PTThe rules of the game and innovation in maize seeds
Biosafety laws, plant variety protection, labelling: every country writes its own. The thesis gathers 156 rules from 12 countries, asks five experts to score each one and compares those scores, year by year, with maize seed patents. The question: do rules help or hinder innovation? Explained in 12 screens.
A rule can raise the bar or block the way
For the economist Douglass North, institutions are the rules of the game: they bring stability and reduce uncertainty for investors. But the literature the thesis reviews sees two possible effects of regulation. Switch the mode and tap the ideas:
The circles are made-up research projects. The two readings come from the thesis's introduction and conclusion: regulation as a selection mechanism (McKelvey and colleagues, 2004; the Porter Hypothesis) and as a barrier of costs and uncertainty (Blind, 2012; Aghion and colleagues, 2023).
Each sector innovates in its own way, and rules are one of the pieces
The thesis uses the idea of a sectoral system of innovation (Malerba): in a sector like seeds, what gets innovated depends on the available knowledge, on who takes part, on the rules and on who buys. Tap a block; the red one is the thesis's focus.
Sections 2.1 and 2.2. The literature's points of agreement are in section 2.5.
Patents on one side, scores for the rules on the other
To compare the same sector under different rules, the thesis chose maize seeds and followed four steps. Tap each one:
China and the United States hold 93% of the patents
From 2005 to 2023, the 12 countries add up to 39,357 maize seed patents. China overtakes the United States in 2014 and reaches 2,482 patents in 2017; the other ten stay below 100 a year. Change the year, take the two giants off the scale and tap a column:
Table 2, Derwent Innovations Index. The grand total and 93% (36,654 of 39,357) are our calculations. The rightmost position of the slider shows the 2005–2023 total.
Score it like an expert
Five real excerpts from Brazilian rules, with the theme the thesis placed them under. From 1 to 5: how much does the excerpt affect maize seed research? After your score, the five experts' scores appear.
Appendix I, Brazil sheet: excerpts and the experts' scores (1 = very negative effect; 3 = no effect; 5 = very positive effect on the performance of firms and researchers). Excerpts shortened as in the spreadsheet; our translation.
Scores stay near the middle and change in steps
Each country got five themes of its own, drawn from its rules. A theme's score in a year is the average score of the rules in force. Almost everything sits between 2.6 and 4.7: neither strongly against nor strongly for. And the score only changes when a rule comes in or goes out. Choose the country and move through the years:
Tables 4 to 15. The floor is score 1 (very negative) and the open box goes up to 5 (very positive); 3 means the rule does not affect research. Themes were defined for each country with the Iramuteq software, so they do not repeat across countries.
The question is whether scores and patents rise and fall together
For each country, the thesis put patents from 2005 to 2023 into a regression with the theme scores. Each theme's result is a coefficient (β). See the United States example in three steps:
Patents from Table 2 and scores from Table 10, each on its own ruler from 0 (lowest value) to 1 (highest) so they fit the same chart (our calculation). The thesis does not describe how it transformed patents; the means of the model without themes (M₀, Tables 17 to 26) match this 0-to-1 ruler (our calculation). β from Table 21, with all themes in the same model.
In each country, a different theme moved with patents
In eight countries, one theme had a clear link with patents (in South Korea, only according to the text); in two, only a hint; in two, nothing clear. The bars say whether years with a better score had more or fewer patents. Choose the country and tap a theme:
Standardised β from Tables 16 to 26; South Korea only in the text (the thesis has no table). Bars cut at ±2.5. "Clear" = p < 0.05. For Brazil and Russia, the table prints p = 0.0229 and 0.0227, but the t in the same row (−1.26) corresponds to p ≈ 0.23 (our calculation).
No rule works the same everywhere
The thesis concludes that the influence of rules is specific to each country and rejects universal best practices. Grouping similar themes, the same kind of rule went one way in one country and the other way in another. Choose a group:
Our grouping of the thesis's themes into six groups. Height = how clear the result is (tall: p < 0.05; medium striped: hint; flat grey: no clear effect), not the size of β. Blue = more patents in years with a better score; red = fewer. When a country has two themes in the same group, the block shows the clearer one.
Not every well-rated rule came with more patents
A high score means the experts thought the rule was good for research. It would be natural to expect more patents in years with a better score. Place your bets and flip the cards:
Predictable rules matter more than identical rules
For the author, a rule's effectiveness lies not in its rigidity but in its predictability, clarity and fit with the country's economic system.
For governments and regulators
- In long, risky research such as biotechnology, a stable, predictable, science-based process is worth more than direct subsidies, says the author.
- Regulate the product (its safety and efficacy) rather than the process that created it.
- Design policies that fit the country's strengths, without copying models.
For agritech companies
- Assess each country's "regulatory risk" as a central part of international strategy, beyond market size.
- Talk to regulators early and often, as partners, according to the author.
For researchers
- Interview the experts to understand the scores behind counter-intuitive results, such as intellectual property in Canada.
- Apply the method to other heavily regulated sectors and look beyond patents (startups, venture capital).
What this research does not say
- Patents are not all of innovation. In agribusiness, much protection comes from plant varieties and trade secrets. · according to the author
- Few years per country. Each regression has 19 years and up to five themes, and scores change in steps (in Japan, the same from 2005 to 2018). · our observation
- Intertwined themes. Standardised coefficients above 1 (in Australia, from −4.3 to +3.5) point to themes that move together; read the direction, not the size. · our observation
- Moving together is not causing. Scores and patents may simply follow the same rhythm over time. · our observation
- Five viewpoints. The scores come from three researchers at a Brazilian federal institution and two agronomists at Chinese multinationals. · our observation, from the thesis's method
Credits
OLIVEIRA, José Henrique de. As instituições regulatórias e o desempenho dos sistemas setoriais de inovação agrícola: um estudo transnacional. Doctoral thesis (Business Administration) — Universidade Presbiteriana Mackenzie, São Paulo, 2025. Advisor: Prof. Dr. Walter Bataglia.
Thesis approved on 6 August 2025. The author thanks UDC and the Federal Institute of Paraná (IFPR). Patents from the Derwent Innovations Index (Web of Science); thematic analysis in Iramuteq. This version simplifies the work for a general audience.