Drawing on data showing a decline in academic freedom over the past decade, David Audretsch, Christian Fisch, Chiara Franzoni, Paul P. Momtaz and Silvio Vismara, analyse the relation of academic freedom to technological innovation, as represented by patents, finding a quantifiable causal link between reduced academic freedom to lower levels of innovation.
Academic freedom is characterized by “the right to choose one’s own problem for investigation, to conduct research free from any outside control, and to teach one’s subject in the light of one’s own opinions”. Over the last century, liberal democracies have championed academic freedom to unleash the full creative potential of science. For example, in 1946, Oppenheimer, quoting Fermi, declared that “an intensive freedom of the individual scientific worker [..] is the only way to insure that no important line of attack is neglected”(p.303), adding that “a too-strict organization is liable to stifle the imagination”. Yet, global academic freedom has declined in the last decade after many years of steady improvement.
Although this decline is concerning for multiple reasons, in our recent study we focus on the tangible costs that it may impose on technological change and ask if diminished academic freedom hinders the inventiveness of those countries burdened by a deterioration in academic freedom?
Image Credit: Detail of The Death of Socrates, Jacques Louis David, The Met (Public Domain).
Freedom, generally, has long been recognized in economic thought as a source of growth. Free-market societies develop faster because freedom spurs innovation. Institutions that allow for free cooperation and competition promote knowledge production and exploitation. Independence from authority and hierarchy fosters an atmosphere of information exchange and tolerance to failures that spurs idea circulation, experimentation, diversity, and creativity, all ultimately inducing innovation. Academic freedom is similarly embedded in ‘scientific norms’. Norms that are key to promoting the kind of unconditional exploration that would not be possible in the private sector, as highly uncertain economic returns and limited appropriability discourage profit-oriented researchers.
Waning Academic Freedom
To gain insight into the decline of academic freedom, we use the 2022 Academic Freedom Dataset from the V-Dem Institute of the University of Gothenburg, the primary and most credentialed international source providing an explicit measurement of the concept of academic freedom. The Academic Freedom Index (AFI) is a country/year metric of academic freedom provided in the 0–1 range, for 175 countries over more than 120 years.
In the last decade, worldwide academic freedom has declined by an average of -0.034. To gain additional insights, we sampled the top 25 leading countries in all sciences. These 25 countries exhibited similar levels of academic freedom in 2021. Looking at the AFI in the same 25 countries ten years earlier (2011), we see that academic freedom improved only in South Korea (+0.07). Fourteen countries remained stable and ten countries experienced a decrease in academic freedom by more than 0.02. Academic freedom plummeted dramatically in the last decade in Brazil (-0.56), Turkey (-0.43), India (-0.39), and Russia (-0.25). Decreases were also exhibited in the USA (-0.15), the UK (-0.13), and China (-0.12). The average level of AFI in the 25 countries increased by +0.42 from 1941 (0.38) to 2001 (0.80), plateaued off around 2002–2007, then decreased by about -0.10 from 2008 to the present. The level of academic freedom registered in 2021 by the 25 leading countries (0.70) is equivalent to that of 1985.
Threats to academic freedom in recent years have come from three main sources. First, some countries have regulations that allow political power to exert direct control over universities. For example, in China, all department chairs and deans are centrally-appointed, and a regulation that mandates a Party-appointed leader in all departments has been enforced since 2013 . Second, interest groups moved by moral, religious, or ideological agendas voice concerns and attack professors on social media for their opinions, research, or teaching in controversial areas. Examples include stem cell research, the use of animal models, and, more recently, views on the COVID-19 pandemic. Finally, in some countries, the pursuit of for-profit opportunities in higher education institutions is progressively shifting the governance structures of academia, away from the collegiate model and toward a managerial model, which is more akin to a corporate research culture. This may induce academics to conform to institutional priorities and to eschew research themes that may be disliked by powerful donors and constituents, thus constraining research exploration.
Diminished academic freedom hinders innovation
Our study examines the effect of changes in academic freedom on innovation output, as represented by patents. Patent data are drawn from PATSTAT, the most comprehensive database on worldwide patent activity maintained by the European Patent Office. We measure innovation quantity by the number of patent applications filed in a country/year and innovation quality by the number of citations received by the patents of a country/year in the first 3 years after issuance. The final dataset comprises information on 62.8M patent applications and 36.8M citations related to the 157 countries for which the AFI is available in the time window 1900–2015. (Our dataset is available in the Supplementary Material of our paper).
The association between AFI and innovation quantity and quality is strong: A one standard deviation increase in academic freedom translates into 300 more patent filings and 50 more patent citations per one million inhabitants.
We model the relation between the level of academic freedom by country-year and innovation quantity (Table 1a) and quality (Table 1b), with year and country fixed effects capturing global trends and country differentials in all specifications. To assess the causal impact of academic freedom on innovation quantity and quality, we estimate four versions for each of the two measures of innovation output using 2SLS. In Models (1) and (2), we address concerns about spurious correlations, e.g., that the general level of freedom confounds the measured effect of academic freedom on the innovative performance in a country, and instrument academic freedom with an index of country-year liberal democracy, following Edgell et al. Model (1) uses as instrumental variable the 5-years-lagged country indicator of democracy and Model (2) uses the stock of democracy over the previous 10 years with a depreciation rate of 1%. Models (3) and (4) use values of the academic freedom index in previous years as an instrument to address concerns of reverse causality, i.e., that the inventiveness of a country induces more academic freedom. Model statistics show that we are able to explain four-fifth and two-thirds of the country-year level variation in patent quantity and quality, respectively. Instrument statistics show that lagged and stock liberal democracy and lagged and stock academic freedom are strong instruments. The magnitude of the effect is relatively stable across the models and specifications.
Table 1. Main analyses of the effect of academic freedom (AF) on innovation quantity (Table 1a) and quality (Table 1b).
Notes: The estimates come from 2SLS regressions. Models (1) and (2) use democracy as an instrument for academic freedom. Democracy is lagged by 5 years in Model (1). Model (2) uses the stock of democracy over the previous 10 years with a deprecation rate of 1%. Models (3) and (4) use values of the academic freedom index in previous years as an instrument. Robust standard errors (SEs) are in parentheses. *** p < 0.01.
We estimate a series of additional models to assess the robustness of our results. First, the results are robust to alternative model specifications, such as adding control variables of migration rate and changing the lags of the dependent variables. Second, we check the robustness to possible measurement errors in the AFI measures due to long time-windows and subjective measure. The results are unchanged if we restrict the estimates to the post 1980 period and if we replace the AFI with an objective metric, which captures the state of de jure academic freedom granted by national laws and regulations.
Academic freedom had progressively increased from the 1940s to the 2010s, but it reversed and started to decline in the last decade both at the global level and in the 25 leading countries in science. Our analyses show that academic freedom has a causal impact on innovation. Based on the estimates, the global decline in academic freedom that occurred in the last decade has resulted in a global loss quantifiable in the range of 4.0 to 6.7% fewer patents filed and 5.9 to 23.5% fewer patent citations.
Overall, our study is the first to link academic freedom to innovation. Given that economic growth results from technological progress, policymakers concerned about inventive productivity and economic growth may want to improve academic freedom to promote breakthrough inventions in their domestic universities that will ultimately spill-over to industry.
The content generated on this blog is for information purposes only. This Article gives the views and opinions of the authors and does not reflect the views and opinions of the Impact of Social Science blog (the blog), nor of the London School of Economics and Political Science. Please review our comments policy if you have any concerns on posting a comment below.
Image Credit: Detail of The Death of Socrates, Jacques Louis David, The Met (Public Domain).