domestic producers is mainly due to the stiff competition
between these highly competitive producers and not due to governmental protection. The se
domestic producers are also very competitive on the world market.

All businessmen love the free open market. There just seems to be one exception; their own
market segment! So they often tries to restrict it by monopolizing it, creating trusts or to put
pressure on the government to restrict it, etc. A recent example is from the car industry , some
car manufacturers accuse the Japanese to protect their car market. The main evidence of thi s
alleged protection seems to be that these car manufacturers are not able to sell cars in any num-
bers in Japan. Some argues that this is a reaso n for their countries to protect their domestic car
markets . These m anufacturers arc mainly the same that have traditionally relied on a big
domestic market and have had difficulties to compete on the world market.
 .. 209

QUANTIFICATION PROBLEMS

A practical problem concerning the proposed definition is the difficulties to quantify the
variables. The true costs and revenues of a product or production system are often not known
to the company. The accounting, infonnation, control and decision support systems are often
not designed for this purpose. It is usually much easier to extract data about number of man-
hours, units, workers/operators, etc. But there is often reason to question even these data. Il is,
for example, not uncommon that the true number of man-hours for a product are much high er
than the number the infonnation system gives you [Bjorkman, Ekdahl].

A lot of companies are introducing new types of accounting and illfom1ation systems/methods
that satisfy the demand of true costs and revenues of a product or production system. The best
known system/method of this kind is perhaps the Activity Based Costing (ABC) method
[Kaplan]. But there are a lot of work left to be done in both the field of research and practical
implementation.

The differences in routines and accounting, infonnation and control systems make it difficult to
make comparisons between different companies. This is valid even if the same definitions of
the productivity concept are used.

The difficulty to quantify revenues and costs of a product is not a reason to use theoretically
unsuitable definitions. Unsuitable definitions may give the wrong signals to the management . Ir
the management is not aware of the limitations of the definitions it may lead to incolTect deci -
sions. The practical quantification problems are often possible to overcome or at least red lice to
a manageable level.

The quantification problems probably make it necessary to lIse computer based accounting.
information, control and decision support methods and systems, if the production is not of a
very limited complexity.

CONCLUSION

There exists no "correct cmiform universal definition" of the concept of productivity; there exist
many different definitions that are suitable under different conditions. It is necessary to analyse
the purpose of using the concept and to choose a definition that is theoretically and practically
suitable for the specific situation.

Productivity can be defined verbally or mathematically . The verbal definitions can pl ay an

important role as strategies or goals for the companies. This has been the case in many SLl ccess -
ful Japanese companies that have used the definition from the Japan Productivity Centre.
"Above all else, productivity is considered as an attitude of mind" in this definition [HellingJ.

Productivity is mathematically normally defined as the ratio between the output of production
and the input of factors/means of production. The problem is to choose relevant measures of the
production output and the input of production factors/means. There eXIsts no definition that is
best in all situations, each specific situation has to be analysed so that the most suitable defini-
tion can be choosen.

The most common definitions are often not suitable for general evaluations and comparisons.
NonnaIly, they are specific to certain companies, production systems or produc ts . These defi-
nitions are suitable to use when evaluating and sontrolling the development of an activity: it s
evolution.

There is also a need for a more general definition which is suitable to use even if the activity is
not just undergoing an evolution but a revolution , for example when replacing a mant:al
system by an automatic production system. It is neces sary to use general measurements of the
input of production factors/means and of the output of the production system as variables in
 210

order to achieve amore general definition of productivity. The definitions (8) - (10) are atte mp ~s
to create such definitions.

In the definition (10) the difference in nature of the main production factors/means of manual
and automatic production systems is considered. The definition measures the utilizatio n of
given resources in the production system. The possible difference in revenues generated from
the systems is also considered.

The total cost includes all the costs that are tied to the production of the product, not only the
direct production costs, but also indirect costs that emanate from the product.

If the purpose of the evaluation is to measure the total productivity of the production of a
product, the total cost may be used instead of the processing cost as in definition (8). The
choice of direct material can be considered as a part of the total productivity. It is more produc-
tive to use a less expensive material under the condition that the other parameters are not
affected by the choice of material.

Perhaps the most controversial variable in the definition (10) is the market value/revenues. The
only truly objective way to measure the production value of a product is its price on an open
free ideal market. The value of a product is the price the market is prepared to pay! To produce
something that there is no demand for is not productive, no matter how rational and efficient
the production seems to be. The purpose of the production is to satisfy the market demands.
Mentally it is vital not to de-couple the production and the productivity from the market.

The difficulty to quantify revenues and costs of a product is not a reason to llse theoreticall y un-
suitable definitions. Unsuitable definitions may give the wrong signals It) the management. [I
the management is not aware of the limitations of the definitions it may lead to incorrect
decisions.

REFERENCES

Bjorkman, M., Ekdahl, B.: En utviirdering av ettflexibelt alltomatiskt mOflterillgss),stcm vid

AB ROBO i Linkoping, LiTH-IKP-R-540.
Craig, C. E., Harris, R. C.,: Total Productivity Measurment at the Firm Le vel , Slpa iJji
Management Review, Spring 1973.
Dagens Nyheter, the 27 of February, 1992
DeGarmo, E. P., Black, J. T., Kohser, R. A.: Materials alld Processes in Manufacturing,
Sixth Edition, Macmillan Publishing Company, New York, 1984.
Deming, W. E.,: Quality; Productivity, and Competitive Position, Massachusetts Institute of
Technology, ~entre for Advanced Engineering Study, Cambridge, MA 02139,1982.
Hallgren, 0.: Finansiell metodik, Studentlitteratur, Lund 1977, ISBN 91-44-40096-9.
Helling, J.: Viirldsmiistarna, Seilin & Partner forlag AB,1991, ISBN 91-7055-036-0
Kaplan, R.: Measuring Manufacturing Performance, A New Challenge for Managerial
Account Research , The Accountin g Review LVIII, October 1983.
Richardson, P. R., Gordon, J. R. M.,: Measuring Total Manufacturing Performanc e,
Sloans Manae;ement Review, Winter 1980.
Sandholm, L.: Kvalitetsstyrning, ISBN 91-44-17182-x.
Son, Y., Park, c.: Economic Measure of Productivity. Quality and Flexibility in Advall ced
Manufacturing Systems , Journal of Manufacturing Systems, Vol. 6, No. 3, 1987.