P1: IML/FFX P2: IML/FFX QC: IML/FFX T1: IML
Software ̇Piracy ̇OLE ̇revised WL040/Bidgolio-Vol I WL040-Sample.cls June 20, 2003 13:8 Char Count= 0
SCOPE ANDIMPACT OFPIRACY 301the exception of Eastern Europe, all of the countries with
piracy rates over 70% are Third World countries. In most
of these countries there is little indigenous software pro-
duction, and so for them piracy has a net economic benefit
without adversely impacting their own local industry. In
this regard it is noteworthy that India, which has become
a major player in the software industry, reduced its piracy
rate significantly, to a low of 61% in 1999 from 79% in 1994
(although the rate has climbed again since then, to 70%
in 2001). China’s piracy rate, on the other hand, declined
only slightly, to 92% in 2001 from 97% in 1994. China
was seeking World Trade Organization membership dur-
ing this period and was consequently under pressure to
improve its protection of intellectual property rights, but
its efforts in this direction evidently had little effect.
Eastern Europe has consistently had the highest piracy
rate of the six regions of the IPR study for each of the
years shown in Figure 1. However, if we exclude Rus-
sia and most other countries of the former Soviet Union,
the rest of the Eastern European countries (including the
Baltic States) have made substantial progress recently,
reducing their software piracy rates by at least 15 percent-
age points over the period 1994–2001. Their current piracy
rates still have room for improvement, ranging from a
high of 75% (Bulgaria and Romania) to a low of 43% (the
Czech Republic), and averaging about 54%. The large re-
ductions in piracy rates in these countries probably reflect
the emergence of free-market economies, a process that
is normally accompanied by improvements in protection
of intellectual property (Software and Information Indus-
try Association, 2000). This transition has apparently not
happened in the countries of the former Soviet Union,
which had a combined piracy rate of 87% in 2001, only a
modest decline from 95% in 1994.Methodology of the Study
In order to calculate the piracy rate for a given country or
region, the IPR needs to determine the number of pirated
copies and the number of legitimately purchased copies
of software applications in use in that region. The IPR
has access to proprietary sales information from the BSA
member companies, and so the figures for legitimate soft-
ware sales are readily obtained. The number of pirated
copies is inferred as the difference between this number
and the “demand” for software, i.e., the number of appli-
cation programs that one would expect to be loaded onto
the computers in use. The IPR uses hardware sales data
to determine the number of new machines sold each year.
A correction is applied for turnover as older machines are
taken out of service and replaced by new ones, to obtain
an estimate of the number of machines currently in use.
From this figure the demand is calculated based on the
average number of application programs that would nor-
mally be installed on each computer. The IPR calculation
of demand takes into account differences between home
and business computers and different levels of technologi-
cal development in different regions, as well as differences
among the various categories of software.
It is important to keep in mind that much software
is excluded from the IPR estimates. Only packaged (as
opposed to custom or in-house developed) software isanalyzed, and only packaged business software applica-
tions are included in the final piracy figures. According to
PricewaterhouseCoopers (1999), packaged business soft-
ware currently represents less than 20% of all software
revenue. Freeware and shareware also escape the analy-
sis because there are no sales records for them, so they will
be counted among the pirated applications. Freeware is
software that its producer makes available free of charge,
and shareware is software that is distributed for free with
a request that users who like it should send the author
a contribution. At present freeware and shareware repre-
sent an insignificant fraction of the market.
Despite all of the uncertainties involved, the IPR piracy
estimates are probably the best obtainable under the cir-
cumstances. Marron and Steel (2000) performed a regres-
sion of the IPR estimated piracy rates on an independently
estimated measure of patent protection in various coun-
tries and found a strong correlation, as would be expected.
This test gives some confidence in the basic validity of the
IPR data.Financial Impact of Piracy
The estimates of the dollar value of the revenue losses to
piracy shown in Figure 2 were obtained simply by multi-
plying the number of pirated copies by the average market
price of a copy. The software categories not included in
the IPR figures probably have different piracy rates from
those for packaged business software. For instance, the
study by Cheng et al. (1997) suggests that piracy rates for
game software are probably much higher than those for
productivity applications such as word processors. In any
case it is likely that piracy in the excluded categories rep-
resents a large additional amount of lost revenue. Thus,
the calculated losses are probably a considerable under-
estimate of the total revenue lost to piracy by the software
industry. At any rate, it is reasonable to assume that the ra-
tios of the revenue losses in Figure 2 represent the relative
financial impacts of piracy in different regions.
Of course, calling the revenues in Figure 2 “losses” im-
plicitly assumes that all software now pirated would be
purchased through legitimate channels at current prices
if all piracy were stopped. But it is likely that some users
who currently pirate software would choose not to pur-
chase the software at all, if piracy were somehow made
impossible. Also, very probably prices would change in
such an altered market. Thus it would be more precise to
refer to this quantity as the “market value” of the pirated
software.
Economic models have been developed to consider
the overall effect of illicit copying on software produc-
ers’ revenues. Slive and Bernhardt (1998), among others,
describe how, in some situations, piracy of a software
product can actually increase the total profits of its manu-
facturer, through what economists call “network external-
ities.” In essence, the value of a particular software prod-
uct is increased by having a large community of users. For
instance, the users enjoy the convenience of being able to
interchange files in the format used by the application.
Also, they may invest considerable time in learning to use
the application, and can then move more easily to a new
employer where the same software is in use than to one