6.1.1 The Data Set and Methodology

The Study analyzed the over 14 million copyrighted works registered with the USCO between 1978 and 2012.Footnote ²² Registration records contain information about the work, such as its title, date of application, publication status, and creative category.Footnote ²³ Usefully, the Study converted all work types into six basic categories: text, music, art, movies, drama, and software.Footnote ²⁴

In analyzing gender demographics, the Study focused on works by known, named individual authors.Footnote ²⁵ Registrants are not asked to self-report demographic information in their copyright registration applications. Using gender-related data extracted from the 1990 Census, the Study could assign to authors a probabilistic gender based on their first names.Footnote ²⁶ First names are highly indicative of gender: 82 percent of the authors had a 99 percent or higher likelihood of being either a man or a woman.Footnote ²⁷ Works created by business associations and anonymous and pseudonymous works were excluded from the gender statistics because the author’s gender could not be determined probabilistically.Footnote ²⁸

6.1.2 Key Findings

This study revealed a wide variation in gender parity across time, type of work, and publication rates, among other aspects. Strikingly, in 2012, 64 percent of authors were men.Footnote ²⁹ While this was a decrease from 1978, when 70 percent of the authors were men, it was still well above their labor market share in 2012.Footnote ³⁰ The percentage of women in the labor market increased from 41.7 percent in 1978 to 46.9 percent in 2012, while the share of women authors has only increased from 30 percent in 1978 to 36 percent in 2012.Footnote ³¹

6.2.1 The Patent Gender Gap

In 1977, only 2.6 percent of granted U.S. patents listed at least one woman as an inventor.Footnote ⁴⁶ More recently, in 2016, women were listed on 21 percent of granted patents.Footnote ⁴⁷ However, this remains below women’s share of the STEM labor force.Footnote ⁴⁸ Furthermore, the total inventor rate – that is, the percentage of all inventors listed on granted patents who are women – has only increased from 10 percent in 2000 to 12 percent in 2015.Footnote ⁴⁹ If we disaggregate some of these patent statistics, we find that the degree of gender disparity varies widely over time and by scientific field. For example, while some fields, such as life sciences, have achieved near gender parity, other fields, such as engineering, have remained vastly disproportionate.Footnote ⁵⁰ This trend is reflected in STEM educational data, labor force makeup, and patent statistics.

Unlike trademarks and copyrights, the expertise required to invent something patentable would often necessitate some formal education or expertise in STEM, a high level of education, or at least a background in STEM.Footnote ⁵¹ As a result, gender disparities in a degree program at both the undergraduate and graduate levels may explain part of the gender disparity in patent rates.Footnote ⁵² For example, one study has argued that individuals with mechanical and electrical engineering degrees produce more patents than most other STEM degrees, but women have historically received less than 10 percent of the mechanical and electrical engineering degrees awarded to men. Because individuals with these STEM degrees have a high likelihood of patenting compared to other degree programs, a lack of women in the degree program could affect the disparity in overall patent rates between men and women.Footnote ⁵³ On the other hand, women have been more proportionally represented in biological and physical science degree programs, and in 2015, women represented 48 percent of biological and life scientists. These STEM degrees are much less likely than mechanical and electrical engineering to lead to patent activity, but while men and women are nearly equally employed as biological and life scientists, in 2015, only 25 percent of the inventors for biotechnology patents and 23 percent of inventors for pharmaceutical patents were women, suggesting educational attainment alone does not determine patent rates.Footnote ⁵⁴

6.2.2 The Trademark Gender Gap

Unlike patents, there has been relatively little research done on women’s trademark ownership. Still, available data suggest a similar gap exists between the number of men and women who own trademarks, although there are signs of greater equality over time. In 1980, women received 16.7 percent of trademarks granted, and by 2010 they still received only 33.31 percent.Footnote ⁷⁸ Unlike patents, no set of educational indicators is associated with people who will apply for a trademark. Trademarks have occasionally been connected to entrepreneurial activity, but a 2015 survey conducted by the U.S. Census Bureau found only a marginal difference in the share of women-owned businesses with a trademark compared with men-owned businesses.Footnote ⁷⁹ That being said, women as a whole are less likely to become entrepreneurs, and in that study, only 20.8 percent of the surveyed businesses were women-owned.Footnote ⁸⁰

Like copyright and patent rates, women appear to specialize in different areas than men. In 2010, women were granted the most trademarks in the following classes: advertising and business, clothing, education and entertainment, and paper goods and printed matter.Footnote ⁸¹ While these categories were among the most frequently used trademark classes between 2004 and 2013, notably missing are the electrical and scientific apparatus and computer and scientific classes, the first and fourth most-used classes in that period.Footnote ⁸²

Interestingly, women’s trademark applications were accepted more frequently than applications by men between 1988 and 2000. However, the ratio of successful trademark applications made by women and men has been nearly equal since 2000.Footnote ⁸³ Therefore, women are likely underrepresented as trademark owners due to a lack of applications compared with men. There has not been much research done on why women file fewer applications than men. But one study has suggested that women may be hesitant to apply for trademarks due to a lack of knowledge about how they work and uncertainty about the financial value of obtaining a trademark.Footnote ⁸⁴

6.3.1 Gender Disparity in Property Ownership and Wealth Accumulation

6.3.2 Entrepreneurial Success and Corporate Power

6.4.1 Book Industry

Women have traditionally been in the majority at every stage of book publishing.Footnote ¹²² In 2012, roughly 55.6 percent of self-reported writers or authors were women.Footnote ¹²³ That same year, 52.3 percent of periodical, book, and directory publishers were women as well.Footnote ¹²⁴ Nevertheless, despite authors being majority women, the majority of authors (54.17 percent) of registered textual works in 2012 were men.Footnote ¹²⁵

While women for all purposes appear to be in the majority representation-wise, there are a set of gendered patterns that may discourage women authors or otherwise act to decrease the representation of women within the industry. First, although women are overrepresented in the publishing industry, some of the largest publishing houses do not publish books written by women at the same rate as they publish books written by men.Footnote ¹²⁶ One study found that out of thirteen publishing houses, only two had more than 30 percent of their published books authored by women.Footnote ¹²⁷ With over 50 percent of self-reported authors being women, it is difficult to argue that this unequal representation in publishing houses is due to women’s lack of interest or ability.

Even when women’s books are published, on average, they sell for less, and their publishers invest less in their promotion compared with books written by men.Footnote ¹²⁸ As a result, women profit less from their textual creativity than men. Furthermore, books written by women are comparatively less likely to be reviewed in magazines.Footnote ¹²⁹ Men also have historically had a greater likelihood of having their books appear on The New York Times’ bestseller list.Footnote ¹³⁰ While these statistics could be tied to lower numbers of books written by women being published in the first place, book reviewers have also predominantly been men.Footnote ¹³¹ With this background, it is perhaps unsurprising that women may not be encouraged to write and register their work as much as men.

6.4.2 Music Industry

The music industry has long struggled with a lack of women both as performers and in music production. In 2018, 35 percent of self-reported musicians and singers and only 12.4 percent of broadcast and sound engineers were women.Footnote ¹³² This represented an improvement for the industry from 2010 when only 31.9 percent of musicians and singers and 9.9 percent of broadcast and sound engineers were women.Footnote ¹³³ However, despite these improvements, women remain underrepresented as musical artists and in music production. In a study of representation on the Billboard Hot 100 Year-End Charts between 2012 and 2019, only 21.7 percent of artists, 12.5 percent of songwriters, and 2.6 percent of producers were women.Footnote ¹³⁴ This disparity varies by genre, with women in pop representing 32.6 percent of artists but only 11 percent of artists in alternative music.Footnote ¹³⁵ This lack of representation is reinforced in awards show nominations. Between 2013 and 2020, 88.3 percent of Grammy Award nominees were men, though the number of women nominees has been trending upward over time.Footnote ¹³⁶

While it is unclear whether this is a cause or an effect, the music industry has been plagued by allegations of gender-based discrimination.Footnote ¹³⁷ In one study that surveyed women in different parts of the music industry, 78 percent of respondents reported that they had been treated differently due to their gender.Footnote ¹³⁸ In another study, women reported that their colleagues did not take them seriously and that their contributions both as producers and performers were not seen by their colleagues as important or valuable.Footnote ¹³⁹ While these barriers would not necessarily prevent a self-recording artist from creating and registering new work, they may be part of why men authored 75.98 percent of copyright registrations in the music category.Footnote ¹⁴⁰

6.4.3 Film Industry

Men hold the vast majority of roles in the film industry. In 2010, women represented only 7 percent of directors, 2 percent of cinematographers, and 15 percent of executive producers for the top 250 grossing films of the year.Footnote ¹⁴¹ By 2019, these numbers have improved, though there is still a way to go. Of the top 250 grossing films of 2019, women represented 13 percent of directors, 5 percent of cinematographers, and 23 percent of executive producers.Footnote ¹⁴² This inequality is mirrored among film actors and actresses as well. In 2010, only 30 of the top 100 movies for the year featured a woman as a lead or colead. By 2019, this had increased to 43 out of the top 100 movies.Footnote ¹⁴³ However, for actresses, this inequality does not appear to be tied to a lack of interest. One of the most popular college majors for professional actors and actresses is visual and performing arts, and in 2017, women received 60.1 percent of visual and performing arts bachelor’s degrees.Footnote ¹⁴⁴ The lack of gender representation goes all the way up to the top of movie studio leadership, with one study finding that 82 percent of chief executives for some of the country’s largest movie studios are men.Footnote ¹⁴⁵

While it is unclear whether some women choose to steer clear of the film industry or whether the industry as a whole creates additional barriers to entry for women, the film industry has faced allegations of gender-based discrimination for years.Footnote ¹⁴⁶ Actresses have often been paid less for their creative labor than their costars, underscoring the finding that in a list of the top ten highest-paid actors and actresses for 2019, only two out of the top ten were women.Footnote ¹⁴⁷ In addition, stories about men refusing to work with women directors and women’s contributions behind the scenes being challenged by their coworkers suggest that there may be social barriers that could discourage women from entering or remaining in the film industry.Footnote ¹⁴⁸

6.4.4 Theatre Industry

Like the film industry, men predominate in almost every role, both on stage and behind the scenes for theatrical works. In a study of theatrical productions on Broadway between 2018 and 2019, 13 percent of directors, 24 percent of choreographers, 13 percent of writers, and 32 percent of named characters were women.Footnote ¹⁴⁹ However, women are in the majority or close to the majority as costume designers (52 percent), stage managers (47 percent), and company managers (57 percent).Footnote ¹⁵⁰ These trends have been mirrored in Off-Broadway productions as well. In 2010, 25 percent of directors and 31 percent of playwrights were women.Footnote ¹⁵¹ Notably, some theaters have had years where they did not feature a single play written by a woman, and in the 2013–14 season, women wrote only 29 percent of the plays produced in Off-Broadway theaters.Footnote ¹⁵²

Women playwrights may be particularly discouraged from writing because their lack of exposure and access to theaters is a nationwide problem.Footnote ¹⁵³ Between 2011 and 2014, a study of 153 theaters across the United States found that women wrote only 22 percent of the plays produced despite making up around two-thirds of theater audiences.Footnote ¹⁵⁴ While some of this disparity may be tied to theaters relying on a certain number of revived productions year to year, men also hold an estimated 80 percent of artistic director positions in regional theater companies, and these individuals often have the greatest influence on the plays that are selected.Footnote ¹⁵⁵ The data may however suggest a relative female inclination toward originality: While men wrote 84 percent of the revivals produced in 2016–17, women wrote 35 percent of the “new” plays produced that year.Footnote ¹⁵⁶

6.4.5 Other Copyright-Related Industries: Art, Dance, and Architecture

Despite making up roughly 46 percent of artists in the United States between 2012 and 2016, women in the arts are not given the same opportunities as men and are compensated less for their work.Footnote ¹⁵⁷ While this statistic suggests that men and women have a roughly equal interest in the arts, works produced by men constituted 87 percent of U.S. museum collections in 2018.Footnote ¹⁵⁸ While 51 percent of visual artists in recent years have been women, women have received only 27 percent of solo exhibitions between 2013 and 2020.Footnote ¹⁵⁹ This lack of exposure in museums is effected partly by women being underrepresented in art markets and fairs around the world, as well as by potential bias within museum leadership.Footnote ¹⁶⁰

Part of the earnings gap may be tied to differences in occupation selection. For example, between 2012 and 2016, women represented 81.4 percent of dancers and choreographers but only 25 percent of architects.Footnote ¹⁶¹ While architects in this period made, on average, US$76,680 per year, dancers and choreographers made only US$31,150.Footnote ¹⁶² Even within the same field, women are still paid less than men. One study found that women across art professions make around US$20,000 less per year than men in the same field.Footnote ¹⁶³ In addition to their undercompensated labor, women’s art is not valued as highly as art produced by men. In the fine art industry, one study found that, on average, women’s artwork sells at auction for 47.6 percent less than artwork created by men.Footnote ¹⁶⁴ To put this into perspective, between 2008 and the first half of 2019, US$196.6 billion was spent at international art auctions, and only 2 percent of this figure, US$4 billion, was spent on works created by women.Footnote ¹⁶⁵ Works by just five women accounted for US$1.6 billion of the total US$4 billion spent on women artists.Footnote ¹⁶⁶ Part of the problem with women’s art valuation is that rather than comparing works by women with art produced by both men and women, it may be common in the industry to compare art prices only to works created by other women artists, which may lead to lower price growth and valuation over time.Footnote ¹⁶⁷

6.5.1 Inequality or Discrimination?

Given the undeniable disparity between authorship rates among men and women, the question remains whether this difference is tied to existing forms of inequality between men and women or whether women register works at lower rates than men due to gender-based discrimination.

7.2.1 Gender Inequality Index: Brazil, India, and Mexico in the Global Context

According to the Gender Inequality Index (GII),Footnote ⁵ Brazil and Mexico stand out for nearly converging in the 2014 GDI (with index scores of 0.997 and 0.943, respectively), whereas India has a lower index score of 0.795 (Figure 7A.1). India, however, rose in the GDI in 2018; Mexico and Brazil had marginal improvements. The Latin American countries’ GII scores rank between very high human development (VHHD) and high human development (HHD), while India ranks in the low human development category.

In terms of average years of schooling, gender advances were observed between 2014 and 2018 in the three emerging countries studied (Figure 7.1). Compared with VHHD and HHD countries, the average years of schooling in Brazil, India, and Mexico were lower for both sexes, especially in India. In Mexico, there was a marginal increase in average years of schooling from 8.2 to 8.4, while the corresponding figure remained 8.8 years for men. In Brazil, women exceeded the average years of schooling for men: 8.1 versus 7.6 average years, respectively, in 2018. Finally, there were more significant differentials in India, with only 4.7 average years of schooling for women and 8.2 for men.

Figure 7.1 Average years of schooling by gender, 2014 and 2018: Mexico, Brazil, and India.

Source: U.N. Development Programme (n.d.-c).

7.2.2 Human Capital Specialization by Gender: Toward Which Scientific Disciplines Are Women in Higher Education Oriented?

The graduate gender gap in the scientific field has different dimensions across countries. Having gender gaps in engineering, manufacturing, and construction, as well as in STEM, is a fact for all three countries (Figure 7.2). There has been very little improvement from 2014 to 2017 – with none, in fact, in Mexico in the STEM field. More men than women in Brazil and Mexico graduate in information and communication technologies, and the same is true for India’s agriculture, forestry, fisheries, and veterinary sciences. These gaps have even increased from 2014 to 2017. However, across all three emerging countries, graduation percentages for women are higher in natural sciences, mathematics, and statistics and are robust in health and welfare. Only minor changes have occurred (UNESCO, 2016, 2018).

Figure 7.2 Percentage of female graduates by STEM career categories, 2017.

Sources: UNESCO (n.d.); UNESCO, UIS Statistics. Distribution of tertiary graduates by field of study Years selected. http://data.uis.unesco.org/index.aspx?queryid=3830

Participation of female graduates in STEM careers in 2017 registered significant and differentiated progress across countries. In India, female graduates reached parity in information and communication technologies. In health and welfare, women surpassed men in all three countries, especially in Brazil (75 percent women). Women reached parity in natural sciences, mathematics, and statistics, especially in Brazil (59.5 percent). However, women still lagged behind men in traditional fields (Figure 7.2).

7.2.3 Inclusion of Women in S&T Research

The participation of women within the total group of researchers in Brazil and MexicoFootnote ⁶ improved between 2011 and 2015, when compared with the period between 1996 and 2000 (Figure 7.3). In general, the number of researchers in Mexico increased more for men (an increase of 39,000) than for women, from 8,100 to 34,400 researchers (an increase of 26,300). As for Brazil, the increase was noteworthy and even higher for women (136,000 women researchers compared with 129,000 men researchers).Footnote ⁷ The differential between Mexico and Brazil (89,500 compared with 312,800 researchers) is most likely associated with the fact that Mexico has had little GDP expenditure on research and development (R&D) (an average of 0.4 percent). In contrast, Brazil intensified its efforts in this area (1 percent) (Table 7A.3).Footnote ⁸

Figure 7.3 Number of researchers by gender* in Mexico and Brazil, 1996–2000 and 2011–2015 (thousands of researchers).

* Among named and gendered author profiles.

Source: Elsevier Research Intelligence (2017).

When identifying researchers by scientific areas in the two Latin American countries between 2011 and 2015, diversification and greater integration of women in the various scientific fields are observed. Although Brazil and Mexico, respectively, have 25 percent and 22 percent of women in medicine, the differential between the two countries in the number of women researchers in this field was substantial (see Figure 7.3). The count for Brazil was 80,600 during 2011–2015, an increase of 9.4 times from 1996 to 2000, and the corresponding figure for Mexico was 17,300 researchers during 2011–2015, an increase of 4.6 times. Other areas of importance are agriculture and biological sciences, biochemistry, genetics, and molecular biology (an average of 10 percent each) and, to a lesser degree, immunology, and microbiology (5 percent), with only marginal percentages in other fields.

7.2.4 Launch of Knowledge Empowerment Policies

We find STEM and Gender Advancement (SAGA) among the international projects. Efforts have focused on pushing for reforms to close gender gaps in STEM at the education and research levels – and also drawing lessons from evidence and the policy mix to improve national science, technology, and innovation (STI) policies. These efforts encourage the incorporation of women into the knowledge economy. In addition, they develop a better understanding of women and girls in science. Finally, they provide estimates and help build sex-disaggregated data, and design and implement the STI policy instruments that affect gender equality in STEM (UNESCO, 2016).

As noted in the UNESCO study, there are several important facts about women researchers: (1) Women are slightly less likely than men to collaborate across academic and corporate sectors on research papers; (2) there is a relatively slight variation between comparator countries and regions in the percentage of cross-sector collaboration between academia and industry; (3) the proportion of scholarly output resulting from the academic–corporate collaboration is slightly lower for women researchers than for men researchers; (4) women tend to have a slightly higher share of the top 10 percent of interdisciplinary scholarly output relative to their total scholarly output than men; (5) among researchers, women are generally less internationally mobile than men; and (6) the highest citation impact is associated with transitory researchers (those who move internationally for periods of less than two years) (Elsevier Research Intelligence, 2017: 7).

7.3.1 Nature and Dynamics of Female Inventive Activity: Data

Our analysis of emerging female inventors focuses on three countries in the study: India, Brazil, and Mexico. We have used the USPTO database of patents granted to the holders in three countries. The period for each country is different: Mexico, 1980–2015; Brazil, 1997–2013; and India, 1997–2010. We identified those patents with at least one woman inventor and then organized the list of women inventors, associating with them the information on the patent(s).

We observe different growth dynamic paths for USPTO patents granted to the three countries. On the one hand, India was favored by a ten-year transition period following the adoption of the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement), showing an increasing dynamic in the case of pharmaceuticals. From 1997 to 2010, 2,685 patents were granted to Indian holders, of which 1,219 involved at least one woman inventor and 1,416 involved only men inventors (see Figure 7.4). India registered an annual average growth rate (AAGR) of 20.9 percent, which was higher for those patents with at least one woman inventor (24.7 percent). Two subperiods stand out in the evolution of patents granted to Indian holders. The first, 1997–2003, showed an AAGR of 40 percent on total patents, and patents with female collaboration were notably higher (54.2 percent). The second subperiod, 2004–2010, had a lower AAGR of 8.34 percent, particularly for patents with women inventors (6.7 percent) (Figure 7.4). Perhaps in the following years, patent expansion grew and was equaled by that of at least one woman.

Figure 7.4 India: Evolution of USPTO patents granted (total and those having at least one woman inventor), 1997–2010.

Source: Authors’ own elaboration, based on USPTO data.

On the other hand, we have Mexico with lower growth. Indeed, 1,193 USPTO patents were granted to Mexican assignees from 1986 to 2015, with an AAGR of 8.4 percent. Although female participation in total patents is characterized by a large gap, the 108 patents identified with at least one woman inventor were slightly higher (AAGR of 8.7 percent). Women’s collaboration increased in 2007, with seven patents, and reached seventeen in 2014 (Figure 7.5). Innovative activity in Mexico remains low because of a lack of entrepreneurial and institutional investment in R&D activities that target innovation and patenting.

Figure 7.5 Mexico: Evolution of USPTO patents granted (total and those having at least one woman inventor), 1980–2015.

Source: Authors’ own elaboration, based on USPTO data.

Concerning Brazil, of the 1,434 patents granted to Brazilians between 1997 and 2013, 388 (27 percent) had at least one woman inventor on the team of researchers. The growth path was very similar to that for all patents. The country has a small and volatile trajectory, but this trajectory is growing as the R&D/GDP percentage increases (Figure 7.6). Although the most significant proportion of patents with female involvement in Brazil is individual, teams with two or more inventors have become the practice since the end of the 1990s.

Figure 7.6 Brazil: Evolution of USPTO patents granted (total and those having at least one woman inventor), 1997–2013.

Source: Authors’ own elaboration, based on USPTO data.

In short, the evolution of granted patents analyzed shows that even if the participation of women inventors in each country is still marginal, it has been increasing, especially in India. It is worth pointing out that men are also on those inventor teams with at least one woman inventor. The differentiated abilities across genders seem to strengthen the scope of innovation.Footnote ⁹ We assume that the increase of women in inventor teams is associated with the increasing incorporation of female inventors in STEM careers and related research activities. This phenomenon is more evident in India and Brazil. There is growth in Mexico but at a lower level.

7.3.2 Nature of Innovation

We proceed to identify the features of patents involving women inventors. It is interesting to compare the patent number by team size in the three emerging countries in this study. In general, the teams are mixed, except in patents with only one inventor. India leads with 45 percent of patents with groups including women, followed by Brazil with 24 percent and Mexico with just 21 percent. During the total period analyzed, the whole patents with teams of two to five inventors predominate in all three countries: two-thirds in India and just over 70 percent in Mexico and Brazil.

Regarding teams with more than six researchers, India stands out, with almost one-third of the patents, and the two Latin American countries have around 10 percent. Individual women’s participation in patenting is relatively marginal. This is especially the case for India, where 3.4 percent of patents are characterized as individual patents. Mexico and Brazil slightly exceed 15 percent. In contrast, for patents in which only men participate, individual participation is significant in Brazil and Mexico, but the figure is lower in India (see Table 7.1).

Regarding patent assignees, institution holders are prominent in India (65 percent) and Mexico (54 percent). In contrast, firms have greater weight in Brazil (73 percent), although there has been less participation in patents with women inventors in Mexico (44 percent) and India (34 percent) (Figure 7.7).

Figure 7.7 Distribution of women in patent assignee type and by technological field in Mexico, Brazil, and India (by percent).

Source: Authors’ own estimation, based on USPTO data.

During 1997–2013 in Brazil, 401 female inventors were identified, representing 18.5 percent of all inventors. In 2003, the Natura companies and the Johnson & Johnson subsidiary in Brazil had the greatest participation, in many cases with inventor teams composed of several researchers. In 2006, a similar situation was seen with the Foundation for Research Support in the State of São Paulo. There is a clear growing trend, which has been most noteworthy in the past few years and seems to be associated with a higher incidence of larger inventor teams.

During 1980–2015 in Mexico, sixty-one female inventors were identified, representing 5.3 percent of the total number of inventors. Outstanding for their participation are Sabritas (six of the sixty-one), Mabe (three), El Instituto Mexicano del Petróleo (nine), and the Universidad Nacional Autónoma de México (seven). As in Brazil, the trend is for working in relatively large teams. The average number of researchers per patent is three, but as many as five in the chemical, pharmaceutical, and medical sectors. Individual patents are registered only in the electrical and electronic sectors.

From 1997 to 2010 in India, 219 female and 659 male inventors were identified in the 1,208 patents with at least one woman. Over half (53 percent) of the patents had the Council of Scientific and Industrial Research as the patent assignee. Companies that had more than seven patents were Aurobind–Pharma (seven), Dr. Reddy’s Laboratories (nineteen), Hetero Drugs (twelve), Indian Oil Corporation (twelve), Ittiam Systems (fourteen), and Reliance Life Sciences (forty-one). A trend toward working in increasingly larger teams of inventors has been recorded in India, as it is the country with the most effective teams – in fact, with ten inventors in the “others” category and nine inventors in chemical. In Brazil, the largest teams of inventors are in pharmaceutical and medical (nine inventors), chemical (seven inventors), and “others” (seven inventors).

For the field of technology, where patents are classified, we underscore the following evidence. In all three countries, women’s participation stands out in pharmaceutical, medical, and chemical patents. Specifically, nearly two-fifths of India’s women inventors are in the pharmaceutical and medical sector, and 42 percent are in the chemical sector, according to the higher percentage of women graduates in these sectors and their integration as researchers (Table 7A.3). Studies have found a positive correlation between the increasing number of women graduates in STEM careers and research activities and a higher number of female inventors (Giuri et al., Reference Giuri, Grimaldi, Kochenkova, Munari and Toschi2020; Kuschel et al., Reference Kuschel, Ettl, Díaz-Garcia and Alsos2020; Lissoni et al., Reference Lissoni, Montobbio and Zirulia2013). However, this study did not explore such correlations due to a lack of information on each woman.

The distribution of women inventors in Brazil and Mexico is similar in the pharmaceutical and medical sectors – 37 percent and 29 percent, respectively. In the chemical sector, the distribution is 31 percent and 26 percent, respectively. Few women inventors, however, are found in mechanical, electrical, and electronic sectors, with percentages ranging between 2 percent and 4 percent. In the “others” category, differentials are observed among the three countries, with Mexico – where women inventors collaborate on a third of the patents – standing out.

In every patent document, previously consulted patents are recorded as Backward Patent Citations (BwPatCit) to show that the work done by the researchers is state of the art in their technological field. We consider this variable the stock of previous technological knowledge. We estimate this indicator based on the average of BwPatCit per patent where a woman is present. Mexico has the highest average (forty-seven citations per patent), whereas India and Brazil have averages of fifteen and sixteen citations, respectively.

Regarding patent value, we use the number of forward patent citations (FwPatCit) per patent as a proxy variable. We observe that, in all three countries, the importance is still relatively low, especially in India.

India has a notably higher percentage of claims in the pharmaceutical and medical sector and the chemical sector (a combined 83 percent). Both sectors are essential for the other two countries, but with a lower proportion: 66 percent for Mexico and 68 percent for Brazil. The percentages for the “others” category are 29 percent and 19 percent, respectively. There is marginal diversification in the other technology categories.

Table 7.2 shows the distribution of patents with at least one woman inventor by variable, which we consider the nature of innovation, and by technological field. It summarizes what was previously said.

Table 7.2 Women’s participation according to innovation nature variables

		Mexico	Brazil	India
TechField		%	%	%
	Chemical	26	29	42
	Computer and communication	3	4	8
	Drugs and medicals	31	37	40
	Electrical and electronic	3	2	2
	Mechanical	2	4	3
	Others	34	24	6
	Total	100	100	100
$\dot{A}$		%	%	%
	Chemical	51	28	4
	Computer and communication	1	2	9
	Drugs and medicals	8	46	75
	Electrical and electronic	2	2	2
	Mechanical	0	5	2
	Others	38	17	8
		100	100	100
ValuePat		%	%	%
	Chemical	34	29	28
	Computer and communication	0	5	3
	Drugs and medicals	18	45	57
	Electrical and electronic	1	0	2
	Mechanical	0	2	3
	Others	47	18	7
		100	100	100
TechInnScope		%	%	%
	Chemical	38	29	42
	Computer and communication	3	4	7
	Drugs and medicals	28	39	41
	Electrical and electronic	2	3	1
	Mechanical	0	6	3
	Others	29	19	5
	Total	100	100	100
AssigPat		%	%	%
	Individual	2	1	1
	Institution	54	26	65
	Firm	44	73	34
	Total	100	100	100
MobInv		%	%	%
	1	93	90	99
	More than 1	7	10	1
	Total	100	100	100
SizeRT		Average	Average	Average
	Chemical	5	7	9
	Computer and communication	2	6	5
	Drugs and medicals	5	9	9
	Electrical and electronic	1	4	1
	Mechanical	3	6	6
	Others	4	7	10
	Total average	3	7	7
Women by team participation		%	%	%
	Chemical	50	45	35
	Computer and communication	75	43	47
	Drugs and medicals	67	50	40
	Electrical and electronic	100	59	56
	Mechanical	66	49	37
	Others	60	64	36
	Total average	70	52	42

Source: Authors’ own estimation, based on USPTO data.

7.3.3 Hypothesis Research and Model Specification

Considering the historical marginalization of women in S&T activities, we expect that women possess potential abilities for invention-innovation and that their inclusion in the sphere, in co-participation with men, will positively affect the creation of new processes and technological products. The factors associated with innovation that especially affect this propensity are (1) the stock of previous knowledge, (2) the type of patent assignee, (3) the technological field of the patent, (4) the invention scope of each patent, (5) international inventor mobility, and (6) patent value. Therefore, we propose a regression model, which is estimated for each woman inventor in Mexico, Brazil, and India and is specified in the following equation:

$W{m}_{j}PropIn{v}_{ij} =\dot{A},SizeRT,AssigPat,TechField,TechInnScope,ScTech\_links,MobInv,ValuePat$

Where the dependent variable:

$W{m}_{j}PropIn{v}_{ij} =\text{women}’\text{s propensity to invent. It has been estimated as follows:}$

$W{m}_{j}PropIn{v}_{ij} =\text{number of women inventor patents/number of patents where }\text{there is at least one woman/total of patents granted by the }\text{USPTO to Mexican, Brazilian, or Indian holders.}$

According to this estimate, a reduced Wm_jPropInv_ij is shown in the three countries (Table 7.3). Compared with Mexico, India and Brazil are relatively higher. They are similar in all technological fields – except that India is higher in the electrical and electronics sector while Brazil is higher in the pharmaceutical and medical sectors. Mexico is notably lower in almost all sectors, although less so in the pharmaceutical, medical, and chemical sectors.

The independent variables are taken from USPTO patent information. They characterize the inventions. Based on the literature reviewed, these variables allow us to assume how each invention influences Wm_jPropInv_ij (Table 7.4).

Table 7.4 Independent variables and hypothesis

Variable	Variable proxy	It is expected that
$\dot{A}$	Stock of previous knowledge. Number of BwPatCit as a proxy variable	… the higher the BwPatCit, which reflects increased R&D expenditure, the higher is the propensity of women to become inventors (Aldieri et al., Reference Aldieri, Carlucci, Paolo Vinci and Yigitcanlar2019; Duguet and MacGarvie, Reference Duguet and MacGarvie2005; Guzmán et al., Reference Guzmán, Brown and Acatitla2020; Hall, Reference Hall, Fagerberg and Mowery2005; Jaffe et al., Reference Jaffe, Trajtenberg and Henderson1993)
SizeRT	Size of research teams. Number of inventors involved in the generation of the patent	… a larger research team, by incorporating women, increases the diversity of new technological ideas and therefore the propensity of women to become inventors (Bianco and Venezia, Reference Bianco and Venezia2019; Breitzman and Thomas, Reference Breitzman and Thomas2015)
AssigPat	Patent assignee. 1 = Firm; 2 = Institution; 3 = Individual; 4 = Co-patent firms; 5 = Co-patent firm-institution	… the higher number of firms patents a greater probability to scale at the industrial level, to eventually commercialize, and for government and university innovation efforts to crystallize in patents, and therefore contributed to a higher propensity of women to become inventors (Chatterjee and Ramu, Reference Chatterjee and Ramu2017; Giuri et al., Reference Giuri, Grimaldi, Kochenkova, Munari and Toschi2020; Meng, Reference Meng2018; Murray, Reference Murray2004; Whittington and Smith-Doerr, Reference Whittington and Smith-Doerr2008; Woolley, Reference Woolley2019)
TechField	Technological field of the patent. 1 = Chemical; 2 = Computer and communication; 3 = Drugs and medical; 4 = Electrical and electronic; 5 = Mechanical and 6 = Others	… the higher distribution of women inventors by technological fields will be differentiated among countries according to the importance of those fields and the advances in science and technology in each case (Cook and Kongcharoen, Reference Cook and Kongcharoen2010; Hunt et al., Reference Hunt, Garant, Herman and Munroe2012; Kahler, Reference Kahler2012; Maldonado Carbajal et al., Reference Maldonado Carbajal, Chávez and de Jesús Peredo2015; Martínez et al., Reference Lax-Martínez, Raffo and Saito2016)
TechInnScope	Invention scope of each patent. Number of claims as the proxy variable	… the higher the number of claims, the higher is the propensity of women to become inventors (Jensen et al., Reference Jensen, Kovács and Sorenson2018)
MobIn	International inventor mobility. Dummy variable, where 0 = inventors of the same nationality; 1 = foreign inventors	… the higher mobility favors the spillover of codified and tacit knowledge, and therefore contributes to as higher propensity of women to become inventors (Bianco and Venezia, Reference Bianco and Venezia2019)
ValuePat	Value of the patent. This variable specifies the number of patent citations made in successive patents. A proxy variable is the number of FwPatCit obtained	… the larger the FwPatCit, which suggests a wider diffusion of new patents and increased importance of new knowledge, the higher is the propensity of women to become inventors (Bransteter, Reference Branstetter2003; Branstetter and Ogura, Reference Branstetter and Ogura2005; Breschi et al., Reference Breschi, Lissoni and Montobbio2005)

Source: Authors’ own elaboration, based on USPTO data.

7.3.4 Analysis of Outcomes

According to the results of each model, we have partially verified our hypothesis, with a different pattern in each of the three countries. The independent variables – the stock of prior knowledge, the size of inventor teams, the type of patent holder, technological field, and presence of foreign influence – impact positively on women’s propensity to invent (Wm_jPropInv_ij) in a differentiated manner in each country. In some cases, they did not. These results are detailed subsequently and shown in Table 7.5.

Table 7.5 Empirical model outcomes: Factors affecting the propensity of women to invent

Variable	Mexico	Brazil	India
$\dot{A}$	0.0004**	0
SizeRT	0.020***	0.051***	0
AssigPat
Institution	0.020***
Firm	0.081***	0.006	0.0442***
Individual		0.104**
TechField
Computer and communication;	-0.004	-0.023	0.03
Drugs and medicals	-0.007	0.055**	0.024
Electrical and electronic	-0.035	0.157***	0.008
Mechanical	0.039**	0.017	-0.023
Others	-0.031	0.130***	0.019
TechInnScope		0	0.0002***
MobInv	0.140***	-0.083	0.017
ValuePat	0	0	0.045
_cons	0.046	-0.073	-0.05
	0.236	0.163	0.477
N	61	299	728
R2	0.67	0.76	0.68
F	6.89	9.14	3.06

Note: p < 0.01*** p < 0.05** p < 0.010*

Source: Authors’ own estimation based on model proposed.

Concerning the stock of previous knowledge, $\dot{A}$ has been statistically significant only in Mexico. According to our elasticity estimation (0.012), if the number of BwPatCit increases by 10 percent, the propensity for women to invent grows by 0.12 percent. In the tradition of Griliches (Reference Griliches1990), several authors have used BwPatCit to study knowledge flows and FwPatCit to analyze the invention value (Gay et al., Reference Gay, Latham and Le Bas2008; Lerner and Seru, Reference Lerner and Seru2017, among others). An essential aspect to consider, however, is when it is used as a proxy value for knowledge links among inventors to explore the nature of knowledge flows and the factors affecting these flows (Jaffe and Rassenfosse, Reference Jaffe and de Rassenfosse2017). In this research, we consider BwPatCit as the stock of previous knowledge upon which inventors develop their new ideas. As references to prior technology that has been used or on which current patents build (Hall, Reference Hall, Fagerberg and Mowery2005), BwPatCit involves an R&D effort, which, when increased, makes it possible to include women in inventive activities. The fact that it has positively impacted Wm_jPropInv_ij is a reminder of the need to disclose patents, just as they are meant to be.

Relative to SizeRT, there is a positive influence in Mexico and Brazil. When the research team is 10 percent bigger, Wm_jPropInv_ij increases 4.9 percent (Mexico) and 10.9 percent (Brazil). Bianco and Venezia (Reference Bianco and Venezia2019: 14) state that “the presence of more members on a team certainly brings more diverse and variegated knowledge and can therefore produce better results.”

Contrary to the importance of institution holders in Mexico and India, firms have a positive influence on Wm_jPropInv_ij in both countries (0.08 and 0.04, respectively). Institutions have been positively significant for Mexico as well, but only individual assignees are positively significant in Brazil. The reason for a positive effect is understandable considering the importance of public R&D expenditure in Mexico and India, even though such expenditure remains marginal in Mexico. Women academic inventors are key to understanding their involvement in scientific research – and eventually the discovery of new products and processes that could make it to the productive sphere, depending on the degree of links between firms and institutions. According to Murray (Reference Murray2004: 643), “The first element that the firm may leverage is the academic’s local laboratory network – a network to current and former students and advisors established by the inventor through his laboratory life.”

Although Martínez et al. (Reference Lax-Martínez, Raffo and Saito2016), in their study of women inventors from 182 countries with PCT patents, do not analyze the causal effects of ownership on Wm_jPropInv_ij, they have identified that an average of 48 percent of women participate in the academic sector, while only 28 percent are in the business sector. Their findings coincide with previous studies (Whittington and Smith-Doerr, Reference Whittington and Smith-Doerr2008). It is pointed out that China, Brazil, and Spain have higher percentages of PCT patent applications with women inventors in the academic sector (around two-thirds). In particular, Mexico has 69 percent participation in the academic sector and 26 percent in firms (Martínez et al., Reference Lax-Martínez, Raffo and Saito2016). Unlike India, Mexico has few entrepreneurial businesspeople, and technological dependence dominates every sector. Women in India have increased their presence in business activities, and they discuss the challenges that they have to face to reduce inequalities (Chatterjee and Ramu, Reference Chatterjee and Ramu2017). In Brazil, however, the increase in R&D efforts surpasses the inventive activity of institutions toward companies (Maldonado Carbajal et al., Reference Maldonado Carbajal, Chávez and de Jesús Peredo2015).

Regarding the technological field, the mechanical sector is statistically significant for Mexico, while the fields of pharmaceutical and medical, electrical and electronic, and others are statistically significant for Brazil. In India, however, not a single field has an impact on Wm_jPropInv_ij. We have confirmed the results of previous research on the positive influence of the pharmaceutical and medical sector in Mexico and Brazil, but we are now modeling each woman inventor and not each patent (Guzmán and Orozco, Reference Guzmán, Orozco and Perrotini-Hernández2011; Maldonado Carbajal et al., Reference Maldonado Carbajal, Chávez and de Jesús Peredo2015). The current study is also connected to the importance achieved by female graduates and researchers in medicine and related disciplines. Our findings coincide with the study by Cook and Kongcharoen (Reference Cook and Kongcharoen2010), where advancement by women in life sciences places them in the realms of innovation and marketing. The involvement of women inventors in different technological fields is one of the topics that have been addressed more in studies, especially with a focus on identification. Some of these studies find that technological fields differ across countries (Martínez et al., Reference Lax-Martínez, Raffo and Saito2016) and are probably associated with the country’s technological specialization, as in certain Latin American countries where chemistry and metallurgy are the main sectors (Sifontes Fernandez and Morales Valera, Reference Sifontes Fernandez and Morales Valera2014). Few studies have thus far analyzed the impact of factors relating to the nature of innovation on the propensity of women to invent, let alone provide analysis based on a micro-level model that considers each woman inventor. Hunt et al. (Reference Hunt, Garant, Herman and Munroe2012) find that to close the gender gaps in patenting, it is essential for the participation of women in physics and engineering to grow. That could increase the GDP by 2.7 percent if we keep in mind GDP growth among countries in the long run and that patents are found among explanatory variables of the country.

In economic literature, mobility has been detected as a factor favorably affecting innovation. In Mexico, we find that the presence of foreign inventors has a positive influence on research teams; when there is a 10 percent increase in foreign inventors, Wm_jPropInv_ij grows by 0.5 percent. The results reinforce the finding of Bianco and Venezia (Reference Bianco and Venezia2019: 14) that “the presence of external inventors broadens the scope of the patent, [and] longer working experience of inventors affects technological and market value, whereas inventors who have already patented in the past develop new product architectures, with broader scope and higher scientific value.” The results also confirm consistency with contributions by previous authors, underscoring that openness and experience positively influence the innovative capabilities of teams. In India, female researcher mobility tends to occur outwardly, and knowledge spillovers take place in research centers with other colleagues. The need for such mobility finally finds channels of communication in India.

We cannot, however, confirm the influence of patent value for any country. The fact that the countries studied are emerging may mean their innovations are essentially incremental as they follow leading countries, so they are not as well recognized as industrialized countries yet. Patent value is still low, but it might be interesting to study Indian, Brazilian, and Mexican inventors collaborating on patents in developed countries and identify the contributions of women therein. New technology patented by mixed teams is often cited in the following patents. This suggests that diversity of ideas could lead to the development of patents that are more useful and consequently more successful (Hunt et al., Reference Hunt, Garant, Herman and Munroe2012).

8.2.1 The Proposed Model

The proposed model would offer inventors a novel form of protection with no need for registration. Inventors who wish to secure full patent protection would still have to register their inventions and undergo an examination process, just as they do today. The central difference between the existing form of registered rights and the proposed form of unregistered rights is the scope and duration of protection afforded under each. The proposed unregistered patents would protect subject inventions for only three years and against only knowing and direct copying. Thus, unlike registered patent rights, these unregistered rights would provide relatively short protections and no protection against independent creation.

Note that the distinction in rights between unregistered and registered patents follows naturally from the rationale behind the registration and examination of traditional patents. Registration has long been seen as providing the public with notice of in rem rights.Footnote ⁵⁰ Because the proposed unregistered patent system by definition would not make registration and description of patent rights public, it would be inefficient and arguably unfair for potential infringers to be held unknowingly liable for infringement. Those who directly copy an invention, by contrast, obviously would know that it is another’s creation and, therefore, would notice that another person could hold unregistered rights in the invention, despite the lack of registration of those rights.

Moreover, inventions eligible for our proposed unregistered rights still would need to meet the existing patentability requirements of subject matter eligibility,Footnote ⁵¹ novelty,Footnote ⁵² utility,Footnote ⁵³ and nonobviousness.Footnote ⁵⁴ Only for inventions meeting these traditional substantive standards would unregistered protections become available. Those protections would then automatically arise as soon as the invention becomes available to the public. Public availability would be measured under the same standards used for novelty under current U.S. patent law, including public availability of a description of the invention, public use of the invention, or the invention becoming otherwise available to the public.Footnote ⁵⁵

Unlike the twenty-year term for a registered patent, however, the duration of an unregistered patent under our proposal would run for only three years from the date the subject invention first became publicly available. Inventors who wish to extend their rights for more than three years would have the option to do so but would have to undergo the usual registration and examination with the applicable patent office, even if their unregistered patent rights had already been successfully enforced in court. Inventors would have to file for such registered rights no more than a year after their unregistered rights attach; inventors who do not file applications for registered rights within a year of acquiring unregistered rights would forfeit registered patent protection altogether, leaving them with only the remaining two years of their unregistered patent rights.

In this way, the proposed unregistered rights regime would comport with what many patent systems already do under their novelty (or statutory bar) provisions. Almost all countries allow inventors to disclose their inventions publicly up to a year before filing their patent applications without anticipating themselves or otherwise being barred from patent eligibility.Footnote ⁵⁶ The law in the United States, for example, establishes a one-year grace period that allows inventors to disclose their invention publicly before filing, in effect measuring novelty and nonobviousness as of the disclosure date rather than the filing dates.Footnote ⁵⁷ Our proposal applies this “first to file or to publicly disclose” approach to both registered and unregistered rights holders.Footnote ⁵⁸ Granted, even under U.S. patent law, patentees cannot sue others for infringement occurring in that first prefiling year the way they would be able to under our proposal. Nonetheless, the effect is that under both systems, an inventor who applies for registered patent rights within a year of publicly disclosing her invention can exclude others until the twenty-first year after that initial prefiling disclosure.

Finally, our proposal would allow inventors only a limited presumption of validity of their unregistered patent rights. In bringing an enforcement action, inventors initially would not need to prove that their inventions meet the standard patentability requirements; instead, the inventors would have to establish only the date their invention became public. They would, however, at all times bear the burden of proving direct copying by the defendant. Once the inventors establish these basic elements of their claim, the defendant would have the right to challenge the validity of the inventors’ unregistered rights by showing that their invention fails to meet one or more of the patentability requirements. Importantly, the defendant would need to rebut the presumed validity of an unregistered patent by mere preponderance of the evidence, unlike the clear and convincing evidence required under U.S. law to invalidate a registered patent.Footnote ⁵⁹ The lower preponderance of the evidence standard would reflect the fact that no patent office had yet vetted the invention’s patentability and that a court would thus have no need to defer to the patent office’s expertise.Footnote ⁶⁰ If the proposed unregistered patent rights were found to be valid as well as infringed, the remedies for infringement would include both injunctions and damages.

However, it may seem odd to grant unregistered patents any presumption of validity, given that neither unregistered trademarks nor copyrights enjoy such a presumption. In both trademark and copyright law, owners of unregistered rights bear the burden of proving the validity of their rights.Footnote ⁶¹ Shifting the burden to alleged infringers may be more efficient in the patent context, however. The exceedingly rigorous standards for patentability make it relatively easy for defendants to identify prior art references that prove unpatentability, even with regard to registered patents previously examined by a patent office.Footnote ⁶² Allowing a defendant to proceed immediately to its invalidity argument may thus lead to faster and less costly disposition of infringement proceedings.

8.2.2 The Costs and Benefits of Unregistered Patents for Gender Equality

Two critical questions our proposal presents are whether unregistered patent rights would in fact help female inventors and, if so, whether the benefits would outweigh the costs of creating such a system. As noted earlier, simply decreasing the barriers women face in obtaining patent protections would go far in helping them to capitalize on their innovations and to participate in technological industries generally. While our proposal would increase the number of patent rights overall, the proposed unregistered rights would be carefully cabined to avoid creating excessive drag on future innovation. Our proposal is thus tailored to ease many of the disadvantages women face in protecting their inventions while avoiding undue burdens on technological progress overall.

First, an unregistered patent regime would enable inventors to obtain patent protection without the oft-prohibitive costs of registration, examination, and maintenance. Women would thus have easier access to the patent system despite frequent difficulty in securing funding. Similarly, unregistered patent rights under the proposed regime would be automatic and therefore would not favor those who have access to a network and other support for guidance on the patenting process. This also will help women and others who lack such connections.

Of course, inventors would have to bear the costs of enforcing their patents, including any litigation costs. The policing costs for infringement of unregistered patents would be lower than those for registered patents, however, as monitoring for only purposeful copying is presumably cheaper than monitoring for both knowing and “innocent” infringement, such as independent design. More importantly, enforcement costs would have to be born only ex post, after direct copying has been detected, and the value of the invention thereby demonstrated. That value – along with the protection the invention would automatically enjoy under unregistered patent rights – could in turn help attract investors. As with any other invention, venture capital and other investment funds can be used for defending rights in the invention through litigation or other means, filing for registered rights and paying for subsequent maintenance fees, and commercialization and licensing generally.

All the same, investors may be hesitant to invest even in innovations valuable enough to have been copied if they lack registered patent rights. The proposed unregistered rights would not enjoy the cachet of vetting by a patent office or the presumption of validity rebuttable only by clear and convincing evidence that comes with such vetting. The proposed rights also would last for only three years. As a result, unregistered patents might not have the same signaling effect in attracting investment.

Investing in unregistered patents may not be significantly riskier than investing in registered patents, however. Investing in technological innovation is always risky, as even registered patents have only speculative economic value and probabilistic legal validity.Footnote ⁶³ Under our proposal, outside investors also would know that even unregistered inventions would have at least some protection, compared to their current lack of protection. Such protection could be converted to registered protection if still within the one-year grace period. Moreover, many technologies do not need the full twenty years of registered protection. The three years of protection under our proposed unregistered rights could be more than enough to recoup investments in industries such as computer software and electronics.Footnote ⁶⁴

A second and equally important benefit is that automatic, unregistered patent rights would allow women to bypass many of the biases inherent in the patent system. The most obvious are patent examiner biases against patent applicants with recognizably feminine names. Less obvious are workplace biases leading to less support for women in filing for patent rights. Automatic rights would make these biases less important in barring access to patent protection.

That being said, unregistered rights holders seeking to enforce their rights may need to rely on the court system, which itself is subject to bias. Allowing courts to recognize unregistered intellectual property rights nonetheless could be more equitable than forcing inventors to go through the patent registration and examination process or forgo their rights altogether. Like investors, courts may be more apt to recognize the value of an invention once others have affirmatively copied it and therefore be more inclined to protect the invention against appropriation. Courts also may be more inclined to look to broader economic and social values to evaluate patent rights with more focus on equity than a patent office would.Footnote ⁶⁵ Administrative agencies such as patent offices, by contrast, lack many of the procedural justice and due process constraints to which courts are subject.Footnote ⁶⁶ Members of the judiciary may be less subject to the biases affecting the science, technology, engineering, and mathematics (STEM) fields from which patent offices typically draw their examiners.Footnote ⁶⁷ While administrative agency decisions are subject to judicial review, this review often tends to be somewhat deferential.Footnote ⁶⁸ Thus, using courts to recognize unregistered intellectual property rights would at least provide an alternative venue for women and other inventors to protect their interests.Footnote ⁶⁹

What are the costs and pitfalls of the proposed new regime of unregistered patent rights, however? The fact that no jurisdiction has ever provided unregistered, automatic patent rights reflects a discomfort with the idea. Many have argued that excessive patent rights would impose much greater burdens on future innovation and creation than would similarly excessive grants of copyright, trademark, design, and trade secret rights. The breadth and robustness of patent rights mean that granting patents too lavishly or easily could lead to hold-outs, patent thickets,Footnote ⁷⁰ patent “trolling,”Footnote ⁷¹ and other phenomena that wastefully deter other inventors from valuable research. Technologies in which development and commercialization are inherently cumulative and complementary are particularly prone to these issues.Footnote ⁷²

In contrast to other forms of intellectual property right, patent law also has very few safety valves to protect inventors from opportunistic claims of infringement that might over-deter them from building on existing technologies.Footnote ⁷³ Unlike copyright and trade secrecy,Footnote ⁷⁴ patent law has no independent creation defense to infringement liability,Footnote ⁷⁵ nor does patent law have the fair use defense seen in copyright and trademark law.Footnote ⁷⁶ Even the experimental use exception available in the patent systems in many countries applies only in limited circumstances, such as regulatory approval of pharmaceuticals or use of medical treatment methods.Footnote ⁷⁷ Although prior user rights also are common in many countries, they are limited to only those using an invention commercially before the patentee filed its patent application.Footnote ⁷⁸

Instead, the patent system relies on its comparatively stringent limitations to reduce both the number of patents and breadth of patent rights.Footnote ⁷⁹ Patent law has far more limited terms than copyright, trademark, design, or trade secret law,Footnote ⁸⁰ as well as novelty and statutory bar provisions that force inventors to file their patent applications or lose their eligibility to patent.Footnote ⁸¹ These limitations serve to release inventions into the public domain as soon as possible. Unconditional patent rights granted without examination or even registration appear at first glance to completely dismantle patent law’s carefully calculated system for cabining patent law’s otherwise robust rights of exclusivity.Footnote ⁸²

The unregistered rights proposed here would be very closely limited to avoid unduly burdening overall innovation, however. Unlike registered patent rights, the proposed unregistered rights would be subject to a number of safety valves to protect downstream innovation. In addition to applying all of the standard patentability requirements, the proposed unregistered patent regime would protect only against copying, not independent creation, and for only three years from the date the subject invention first became publicly available. These constraints establish significant safeguards against placing undue burdens on others.

Some might argue that unregistered patent rights would exacerbate the innovation bottlenecks created by “bad” or “weak” patents – that is, patents of suspect validity.Footnote ⁸³ As with registered patents, however, the proposed unregistered rights would be subject to judicial review for validity. Our proposal in this way invokes the same efficiencies of various proposed “soft-look” registration approaches. Under these latter proposals, patent applications would undergo only minimal administrative examination prior to issuance, and only those that were important enough to give rise later to litigation would then face more substantive judicial examination.Footnote ⁸⁴ Such judicial review ex post, only when needed, is more resource-efficient than the current system of administratively examining all patent applications ex ante. In addition, judicial review ex post would have the benefit of not only the defendants’ field-related expertise and motivation to invalidate patents but also a presumption of validity rebuttable by a mere preponderance of the evidence.Footnote ⁸⁵ Registration-only patents rest on the idea that the high costs of eliminating bad patents administratively ex ante outweigh the social costs of waiting to eliminate those patents judicially ex post.Footnote ⁸⁶ Similarly, the unregistered patent regime proposed here rests on the idea that the cost of patent registration and examination and how it sidelines women and other inventors outweighs the cost of potentially bad unregistered patents.Footnote ⁸⁷

Regardless of whether patents are “good” or “bad,” simply adding to the overall volume of patent rights still could impose a drag on innovation. Technology may be more incremental, cumulative, or complementary than expressive works, trademarks, or trade secrets,Footnote ⁸⁸ and inventive concepts are often more difficult to design around.Footnote ⁸⁹ The fear of litigation and the uncertainty of proving invalidity therefore often causes competitors to fear innovating in heavily patented technologies.Footnote ⁹⁰ This in terrorem effect can over-deter competitors from innovating near or around others’ patents, particularly given the patent system’s lack of independent creation, fair use, and other defenses available in other forms of intellectual property right.Footnote ⁹¹ In this way excessive patent rights may be more likely than other intellectual property rights to over-deter future innovation.

The proposed unregistered patent regime would provide a number of safe harbors to address this problem, the most obvious of which would be for independent creation. Cumulative or complementary innovation might still be unduly deterred from copying needed components, however. So unlike the current patent regime in the United States, the proposed regime would therefore not impose treble damages for knowing infringement, especially if only for experimental use. Perhaps most importantly, the proposed unregistered rights would be very limited in duration; other innovators could thus just wait three years to copy an invention with impunity. The limited duration and scope of these unregistered rights would greatly lower the risk of patent trolling, nuisance suits, patent thickets, and holdouts. The proposed unregistered rights would thus broaden access to patent protection without either increasing or prolonging them.

As an unregistered regime, our proposed patent rights could undermine the public-notice function of patent registration, however, thereby increasing the informational costs of establishing freedom to operate within a particular technological space. The breadth of patent rights and the relatively limited defenses to patent infringement call for clear public notice of what constitutes such infringement. Indeed, this is one of the foremost values of registering patent rights and perhaps the primary reason that, in contrast to copyright, trademark, and trade secret law, unregistered rights in patent law never emerged. The patent registration system already falls short of this public-notice ideal, as the clearance costs of identifying and interpreting all the patents potentially relevant to one’s project are often astronomical.Footnote ⁹² Injecting unregistered, unrecorded rights into the existing system could exacerbate these costs exponentially. Under the proposed unregistered rights system, by contrast, infringers would have demonstrated actual notice of the protected invention by virtue of copying it. Public notice, therefore, would not be an issue. Downstream inventors would have to determine the date the invention became publicly available, of course, which could be costly, but this cost is already part of the existing registered rights system.Footnote ⁹³

One final objection could be that the proposed unregistered regime would undermine the peripheral-claiming system.Footnote ⁹⁴ In the United States, peripheral-claiming mandates that registered patents include claims “particularly pointing out and distinctly claiming” the subject invention.Footnote ⁹⁵ Peripheral claiming is designed to delineate a patent’s boundaries as precisely as possible to give the public detailed notice. An unregistered patent system, by contrast, would give the public only the inventions themselves as indicators of their patent boundaries in a way resembling the now-obsolete central-claiming system. Under the central-claiming system in the United States, courts could only compare an allegedly infringing device with the patented invention, without the benefit of claims or other express boundaries, often leading to surprise and uncertainty.Footnote ⁹⁶ Reversion to a central-claiming-like system may thus seem inefficient,Footnote ⁹⁷ but in practice the modern-day peripheral-claiming system introduces its own inefficiencies.Footnote ⁹⁸ Peripheral claiming has led to the notoriously difficult problems of construing patent claims and widespread complaints about the uncertainty to which they lead.Footnote ⁹⁹ Indeed, some commentators have even called for a return to the central-claiming system as a more equitable and flexible approach to patenting.Footnote ¹⁰⁰

Lastly, none of what is proposed here should be taken as urging the abolition of existing registered patent systems. Without a doubt, registration and examination of patents yield many benefits, including greater public notice and independently vetted patentability. Registration and examination come at a high cost for inventors, but for many inventors – and their employers or investors – this cost will not pose an obstacle to patenting inventions perceived to be worthwhile.Footnote ¹⁰¹

For all its benefits, however, the high cost of the patent application and examination process falls disparately on women and other disadvantaged inventors. Predicting the commercial value of an invention is immensely difficult, and deciding whether to invest in the process of applying for patent protection, even if only provisionally, is a gamble.Footnote ¹⁰² To assume that investors and employers evaluate inventions from a purely rational, rent-maximizing perspective is undoubtedly inaccurate, especially for inventions by female inventors. As a result, a disproportionately large number of female inventors are disenfranchised by the costs and complexities of patent registration and examination and their inherent biases.Footnote ¹⁰³ Adding an unregistered regime such as the one proposed here would help ameliorate the gender gap in patent rights without creating undue burdens on innovation.Footnote ¹⁰⁴ The benefits of unregistered patent protection, even if for a relatively brief period, could be quite significant and thus outweigh the potential risks of such a regime.

9.2.1 The Problems in a Centralized Patent Regime

The most apparent distributive concern in the patent system involves high entry costs.Footnote ³⁴ Patent prosecution has become increasingly expensive, partially due to the prolonged processes at the PTO, which compound legal and other expenses, and partly because the complexity of these processes requires expensive professionals to navigate them.Footnote ³⁵ Patent prosecution may be even costlier for small inventors, who, among other things, do not have in-house patent attorneys.Footnote ³⁶

The financial barrier to the patent system prompted Congress to lower the filing fees for small firms and micro-entities through the Leahy-Smith America Invents Act.Footnote ³⁷ This statute also entitles small entities and independent and unrepresented inventors to technical, pro bono, and pro se assistance.Footnote ³⁸ Despite such measures, inequality seems to worsen with time.Footnote ³⁹ The reduced fees have not moved the needle, most likely because these fees are negligible in the overall expenses of patent prosecution. Also, the rate of abandoned applications is more than twofold for pro se applicants than for represented applicants.Footnote ⁴⁰

High entry costs confine small companies and individuals more than well-resourced corporations. The most obvious result of costly patent prosecution is that small inventors give up on patents. While avoiding patenting may be rational for small entities,Footnote ⁴¹ it exposes their inventions to copying and may yield an inferior position in the marketplace.Footnote ⁴² Indeed, the rise of trade secrets protection provides alternative, cheaper protection that often suits better the financial capabilities, focus, and business of small entities and individual inventors.Footnote ⁴³ Yet, if patents exist as a tool to generate an incentive to create, then the exclusion of certain groups from patenting adversely affects their incentives to invent.Footnote ⁴⁴ But high costs can be even more devastating; small inventors depend on investments, but investors are reluctant to fund patent prosecution and often delay investments until patents are secured.Footnote ⁴⁵ Consequently, inventors may be forced to relinquish their ventures altogether or settle for lower valuations than their invention is worth.Footnote ⁴⁶

Consider now the backlog of the patent system. Currently, the PTO examines all patent applications alone, generating severe delays.Footnote ⁴⁷ A high pendency rate has a differential effect on the market. For small players, pendency shrinks the duration and thus the value of patent protection.Footnote ⁴⁸ Big players, in contrast, can make lemonade out of the long application process lemon. Such players can file numerous broad patents, including for embryonic ideas that did not go through full development or feasibility testing. They can then strategically use the “patent pending” status of their applications to persuade others to back away from their asserted territory during the long examination period, even if the application would eventually be denied or reduced in scope.Footnote ⁴⁹

The backlog at the PTO potentially affects the incentives of examiners as well.Footnote ⁵⁰ While probably unintended, proxies for validity – such as inventors’ names, former acquaintance with the representing agent, and institutional affiliation – are almost inevitable considering the severe examination backlog at the PTO and the time constraints imposed on examiners.Footnote ⁵¹ These biases are compounded by the notorious public choice problems at the PTO that further the interests of large corporate patent filers and other repeat players at the expense of smaller inventors.Footnote ⁵² Institutional capture easily penetrates into the individual level of examiners, who may be contemplating their next career step as patent prosecutors or major patent holders.Footnote ⁵³

The third challenge that the extant regime poses to equality concerns the prevalence of low-quality patents. The issue of patent quality has long troubled patent scholars.Footnote ⁵⁴ From the equality prism, the concern is that the PTO is biased toward granting patents, de facto disadvantaging small inventors who file fewer patents and must penetrate a dense screen of patents to function. Remarkably, Alberto Galasso and Mark Schankerman find that innovation by small firms is often triggered by invalidating large firms’ patents, suggesting that those patents thwart innovative efforts.Footnote ⁵⁵

A key reason for the over-grant bias is that the PTO has limited access to the information and proficiency required to disqualify patents and insufficient time to acquire such information and proficiency.Footnote ⁵⁶ Indeed, patent applications must be granted absent reasons to deny them.Footnote ⁵⁷ But assessing applications’ novelty and nonobviousness typically requires highly contextual, fact-intensive, industry-specific information (to disqualify patents based on a lack of novelty) and very specific expertise (to deny patents based on their obviousness).Footnote ⁵⁸ Such knowledge is naturally dispersed in society rather than concentrated in one agency.Footnote ⁵⁹

What is more, the incentives of the PTO are set up to make it an ineffective gatekeeper for patent quality. The PTO is largely self-funded by patentees’ fees, generating an institutional disincentive to reject patents.Footnote ⁶⁰ Studies even indicate that the PTO grants more patents in areas where fees are higher.Footnote ⁶¹ To avoid time-consuming disputes, a swamped PTO is also better off granting borderline patents, particularly to applicants who will likely challenge rejections.Footnote ⁶²

Finally, another way that a centralized patent system contributes to inequality concerns information problems in the patent record. Naturally, a PTO-operated patent record can only include information that the PTO possesses at the time of publication.Footnote ⁶³ But such a record is gravely deficient. The current record lacks information about the background science of the invention and any developments since it was filed. Neither does it include information relating to litigations or transactions, such as available licenses and prices.Footnote ⁶⁴ However, its immense value, including robust information in the patent record, is unrealistic under a centralized model. The PTO has no access to the information that is most relevant to make the record useful for the inventive community. Much of this information is also dynamic and is changing over time. A central agency cannot realistically be responsible for collecting and registering updates to the record.

An inhibited record has distributive effects. Highly capitalized inventors can overcome information deficiencies by investing in other methods to acquire information, which is cost-prohibitive for small players.Footnote ⁶⁵ But running an underinformed business involves market and legal risks. The main risk concerns exposure to “patent trolls” – nonpracticing entities who accumulate vast portfolios of dubious patents and use them for strike suits against innocent infringers.Footnote ⁶⁶

These problems did not go unnoticed, and Congress, courts, and the PTO have attempted to tackle some of them.Footnote ⁶⁷ Yet, high costs, congestion, information problems, and capture are inherent to a centralized PTO and will render any solution that maintains the current structure partial at best. Next, I explore whether decentralizing some PTO functions can provide a better way forward.

9.2.2 Can Decentralization Provide a Solution?

Imagine that the power to write information into the record was not exclusive to the PTO but distributed among state and market actors. The patent record would be built on a blockchain or another platform that enables the autonomous sharing of information by authorized parties.Footnote ⁶⁸ Such parties would be able, and in some cases obliged, to update the record with relevant information during the examination process and throughout the duration of the patent. Under a decentralized model, third parties and scientists could submit prior art and weigh in on obviousness for the use of examiners during the examination period. After patents are issued, patentees could include in the record commercial information, such as licensing and pricing information, and even plug-in smart contracts. Courts would list decisions and outstanding cases that pertain to the patents. These updates would inform the patent community whether patent claims are in question, whether the patent holder is prone to litigation, and other useful information that is costly to obtain under the extant regime. The PTO would be responsible for reviewing the patent record and fixing errors.

In my previous work, I tackled the practical aspects of this proposal and the related concerns.Footnote ⁶⁹ The main concern revolves around whether there are adequate incentives for market players to contribute to a decentralized record. This is a critical point. After all, a functioning patent record is a public good. Without private incentives to participate in it, the record would be unable to achieve its goals.

The first answer to that concern is that some contributions to the decentralized record would be automatic or mandatory and would be independent of the participants’ goodwill. For example, patent applications would be fully published, with automatic references to similar materials on the record;Footnote ⁷⁰ information about litigations would be added by courts as created, and patentees would be obliged to maintain the accuracy of the information on available licenses. This information would already be a substantial improvement over the current regime.

Consider now the incentives for market players to participate in the system voluntarily. It is relatively easy to see the incentives for patentees to add licensing information to the record. Adding such information would enable patentees to commercialize their inventions without much effort. Some patentees, such as big pharmaceutical companies, may prefer to continue licensing their patents individually.Footnote ⁷¹ Yet, this functionality can be extremely valuable for small companies or individual patentees that do not have resources to negotiate each license individually, as well as for owners of vast patent portfolios who can build a cost-effective licensing model.

How about the interests of third parties to contribute information to the record? During the examination process, firms would be incentivized to question the validity of patents that would curb their free operations. From their point of view, this mechanism would be a cheaper, easier way to battle a patent than postreview examination, let alone litigation.Footnote ⁷² The concern may be that such parties would be too eager to provide invalidating information in a way that would jeopardize the efficiency of the system. My previous work addressed this concern.Footnote ⁷³

Other contributors, such as academics or scientists, may be motivated to participate by other considerations, such as reputation or exposure to new information in their field. Volunteer reviewers are the regular case in many academic frameworks, such as reviews of submissions to scientific journals and grants. It would thus not be surprising to see scientists weigh in on an invention in their field.Footnote ⁷⁴ At the postexamination stage, technology firms and scientists would be incentivized to update and consult the record as an affordable mechanism to avoid infringements, as well as to learn of new inventions and form collaborations.

Decentralizing the patent record could transform the record from a static database that includes basic filing information that is barely used in the industry into an up-to-date platform that forms a central tool in the innovation economy. This strategy can potentially improve equality in the patent system in three main ways. First, it would tackle distributive concerns in the patent examination process. Second, it would protect small inventors from information problems. Third, it would boost the commercialization of patents, particularly for small inventors.

Consider first how decentralization can cure distributive harms in the examination process. As discussed, the patent prosecution process is done ex parte under the current regime. However, market actors hold superior information, such as knowledge of instances where the invention was offered for sale or disclosed publicly.Footnote ⁷⁵ The PTO alone is entrusted with raising arguments to reject the patent or narrow down its claims.Footnote ⁷⁶ It performs this task in a suboptimal manner, under time and knowledge constraints.Footnote ⁷⁷ Decentralization would make invalidating information handily available to examiners and reduce the rate of low-quality patents, which form entry barriers for small players. A broader information basis would also save examiners search time and attenuate their need to rely on proxies, which have discriminatory effects.

Even greater equality can be found in the postexamination period. An updated patent record would provide a cost-effective means to gain useful technological, transactional, and legal information throughout the duration of the patents. Making robust information available to small players levels the playing field between them and big players, rendering it less likely that entrenched firms could learn about patents and new players in their field more than start-up ventures.Footnote ⁷⁸ Tackling information problems would also enable innovators to avoid infringements, thus battling patent strike suits, which can devastate small entities.

Finally, decentralization of the record would introduce a platform to commercialize inventions globally. Patentees could list licensing options, potential use cases, and prices on the patent record. Once the record contains such robust transactional information, cooperation with commercial bodies to bring the invention to the market is much more likely.Footnote ⁷⁹ A distributed registry could also enable functionality not existing today, such as plugging in smart contracts and allowing third parties to act on available licenses automatically.Footnote ⁸⁰ This functionality can be extremely valuable for small companies or individual patentees that do not have resources to negotiate each license individually.

The benefit of this policy is twofold. First and foremost, the dynamic record can further the goal of the patent system – to accelerate the path of innovation to the market.Footnote ⁸¹ Second, improved commercialization would have a dynamic effect. It could spur innovation and patenting because patents would become an effective reward for inventors.Footnote ⁸²

10.4.1 Asymmetrical Motivations

Constitutional ideas diffuse across borders, systems, and contexts.Footnote ⁷³ David Strang defines diffusion as “the process by which the prior adoption of a trait or practice in a population alters the probability of adoption for the remaining non-adopters.”Footnote ⁷⁴ In their study of the diffusion of 108 rights across 188 countries in sixty-one years, Benedikt Goderis and Mila Versteeg found evidence of diffusion to particular groups of countries. For example, countries with a common aid donor or a common colonizer suggest that constitutions are affected by coercive pressures from aid donors and former colonizers.Footnote ⁷⁵ Diffusion can also be found among countries that compete for foreign capitalFootnote ⁷⁶ or aim to be acculturated into the constitutional norms of world culture.Footnote ⁷⁷ In these situations, unequal power relations dictate the content and structure of many norms that found their way into constitutions, regardless of ideological preferences that aim to protect cultural diversity and local legal history. When intellectual property is at stake, these diffusion channels exhibit how unequal power relations result in countries adding to their formal constitutions rights and duties that they cannot protect.

Various pathways define foreign influences and mechanisms of rights diffusion in constitutional design. The common mechanisms employed are coercion, learning, competition, and acculturation.Footnote ⁷⁸ The logic behind competition suggests that states strategically imitate foreign constitutions in order to attract foreign capital. Competition is the rivalry between two or more states for material benefits.Footnote ⁷⁹ Thus, countries act strategically in order to attract foreign capital, design a set of rules, and establish institutions that signal to investors and buyers that the local market can accommodate their interests, limit economic risks, and provide stability. Constitutional documents and lists of fundamental rights provide one of the ultimate sources of good signals to alleviate foreign investors’ concerns. Contemporaneous observers have found that a high level of protection of basic human rights renders countries more attractive to foreign investment.Footnote ⁸⁰ Investors believe that “regimes with strong human rights records are typically stable ones”Footnote ⁸¹ and that public reception will be better if their company gains a reputation that promotes fair trade, does not engage child labor, or builds where basic income is unavailable. Thus, when governments offer a strong and detailed list of rights in a constitution, they believe it “may attract economic benefits.”Footnote ⁸² Property and intellectual property as constitutional human rights are important to foreign investors. In other words, when a state constitutionalizes intellectual property rights in its highest formal legal script, it invites investors to view its commitment to their rights.

Another example of how constitutional ideas diffuse across the border is the “learning” process. Learning means that “countries borrow each other’s constitutional provisions because the constitutional choices of others have altered their preexisting beliefs: they adopt certain arrangements only when they are convinced that these will be beneficial.”Footnote ⁸³ Learning raises interesting issues with regard to intellectual property laws. Countries are open to learning from foreign sources, mainly because “the new global obligations from the treatment of intellectual property are transmitted from the international to the national level.”Footnote ⁸⁴ Intellectual property rights

Ruth Okediji further reminded us how these processes affect weaker countries:

The above reflects how “states respond to cultural forces”Footnote ⁸⁷ and how, in doing so, they often emulate foreign constitutional templates irrespective of their content, unsuitable dialectics to the local culture, or even their ability to apply them locally. This diffusion mechanism is defined as acculturation – denoting, as Ryan Goodman and Derek Jinks explain, “the general process of adopting the beliefs and behavioral patterns of the surrounding culture.”Footnote ⁸⁸ Acculturation is different from coercion or competition because it explains how states act in order to reap social benefits. The logic behind acculturation is premised on organizational sociology, implying that organizations adopt models “not because of their functional utility but because of their legitimacy and the social relationships they represent.”Footnote ⁸⁹ Acculturation brings states to sign international human rights agreements and adopt environmental policies and other trade treaties without intending to comply with their requirements.Footnote ⁹⁰ For example, as discussed earlier, the number of constitutions that include provisions on gender equality is increasing. However, as MacKinnon has found, the existence of such a provision is disconnected from its practical applicability.Footnote ⁹¹ What states achieve through acculturation is an option to signal to domestic and international audiences that they value integration in world society and comply with cultural norms. Acculturation has been a defining component in contemporary discourse on intellectual property protection for indigenous cultures and native communitiesFootnote ⁹² and on the tension between preserving ancient languages and cultural assimilation.Footnote ⁹³ Acculturation in these ways has impacted biodiversity, cultural diversity,Footnote ⁹⁴ as well as individual acceptance of the effects of acculturation.Footnote ⁹⁵

Many countries, rather than engaging in deliberate processes of adopting foreign constitutional norms and textual choices, are coerced into applying values alien to their national identity. Coercion sets aside the constitutional autonomy of a country and its cultural and legal histories.Footnote ⁹⁶ Under this logic, “weaker states will converge upon the models provided by stronger states.”Footnote ⁹⁷ Powerful countries often pursue coercion unilaterally to overcome resistance to international treaties,Footnote ⁹⁸ respect international norms, and join multilateral institutions. However, coercion can also be used as a mechanism to assist countries in nation-building.Footnote ⁹⁹

Coercion is not an alien concept in contemporary intellectual property discourse. From a multilateral perspective, the TRIPS Agreement has been at the forefront of critiques. The primary complaint is that “there were elements of coercion and questionable trade-offs may have been made between market access for commodities and intellectual property protection.”Footnote ¹⁰⁰ The aggressive imposition of intellectual property norms on developing countries has become associated with concepts such as “recolonization”Footnote ¹⁰¹ and economic imperialism.Footnote ¹⁰² From a unilateral perspective, coercion has become a matter of trade policy in intellectual property. Lack of intellectual property protection risks economic isolation and unilateral acts such as economic sanctions and placement on lists of counterfeiting countries. The U.S. Trade Representative’s Special 301 ReportFootnote ¹⁰³ is an example where a government empowers trade representatives to unilaterally test the level of intellectual property protection in other countries. Another example is the European Union’s annual Report on the Protection and Enforcement of Intellectual Property Rights in Third Countries. In its 2020 Report, the European Commission has remarked that one of the main objectives of the report is to “inform right holders, in particular small and medium-sized enterprises, about potential risks to their [intellectual property rights] when engaging in business activities in certain third countries and thus to allow them to design business strategies and operations to protect the value of their intangibles.”Footnote ¹⁰⁴

The motivations of countries to include intellectual property in their constitutional bills of rights affect the structure of intellectual property constitutional clauses. Such clauses are sometimes shaped as a result of external interests or the influence of strong elites, as these clauses may “bear the imprint of the powerful.”Footnote ¹⁰⁵ These imprints, as discussed earlier, disturb constitutional cultures that countries aim to preserve. These imprints project a desired and unequal cultural hegemony that conflicts with the legal history and political autonomy of constitutional regimes and affects the actual protection these constitutions can afford to adopt on the ground.

10.4.2 Empirical Results

A comprehensive and original dataset was created to provide modest empirical support to the main claim of this chapter. The findings show how asymmetrical power relations may bring countries to constitutionalize intellectual property protection. In reality, however, constitutionalization cannot guarantee protection on the ground – and, in the case of developing countries, this guarantees the opposite. The dataset covers all constitutional provisions since 1801 that protect intellectual property as a fundamental right. Compiling the dataset involved a range of methodological choices. A choice has been made to focus on formal, written constitutions.Footnote ¹⁰⁶ A search has been conducted in numerous databases using keywords pertaining to intellectual property rights.Footnote ¹⁰⁷ The findings were cross-referenced with the WIPO Lex database constructed by the World Intellectual Property Organization (WIPO),Footnote ¹⁰⁸ the Comparative Constitutions Project,Footnote ¹⁰⁹ Constitution Finder,Footnote ¹¹⁰ Oxford Constitutions of the World,Footnote ¹¹¹ HeinOnline, and Constitutions of Nations.Footnote ¹¹² The data includes information gathered from sources such as the U.S. Trade Representative, which publishes an annual report on intellectual property protection in designated countries.Footnote ¹¹³ Two indices were used for the purpose of evaluating the level of de facto intellectual property protection given by the relevant country: the Intellectual Property Rights Index constructed by the Property Rights Alliance (IPR)Footnote ¹¹⁴ and the U.S. Chamber International Intellectual Property Index (GIPC).Footnote ¹¹⁵ It is important to mention again that the data collected for this chapter deals with intellectual property as a fundamental right in constitutional provisions; it does not thoroughly evaluate provisions referring to intellectual property as a legislative empowerment clause. The dataset is updated to January 2017.

As Figure 10.1 shows, the number of countries adopting intellectual property as a fundamental constitutional right has risen. Rapid growth can be seen since 1974 when Sweden adopted intellectual property rights in its formal constitution. Since then, fifty-one countries have followed suit. Haiti was the first country to adopt intellectual property as a fundamental constitutional right in 1801,Footnote ¹¹⁶ and Nepal is the most recent country to have adopted such a provision in 2015. Between 1895 and 1973, only nine countries adopted an intellectual property clause. Colonial histories determined the structure of the constitutional design, and states in Central and South America were the first to adopt an intellectual property clause as a fundamental right in the nineteenth century after Portugal’s adoption of such a clause in 1826. Since the early twentieth century, including intellectual property in bills of rights has become more common among countries. As presented later, the major growth occurred during the last two decades of the twentieth century, shortly after Sweden adopted such a provision in its formal constitution in 1974. Another interesting insight is that many countries that adopted intellectual property as a fundamental constitutional right during that period were developing countries. There are currently seventy-eight countries with a clause protecting intellectual property as a fundamental constitutional right.

Figure 10.1 Period of adoption of intellectual property as a fundamental constitutional right

Countries refer to intellectual property in their constitution in two ways. First, they adopt an authoritative/empowerment clause – namely, a commitment vested in the state to legislate and regulate in the field. An example of such a clause is the U.S. Constitution, which empowers Congress “[t]o promote the Progress of Science and useful Arts,” as discussed at the beginning of the third section. Second, and more substantively, countries treat the right to own and protect intellectual property as a fundamental socio-economic right and make it part of a given bill of rights.

Some countries refer to intellectual property in both ways. As illustrated in Figure 10.2, twenty-two countries (thirteen of which are developing countries) adopted only an authoritative/empowerment clause;Footnote ¹¹⁷ sixty-five countries (fifty-one of which are developing countries) adopted a substantive clause;Footnote ¹¹⁸ and fourteen (all of which are developing countries) adopted an authoritative as well as a substantive clause.Footnote ¹¹⁹ It can be assumed that the status of a country as developed or developing will impact the way the country adopts such a clause.

Figure 10.2 Types of constitutional intellectual property clauses

Including intellectual property protection in constitutional bills of rights does not guarantee enforcement of these rights on the ground. That is, the mere existence of the right in a constitution can sometimes amount to a false signal: “False can be detected by a conspicuous absence of real enforcement mechanisms.”Footnote ¹²⁰ Applied to the present argument, the adoption of intellectual property clauses sends a false signal if it lacks the possibility of enforcement. This section offers a modest empirical answer to this assumption. It explores the relations between the scope of protection given in a constitution (as measured by the textual ranking index created for this chapter)Footnote ¹²¹ and the level of protection given de facto to intellectual property in the applicable country (as measured by two available comparative data indices commonly used in the literature).Footnote ¹²² Table 10.1 provides the score of intellectual property clauses received, which were then applied to the data gathered from the indices:

Analysis of the textual ranking, alongside the ranking of the state in the two indices examined, reveals a paradoxical reality: The more a constitution expands the scope of the protection by explicitly specifying the main branches of intellectual property rights (e.g., copyright, patents, and trademarks) in the text of the constitution and makes efforts to encompass different intellectual property doctrines and principles (e.g., moral rights) and rights holders (e.g., indigenous people), the less intellectual property protection is given de facto (in the manner the indices capture the latter). For example, Venezuela, a country with a broad constitutional language for de jure protection of intellectual property rights,Footnote ¹²³ was given a maximum textual ranking of 9. Venezuela is also the country with the minimal overall GIPC score, at 6.88 out of 35. Similarly, Azerbaijan, which specifically includes in its constitutional language protection for “copyright, patent rights and other rights for intellectual property” as well as safeguarding “the right for intellectual property,”Footnote ¹²⁴ achieved almost the maximum textual ranking score (8), but minimal de facto overall IPR score at 2.8 out of 10. Egypt, a country that provides constitutional protection to “all types of intellectual property rights in all fields,”Footnote ¹²⁵ scored a similar textual ranking of 8 and low de facto intellectual property indices scores (overall IPR score of 4.4 and overall GIPC score of 9.4 out of 35). In contrast, the textual ranking score of Sweden, a leading nation in intellectual property de facto protection, is only 2 even though it has a GIPC overall score of 30.99 and overall IPR score of 8.2.

These findings uncover the apparent paradox concerning the constitutional protection of intellectual property rights, where countries with strong de facto protection of intellectual property rights do not offer broad explicit protection of those rights in their formal constitutions. The majority of the IPR and GIPC scores – Singapore, Switzerland, France, Belgium, Denmark, the Netherlands, Finland, Japan, Germany, the United Kingdom, and the United States (except Sweden) – do not refer to intellectual property as a fundamental socio-economic right in their constitutions. The geographical spread of textual ranking worldwide, illustrated in Figure 10.3, explains the almost inevitable pattern: Many countries with the highest score in the Textual Ranking Index are developing countries.

Figure 10.3 Textual ranking of constitutions with intellectual property as a fundamental right

As Figure 10.4 illustrates, the textual ranking of developing countries (4.431) is, on average, higher than those of developed countries (2.929). This is, of course, not surprising, given that most countries that adopted a substantive clause in their constitutions are developing countries, as Figure 10.2 shows. Figure 10.4, therefore, emphasizes this through a more detailed explanation. These findings lend credence to the conflicting motivations of countries to adopt intellectual property as a fundamental constitutional right. Figure 10.4 confirms that regimes pay lip service or send false signals to appease the international community, certain powerful states, and potential foreign investors without an ability to honor these rights or the ability to enforce them due to various reasons such as political unrest and economic instability.

Figure 10.4 Average of textual ranking for developing and developed countries