Information technology is a key driver in the globalization and growth of the world economy. The total worldwide package software market has been estimated at $135 billion . Worldwide expenditures on software are expected to increase to about $220 billion by the year 2002 . The U.S. software industry is reaping the benefits of this hyper growth, having captured 70% of global software sales. However, the specter of piracy haunts the U.S. software industry and other digital mediums because it diminishes revenues and retards investment in R&D, the lifeblood of the high technology endeavors.
While piracy causes significant losses in the U.S. (estimated at $2.8 billion for business application software in 1997), it is particularly problematic internationally. According to the Software Publishers Association (SPA), the worldwide revenues of business-based PC applications was $17.2 billion, but global revenue losses due to piracy in the business application software market were calculated at $11.4 billion . The piracy rates in several countries exceed or approach 90% (for example, Kuwait, Pakistan, Indonesia, Bulgaria, China, Turkey, Egypt, Bolivia, and El Salvador).
Software has the distinctive characteristic of digital goodsit is expensive to produce for the first copy (high-fixed costs) and inexpensive to reproduce and distribute for subsequent copies (very low, approaching zero, variable costs). It exhibits the classic characteristics of a public good in that sharing with others does not reduce the consumption utility of the product. These idiosyncratic traits of software and related digital products have facilitated their illegal distribution worldwide. Several countries in the world are referred to as "one-legitimate-copy" countries, where one legal copy of a software package is sufficient to meet the demands of an entire nation. The advent of the Internet has amplified the problem. There are numerous warez1 Web sites that permitcustomers to select from a menu of pirated software up to the 650MB capacity of a CD. For around $30$60, payable via credit card, the warez vendor will burn the CD and deliver it to the customers doorstep.2
Global Piracy Campaign
Industry groups such as the SPA and the Business Software Alliance (BSA) have spearheaded the campaign against global software piracy. Since 1984, the 1,200 member SPA has taken a leading role in combating global software piracy by lobbying the U.S. Trade Representative's (USTR) office to enforce and protect intellectual property rights, and by engaging in domestic and international legal and education activities to achieve compliance and fight software piracy.
The SPA3 prepares an annual list for the USTR of the nations that deny adequate and effective protection of intellectual property rights. The USTR solicits input from industry groups, Congressional leaders, U.S. embassies, and trading partners in the preparation of a yearly report on intellectual property rights violations. The USTR has both domestic and international legal authority to instigate investigations and to call for trade sanctions against flagrant offenders of intellectual property. The SPA has been very effective in its efforts to lobby the USTR. As illustrated in Tables 1 and 2, many of the countries cited by the SPA as problem countries have also been targeted by USTR as part of Special 301 provision of the Omnibus Trade and Competitiveness Act of 1988. As is typical in most political processes, the Special 301 provision is used in a variety of ways to engender countries to act according to U.S. interests to protect intellectual property, but it is also used as a way to prod and reward countries for taking steps to promote other trade-related activities.
In the fight against piracy, the legislative and educational weapons may win a few battles, but the overall war against piracy cannot be won without addressing the current draconian pricing policies.
The growing importance of software piracy and the enormous resources at stake have recently spurred research on the behavioral and economic understandings of software piracy activity (see, for example  and ). Studies have reported that females pirate less, older individuals (as opposed to younger college students) pirate less, and that individuals with an ethical predisposition towards legal justice (a primarily western notion; less important in the moral makeup of the eastern cultures) tend to pirate less. Key economic findings related to software piracy include:
- Deterrent controls in contrast to preventive controls result in higher profits to software publishers and higher levels of the welfare function. Deterrent controls include government-to-government negotiations, educational campaigns, and legal activity related to expanding domestic copyright laws and seeking to enforce those laws. Preventive controls include software and hardware schemes to prevent the actual copying of the software. Few, if any, publishers still employ such preventive controls.
- The price of the software is inversely related to the degree of piracy.
- The incentive for governments to enact and enforce copyright laws is related to the size of the domestic software industry.
Many countries, particularly those without a viable local software industry, find it counterproductive to thwart piracy. Software piracy allows them to quickly and inexpensively acquire the latest software products, and keep up-to-date with the advancing technologies. While the threats of trade sanctions have arm-twisted many countries to enact copyright laws, they are rarely enforced. Even when enforced, many suspect the realpolitik is not in safeguarding the intellectual property rights of the software publishers. China, for example, has recently cranked up raids against CD factories that manufacture bootlegged copies of software and confiscated the CD presses (each worth as much as $1 million). The suspicion, however, is these presses are put back into service in government-owned factories to manufacture CDs containing pirated software .
The Income Effect
Most software developers and also the SPA are aware that income levels can influence the ability of consumers to purchase software. However, the SPA is not convinced that per capita income is an issue as is evident by its following proclamation:
- "Although some argue that lower levels of personal income justify software piracy, this is misleading. In most developing countries, computer software is only used by a relatively small group of individuals and organizations affluent enough to purchase computers-not the average citizen. If individuals and organizations can afford to buy computer hardware, they have no excuse for pirating software. "
This issue is definitely debatable. Consider the empirical evidence in Figure 1, which correlates the national piracy rates compiled by the SPA with the per capita GNP for 65 countries in the year 1997. Higher software piracy rates are heavily skewed towards countries with low per capita GNP. The statistical regression relationship between piracy and GNP in Figure 1 is:
- Piracy rate = 77.5 - 0.0013 * per capita GNP 4
In essence, the results indicate that for every $1,000 increase in GNP there will be a 1.3% decrease in piracy rates. A closer examination of the data reveals an inflection point at about $6,000, where income levels below the inflection point exhibit a different relationship with the piracy rates. Interestingly, this inflection is around the midpoint of the income range classified by the World Bank as "upper middle income" ($3,126$9,655). The statistical analysis for the two data segments yields (see Figures 2 and 3):
- Piracy rate = 88.7 - 0.0059 * per capita GNP (less than $6,000)
- Piracy rate = 67.1 - 0.0009 * per capita GNP ($6,000 or more)
The effect of GNP is much more pronounced for the countries with GNPs less than $6,000. Each $1,000 increase in per capita GNP is associated with a nearly 6% decrease in the piracy rate. For countries with per capita GNPs greater than $6,000, an increase in the per capita GNP of $1,000 will not even yield a 1% reduction in the piracy rate. These results indicate a significant income effect on the global piracy rates, particularly in the poorer segments of the world.
We are not suggesting those policymakers in government and industry simply standby and wait for the GNP levels to rise. Our contention is that an understanding of the factors that contribute to and strengthen this correlation may assist policy makers in devising strategies to more effectively combat global software piracy. To be viable, any strategy must both improve the profitability of U.S. software publishers, and at the same time benefit the concerned countries and their citizens. To help devise such a strategy we explore a key link in the piracy-income equation: the software pricing mechanism.
A recurring theme in newspaper analyses and investigative reports on the piracy problem in developing countries is the extremely high price of legal software programs that propel individuals to purchase counterfeit software. In Mexico , for example, a copy of Microsoft Windows95 costs roughly $255 (in 1998), whereas it costs $110 in the U.S. Contrast the pricing of the software with the income differentials (per capita income in the U.S. is about eight times that in Mexico), and that a pirated version can be obtained for under $15. In Thailand (with one-tenth the income levels of the U.S.) one can buy Oracle's database system for $25, whereas the legal copy costs around $20,000 . The motivation to pirate is the significant price differential between legal and pirated versions, which is exacerbated by the relatively low income levels in these developing countries. The key problem is that prices are set at the U.S. levels, which are significantly higher than individuals in most countries can afford. In a number of countries, legitimately purchasing software is not an option that most individuals consider seriously. A software vendor in Thailand echoes this sentiment succinctly: "If I sell the real software, I wouldn't even sell 10% of what I am selling now" . According to World Bank statistics, there are approximately 69 countries in the world with a per capita GNP of less than $1,000.
In is regard, the rationale behind the unwillingness of most governments in enforcing and enacting copyright laws becomes clearer. Stricter copyright laws, coupled with high prices of legitimate software, will severely restrict software usage by a large majority of the general and even computer literate population. Hence the hesitation, despite the fact that lack of intellectual property rights impedes the growth of the domestic software industry and lowers tax revenues.
In a recent study of 17 developed countries, IT was found to have a dramatic impact on economic output and productivity . Empirical evidence suggests that IT capital stock (includes hardware, data communications, software, and services) provides an impressive 70.6% return on investment. Even if depreciation is factored in, the return on investment is still a very substantial 40%45%.5 This compares to labor productivity figures of between 1 and 2% for the 1990s. Governments are keenly aware of the enormous economic benefits of acquiring information technology. However, as long as the price of software remains high relative to local income levels, it makes economic sense to shift resources to hardware and infrastructure investments and rely on bootlegged versions for software needs where possible.
Thus, it is apparent that focusing purely on enforcing intellectual property rights will have limited success. In the fight against piracy, the legislative and educational weapons may win a few battles, but the overall war against piracy cannot be won without addressing the current draconian pricing policies. We argue the economic principles of price discrimination need to be adopted by the software publishing industry to realize significant reductions in the global piracy rates.
Price Discrimination and Piracy
It is a fundamental economic principal that the way to maximize profits is to charge a price that equates to the value of the product to each consumer, instead of a uniform price that is charged to all consumers. Price discrimination involves selling the same good at different prices to different consumers. (Figure 4 illustrates the potential additional revenues that would accrue to a monopolist practicing perfect price discrimination.) The idea is to identify each consumer's willingness-to-pay function and set prices accordingly . In reality, perfect price discrimination is difficult because it requires the effective measurement of consumer preferences, it needs a mechanism for preventing arbitrage, and the producer must have sufficient monopolistic market power . Arbitrage represents the opportunity to consumers to exploit the discrepancies in the market for profit. In particular, arbitrage is the process of buying a commodity in one market and immediately selling them in another market for a higher price.
Price discrimination is already found throughout the software publishing industry through the granting of site licenses and customer segmentation. It is a common industry practice to enter into contracts where a certain number of copies of a title are sold at a reduced cost. Similarly, students can buy copies of many popular software packages from school bookstores at steeply discounted prices. In both instances, there is some level of control related to identifying the consumer and controlling the distribution of the software. Arbitrage is not eliminated, but it is controlled via the distribution channel (you must be a company employee or you must have a student ID) and some type of product versioning based on features, capability, and support .
Surprisingly, most software publishers have not adopted price discrimination strategies based on national and geographical criteria. Pricing policies are often set with the U.S. market as the target; international markets, especially in developing countries, are but an afterthought. U.S.-based software publishers need to view the developing countries as an integral part of their consumer base, albeit as market segments where the typical reservation price is sharply lower than in the U.S. The low value of willingness-to-pay among these consumers requires pricing policies comfortably afforded by this consumer base. Otherwise, these consumers who view current pricing mechanisms as draconian will continue to practice software piracy en masse. Price discrimination presents an immediate and effective strategy to make a significant dent in the current software piracy rates. Properly implemented, it can result in substantially increased revenues to software publishers.
While a number of mechanisms of price discrimination can be implemented, perhaps the simplest approach would be to index the price of the software to the per-capita GNP level in the concerned country. National income levels broadly capture the willingness-to-pay function, and would provide a good starting point to implement the global price discrimination strategy. U.S.-based software publishers certainly possess sufficient monopolistic market power to adopt this strategy as they dominate the world's software industry. Any arbitrage issues that arise would imply illegal transactions that flow from poorer to richer countries. This should not present major additional impediments as more developed countries have stricter enforcement of copyright laws, and the tradition of legitimately purchasing software is more strongly entrenched. In any event, the disincentives for arbitrage can be substantially enhanced with increased enforcement of the copyright laws.
Governments across the world would be more willing participants in the enforcement of intellectual property rights, provided country-specific pricing policies are adopted by the U.S. software publishers. Once the legal price of the software is within the affordability range of its citizens, the government's incentives for enactment and enforcement of intellectual property rights increase substantially. The governments would pay more attention to the protection of intellectual property rights in order to accelerate the growth of the domestic software industry, and to increase the tax revenues from legal software sales.
A New Combat Strategy for Global Price Discrimination
The oft repeated mantra "piracy is wrong no matter what the price of the software," has its skeptics, especially in the poorer segments of the world. It has clearly not proven effective, as is starkly evident from the global piracy rates. The effect of price-to-income on purchase decisions, particularly when the good involves international exchanges rates, is considerable.6 This article recommends a new strategy to combat software piracy-global price discrimination. To wage an effective campaign against software piracy, it should be used as a first line of defense.
As Figure 5 illustrates, global price discrimination is a key driver that when properly implemented substantially alters the incentive mechanisms for governments and their citizens. Individuals around the world would begin to cultivate the habit of purchasing legal software as it becomes more affordable, and the climate of respecting intellectual property rights gains a foothold. Governments would begin to utilize intellectual property rights to foster the development of the local software industry, and to increase the tax revenues. U.S. software publishers would be the significant beneficiaries, and the fruits of the new strategy will soon begin to improve their bottomline.
Finally, the U.S. needs a cogent and consistent strategy for dealing with software piracy. Column five in Table 1 contains expected piracy rates for selected countries as a function of per capita GNP. Only countries with actual piracy rates one standard-deviation away from their expected piracy rates are presented. Countries such as Australia, Columbia, India, and South Africa have piracy rates significantly below their predicted values as a function of GNP. These countries should probably not be on the USTR's watch list. In contrast, countries such as Bulgaria, China, Greece, Hong Kong, Indonesia, Ireland, Paraguay, Russia, Singapore, Thailand, Turkey, and Vietnam have piracy rates that deserve closer attention. A concerted effort by the USTR to implement a piracy model based on sound economic principles will facilitate the development of consistent trade policies.
Consider this scenario repeated constantly and internationally:
- Cedric has a small business in a country with a per capita GNP of less than $1,000. Cedric is interested in purchasing the latest version of Sweet 2000 because of the advanced and very useful data management features in the product. Cedric can purchase the product locally for about $500 or he can obtain a newly burned copy from a local bookstore for about $20.
Many consumers around the world realize that piracy has a deleterious effect on the incentive to produce intellectual property and is wrong from both a legal and moral standpoint. They also realize the very positive benefits derived from having supported software. However, if the cost of the software is not close to the consumer's reservation price-the willingness-to-pay-then the warez vendor will get the sale. Price discrimination, along with improved enforcement of international copyright laws, will potentially help transform this turnip into a cash cow.
1Colloquial reference to pirated and cracked software. According to the New Hacker's Dictionary: warez /weirz/ /n./ Widely used in cracker subcultures to denote cracked version of commercial software, that is versions from which copy protection has been stripped. Hackers recognize this term but don't use it themselves. The dictionary is available at www.tuxedo.org/~esr/jargon/
3In January 1999 the SPA merged with the IIA to form the Software & Information Industry Association (SIIA). The IIA represented companies involved in creating and distributing print in digital formats.
4The sample size of 65 was a function of data availability for piracy statistics from the SPA and the per capita GNP statistics from the World Bank. The F statistics for regression equations 1, 2, 3 and the regression t-values for the regression coefficients are all significant at the .001 level. These results suggest income levels have a significant impact on piracy rates.
5The returns for less developed countries could be lower. There is the issue of technological sophistication and readiness and the ability to translate investments into increased levels of GNP. A certain level of infrastructurehardware, data communications networks and the availability of know-howis probably necessary to realize the same level of returns. Software and infrastructure are complementary and interdependent components of IT productivity.
6Price discrimination and software piracy could be viewed as a transfer of payments to less developed countries, similar to foreign aid. Given that the current level of piracy for business application software is about $12 billion and the U.S. publishers deliver 70% of it, then the U.S. publishers are in effect donating about $8 billion to the world.
©2000 ACM 0002-0782/00/0900 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2000 ACM, Inc.