By Wesley D. Markham*

A pdf version of this article may be downloaded here.

I.  Introduction

This paper takes an empirical approach to a policy-based question: how long should patents last in the United States, especially given changes in the international patent law regime?

The overarching, even constitutionalized, policy behind the U.S.’s patent system is to promote the progress of science and the useful arts.[FN1] This is a laudable goal, but the devil is in the details.  Utilizing an intellectual property regime to maximize innovation requires a delicate balancing act.  On the one hand, inventors and the firms for which they work need an incentive to innovate.  In the United States, one such incentive is a limited monopoly in the form of patents for new and useful inventions.  On the other hand, every patent takes something away from the public domain, thereby making it more difficult for others to build on prior discoveries.  In other words, patents both encourage and discourage innovation.  The key to a successful patent system is determining the correct balancing point, i.e., the smallest reward necessary to spur inventors to invent.

The term of patent protection embodies a primary variable in this balancing act.  A relatively long patent term greatly rewards an inventor, but it also imposes a high burden on others who wish to use or improve upon patented technology.  Under both U.S. and international law, patent terms generally run for approximately twenty years.[FN2]

But while U.S. patent terms have remained fairly – though not entirely – stable for the last fifteen years, international patents have become increasingly lucrative.  To remain relevant, patent law in the U.S. must adapt to this patent globalization, in part by recalibrating the duration of the patent monopoly.  I tackle this issue by asking whether we should reduce the patent term in the U.S. to compensate for the enhanced potential for patent exclusivity outside the U.S.  Specifically, I developed a new metric, termed “global patent term” (GPT), to assess whether the average “reward per invention” has significantly increased over the last fifteen years given an expanding global patent regime.  If it has become relatively easy for firms to secure a high level of protection for each invention, e.g., by receiving multiple patents on the same invention in many different countries, the “reward per invention” is high.  In a well-functioning patent system, a high “reward per invention” should result in increased innovation.  If the “reward per invention” has increased but the level of innovation has remained stagnant or decreased, then we might conclude that the current level of patent protection is too high and should respond by reducing the patent term in the United States.  A shorter patent term would allow patented inventions to fall into the public domain more quickly, enabling others to take full advantage of the patented technology.

Furthermore, I conducted this analysis on three firms in three very different industries – Pfizer, the pharmaceutical giant; International Paper Company, a worldwide leader in paper products; and UNISYS, a large technology services provider.  These three case studies indicate that patent globalization discriminates based on technology, i.e., rewards firms in some industries but not others.  Given my results, I conclude that we should seriously consider implementing technology-specific patent terms in the U.S.  For industries in which patent globalization has made a real difference, such as the pharmaceutical industry, the increased “reward per invention” outside the U.S. should offset any innovation incentives lost by shortening the duration of U.S. patents.  For industries that depend on patent exclusivity in the U.S. but not elsewhere, a longer U.S. patent term may provide the optimum innovation incentives.

II.  A Short History of the Patent Term in the United States & Abroad

A. The U.S. Story

The following history of the patent term in the U.S. serves to illuminate two important points.  First, the chosen length of patent protection in this country is, and has always been, rather arbitrary, a product of compromise and historical accident instead of sound economic analysis.  Second, as a practical matter, any proposal that might be interpreted as weakening the patent system in the U.S., e.g., an attempt to shorten the patent term, is likely to meet serious political resistance.[FN3]

Like so much of American law, the original duration of patent protection in the United States was borrowed from England.[FN4] The Patent Act of 1790 provided that the exclusive rights associated with a patent shall last “for any term not exceeding fourteen years.”[FN5] In 1836, a revised Patent Act made available a seven year patent term extension, above and beyond the original fourteen year term, in certain circumstances.[FN6] This lasted until 1861, at which time Congress declared that “[a]ll patents hereafter granted shall remain in force for the term of seventeen years from the date of issue; and all extension of such patents is hereby prohibited.”[FN7] One commentator proposes that the change from a fourteen year patent term to a seventeen year patent term was made, in part, because extensions were so common under the old “fourteen plus seven” system.[FN8] Yet another opines that Congress selected the seventeen year term because seventeen is “the number midway between 14 and 21.”[FN9]

Between 1790 and 1875, Congress also passed approximately seventy-five private bills extending the patent terms of particular inventors who successfully lobbied Congress that the generally applicable statutory term had not afforded time to generate sufficient income from their inventions for one reason or another.[FN10] Although many of these private bills extended the term of patents that had already lapsed into the public domain, a practice held categorically unconstitutional in Graham v. John Deere Co. of Kansas City,[FN11] Congress’s historical practice of enacting discriminatory, inventor-specific patent terms adds a degree of constitutional legitimacy to the more principled, industry-specific terms I ultimately propose.

The patent term of “seventeen years from issue date” remained in effect until 1994, when Congress brought U.S. patent law into compliance with the TRIPS Agreement.[FN12] Implementation of TRIPS modified the term of patent protection in the U.S. from “seventeen years from issue date” to “a term beginning on the date on which the patent issues and ending 20 years from the date on which the application for the patent was filed in the United States…”[FN13]

The change in patent term from “seventeen years from issue date” to “twenty years from filing date” sparked controversy.  Opponents of the new “twenty years from filing date” patent term stressed that pioneering technology typically takes a long time to evaluate and approve, which would result in shorter terms under the new system.[FN14] Indeed, the biotechnology sector would be particularly hard hit “because the most commercially attractive patents can take over fourteen years to issue.”[FN15] Weaker patents, the opponents continued, would in turn reduce the royalties that are the lifeblood of private research and development (R&D) funds.[FN16]

To address these concerns, Congress passed the American Inventors Protection Act of 1999 (AIPA). [FN17] Specifically, the AIPA adds a new provision to compensate applicants fully for PTO-caused administrative delays, and, for good measure, includes a new provision guaranteeing diligent applicants at least a 17-year term by extending the term of any patent not granted within three years of filing.  Thus, no patent applicant diligently seeking to obtain a patent will receive a term of less than the 17 years as provided under the pre-GATT3 standard; in fact, most will receive considerably more.  Only those who purposely manipulate the system to delay the issuance of their patents will be penalized…[FN18]

Today, the duration of patent protection in the United States remains “twenty years from filing date,” subject to the aforementioned patent term adjustment provisions.[FN19]

B. Patent Globalization Under TRIPS and the PCT

Patenting has gone global.  The Patent Cooperation Treaty[FN20] (PCT) makes it possible for inventors and firms to seek patent protection simultaneously in many different countries by filing “international” patent applications.[FN21] As of 1980, only 30 countries adhered to the PCT.[FN22] By 1995, the number of PCT contracting states had grown to 82.[FN23] Today, 142 countries have signed on to the PCT.[FN24] According to the World Intellectual Property Organization (WIPO), the body that administers the PCT, the PCT “postpones the major costs associated with international patent protection” and “brings the world within reach” (emphasis added).[FN25] In effect, the proliferation of PCT contracting states makes it relatively easy for a patent applicant to pursue de facto worldwide patent protection on her invention.

Not only has it become easier to patent around the globe, but the minimum level of patent protection afforded by most countries has never been higher.  The 1994 Agreement on Trade- Related Aspects of Intellectual Property Rights, or TRIPS, sets a floor below which intellectual property standards shall not fall.[FN26] For example, TRIPS requires that “patents shall be available for any inventions, whether products or processes, in all fields of technology,” with only limited exceptions.[FN27] Additionally, TRIPS mandates a minimum patent term of twenty years, measured from the patent application filing date.[FN28]

III:  Methodology:

GPT as a Proxy for “Reward Per Invention”

R&D Expenditures as a Proxy for “Amount of Innovation”

In this article, I investigate whether the average “reward per invention” has significantly increased over the last fifteen years, due in large part to the 1994 TRIPS Agreement and the proliferation of PCT contracting states.  More specifically, I evaluate whether patent applicants are actually seeking and obtaining more widespread patent protection today than they were at the time of the TRIPS Agreement fifteen years ago.  As stated above, if it has become relatively easy for firms to secure a high level of protection for each invention, e.g., by receiving multiple patents on the same invention in many different countries, the “reward per invention” is high.  In a well-functioning patent system, a high “reward per invention” should result in increased innovation.  If the “reward per invention” has increased but the level of innovation has not, then the prevailing level of patent protection may be too high, and we should respond by reducing the patent term in the United States.  A shorter patent term would bring patented inventions into the public domain more quickly, enabling others to take early advantage of the patented technology.

To measure the “reward per invention,” I developed a metric called the “global patent term,” or GPT.  The GPT is the sum of the terms of protection of all the patents, worldwide, on a single invention.  Of course, not all patents are created equal.  For example, a twenty year monopoly in the United States is likely more valuable than a twenty year monopoly in Belgium.  To account for this difference in value, each patent term in the GPT calculation is scaled according to the gross domestic product (GDP) of the corresponding country, using the United States GDP as a baseline.

A short hypothetical will illustrate the calculation.  Suppose that in 1995, Company XYZ typically applied for patent protection for its inventions only in the United States.  In this case, the GPT equals the U.S. patent term, or twenty years.  Now, in 2009, suppose that Company XYZ typically applies for patent protection in the United States, China, and Canada.  The GPT equals [(U.S. patent term)(U.S. GDP / U.S. GDP) + (China patent term)(China GDP / U.S. GDP) + (Canada patent term)(Canada GDP / U.S. GDP)], or [(20 years)($14,204 billion / $14,204 billion) + (20 years)($4,326 billion / $14,204 billion) + (20 years)($1,400 billion / $14,204 billion)], or [20 years + 6.1 years + 2.0 years], or approximately twenty-eight years.

In this hypothetical, the “global patent term” for Company XYZ’s invention has risen from twenty years in 1995 to twenty-eight years in 2009, an increase of 40%.  If the patent system is working properly, we would expect to see a higher level of innovation from Company XYZ as a result of this 40% increase in “reward per invention.”  To test whether this is actually the case, we can examine the R&D spending of Company XYZ during the relevant timeframe.

The TRIPS Agreement and the steady growth in number of PCT contracting states have arguably facilitated patenting around the globe.  If the reward for inventing, as measured by the “global patent term” metric, is greater than it has ever been before, then the level of innovation, as measured by R&D spending, should be similarly high.  However, if innovation lags in the face of ever-increasing patent protection, the extra rewards are not fostering innovation but rather unjustly enriching some patent holders.  If this is the case, we should consider shortening the patent term in the United States.

IV:  Three Case Studies: Pfizer, International Paper, and UNISYS

To shed light on these difficult issues, I analyzed the “global patent terms” and R&D spending of three different firms in three very different industries: Pfizer, a pharmaceutical giant; International Paper Company, a global paper product producer; and UNISYS, a worldwide information technology provider.  In particular, I compared each firm’s average GPT and R&D spending in 1995, just after TRIPS was enacted, and 2009, after the proliferation of PCT contracting states and the entrenchment of the TRIPS regime.  The results vary widely among the three firms.

Of the three firms, Pfizer has most clearly taken advantage of the increased availability of patents around the globe.  The average (mean/median) GPT of Pfizer’s 1995 U.S. patents was 95.7/85.4 years.  The average GPT rose significantly to 131.4/108.7 years in 2009.  Even excluding U.S., EP, and WO filings, which heavily influence the overall GPT calculations due to their high GDPs, Pfizer’s average GPT increased from 23.7/22.6 years in 1995 to 29.9/25.7 years in 2009.  Not surprisingly, the average number of countries in which Pfizer filed for patent protection also increased, from 13 in 1995 to 16 in 2009.

Additionally, between 1995 and 2009, Pfizer drastically altered its approach regarding where to file for patent protection, turning away from filing in many Western European countries such as Austria, Germany, and Denmark, and towards filing in Central and South American countries such as Mexico, Argentina, and Brazil.  For example, 66% of Pfizer’s 1995 U.S. patents had counterparts in Germany.  That number fell to just 13% in 2009.  On the other hand, only 14% of Pfizer’s 1995 U.S. patents had counterparts in Brazil, but by 2009, 72% of Pfizer’s U.S. patents had related Brazilian applications.

The aforementioned data strongly suggests that Pfizer reaps the rewards of the strong global patent system buttressed by the Patent Cooperation Treaty and the TRIPS Agreement.  In turn, we should see enhanced innovation in the form of increased R&D spending by Pfizer if the patent system is functioning properly.  In one respect, we do.  Pfizer greatly increased R&D spending from approximately $1.4 billion in 1995 to approximately $7.8 billion in 2009.  However, measured as a percentage of total revenues, the increase in Pfizer’s R&D spending from 1995 to 2009 was minimal – 14.4% to 15.7%.

If Pfizer were the only company playing the patent game, then it might make sense to reduce the term of patent protection in the United States.  After all, the data shows that Pfizer gets significantly more worldwide patent protection for each invention today than it did before the TRIPS Agreement took effect.  In return, Pfizer has only minimally increased its R&D spending as a fraction of its total revenues.  This looks a lot more like unjust enrichment than enhanced innovation.  And for a company like Pfizer, which relies heavily on patenting both inside and outside the United States, the increased availability and strength of patents around the globe will likely offset any harm to innovation that may come from reducing the patent term in the U.S.  But Pfizer is not the only company in town.  Other firms, such as International Paper Company and UNISYS, that rely more heavily on the U.S. patent system and not on patenting in foreign countries, do not benefit nearly as much as Pfizer does from the worldwide patent regime ushered in by the Patent Cooperation Treaty and the TRIPS Agreement.

While International Paper’s average GPT (mean/median) rose from 42.9/20.0 years in 1995 to 66.1/40.0 years in 2009, most of that increase can be attributed to multiple related U.S. patents on the same technology, not higher global patenting activity on the part of the company.  Excluding U.S., EP, and WO filings, International Paper’s average GPT remained relatively flat, rising slightly from 4.0/0.0 years in 1995 to 5.6/0.0 years in 2009.  The average number of countries in which International Paper filed for patent protection was essentially flat as well – 3.2/1.0 in 1995 and 3.4/1.0 in 2009.

In stark contrast to Pfizer, International Paper does not appear to have benefited from heightened levels of global patent protection.  The company simply does not pursue patents in enough countries outside the United States for the TRIPS Agreement and the proliferation of PCT contracting states to make an appreciable difference in terms of innovation incentives.  Therefore, reducing the patent term in the United States may have a deleterious effect on innovation at a company such as International Paper, which has already seen R&D expenditures plummet from $111 million (0.56% of net sales) in 1995 to $13 million (0.06% of net sales) in 2009.

UNISYS finds itself in a similar position to International Paper when it comes to patenting on a global scale – it didn’t do it in 1995, and it doesn’t do it now.  UNISYS’ average GPT (mean/median) actually fell from 54.1/20.0 years in 1995 to 42.0/20.0 years in 2009.  Excluding U.S., EP, and WO patent application filings, UNISYS’ average GPT was flat and basically non-existent – 1.8/0.0 years in 1995 and 2.8/0.0 years in 2009.  In general, UNISYS only seeks patents for its inventions in the United States, so the enhanced global patent protection made possible by TRIPS and the PCT has little effect on the innovation decisions and R&D expenditure incentives at UNISYS.  This becomes painfully obvious when one considers that R&D expenditures at UNISYS have fallen from $405 million (6.4% of revenue) in 1995 to $102 million (2.2% of revenue) in 2009.

V:  A Modest Proposal

To summarize, not all companies benefit from the availability of strong global patent protection facilitated by the Patent Cooperation Treaty and mandated by TRIPS.  In fact, not even all companies that pursue patents benefit from the aforementioned regime.  Easy access to strong patents around the globe favors firms like Pfizer that seek protection for their inventions in many different countries.  For companies like International Paper and UNISYS that file for patents primarily in the United States, the theoretical presence of strong patent regimes in other countries has little if any practical effect on the companies’ incentives to innovate.

Therefore, even if it were possible under TRIPS to decrease the U.S. patent term to less than twenty years, an across the board patent term reduction would not be advisable.  Such a move would likely have little influence on innovation at firms such as Pfizer because the enhanced innovation incentives provided by increased patent protection in the rest of the world would counterbalance the effect of a shorter U.S. patent term.  However, for firms that believe it to be in their best interest to seek patents only or primarily in the United States, reducing the U.S. patent term might stifle their drive to innovate.

To solve this problem, we should consider setting different patent terms for different technology sectors.  A relatively short U.S. patent term might provide the ideal level of innovation for industries such as pharmaceuticals in which the typical inventions (drugs) have worldwide profit-making power.  In other industries, such as information technology and software, patenting in many less-developed countries just does not make economic sense because the market is so limited.  A longer term of patent protection in the U.S. might be necessary to achieve the maximum amount of innovation in technology sectors such as these.  By adopting technology-specific patent terms, it may be possible to increase innovation and to bring patented subject matter into the public domain more quickly.

* I would like to thank Rochelle Dreyfuss, Barton Beebe, Oren Bar-Gill, and all those who participated in the Lederman/Milbank Fellowship workshop and the Colloquium on Innovation Policy for their insightful comments and critiques.  All errors are, of course, my own.

[FN1] U.S. Const. art. I, § 8, cl. 8.

[FN2] See 35 U.S.C. § 154 (2006) (subject to certain exceptions, patents run for 20 years from application filing date); Agreement on Trade-Related Aspects of Intellectual Property Rights, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, Legal Instruments – Results of the Uruguay Round, 33 I.L.M. 1197 (1994) [hereinafter TRIPS Agreement or TRIPS], art. 33 (prescribing a minimum term of patent protection of 20 years from application filing date).

[FN3] The intense debate over recently-proposed patent reform legislation exemplifies this point.  See, e.g., Amy Schatz, Patent Overhaul Gets Close, Draws Opposition, Wall St. J., Mar. 28, 2011, at B3 (noting that Corporate America is divided over Congress’s current patent law reform efforts, despite “calling for an overhaul of U.S. patent laws for years”).

[FN4] See David S. Abrams, Did TRIPS Spur Innovation An Analysis of Patent Duration and Incentives to Innovate, 157 U. Pa. L. Rev. 1613, 1618 (2009).

[FN5] Patent Act of 1790, ch. 7, § 1, 1 Stat. 109, 110.

[FN6] See Patent Act of 1836, ch. 357, § 17, 5 Stat. 117, 124-25.

[FN7] See Act of Mar. 2, 1861, ch. 88, § 16, 12 Stat. 246, 249.

[FN8] See Dana Rohrabacher and Paul Crilly, Congressional Commentary: The Case for a Strong Patent System, 8 Harv. J. Law & Tech. 263, 264 (1995).

[FN9] Eric E. Johnson, Calibrating Patent Lifetimes, 22 Santa Clara Computer & High Tech. L.J. 269, 283 (2006).

[FN10] See, e.g., An Act for the Relief of Oliver Evans, 6 Stat. 70 (1808).  See also Eldred v. Ashcroft, 537 U.S. 186, 234-35 (2003) (Stevens, J., dissenting) (providing historical analysis of early U.S. patent and copyright law).

[FN11] 383 U.S. 1, 6 (1966) (“Congress may not authorize the issuance of patents whose effects are to remove existent knowledge from the public domain”).

[FN12]See S. Rep. No. 103-412 (1994) (discussing Section 532 of the Uruguay Round Agreements Act).

[FN13] See Uruguay Round Agreements Act § 532, Pub. L. No. 103-465, 108 Stat. 4809 (1994).

[FN14] See H.R. Rep. No. 104-887 (1997) (discussing Hearing Volume No. 104-58, particularly the testimony of Ms. Gardner).

[FN15] Patricia Montalvo, How Will the New Twenty-Year Patent Term Affect You?  A Look at the TRIPS Agreement and the Adoption of a Twenty-Year Patent Term, 12 Santa Clara Computer & High Tech. L.J. 139, 160 (1996).

[FN16] See Dana Rohrabacher and Paul Crilly, Congressional Commentary: The Case for a Strong Patent System, 8 Harv. J. Law & Tech. 263, 264 (1995).

[FN17] See generally American Inventors Protection Act of 1999, Pub. L. No. 106-113, 113 Stat. 1501.

[FN18] Id. at 49-50.

[FN19] See 35 U.S.C. § 154 (2006).

[FN20] Patent Cooperation Treaty, June 19, 1970, 28 U.S.T. 7645, 1160 U.N.T.S. 231.

[FN21] About the Patent Cooperation Treaty, (last visited May 5, 2010).

[FN22] See PCT Contracting States (2010),

[FN23] See id.

[FN24] See id.

[FN25] World Intellectual Prop. Org., Protecting Your Inventions Abroad: Frequently Asked Questions About the Patent Cooperation Treaty (PCT) 16 (2006), available at

[FN26] See generally TRIPS Agreement, available at

[FN27] TRIPS art. 27.

[FN28] TRIPS art. 33.