CC BY
1DOI: 10.29141/2658-5081-2021-22-4-2 JEL classification: B52, O31, Z13
Vyacheslav V. Volchik Southern Federal University, Rostov-on-Don, Russia
Elena V. Maslyukova Southern Federal University, Rostov-on-Don, Russia
Sophia A. Panteeva Southern Federal University, Rostov-on-Don, Russia
Innovation indicators in the context of narrative economics
Abstract. The analysis of innovation systems is a demanding task, which needs to be tackled comprehensively. Their modelling provides an indication of the formal innovative performance while the narrative analysis helps to examine relevant judgments about the rules and institutions that are key to the development of innovations. The study of the complementarity and non-complementarity of the conclusions obtained using both approaches is likely to contribute to a deeper understanding of the Russian innovation system functioning. The paper aims to analyse how the content and spread of the national innovation system narratives are interrelated with the modelling of the Russian innovation system and the most important indicators of its functioning. Methodologically, the research takes advantage of the methodological synthesis of narrative economics and original institutionalism. The research methods include narrative analysis, cointegration analysis of mutual influence between narratives and innovation indicators. The authors identify peculiar features of innovation systems' modelling, analyse the texts, stories and other sources containing narratives about the Russian innovation system development, and based on this discover the relationship between the spread of certain narratives (the frequencies of word groups "innovative technologies", "technology development", "innovation activities", "technology implementation", "scientific schools", "innovation policy", "science and technology") and innovation indicators (share of innovative goods, works, services in the total volume of goods shipped, works performed, services provided; share of expenditure on innovation activities in the total volume of goods shipped, works performed, services provided). The analysis also shows cointegration between the frequency of search queries "use of patents" and innovation indicators "number of patent application" and "number of patents issued". The authors point to a number of limitations related to the innovation system modelling. There are some typical limitations such as availability and reliability of statistical data, methodological heterogeneity of statistics, model specification, etc., as well as some specific ones including a broad, ambiguous definition of the national innovation system, differences in the approaches to defining principal determinants, innovations immeasurability, considerations of the system connections and economies of scale, etc. Thus, the findings confirm the need for a narrative study in order to achieve an in-depth understanding of complex economic processes.
Keywords: economic growth; national innovation system; innovations; narrative economics; Russian innovation system; cointegration analysis; economic policy.
Acknowledgments: The research is funded by the grant of the Russian Science Foundation (RNF) no. 21-18-00562 "Developing the national innovation system in Russia in the context of narrative economics" (https://rscf.ru/en/project/21-18-00562/).
For citation: Volchik V. V., Maslyukova E. V., Panteeva S. A. (2021). Innovation indicators in the context of narrative economics. Journal of New Economy, vol. 22, no. 4, pp. 24-44. DOI: 10.29141/2658-5081-2021-22-4-2 Received September 24, 2021.
Introduction
The development of the innovation system is a key factor in economic growth. Innovation processes are complex and conditioned by the economy's institutional structure. The state and quality of this structure affect the ways of generating and using knowledge in society, mechanisms for financing and introducing innovations [Stiglitz, Greenwald, 2014, p. 140].
Models and indicators reflect the analytical perception of the development patterns of the national innovation system (NIS). In turn, narratives provide information about everyday practices or common simplified proto-economic models that are used by actors to explain and understand the processes [Shiller, 2019b].
Attention to narratives in economics has increased due to the widespread approach that Shiller, Nobel laureate, called narrative economics or narrative economic theory [Shiller, 2017; Shiller, 2019a]. This research program followed on from the study of institutional and social factors affecting economic behaviour and economic processes. In many ways, the provisions of narrative economics are complementary to the research approaches of the original institutionalism [Efimov, 2016]. For economists, narratives are important primarily as a significant source of qualitative data on social processes, rules, institutions and behavioural patterns that are relevant to actors in a particular field of activity [Volchik, 2017].
In economic theory, the number of studies on narrative economics has noticeably increased. Such studies consider important to look into social norms, identity and narratives to achieve an in-depth understanding of complex economic processes [Collier, 2021]. Although the term narrative economics emerged in 2017, the relevant scientific research had appeared before. For example, Akerlof and Snower [2016] gave an extensive analysis of narratives as simplified schemes of economic events, which allowed to use historical narratives or even humorous anecdotes to explain patterns
in the economy of the Soviet Union. A slightly different approach is taken by Collier [2016], who considers narratives in connection with beliefs and norms governing economic behaviour. An analysis of this phenomenon with regard to business processes shows that the use of vivid narratives can lead to an increase in the value of the company [Damodaran, 2017]. Currently, we observe a growing number of papers that apply the theoretical framework of narrative economics. For example, 29 papers have been published in the Web of Science database since 2017 that include the concept of narrative economics. Moreover, research within this theoretical approach covers a wide range of problems: from the economics of sports [Newman et al., 2020] to the analysis of financial crises [Willett, 2021].
The most important aspect related to the study of the narratives' role is the viral spread of stories in the media and the Internet, since it promotes the spread of ideas that make it possible to comprehend certain processes in the economy. These ideas, in turn, have a significant impact on the formation of the institutional environment and institutions [Markey-Towler, 2019], which structure repetitive interactions in various fields.
The purpose of the study is to demonstrate how narratives about the national innovation system, their content and spread are interrelated with the modelling and the main indicators of this system. To achieve this purpose, we will consider the specifics of modelling of innovation systems, as well as narratives about innovation activities, ways to identify and analyse them; and reveal links between the spread of certain narratives and indicators of innovation activities.
Models and indicators of the national innovation system
Quantitative analysis of both the national innovation systems and the economic growth associated with their development engenders a number of problems related to the formalisation and availability of relevant quantitative data. Thus, when modelling these economic phenomena, a number of typical (characteristic of all econometric models) and specific (inherent in the categories under consideration) problems arise. Understanding the complexities of quantitative analysis makes it possible to identify weak points and limitations of the modelling and the mathematical models, indicate key aspects that cannot be described using them in principle or at this stage of development. This is important, first of all, for the development of a methodology to study national innovation systems.
We analysed a number of papers, thus revealing the relevant problems (Figure 1).
Typical difficulties of mathematical modelling include problems of statistics availability. Thus, in the overwhelming majority of the papers under consideration, it is noted that their absence leads to the limitation of quantitative concepts. At the same time, the information may lack in time samples (for a period or by periods; for
example, there is only one indicator for a year, whereas the researcher needs indicators for a quarter, etc.) or spatial data (as a rule, there are no observations about developing countries). The missing values are filled in by interpolation methods, with averages or omitted, however, as a result, the problem of a small sample size arises, which does not allow building reliable models. The researchers also point to the problem of the statistics reliability. Methodological difficulties are also noted when the same indicators are measured by different bodies or organisations in different ways, as a result they vary significantly [Khrustalev, Slavyanov, 2016].
Fig. 1. Limitations of the NIS modelling
Paradoxically, at the same time as there is a lack of relevant data, there is often an excess of available information. A large number of available variables can be useful for determining system relationships, however, the inclusion of all indicators in the model is impossible due to obvious modelling difficulties associated with both the required number of observations and the increasing time of parameter selection. The multicollinearity problem should also be mentioned here. The most common approach to eliminating these difficulties is to perform a factor analysis that allows to convert a set of correlating variables into several independent ones.
Another problem characteristic of any objects of modelling is the presence of outliers. When analysing at the level of countries or regions, this complicates the work especially. When excluding such observations, the problem of data insufficiency worsens and we need to justify their exclusion, since they may belong to other clusters or be an example of so-called positive outliers demonstrating strikingly different successes (or failures) in comparison with other objects with similar other parameters, which may indicate factors not taken into account in the study and, thus, represent a valuable source of information [Peiffer, Armytage, 2019].
The basic question that arises in modelling, regardless of the object of study, is the choice of the model specification. Of course, the sensitivity of the model to the specification determines the accuracy of the results, their practical significance and applicability, and hence the importance of this stage.
Another group of the highlighted constraints includes specific ones, i.e. specific to the modelling of national systems and economic growth (they also may arise when considering other topics, but are not typical for all econometric models).
The national innovation system is a broad concept that, depending on the interpretation, may include a large number of elements, from individual actors to technologies, infrastructure, institutions and system interconnections, as a result of which uncertainty arises when modelling this system. An attempt to bring such lengthy definitions to a quantitative denominator is not an easy task [Fagerberg, Srholec, 2008]. Moreover, if different theoretical concepts imply the inclusion of different determinants in the model, then it is expected that the empirical results will also differ, and this makes their comparison difficult. For example, if Romer [1990] omits the institutional variable in his analysis of innovations, other researchers, on the contrary, add it [Matrizaev, 2019].
The fundamental complexity for modelling technological progress is created by the very nature of innovations. Being abstract, often insubstantial and at the same time comprehensive, they are difficult to measure. Heterogeneity of aspects makes it impossible to reflect innovations in a single indicator [Archibugi, Coco, 2005]. The reason lies in the qualitative essence of not only innovations, but also a number of other components of NIS: researchers point to the absence of a way to include indicators of international relations and economy's openness in the models [Tseluiko, 2017]. On the other hand, even for those phenomena that can be quantified, it is not always possible to find relevant factors [Nasierowski, Arcelus, 2003].
A significant obstacle that arises due to the abstractness of the NIS concept, as well as the difficulty of measuring its components, is traditionally smoothed out by using proxy variables. Nevertheless, they do not sufficiently compensate for these shortcomings. Thus, traditional proxies for innovation (the number of patents, indicators of high-tech exports, R&D expenditures) make it possible to judge NIS only in a narrow sense, and therefore are ambiguous and not sufficiently capacious to explain innovation, technology and economic growth [Sesay, Yulin, Wang, 2018]. For example, research and development costs as an indicator of innovation reflect only the financial aspect, ignoring other important components [Gackstatter, Kotzemir, Meissner, 2014]. Many other indicators can also be called incomplete, only indirectly indicating real phenomena [Matrizaev, 2019]. It is often noted that the creation of a unified, integrated system for measuring innovations, dynamic components of NIS, and its performance is an unresolved problem [Pyastolov, 2012]. As a result, there is no
possibility of both a complete objective assessment of innovation systems and ensuring comparability of subjects in this parameter. The use of composite indexes, which are a popular choice when modelling national innovation systems, also does not solve this problem. They yield little information because they consist of many variables, depend on the weights assigned to sub-indexes and the quality of information from primary sources [Gackstatter, Kotzemir, Meissner, 2014].
The NIS development is systemic in nature, therefore, it has all the properties of systems, including emergence, which significantly complicates modelling due to the difficulty of assessing it. Many authors call the consideration of the NIS performance independently of the scientific sphere a frequently faced problem of analytical research, which does not allow taking into account their mutual influence [Balatsky et al., 2017]. A similar omission is noted in the study, where the overall NIS performance index demonstrated independence of its components, which may be a consequence of omitting unobservable connections [Matrizaev, 2019]. However, synergetic effects are one of the most important factors in the development of NIS [Bogdanova, Ibrahim, 2018], which justifies the importance of taking them into account. Therefore, when modelling innovation systems, an integrated, including interdisciplinary, approach is required.
We can reasonably assume that not all investment inflows into innovations produce results in the short term, since research, both applied and fundamental, may require a long time period to be completed and implemented. If indicators with some lag should be used in modelling, then a question about the order of lags is expected, but there is no straight answer here. Innovation performance indicators are not easy to analyse, because by the moment 'today's' data become available, the entire chain of technological development is difficult to track due to long and unstable lags, as well as the influence of many other factors during this period [Griliches, 1998]. In this regard, we attribute the search for a suitable lag structure to significant limitations of research [Gackstatter, Kotzemir, Meissner, 2014], and point to the lack of adequate theoretical and methodological studies on this issue [Nasierowski, Arcelus, 2003].
Modelling challenges also include assessment of the economies of scale. Most empirical studies of economic growth and innovation development suppose constant economies of scale, but at the same time there is applied evidence that economies of scale change over time [Nasierowski, Arcelus, 2003]. Although both decreasing and increasing economies of scale are evident [Artz et al., 2010], ignoring them in models seems to be a significant omission.
Narratives about the innovation system
The described limitations naturally make the reflection of innovative reality in mathematical modelling not just incomplete, but often even contradicting the experts'
opinions regarding the functioning of NIS. Modelling in such cases must be supplemented with qualitative data, sourced from the narratives of the main actors of the innovation system. Narratives, in turn, can be considered as proto-economic models [Schiller, 2019b, p. 477], which change along with the agents' heuristics [Bookstaber, 2021, p. 236]. To identify discrepancies of this kind, data arrays of 43 mass media and Internet sources were analysed, selected using Federal Media: 2020, the ranking of the Medialogia company1 for the period from January 1, 2010 to July 1, 2021.
When modelling NIS, researchers, as a rule, use a number of standard indicators to reflect its results, including the number of patents, the GDP produced, its growth rate or per capita, as well as labour productivity, various innovation indices, foreign direct investment, indicators of publication activity and citation, export volumes and some others. The limitations of this approach are noted as one of the specific problems of NIS modelling, however, the validity of this thesis is also confirmed by the fact that the actors of the innovation process practically do not use a significant part of the listed indicators in their interviews and statements. On the contrary, experts think globally, rarely identifying NIS with specific statistics. Among the indicators of NIS performance used in narratives, we observe indicators of patenting, scientific publications, and innovation ratings of different countries and universities. Experts do not apply FDI, GDP and its derivatives, the volume of exports in this context. This, however, does not contradict the inclusion of these variables as resultants in the model.
Returning to the use of patenting indicators, it is noteworthy that they are among the most popular variables in the mathematical modelling of NIS. However, the research papers do not reflect the fact that these statistics are systematically poorly estimated due to both the uselessness of these indicators and the complexity of their registration.
"The implementation of patents into manufacturing processes and, as a consequence, the need to protect R&D results exist under high economic competition, which we do not have' (Vitaly Kastalsky, Managing Partner of AK Patent Law Group)2. "At the same time, despite the global financial crisis, patent fees in Russia doubled in 2008... It cannot be said that patent fees are too high, approximately 5,000-6,000 rubles per patent. But when we are referring to 50-100 or, for example, 500 patents, the amount is significant for any university'"3 (Boris Korobets, Head of the Jurisprudence Department of the Moscow State Technical University named after N. E. Bauman).
Moreover, the narratives contain information about the low quality of patents. Therefore, at least it is difficult to use this indicator in modelling: "Some patents are
1 Federal Media: 2020. Medialogia. https://www.mlg.ru/ratings/media/federal/10165/ (in Russ.)
2 Dmitrienko I. (2017). The number of inventions in Russia has reached a catastrophically low level. Profile, no. 42, pp. 8-14. (in Russ.)
3 Ivoylova I. Who will give the right to open? Rossiyskaya Gazeta, October 20, 2010. https://rg.ru/2010/10/20/ pravo.html (in Russ.)
of dubious value not only from the viewpoint of science, but also from the standpoint of common sense"1 (Inna Rykova, Head of the Center for Sectoral Economics of the Financial Research Institute of the Ministry of Finance of the Russian Federation).
A similar conclusion follows from the analysis of publication activity indicators in models and narratives. In the mathematical representation, the number of publications in scientific journals and citation indicators are often used, which, however, does not give a clear understanding of NIS and results of scientific activities. The reason for this is expressively indicated in a number of narratives, for example: " We are constantly told how much money has been invested in science, and at the same time we write too few research papers. Papers and publications in leading scientific journals are an important point, but this is the final stage of research. So, when should we think and engage in scientific research if we put the papers writing and their publication in Scopus and Web of Science at the forefront? Let's consider the Mathematical Institute named after Steklov: we have reached the level of 2.5 publications per employee per year. This is a lot, because mathematics is a fundamental science, not applied, a mathematician cannot write more than two good research papers a year, even if you load him with money, one cannot jump over his own head.
But they persistently demand of us to write more and more papers. Moreover, without coordinating the state tasks on the number of publications with the academy, with its specialised departments, which could give their professional assessment. What does this lead to? My colleagues, in order to get out of this absurd situation, suggest: let's split up and publish separate 'episodes' instead of finished, large papers ...
And another problem. As part of the program to increase the global competitiveness of our higher education institutions "5-100", they are tasked with rising in the international university rankings and, therefore, increasing publication activity. Universities that participate in this program compete for additional government funding. As a result, these universities lure our scientists into working for them and pay them only for writing papers and indicating these universities as affiliations. Therefore, the bursts of publication activity in such universities are not the performance of their own university science, as we would like to think and how it is served, but simply the effect of adapting the university to the requirements and conditions set 'from above'. Is this behaviour worthy? I think not, and many of my colleagues think the same"2 (Valery Kozlov, Vice-President of the Russian Academy of Sciences^.
1 Zulina V. Adygea vs. Hirsch. Rossiyskaya Gazeta, October 24, 2014. https://www.gazeta.ru/science/ 2014/10/24_a_6274225.shtml (in Russ.)
2 Zadorozhny A. Our scientists have a feeling of chaos and injustice. Znak, February 8, 2019. https://www. znak.com/2019-02-08/v_den_rossiyskoy_nauki_o_prichinah_ee_krizisnogo_polozheniya_intervyu_vice_prezi-denta_ran_valeriya_k (in Russ.)
The models also often take into account the indicators of the labour force, the economically active population or the population of the country (region) as a whole. In narratives, on the contrary, they are not specified. In terms of the innovation system, it is not the quantity that matters, but the quality of human capital. Therefore, the use of such indicators seems doubtful, since it is not at all obvious what they can bring regarding an analytical understanding of NIS.
On the other hand, it would be wrong to say that the qualitative characteristics of personnel are ignored in mathematical models, since they take into account the number of applicants to universities, students at different levels of education and similar indicators. However, from the narrative analysis we can understand that the use of these variables has a significant omission: the number of graduated persons does not correspond at all to the number of persons who are actually ready and able to create and spread innovations. "But lets look further: young people come to universities. Just within the first two years, a significant number of students is lost for science... Because many of them have moved away from their parents and need to earn their own living... And then it turns out that studying is not necessary to provide for their life, its enough to walk to the nearest crossroads and find a job, not qualified but in-demand, as a sales representative or security guard. And those who were initially focused on science are leaving this trajectory' 1 (Alexander Sergeev, President of the Russian Academy of Sciences, academician).
Moreover, the indicators of the students' number do not take into account the emigration factor or 'brain drain', which is a very common narrative: "More than a half of those educated professionals who leave the country after graduation or in the middle of studying are excellent students. Mediocre students stay at home 'to steer processes', to run the country. We see the results of their management every day' 2 (Georgy Bovt, political scientist). "We are talking about the future, we hear promises that tomorrow will be good. But today, in my laboratory, graduate students are defending their theses and leaving the country, and when I promise them that the situation will improve soon, they say: 'Here's my phone number, when the situation changes, call me, and now it's better not to be here " 3 (Konstantin Severinov, Head of the laboratory of the Institute of Molecular Genetics of the Russian Academy of Sciences and Rutgers University).
From this point of view, it is probably more correct to use indicators of the number of scientific personnel, personnel engaged in research and development, the number of researchers or scientists with a degree, the number of researchers in the field of R&D, etc.
1 Chernyak I. There will be science for us. The President of the Russian Academy of Sciences speaks about the pressing problems and hopes of scientists. Argumety I fakty, December 16, 2020. https://aif.ru/society/science/ budet_nam_nauka_prezident_ran_o_nasushchnyh_problemah_i_nadezhdah_uchyonyh (in Russ.)
2 Bovt G. This is not the time to be smart. Gazeta.ru, February 18, 2019. https://www.gazeta.ru/comments/col-umn/bovt/12190555.shtml (in Russ.)
3 Borisova A. The Ministry of Education and Science is on our side. Gazeta.ru, November 29, 2012. https://www. gazeta.ru/science/2012/11/29_a_4873081.shtml (in Russ.)
Indicators of infrastructure support are also used in the modelling of NIS. Usually among such indicators we can see the number of Internet users, telephone communications, computers, the number of computers with Internet access, the share of organisations using advanced technology products (ATP) for research, etc.
In narratives, on the contrary, they do not talk about engineering and information infrastructures, but they point to social, market and innovation infrastructure: "We have a weak segment of infrastructure support for innovation. Now the work of the institutes is focused on the production development, on its modernisation. And the emphasis should be on innovation. We have weak systemic coordination of innovative development. There is no logical algorithm of interaction'1 (Sergey Smolnikov, Minister of Economic Development of the Chelyabinsk oblast); "But for their prototyping and scaling, we need a living space, technological infrastructure, a generation system (continuous cycle). Without this, it is pointless to count on the emergence of a new complex of companies"2 (Natalia Reshetnikova).
Taking into account institutional specifics in mathematical modelling seems to be a complicated and even not feasible task in some aspects, so we expect that drawing conclusions about them may be difficult. Indeed, the influence of institutional components is ambiguous in models, they are often recognised as insignificant [Volchik, Maslyukova, Panteeva, 2021]. In narratives, the situation is reversed: institutions are among the most significant factors affecting NIS, and we are talking about them in the widest range, starting with the cultural characteristics of the population and ending with formal institutions (legislation, government actions).
Inefficient institutions and related corruption cases, bureaucracy, administrative barriers, and an uncoordinated model of state innovation management are often mentioned: "I do not know how to move ahead otherwise in Russia. After all, the demand for innovation is born only in a competitive environment, where corruption is minimised. <...> And how else? Pay bribes to innovate?"3 (Daniil Livshits, Vice-President for Technology in Innolume Gmbh). "A concrete example of a bureaucratic problem is the allocation of money for grants from the President of the Russian Federation for young scientists. October is already ending, the Ministry of Science and Education has requested a report on the grants allocated, and the money for the current year has not arrived in the accounts of the Russian Academy of Sciences. I want to emphasise once again that the problems described are on the side of the government,
1 Dybin A. They spend more on the Tractor team. Znak, April 13, 2017. https://www.znak.com/2017-04-13/v_ chelyabinske_obsudili_podhod_k_innovaciyam_i_strategiyu_2035 (in Russ.)
2 Reshetnikova N. Cluster under the order. Rossiyskaya Gazeta, February 26, 2015, no. 6610. (in Russ.)
3 Kanygin P. Kovsh, Bugrov, Odnoblyudov. Novaya gazeta, September 8, 2010. https://novayagazeta.ru/ articles/2010/09/08/1699-kovsh-bugrov-odnoblyudov (in Russ.)
not the scientific community"1 (Dmitry Perekalin, Researcher at the Institute of Orga-noelement Compounds of the Russian Academy of Sciences, member of the Council of Young Scientists of the Russian Academy of Sciences); "Sticks are being put in the wheels not only for innovative business, but also for fundamental science. Scientists who have come to work with us from abroad (they have something to compare with) are literally groaning from the dominance of bureaucracy. If you have received a government grant, every three months you will have to write a 500-page report on the work done and the funds spent. In terms of volume, these are two doctoral theses! Moreover, you need to start writing as soon as you have won the grant. When to do science if you have to report on every box of paper clips and a pack of paper?" 2 (Alexander Samkov, engineer, inventor and developer of a unique cardioprosthesis, the world's first tricuspid heart valve); "This is also market competition, which in some cases is the main condition for success: when an entrepreneur just entering the market does not experience administrative pressure, is not blocked from access to loans and finance. Unfortunately, all these conditions are not in the best shape in our economy right now"3 (Sergey Dubinin, ex-head of the Central Bank of Russia, Chairman of the Supervisory Board of VTB Bank, member of the Board of Directors of VTB Capital).
A serious problem of NIS modelling is that the demand for innovations is not taken into account. As with any other economic goods, supply is only half of the picture, whereas without demand, the existence of a market (in this case, the innovation market) is basically impossible.
Narrative analysis makes it possible to compensate for this omission. The study of the media revealed that the demand for innovations is one of the key aspects highlighted by experts: "First of all, we must tell that science in Russia has not died, although it has fallen on hard times. We have excellent R&D results that lie idle in the institute's portfolios, because no one needs them. You are right, business has no demand for science yet"4 (Lev Nikolaev, Artistic Director of the TV company "Civilisation"); "Among the problems that we have, it is quite obvious to mention the low level of cooperation between science and private business, the lack of demand for innovations, which we have not been tired of talking about for the last seven or eight years"5 (Dmitry Medvedev, Prime Minister); "I am critical to Skolkovo. I am fundamentally
1 Borisova A., Podorvanyuk N. Discoveries for free. Gazeta.ru, October 29, 2012. https://www.gazeta.ru/ science/2012/10/29_a_4828873.shtml. (in Russ.)
2 Pisarenko D. et al . Bureaucracy versus Scientists. Argumety i fakty, September 8, 2016. https://aif.ru/society/ science/umnye_no_bez_proryva_chto_meshaet_rossii_stat_nauchnoy_sverhderzhavoy (in Russ.)
3 The increasing number of startups will be followed by the growth of the Russian economy. Izvestiya, May 5, 2014. https://iz.ru/news/571311 (in Russ.)
4 Medvedev Yu. Seducers with integral. Rossiyskaya Gazeta, September 24, 2010, no. 5295. (in Russ.)
5 Khristova K. Medvedev said that innovations in Russia are in little demand. Komsomolskaya Pravda, January 26, 2016. https://www.kp.ru/online/news/2288326/ (in Russ.)
critical, because if we want to build an innovative economy, we had to start with something else. It is necessary to create a demand for innovation. And there is no demand, because the level of competition is insufficient. And if there is no competition, then innovation is not an argument in competition n (Igor Nikolaev, Director of the Strategic Analysis Department of the FBK Audit Group).
On the other hand, in a number of aspects, there is a consistency between the parameters and conclusions of formal models and narratives. First of all, this concerns the financing of R&D, regarding which opposite statements are found in the media. Some of them claim that financial support is available to Russian innovators: "There is enough money, we urge entrepreneurs not to be afraid to come out with their applications'2 (Mikhail Vyshegorodtsev, Minister of the Moscow Government, Head of the Department of Support and Development of Small and Medium-Sized Businesses in Moscow); "Scientists constantly refer to the fact that our science is kept on starvation rations, and is tasked to achieve world-class results. But one of the heads of the Ministry of Education and Science claims that talks about the underfunding of our science are from the evil one"3 (Vladimir Fortov, President of the Russian Academy of Sciences). Other statements point to the insufficiency of funding: "- Is fundamental science being funded worse today than it was during the Soviet era? - Much worse. At our institute, for example, money is mainly spent on public utilities, because we have to pay for electricity, water, garbage collection, for everything related to the building where we work. And somehow we manage to keep a little to pay the salary. Here are two articles that the institute's budget goes to"4 (Evgeny Gordeev, Academician of the Russian Academy of Sciences).
Both approaches demonstrate that there are differences between public and private research funding. Thus, many authors select as variables not just R&D expenditures, but government expenditures on R&D, allocations for civil science from the federal budget, expenditures on technological innovations in the total volume of goods shipped (i. e. the share of expenditures in the production of industrial organisations), R&D costs of business sector, R&D costs of scientific sector, etc.
Narratives also illustrate that the source of funding is important: " We must understand that there are three main links in the research chain. It all starts with fundamental
1 Kotlyar P., Podorvanyuk N. Skoltech is more transparent than the State Duma, the Ministry of Defence or Moscow State University. Gazeta.ru, April 23, 2013. https://www.gazeta.ru/science/2013/04/23_a_5277445.sht-ml (in Russ.)
2 Gurvich V. There is a lot of money - there are few good projects. Moskovsky Komsomolets, June 28, 2010. https://www.mk.ru/economics/article/2010/06/27/512486-deneg-mnogo-horoshih-proektov-malo.html (in Russ.)
3 Medvedev Yu . Not to leave Russian Academy of Sciences. Rossiyskaya Gazeta, December 5, 2014, no. 6550. (in Russ.)
4 Khitrov V. The country's leadership does not know what science is needed for. Novaya Gazeta, June 5, 2017. (in Russ.)
science, for which the state is responsible. Its results are used in search-oriented research. And applied science appears when there is a working prototype and business is ready to invest money in if1 (Alexander Sergeev, President of the Russian Academy of Sciences); "This is the percentage of money coming into science today from the budget and from business. In the USA, this ratio is 20/80, respectively. Our situation is exactly the opposite: 70/30. We currently have science funding of about 1.15 % of GDP. But other states do not invest much more from the budget as a percentage. A record 4 % of GDP in science in Sweden, for example, is divided in the proportion of 1 % from the state and 3 % from business. Therefore, our task is, following the example of high-tech countries, to increase extra-budgetary financing of science and achieve at least 50/50 in the near future'2 (Alexander Sergeev, President of the Russian Academy of Sciences).
The narrative understanding of NIS deepens another of the modelling problems considered earlier among the specific ones. If the question about a lag between the input (for example, financing) and the output (for example, patent registration) of the system is legitimate for modelling due to the physical possibilities of creating innovations, then narratives add volume to this topic, indicating the aspect of the low speed of innovation implementation. The time lag may be longer not only due to the fact that research has a long-term nature, requires a long time to complete it, but also due to the fact that innovations are not immediately sold after being created, transferred to production or registered due to environmental constraints: "The problem of implementation has yet to be solved. We have a lot of good R&D activities at the academy, but their implementation stretches for many years for reasons beyond the control of scientists. In our institute, for example, there are R&D results that we cannot implement for several years, and foreigners buy them, and in six months they have everything ready"3 (Vladimir Fortov, Academician, President of the Russian Academy of Sciences); "The process of converting intellectual property into a realised commercial product is really very difficult. 10 years ago, even current R&D results could be stored for years, waiting for us to sell them. Something was sold in fact, but not much. There was no real mass market launch of our scientific and technical R&D results. So the task of including the universities' intellectual property in the economic turnover of the country was quite acute'4 (Mikhail Shestopalov, Vice-Rector for Scientific Work of the Saint Petersburg State Electrotechnical University "ETU").
1 Potatoes, chickens, sugar - we have such priorities. Kommersant, March 29, 2018. https://www.kommersant. ru/doc/3586265 (in Russ.)
2 Vedeneeva N. Full house in physics: the head of the Russian Academy of Sciences named the most breakthrough areas of science. Moskovsky Komsomolets, January 10, 2019. https://www.mk.ru/science/2018/12/27/na-fizike-anshlag-glava-ran-nazval-samye-proryvnye-napravleniya-nauki.html (in Russ.)
3 Ukolov R. Physics of change: Vladimir Fortov told what changes are expected by RAS. Profile, May 31, 2013. (in Russ.)
4 Rogozin O. The topic of small innovative enterprises was discussed by the heads of the leading technical universities of Saint Petersburg. Izvestia, May 16, 2011. (in Russ.)
We can state that quantitative and qualitative approaches converge in conclusions about the overall performance of the Russian innovation system. Thus, models often indicate its low efficiency and poor competitiveness. This is reflected in many narratives: "The fact that we are able to make a mobilisation impulse and are aimed at self-realisation is important, but insufficient. This fact explains why the Russians invented the hydrogen bomb and sent a man into space, but could not make a competitive car, refrigerator or TV. To produce a mass product, you need the ability to comply with standards, which Russians are not characteristic of, but to create a unique piece, this skill is not required" 1 (Alexander Auzan, Dean of the Faculty of Economics of the Moscow State University); "We have been talking for years about the need to move to the knowledge economy, to innovation, including at the highest political level. However, there is still no effective innovation system in the country. Therefore, "Balakiny", if they achieve success, it is not thanks to the system, but in spite of it" (Andrey Vasyanin)2.
Narratives and indicators of the innovation system development
The most important feature affecting the effectiveness of a narrative is its 'virality', or the degree of prevalence in the media space [Shiller, 2021]. Therefore, along with qualitative analysis of texts, we need data on the coverage of certain narratives by significant mass media and Internet sources. To do this, the keywords that were used at the first stage of the search for narratives were taken as a basis for analysing the prevalence of certain stories about the innovation system development. This analysis was carried out using data for the period from January 1, 2010 to July 1, 2021, posted in the Interfax database. To implement the selection of narratives, a list of Web search queries was compiled, including 33 phrases / terms (30 used in the previous search for full texts and 3 new ones: "Russian innovation system", "innovations" and "innovatics" related to the innovation system development).
To analyse the relationship between the spread of certain narratives and indicators of innovation activity, 10 of the most popular queries from the Interfax database for the specified period were selected (Table 1).
We used the following indicators as statistical measures (data from the Federal State Statistics Service for the period from 2010 to 2020):
y1 - the share of innovative goods, works, services in the total volume of goods shipped, works performed, services provided, %;
y2 - the share of expenditure on innovation activities in the total volume of goods shipped, works performed, services provided, %.
1 Auzan A. The phenomenon of QWERTY. Lenta.ru, April 24, 2015. https://lenta.ru/articles/2015/04/24/auzan-culture/ (in Russ.)
2 Vasyanin A. A killer for a murderer. Rossiyskaya Gazeta, February 3, 2016, no. 6889. (in Russ.)
Table 1. Top-10 of the most popular Web search queries about the Russian innovation system, 2010-2020
Variable Web search query Total of Web search queries
Xi innovations 2 176 734
X2 innovative technologies 621 364
X3 technology development 328 852
X4 research and development 242 717
X5 innovation activities 237 102
X6 technology implementation 177 657
X7 scientific schools 132 702
X8 innovative enterprises 121 648
X9 innovation policy 121 031
X10 science and technology 107 164
For a visual representation, the dynamics of Web search queries and statistical data were previously normalised from 0 to 1 using the min-max procedure:
^ _ Xi Xmin
1 ^max — ^mi
(1)
where zj is the normalised value, xj is the original value, xmin is the minimum value, xmax is the maximum value of indicators.
Figure 2 shows the comparative dynamics of mentioning the corresponding phrases. In the period under consideration, the frequency of all search queries increased, except for the phrase "innovation policy" which reached the maximum frequency in 2015.
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Innovations
Technology development Innovation activities Scientific schools Innovation policy
Innovative technologies Research and development Technology implementation Innovative enterprises Science and technology
Fig. 2. Dynamics of Web search queries The dynamics of statistical indicators is shown in Fig. 3.
1.0 0.8 jS 0.9 •3 0.7
'S 0.6
.3 0.5
0.0
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
— Share of innovative goods, works, services in the total volume of goods shipped, works performed, services provided
— Share of expenditure on innovation activities in the total volume of goods shipped, works performed, services provided, by subjects of the Russian Federation
Fig. 3. Dynamics of statistical indicators of innovation system development
We used the Engle - Granger cointegration test to identify a long-term stable relationship between statistical indicators. This method includes the following steps:
(i) testing the stationarity of a time series xt;
(ii) testing the stationarity of a time series yt;
(iii) assessing the regressionyt = a + 6 x xt, and getting a time series of residuals et;
(iv) testing the stationarity of a time series et.
If the time series xt andyt are non-stationary at the 1st and 2nd steps, and the time series of residuals et is stationary at the 4th step, then the time series xt and yt are cointegrated.
The results of testing, presented in Table 2, demonstrate the stationarity of the time series x1, x4 and x8 (the frequency of Web search queries "innovations", "research and development" and "innovative enterprises"), therefore, these time series were excluded from further analysis.
Table 2. Stationarity test for time series (ADF test with constant and trend)
Variable ADF statistics P-value
Xi -5.3816 3.012e-05
X2 2.16982 0.999998
X3 3.78872 0.999998
X4 -4.10617 0.006102
X5 -2.60302 0.2789
X6 0.778229 0.9998
X7 -0.852289 0.9595
X8 -4.31676 0.002902
X9 -1.105 0.9269
X10 1.13023 0.9999
yi -2.4956 0.3304
72 -1.89253 0.6584
Note: stationary time series are in italics.
The results of testing for stationarity of cointegration regression residuals are presented in Table 3.
Table 3. Stationarity test for residuals of cointegration regressions
Variables ADF statistics P-value
yi ~ *2 -2.39227 0.0i62**
yi ~ X -2.55098 0.0i04i**
yi ~ *5 -i.956i6 0.04827**
yi ~ X -2.580i 0.009575***
yi ~ *7 -2.26786 0.02253**
yi ~ X -i.84787 0.06i6i*
yi ~ Xio -2.48992 0.0i237**
y2 ~ *2 -2.i9694 0.027**
y2 ~ X -2.i888 0.02756**
y2 ~ X5 -2.67522 0.007245***
y2 ~ X6 -2.i8975 0.0275**
y2 ~ X7 -2.2357i 0.02447**
y2 ~ X9 -2.96452 0.002954***
y2 ~ Xi0 -2.22464 0.025i7**
Note: here and further the superscripts *, **, and *** denote statistical significance: *p < 0,1; **p < 0,05; ***p < 0,01.
The data presented in Table 3 confirm the existence of a long-term relationship between Web search queries and statistical indicators.
Additionally, we checked the cointegration between the frequency of the Web search query "use of patents" (x11) and two statistical indicators: number of patent applications, units (y3) and number of patents issued, units (y4) (Table 4).
Table 4. Stationarity test for time series (ADF test with constant and trend)
Variables ADF statistics P-value
xii 0.7ii305 0.9924
V3 i.55234 0.507i
y4 -i.3687 0.5993
The presented time series are also non-stationary. The results of stationarity test for cointegration regression residuals are presented in Table 5.
Table 5. Stationarity test for residuals of cointegration regressions
Variables ADF statistics P-value
y3 ~ Xii -i.94423 0.0496i**
y4 ~ Xii -3.29i78 0.000974i***
The analysis confirms a long-term relationship between Web search queries and statistical indicators (Fig. 4).
Use of patents (xn) Number of patent applications (y3) — Number of patents issued (y4) Fig. 4. Dynamics of Web search queries and statistical indicators
The popularisation of innovations is inextricably linked with the number of stories about the innovation system development published in the media. Such publications contribute to the creation of an information environment that forms ideas about innovation and innovative development. The identification of long-term relationships between the number of mentions in the media of various aspects of innovation activities and quantitative indicators of the innovative development from official sources (Federal State Statistics Service) confirms the influence of narratives on economic behaviour and economic processes.
The analysis of narratives contributes to a more complete and systematic understanding of the Russian innovation system's economic modelling in terms of complementarity and non-complementarity of conclusions.
Conclusion
Using the theoretical tools of narrative economics to examine national innovation system opens up new horizons both for the inclusion of additional qualitative and quantitative data in this analysis, and for the in-depth understanding of innovation processes. The logic of the research made it possible to consider the NIS modelling and narratives about this system, as well as to perform a cointegration analysis of the mutual influence of narratives and innovation activities indicators. The results discover and confirm the existence of a relationship between the number of mentions of various aspects of innovation in the media and quantitative indicators of the innovative development from the official sources.
A review of the publications on mathematical modelling of NIS revealed the limitations of this approach, both characteristic of any econometric models (for example, the availability and reliability of statistical data, methodological heterogeneity of statistics, the choice of model specification), and inherent in the modelling of national systems and economic growth (ambiguous interpretation of the national innovation
system, different approaches to defining the main determinants, immeasurability of innovations, taking into account system connections and economies of scale).
The limitations of economic and mathematical modelling make the reflection of innovative reality not just incomplete, but often even contradicting the experts' opinions regarding the NIS functioning. Understanding and coverage in media of innovation policy can be considered as an indicator of its performance: the better the NIS is functioning, the higher is the consistency between the parameters and conclusions of formal models and narratives.
Currently, there is no such consistency: the paper shows significant discrepancies between the results of quantitative and qualitative methods. Some of them are directly related to the indicators used in mathematical modelling, which either are not reflected in narratives, or are found, but with significant shortcomings and systemic omissions (for example, regarding the number of patents or scientific personnel). The situation can also be opposite: the phenomenon is often described in narratives, but is not used in formal modelling (consider the examples of 'brain drain' and the demand for innovation).
Since qualitative analysis presupposes deeper and more versatile judgments than quantitative analysis, there is a contradiction regarding innovation infrastructure and institutional specifics. However, modelling and narratives have the common ground, and in this case, both methods allow obtaining comparable conclusions regarding the financing of R&D, the lags in innovation cycles and the NIS performance.
References
Balatsky E. V., Ushakova S. E., Malakhov V. A., Yurevich M. A. (2017). Natsional'nye modeli tekhnologicheskogo razvitiya: sravnitel'nyy analiz [National models of technological development: a comparative analysis]. Journal of Institutional Studies, no. 9 (4), pp. 37-51. (in Russ.)
Bogdanova E. L., Ibragim A. (2018). Natsional'naya innovatsionnaya sistema i ee rol' v povyshenii VVP na dushu naseleniya (Vzglyad iz Rossii) [National innovation system and its role in increasing GDP per capita (evidence from Russia)]. Journal of Economic Regulation, no. 9 (1), pp. 33-39. (in Russ.)
Bookstaber R. (2021). Radikal'naya neopredelennost': manifest o prirode ekonomicheskikh krizisov, ikh prognozirovanii i preodolenii [Radical uncertainty: A manifesto on the nature of economic crises, their forecasting and overcoming]. Moscow: Eksmo Publ. 304 p. (in Russ.)
Volchik V. V. (2017). Narrativnaya i institutsional'naya ekonomika [Narrative and Institutional Economics]. Journal of Institutional Studies, no. 9 (4), pp. 132-143. https://doi. org/10.17835/2076-6297.2017.9.4.132-143. (in Russ.)
Volchik V. V., Maslyukova E. V., Panteeva S. A. (2021). Issledovanie podkhodov k mode-lirovaniyu natsional'nykh innovatsionnykh sistem [Research of approaches to modeling national innovation systems]. Ekonomicheskie i sotsial'nyeperemeny: fakty, tendentsii, prognoz = Economic and Social Changes: Facts, Trends, Forecast, vol. 14, no. 5, pp. 135-150. DOI: 10.15838/ esc.2021.5.77.8. (in Russ.)
Efimov V. M. (2016). Ekonomicheskaya nauka pod voprosom [Economic science is questionable]. Moscow: Infra-M Publ. 354 p. (in Russ.)
Collier P. (2021). Budushchee kapitalizma [The future of capitalism]. Moscow: Gaidar Institute. 376 p. (in Russ.)
Matrizaev B. D. (2019). Issledovanie sravnitel'noy effektivnosti natsional'noy innovatsion-noy sistemy i kachestva ekonomicheskogo rosta na primere sravnitel'nogo analiza stran OESR i BRIKS [Study of the comparative efficiency of the national innovation system and the quality of economic growth: The comparative analysis of OECD and BRICS countries]. Voprosy innovatsi-onnoy ekonomiki = Issues of Innovative Economy, no. 9 (3), pp. 673-692. (in Russ.)
Pyastolov S. M. (2012). Effektivnost' natsional'noy innovatsionnoy sistemy: infrastrukturnyy aspekt [Effectiveness of the national innovation system: Infrastructural aspect]. Transportnoe delo Rossii = Transport Business of Russia, no. 6 (1), pp. 18-20. (in Russ.)
Khrustalev E. Yu., Slavyanov A. S. (2016). Innovatsionno-orientirovannaya metodologiya otsenki sostoyaniya i vozmozhnostey rosta natsional'noy ekonomicheskoy sistemy [Innovation-oriented methodology for assessing the state and growth opportunities of the national economic system]. Politematicheskiy setevoy elektronnyy nauchnyy zhurnal Kubanskogogosudarstvennogo agrarnogo universiteta = Polythematic Online Scientific Journal of Kuban State Agrarian University, no. 115 (1), pp. 1-19. (in Russ.)
Tseluyko A. A. (2017). Otkrytaya natsional'naya innovatsionnaya sistema i ekonomicheskiy rost [Open national innovation system and economic growth]. Innovatsionnoe razvitie ekonomiki = Innovative Development of Economy, no. 6, pp. 203-212. (in Russ.)
Akerlof G., Snower D. (2016). Bread and bullets. Journal of Economic Behavior & Organization, vol. 126, pp. 58-71. https://doi.org/10.1016/j.jebo.2015.10.021.
Archibugi D., Coco A. (2005). Measuring technological capabilities at the country level: A survey and a menu for choice. Research Policy, vol. 34, no. 2, pp. 175-194. DOI:10.1016/j.re-spol.2004.12.002.
Artz K. W., Norman P. M., Hatfield D. E., Cardinal L. B. (2010). A longitudinal study of the impact of R&D, patents, and product innovation on firm performance. Journal of Product Innovation Management, vol. 27, no. 5, pp. 725-740. https://doi.org/10.1111/j.1540-5885.2010.00747.x.
Collier P. (2016). The cultural foundations of economic failure: A conceptual toolkit. Journal of Economic Behavior & Organization, vol. 126, pp. 5-24. https://doi.org/10.1016/j. jebo.2015.10.017.
Damodaran A. (2017). Narrative and numbers: The value of stories in business. Columbia Business School Publishing. 296 p.
Fagerberg J., Srholec M. (2008). National innovation systems, capabilities and economic development. Research Policy, vol. 37, no. 9, pp. 1417-1435. DOI:10.1016/j.respol.2008.06.003.
Gackstatter S., Kotzemir M., Meissner D. (2014). Building an innovation-driven economy - the case of BRIC and GCC countries. Foresight, vol.16, no. 4, pp. 293-308. DOI:10.1108/FS-09-2012-0063.
Griliches Z. (1998). Introduction to "R&D and productivity: The Econometric evidence". In: Griliches Z. R&D and productivity: The econometric evidence. University of Chicago Press, pp. 1-14.
Markey-Towler B. (2019). Rules, perception and emotion: When do institutions determine behaviour? Journal of Institutional Economics, vol. 15, no. 3, pp. 381-396. https://doi.org/10.1017/ S1744137418000267.
Nasierowski W., Arcelus F. J. (2003). On the efficiency of national innovation systems. SocioEconomic Planning Sciences, vol. 37, no. 3, pp. 215-234.
Newman J. I., Xue H., Watanabe N. M., Yan G., McLeod C. M. (2020). Gaming gone viral: An analysis of the emerging esports narrative economy. Communication and Sport. https://doi. org/10.1177/2167479520961036.
Peiffer C., Armytage R. (2019). Searching for success: A mixed methods approach to identifying and examining positive outliers in development outcomes. World Development, vol. 121, pp. 97-107. DOI: 10.1016/j.worlddev.2019.04.013.
Romer P. M. (1990). Endogenous technological change. Journal of Political Economy, vol. 98, no. 5, p. 2, pp. 71-102.
Sesay B., Yulin Z., Wang F. (2018). Does the national innovation system spur economic growth in Brazil, Russia, India, China and South Africa economies? Evidence from panel data. South African Journal of Economic and Management Sciences, vol. 21, no. 1, pp. 1-12. http://dx.doi. org/10.4102/saj ems.v21i1.1647.
Shiller R. J. (2017). Narrative economics. American Economic Review, vol. 107, no. 4, pp. 967-1004. https://doi.org/10.1257/aer.107.4.967.
Shiller R. J. (2019a). Narrative economics. How stories go viral and drive major economic events. Princeton University Press. 400 p.
Shiller R. J. (2019b). Narratives about technology-induced job degradation then and now. Journal of Policy Modeling, vol. 41, no. 3, pp. 477-488. https://doi.org/10.1016/JJP0LM-0D.2019.03.015.
Shiller R. J. (2021). Animal spirits and viral popular narratives. Review of Keynesian Economics, vol. 9, no. 1, pp. 1-10. https://doi.org/10.4337/R0KE.2021.01.01.
Stiglitz J. E., Greenwald B. (2014). Creating a learning society. Columbia University Press. 680 p.
Willett T. D. (2021). New developments in financial economics. Journal of Financial Economic Policy. https://doi.org/10.1108/JFEP-04-2021-0113.
Information about the authors
Vyacheslav V. Volchik, Dr. Sc. (Econ.), Prof., Head of Economic Theory Dept., Southern Federal University, 105/42 Bolshaya Sadovaya St., Rostov-on-Don, 344006, Russia Phone: +7 (863) 218-40-00 (ext. 13056), e-mail: volchik@sfedu.ru
Elena V. Maslyukova, Cand. Sc. (Econ.), Associate Prof., Associate Prof. of Economic Cybernetics Dept., Southern Federal University, 105/42 Bolshaya Sadovaya St., Rostov-on-Don, 344006, Russia
Phone: +7 (863) 218-40-40 (ext. 13065), e-mail: maslyukova@sfedu.ru
Sophia A. Panteeva, Research Intern, Southern Federal University, 105/42 Bolshaya Sadovaya St., Rostov-on-Don, 344006, Russia
Phone: +7 (863) 218-40-00 (ext. 13056), e-mail: panteeva@sfedu.ru
© Volchik V. V., Maslyukova E. V., Panteeva S. A., 2021
