148 r-economy.com R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 Online ISSN 2412-0731 Original Paper © Sandler, D.G., Gladyrev, D.A., Kochetkov, D.M., Zorina, A.D., 2022 doi 10.15826/recon.2022.8.2.012 UDC 378.3 JEL I22, I23, H52 Factors of research groups’ productivity: The case of the Ural Federal University D.G. Sandler1, D.A. Gladyrev1 , D.M. Kochetkov1, 2, A.D. Zorina1 1 Ural Federal University, Ekaterinburg, Russia;  d.a.gladyrev@urfu.ru 2 Centre for Science and Technology Studies, Leiden University, Leiden, Netherlands ABSTRACT Relevance. One of the main goals of state university support programs in Russia is to increase the number of scientific publications. In 2021, Project 5-100 was replaced by the program PRIORITY 2030 (Strategic Academic Leadership Pro- gram). The new program increased the significance of the factors affecting the number of publications in universities and the issue of the optimal allocation of funding among research groups. Research objective. This study examines the factors that affect the productivity of research groups at the university. Unlike the majority of other studies on this topic, this study analyzes scientific productivity at the level of research groups. Data and methods. The study was possible due to the availability of data for 79 research groups at the Ural Federal University for the period from 2014 to 2020. The total number of articles and the number of articles in journals with an impact factor of more than two were used as indicators of research groups’ per- formance. To determine the factors influencing these indicators, we used econo- metric models for panel data. We used two separate samples: for social sciences and humanities and for other sciences. Results. We identified the following factors affecting the performance of research groups: the number of participants, the age of the research group, the supervi- sor’s scientific age, and the amount of funding (the possibility of obtaining more funds or being denied funds). The most interesting result is the following: the supervisor’s scientific age and increased funding have a negative impact on the group’s performance. The article provides possible explanations for these results. Conclusion. Since the purpose of creating and funding research groups is pri- marily to increase their productivity, the results may be in favor of younger su- pervisors. University managers may also be interested in the ambiguous impact of increased funding: we suppose that research groups are more motivated not by the actual funding but by the prospective amount they may get. KEYWORDS research groups, university economics, economics of higher education, science management, scientometrics, econometric analysis ACKNOWLEDGEMENTS The research funding from the Ministry of Science and Higher Education of the Russian Federation (Ural Federal University Program of Development within the Priority-2030 Program) is gratefully acknowledged. FOR CITATION Sandler, D.G., Gladyrev, D.A., Kochetkov, D.M., & Zorina, A.D. (2022). Factors of research groups’ productivity: The case of the Ural Federal University. R-economy, 8(2), 148–160. doi: 10.15826/recon.2022.8.2.012 Факторы продуктивности исследовательских групп: пример Уральского федерального университета Д.Г. Сандлер1, Д.А. Гладырев1 , Д.М. Кочетков1, 2, А.Д. Зорина1 1 Уральский федеральный университет, Екатеринбург, Россия;  d.a.gladyrev@urfu.ru 2 Центр исследований науки и технологий, Лейденский университет, Лейден, Нидерланды АННОТАЦИЯ Актуальность. Одной из основных целей программ поддержки государ- ственных университетов в России является увеличение количества на- учных публикаций. В 2021 году Проект 5-100 был заменен программой ПРИОРИТЕТ 2030 (Программа стратегического академического лидер- ства). Новая программа увеличила значимость факторов, влияющих на количество публикаций в университетах, и вопроса оптимального распре- деления финансирования между исследовательскими группами. Цель исследования. В данном исследовании рассматриваются факторы, влияющие на продуктивность исследовательских групп в университете. В отличие от большинства других исследований по этой теме, данное ис- следование анализирует научную продуктивность на уровне исследова- тельских групп. КЛЮЧЕВЫЕ СЛОВА исследовательские группы, экономика вуза, экономика высшего образования, управление наукой, наукометрия, эконометрический анализ https://doi.org/10.15826/recon.2022.8.2.012 https://doi.org/10.15826/recon.2022.8.2.012 mailto:d.a.gladyrev@urfu.ru mailto:d.a.gladyrev@urfu.ru R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 149 r-economy.com Online ISSN 2412-0731 Introduction In any economy, universities and research organizations have limited and usually insuffi- cient resources to provide funding for all possi- ble topics and projects. Every year, universities and academic institutions have to distribute limited funds between their research groups to maximize the overall research performance. Government agencies and scientific foundations are dealing with a similar problem by setting models and rules for funds distribution between organizations, teams, and individual scientists. Sometimes the task is different – how to measure the effectiveness of current funding and reallo- cate funds without negative consequences. There is a need for the data on the factors affecting re- search groups’ performance to allow for more evidence-based decision-making. In this case, it Данные и методы. Исследование стало возможным благодаря наличию данных по 79 научным группам Уральского федерального университета за период с 2014 по 2020 годы. В качестве показателя работы исследователь- ских групп используются показатели её общего числа статей и числа ста- тей в журналах с импакт-фактором более двух. Для определения факторов, влияющих на эти показатели, использовались эконометрические модели панельных данных. Мы использовали две отдельные выборки: по социаль- но-гуманитарным наукам и по прочим наукам. Результаты. Выявлены следующие факторы, влияющие на результаты работы групп: количество участников, возраст исследовательской группы, научный возраст руководителя группы и объем финансирования. Наиболее интерес- ный результат заключается в следующем: научный возраст научного руково- дителя и увеличение финансирования негативно сказываются на результатив- ности группы. В статье приведены возможные объяснения этих результатов. Вывод. Поскольку целью создания и финансирования исследовательских групп является прежде всего повышение их научной результативности, результаты могут говорить в пользу назначения более молодых руко- водителей. Университетских управленцев также может заинтересовать неоднозначное влияние увеличения финансирования: мы полагаем, что исследовательские группы больше мотивированы не фактическим финан- сированием, а будущей суммой, которую они могут получить. БЛАГОДАРНОСТИ Исследование выполнено при финансовой поддержке Мини- стерства науки и высшего обра- зования Российской Федерации в рамках Программы разви- тия Уральского федерального университета имени первого Президента России Б.Н. Ельцина в соответствии с программой стратегического академического лидерства «Приоритет-2030» ДЛЯ ЦИТИРОВАНИЯ Sandler, D.G., Gladyrev, D.A., Kochetkov, D.M., & Zorina, A.D. (2022). Factors of research groups’ productivity: The case of the Ural Federal University. R-economy, 8(2), 148–160. doi: 10.15826/recon.2022.8.2.012 研究小组的科研效率：以乌拉尔联邦大学为例 桑德勒1，格拉德列夫1 ，科切特科夫1, 2，佐丽娜1 1 乌拉尔联邦大学，叶卡捷琳堡，俄罗斯； d.a.gladyrev@urfu.ru 2 莱顿大学科学技术研究中心，莱顿，荷兰 摘要 现实性：俄罗斯大学支持项目的主要目标之一是增加科研成果。2021 年，“5-100大学计划”被“优先2030计划”（战略学术领导力计划） 所取代。新的计划聚焦于大学的科研出版数量，并优化研究小组之间的 科研资金分配。 研究目标：本研究考察了影响大学各研究小组科研效率的因素。与其他 研究相似主题的大多学者不同，我们把目光转向研究小组的科研效率。 数据与方法：本文收集了乌拉尔联邦大学2014–2020年79个研究小组的 数据，这使研究成果具有代表性。数据来源是科研论文的总数和影响因子 大于2的论文数量。为了确定影响科研效率的因素，我们采用了经济面板 数据模式。另外，我们将科研数据分为两块：社会人文学科和其他学科。 研究结果：研究得出了影响科研效率的因素：参与者人数、研究小组的 成立时间、小组组长的科研年龄及研究经费。最有趣的结果如下：研究 小组组长的科研年龄和研究经费的增加对小组的科研结果有消极影响。 本文对这些结果提出了可能的解释。 结论：创建和资助研究小组的目的主要是提高参与者的科研绩效，从而 有利于任命更年轻的组长。大学的管理层可以对科研经费进行多层计 划：我们认为，更能激励研究小组成员的不是实际科研经费，而是未来 可以获得的额度。 关键词 研究小组，大学经济学，高等 教育经济学，科学管理，科学 计量学，计量模型分析 供引用 Sandler, D.G., Gladyrev, D.A., Kochetkov, D.M., & Zorina, A.D. (2022). Factors of research groups’ productivity: The case of the Ural Federal University. R-economy, 8(2), 148–160. doi: 10.15826/recon.2022.8.2.012 https://doi.org/10.15826/recon.2022.8.2.012 mailto:d.a.gladyrev@urfu.ru 150 r-economy.com R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 Online ISSN 2412-0731 is possible to maximize the efficiency of the re- search funding system. We have chosen research groups as the main actor in knowledge generation. Usually the data on research groups are not available and we can find only the data on universities, countries or in- dividual researchers. But since we have access to the performance indicators of research groups at the Ural Federal University, it is possible to con- duct such analysis. The purpose of the study is to determine the factors of research groups’ effectiveness. The number of publications was chosen as the main performance indicator. To achieve this goal, we collected the data on 79 research groups from the Ural Federal University (Ekaterinburg, Sverdlovsk region) for the period from 2014 to 2020 and studied its connection with the regio- nal economy. Another issue to be considered was data representativeness. Based on the data from the Ural Federal University, we have built econo- metric models to study the influence of different factors on research productivity and analyzed the results. Literature review The idea of using econometric methods to study the factors that affect R&D is not new. Such studies were conducted in the second half of the 20th century (Pakes, 1978; Griliches, 1979; Hall, Griliches and Hausman, 1986; Pardey, 1989). Many scholars studied the impact of university research on economic growth (Jaffe, 1989; Acs, Audretsch and Feldman, 1994; Jaffe and Tra- jtenberg, 1996; Martin, 1998; Varga, 1998, 2000, 2001; Fischer and Varga, 2003; Riddel and Schwer, 2003). Evaluations were made of research teams’ effectiveness based on a combination of econo- metric and scientometric methods (Adams et al., 2005). Among other things, these studies raised the question of the size and composition of re- search groups (Perovic et al., 2016). Quite illustra- tive in this respect is the study of the effectiveness of university hospitals in Tehran, performed on the basis of a combination of nonparametric ana- lysis methods-data envelope analysis (DEA) and stochastic frontier analysis (SFA) (Rezapour et al., 2015)in Tehran, Iran. METHODS: This study was conducted in 2012; the research population con- sisted of all hospitals affiliated to Iran and Tehran medical sciences universities of. Required data, such as human and capital resources information and also production variables (hospital outputs. There is a substantial body of research that establishes links between scientometric, econo- mic, and other indicators at the university level (Zinchenko and Yegorov, 2019; Geiger, 2004). In particular, for Russian universities, it was shown that the number of publications is higher in the universities that: 1) are engaged in research in physics; 2) have a higher share of international col- laborations; 3) accept students with a higher en- trance score; 4) have a larger share of Master’s and PhD students; 5) have higher levels of citations; 6) have a higher share of foreign students; 7) have a higher level of salaries in comparison with the re- gion’s average (Sandler & Gladyrev, 2020). A high positive correlation between the number of publi- cations and their quality (usually measured by the level of citations of these articles or the journal in general) has also been revealed by international studies at the level of individual researchers (Mi- chalska-Smith and Allesina, 2017), at the univer- sity level (Hayati and Ebrahimy, 2009), and at the national level (Lawani, 1986). Other studies have shown a positive effect of collaboration (Landry et al., 1996), especially international (Aldieri et al., 2018; Aldieri et al., 2019). A J-shaped impact of government funding was also revealed in some sectors, but there was no impact of business funding (Beaudry & Allaoui, 2012). There is evidence of the positive impact of the long-term university-industry interactions (Garcia et al., 2020). In a study based on the uni- versity data in Leuven (Belgium), the authors have shown higher scientific productivity of female re- searchers and researchers with an academic degree (De Witte & Rogge, 2010). Another study based on the Spanish data, on the contrary, demonstrated a higher scientific performance of male researchers (Albert et al., 2016). Some other studies compared young and older researchers: it was found that the young researchers have a higher level of scientific performance (Levin and Stephan, 1989; Albert et al., 2016). It is also worth noting that all these fac- tors can have a different impact on scientific pro- ductivity, depending on the level of the considered journals (Jung et al., 2017). Data and methods We used the data on the performance of 79 research groups of the Ural Federal University (Ekaterinburg) for the period from 2014 to 2020. The data were provided by the University’s De- partment of Strategic Development and Marke- ting. Due to the fact that not all research groups https://doi.org/10.15826/recon.2022.8.2.012 R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 151 r-economy.com Online ISSN 2412-0731 were functioning during the entire reviewed peri- od, the total number of observations was 438. From an organizational point of view, a re- search group (in the University’s documentation it is referred to as a “competence center”) is a team selected on a competitive basis in order to support its members’ research activities. Commitments to work on a specific topic formulated by the re- search team are recorded in the project passport, which also specifies the planned indicators for the number of publications, the amount of R & D, and additional indicators. Annually, a special commission of reputable researchers (direct conflicts of interest are exclu- ded) evaluates each group’s activities: the dyna- mics of the key indicators and correspondence to the obligations taken. These evaluations are used further by the special council that divides research groups into several funding groups. Groups with better results receive more funding. Every year, from 2 to 5 groups are denied funding for a year or are completely withdrawn from the project. In- stead, several new research groups are introduced on a competitive basis. One of the signs of the project’s success is a significant increase in the University’s publication activity (see Table 1): the total number of pub- lications almost tripled in 6 years and research groups kept more than a half of the University’s articles for almost all of the years (and more than 60% of articles in journals with an impact factor of more than 2). Despite these results, we assume that there is still room for improvement in terms of the funding system’s efficiency. In this study, we took all the variables included in research groups’ reports, with the exception of the number of articles in journals with IF>5 (as only few research groups have such publications). One variable (the supervisor’s scientific age) was collec- ted manually for all research groups from Scopus. The original dataset has eight variables: 1) ARTICLES is the number of articles of the research group indexed in Scopus and Web of Sci- ence in the reporting year. 2) ARTICLES IN IF>2 is the number of ar- ticles of the research group in journals with IF>2 indexed in Scopus and Web of Science in the re- porting year. 3) FUNDING is the amount of funding for the research group in the reporting year, million rubles. 4) PARTICIPANTS is the number of partic- ipants in the research group at the end of the re- porting year. Table 1 Dynamics of the number of articles published by the University’s researchers indexed in Scopus and Web of Science Year Total number of articles Articles of research groups Share of research groups’ articles Total articles in IF>2 journals Articles of research groups in IF>2 journals Share of research groups’ articles in IF>2 journals 2014 1413 836 59.16% 275 201 73.09% 2015 1742 1091 62.63% 387 265 68.48% 2016 2334 1256 53.81% 480 350 72.92% 2017 2930 1482 50.58% 611 391 63.99% 2018 3253 1594 49.00% 710 437 61.55% 2019 3772 1992 52.81% 954 567 59.43% 2020 3946 2001 50.71% 991 639 64.48% Source: compiled by the authors Table 2 Descriptive statistics ARTICLES ARTICLES IN IF>2 FUNDING PARTICIPANTS PROJECT AGE SOCIAL-HUM SUPERVISOR’S SCIENTIFIC AGE R&D Average 23.21 6.507 2.263 19.925 3.753 0.18 23.388 12.969 Median 17 2 1,4 15 4 0 0 0 Maximum 107 68 15.593 112 7 1 53 398.61 Minimum 0 0 0.08 1 1 0 14.76 32.77 Standard deviation 20.465 10.383 2.686 16.956 1.967 0.385 23.39 12.97 Source: compiled by the authors https://doi.org/10.15826/recon.2022.8.2.012 152 r-economy.com R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 Online ISSN 2412-0731 5) PROJECT AGE is the number of the year when the research group received funding (start- ing from 2014, when the program in its current format was launched). 6) SOCIAL-HUM is a binary variable equal to 1 if the research group belongs to social scienc- es and arts & humanities (there are 15 such groups with 79 observations) and 0 otherwise (there are 64 such groups with 359 observations); 7) SUPERVISOR’S SCIENTIFIC AGE is the number of years since the first supervisor’s Sco- pus-indexed article was published. 8) R & D is the declared amount of R&D in- come of the research group, million rubles. The main statistical characteristics of the vari- ables are shown in Table 2. The econometric models took into account the panel data structure; the tests proved that the best model is a model with fixed effects. The main variable is Δ ARTICLES; an addi- tional model also uses the variable Δ ARTICLES IN IF>2. The analysis of the second model is less interesting, since the selected indicator has a very low deviation (a significant number of research groups do not have any articles in journals with an impact factor higher than two). It should be noted that different subject areas have different average impact factors. The impact of the total time that the research group has been receiving organized funding was considered in variable PROJECT AGE. The model also included variables Δ FUNDING and Δ R & D. Using variables Δ ARTICLES, Δ FUNDING and Δ R & D (instead of ARTICLES, FUNDING, and R & D directly) helps us overcome endogeneity and outliers. Taking into account the fact that the effect of funding growth can be lagged, the mo- dels were created with both the current and the previous period value. Since many of the considered dependencies are not strictly linear, preference was given to non- linear dependencies. For this reason, the model did include natural logarithms of PARTICIPANTS and the SUPERVISOR’S SCIENTIFIC AGE. The SOCIAL-HUM variable was used to di- vide the sample into two and create a separate model for each of them. This is done under the assumption that research groups in social sciences and arts & humanities are significantly different from others. Table 3 confirms this assumption: al- most all the key indicators differ in comparison with the research groups specializing in social sci- ences and the humanities. Table 3 Average values by category of research groups Social sciences and humanities (N = 79) Other sciences (N = 359) Average number of articles 12.48 25.57 Average number of articles in journals with IF > 2 0.59 7.81 Average annual funding, mln 1.45 2.44 Average number of partici- pants 16 20.79 Source: compiled by the authors Thus, the following variables were taken as explanatory variables: 1. PROJECT AGE 2. ΔGROWTH 3. ΔFUNDING 4. LOG (PARTICIPANTS) 5. LOG (SCIENTIFIC AGE OF THE SUPER- VISOR) 6. R & D The issue of representativeness should be also considered. Is it possible to use the Ural Federal University’s data to study the performance factors of research groups in general? There is a num- ber of reasons for considering the University’s research groups as a representative sample: the university has a very high scientific performance (it ranks 10th among all the Russian institutions and 7th among universities by the total number of publications in 2015–2020, according to SciVal); it also boasts a diversity of subject areas. It should, however, be noted that the University’s scientific performance is connected with the structure of Sverdlovsk Region’s economy (and to some ex- tent to that of other neighboring regions). At the same time, we can assume that the University’s scientific performance also affects the structure of the region’s economy. The impact of research on the economic development of regional econ- omies is one of the tasks of the federal program “Priority 2030”1. Table 4 shows how the distribution of subject areas at the Ural Federal University differs from the national-level distribution. These differences include a higher share of articles in Physics and Astronomy, Materials Science and Chemistry, and a lower share in Medicine, Environmental Science, Energy and Agricultural and Biological Sciences. 1 https://priority2030.ru/about https://doi.org/10.15826/recon.2022.8.2.012 https://priority2030.ru/about R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 153 r-economy.com Online ISSN 2412-0731 Table 4 Comparison of the share of subject areas of publications of the Ural Federal University and in Russia as a whole Subject fields Share in Russia Share of the University Physics and Astronomy 14.4% 21.4% Engineering 12.2% 12.4% Materials Science 9.7% 16.5% Computer Science 6.6% 5.9% Medicine 6.5% <2% Earth and Planetary Sciences 6.2% 2.6% Chemistry 6.0% 8.6% Mathematics 5.7% 6.0% Social Sciences 4.8% 4.3% Environmental Science 4.6% 2.9% Biochemistry, Genetics and Molecular Biology 3.8% <2% Energy 3.3% <2% Agricultural and Biological Sciences 2.9% <2% Chemical Engineering 2.8% <2% Arts and Humanities 2.6% 2.1% Source: SciVal from 2016 to May 2022 Table 5 shows the differences between the economy of Sverdlovsk region and the national economy. These differences include a lower share of natural resources in Sverdlovsk Region and a higher share of manufacturing. The parallels between the deviations in the University’s subject areas from the national ones and between the deviations of the regional economy from the national economy are shown in Table  6. The main positive deviations in the University’s subject areas are related to physics, chemistry and materials sciences and these devi- ations can be connected with the dominance of the most powerful branch of Sverdlovsk region’s economy – manufacturing. On the contrary, the subject areas corresponding to earth sciences, energy, environmental economics, and agricul- ture at the Ural Federal University are below the national average, which can be explained by the lower (in comparison with the national) share of the region’s economy in mining and agriculture. All of these findings are consistent with the pre- vious studies that noted close links between uni- versities, government, and business in Russian regions (Vlasova & Lyashenko, 2021). Table 5 Industry structure of gross value added in 2019 in Russia Branch Share in Russia Share in Sverdlovsk Region Difference between Sverdlovsk region and country in general Agriculture, forest- ry, hunting, fishing and fish farming 4.1 2.4 –1.7 Natural resources / mining 13.5 2.1 –11.4 Manufacturing 16.8 31.9 15.1 Provision of elec- tric energy, gas and steam; air condi- tioning 2.9 3.9 1 Water supply; water disposal, or- ganization of waste collection and dis- posal, activities to eliminate pollution 0.6 1.1 0.5 Construction 5.4 4 –1.4 Wholesale and retail trade; repair of motor vehicles and motorcycles 14.2 12.7 –1.5 Transportation and storage 7.3 7.5 0.2 Activities of hotels and public catering 1 1 0 Information and communication activities 3 2.4 –0.6 Financial and in- surance activities 0.5 0.2 –0.3 Real estate opera- tions 10 10.4 0.4 Professional, scien- tific and technical activities 4.3 4.2 –0.1 Administrative ac- tivities and related additional services 2.3 2 –0.3 Public administra- tion and military security; social security 5.6 5.7 0.1 Education 3 3.1 0.1 Health and social services activities 4 4.1 0.1 Activities in the field of culture, sports, leisure and entertainment 1 0.7 –0.3 Provision of other types of services 0.5 0.6 0.1 Activity of house- holds as employers 0 0 0 Source: Rosstat: https://gks.ru/bgd/regl/b21_14p/Main.htm https://doi.org/10.15826/recon.2022.8.2.012 https://gks.ru/bgd/regl/b21_14p/Main.htm 154 r-economy.com R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 Online ISSN 2412-0731 Among the factors that speak in favor of the representativeness of the data is the fact that the University was formed relatively recently by merging a classical and technical university (with different cultures of academic activity). The final argument is that the sample includes groups that differ in terms of their research experience and the level of citation. It should be noted that the detected dependencies will be sufficiently re- liable only for the Ural Federal University, and in other universities, due to historical, organiza- tional and subject area differences, the patterns may be different. Some variables were not used for our ana- lysis because their variation was too low. The most interesting of these variables is the super- visor’s gender. Table 7 shows the distribution of research groups by the supervisor’s gender and subject area. Of the 79 research groups under review, 60  are supervised by men and 19, by women. At the same time, among the groups in social sciences and the humanities, women lead 9 out of 15 research groups. Table 7 Statistics of research groups by the supervisor’s gender Social sciences and humanities Other sciences Total Male 6 54 60 Female 9 10 19 Source: compiled by the authors Results The correlation matrix (see Table 8) gives us a basic understanding of the relationships between the variables and helps us make sure that the re- sulting models will not have multicollinearity (high correlation between the factors). It should be noted that an increase in the number of articles does not result in a decrease in their quality. The correlation coefficient between an increase in the number of articles and an in- crease in the number of articles in journals with IF>2 is 0.56. Thus, the goals of increasing the total number and quality of articles are not contradic- tory and even accompany each other. Previously, a similar link was established for Russian univer- sities (Sandler & Gladyrev, 2020), and now it has been demonstrated at the level of individual re- search groups. Our conclusions, however, cannot be interpreted in such a way that an increase in the number of articles will always be accompanied by an increase in their quality. Table 9 shows the results of the first model with fixed effects, where the explained variable is the growth in the number of articles of the re- search group. The most reliable factor determining the growth in the number of articles is the size of the given research group. This means that an increase in the size of the research group leads to an in- crease in the number of scientific articles and this result is not as trivial as it may seem. Often, es- Table 6 Comparison of the differences in scientific performance between the University and Russia and corresponding branches of the regional economy and Russia Branch Difference between Russia and Sverdlovsk region Subject area Difference between Russia and the University Mining Russia: 13.5%SR: 2.1% ↓ Earth and Planetary Sciences Russia: 6.2% UrFU: 2.6%↓ Environmental Science Russia: 4.6%UrFU: 2.9%↓ Energy Russia: 3.3%UrFU: <2%↓ Manufacturing Russia: 16.8%SR: 31.9% ↑ Physics and Astronomy Russia: 14.4%UrFU: 21.4%↑ Materials Science Russia: 9.7%UrFU: 16.5%↑ Chemistry Russia: 6.0%UrFU: 8.6%↑ Agriculture, forestry, hunting, fishing and fish farming Russia: 4.1% SR: 2.4% ↓ Agricultural and Biological Sciences Russia: 2.9% UrFU: <2%↓ Source: SciVal from 2016 to May 2022 and Rosstat: https://gks.ru/bgd/regl/b21_14p/Main.htm https://doi.org/10.15826/recon.2022.8.2.012 https://unicode-table.com/ru/2191/ https://unicode-table.com/ru/2191/ https://unicode-table.com/ru/2191/ https://unicode-table.com/ru/2191/ https://gks.ru/bgd/regl/b21_14p/Main.htm R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 155 r-economy.com Online ISSN 2412-0731 Table 8 Correlation matrix Δ ARTICLES Δ ARTICLES IN IF>2 PROJECT AGE Δ FUNDING LOG(PARTICI- PANTS) LOG(SUPERVISOR’S SCIENTIFIC AGE) Δ R & D Δ ARTICLES 1.00 Δ ARTICLES IN IF>2 0.56 1.00 PROJECT AGE 0.01 0.10 1.00 Δ FUNDING –0.09 –0.10 0.44 1.00 LOG(PARTICIPANTS) 0.20 0.11 0.38 0.04 1.00 LOG(SUPERVISOR’S SCIENTIFIC AGE) –0.02 0.03 0.10 –0.05 0.18 1.00 Δ R & D –0.05 –0.03 –0.03 –0.12 0.03 0.03 1.00 Table 9 Model for the number of the research group’s articles Variable Explained variable – Δ ARTICLES Other subject areas Social sciences and humanities (1) (2) (3) (4) PROJECT AGE –0.456 (0.71) 0.67 (1.01) 2.24 (1.37) 3.83*** (1.2) Δ FUNDING –0.22 (0.49) –1.53* (0.73) Δ FUNDING(-1) –0.55 (0.55) –0.85** (0.36) LOG(PARTICIPANTS) 6.76*** (2.34) 4.67 (3.31) 7.8* (3.75) 9.48*** (2.75) LOG(SUPERVISOR’S SCIENTIFIC AGE) –10.85** (5.22) –10.29* (5.56) –14.28** (6.09) –21.1*** (5.17) Δ R & D –0.036* (0.02) –0.03* (0.018) –0.072 (0.19) –0.11 (0.16) CONSTANT 20.39 (15.68) 18.2 (14.9) –7.59 (7.48) –8.91* (4.95) Panel data model with fixed effects Robust standard errors are shown in parentheses *** significant at the 1% significance level ** significant at the 5% significance level * significant at the 10% significance level pecially when the recruitment of new members of the research group is limited only to university employees, students, and postgraduates, it may seem that new members of the group will not give a significant increase in articles (or will do it only with a lag); and the main growth potential lies in increasing the productivity of the group’s core. The results show that this is not true. An interesting and even paradoxical result connected with the coefficient of the supervisor’s scientific age is as follows: a negative sign and high statistical reliability indicate that the more expe- rienced is the supervisor, the lower is the group’s rate of publication growth; and vice versa. Some reservations, however, should be made regarding the interpretation of this result: it does not mean that groups with an experienced scientific super- visor have a low scientific outcome, but that such groups are less likely to increase their scientific performance, and their potential is already rea- lized. Since one of the main goals of forming re- search groups is increasing their scientific pro- ductivity by using university funding, this result can be used in favor of appointing younger mana- gers. Some previous studies have shown the lower scientific performance of more senior researchers in many subject areas (Levin and Stephan, 1989; Albert et al., 2016). https://doi.org/10.15826/recon.2022.8.2.012 156 r-economy.com R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 Online ISSN 2412-0731 The role of funding growth is also a paradoxi- cal result at first glance. We could expect a reliable direct relationship between increased funding and the growth in the number of articles, but it is not observed both for current and previous fun- ding; moreover, there is some evidence in favor of the inverse relationship. It is fair to note that the statistical reliability of this result is not high. One explanation for this result is the motivation factor: research groups whose funding has been reduced or increased slightly are more motiva- ted to achieve high scientific performance in the hope of receiving higher funding for the next year. The groups that have already received substantial funding can be satisfied with merely maintai- ning the last year’s level of performance. Thus, it is possible that a prospective increase in funding is a stronger motivating factor than maintaining the same level of funding. For the growth in R & D, the results are also interesting: in all the models the dependence is negative (but only in two models this coefficient is significant at the 10% significance level). It means that the higher is the growth in R & D income, the lower is the increase in the number of articles. This may indicate that income-gene- rating research work and scientific publications are not complementary activities, but rather sub- stitutes – at least in terms of the dynamics of the indicators. Table 10 shows the results of the second model, where the explained variable is the num- ber of articles of the research group in journals with an impact factor of more than two. The results of this model show approximately the same results as it was for the first model. The growth in the number of articles in journals with IF>2 is also positively connected with the num- ber of participants in the research group, nega- tively connected with the supervisor’s scientific age (but this result is statistically significant only for social sciences and arts & humanities), and there is weak evidence of the negative impact of increased funding on the growth in the num- ber of articles. Like in the previous model, there is a negative impact of the growth in research volumes for other sciences. We found a significant impact of the project’s period for projects in social sciences and arts & humanities, where the number of publications in high-impact journals tends to be lower (WoS Arts and Humanities Citation Index doesn’t have IF at all). It can be assumed that the accumulated ex- perience and interaction within the team allow research groups to increase their publications in such journals over time. Table 10 Model for the number of research group articles in journals with IF>2 Variable Explained variable – Δ ARTICLES IN IF>2 Other subject areas Social sciences and humanities (1) (2) (3) (4) PROJECT AGE 0.413 (0.48) 0.83 (0.72) 1.65** (0.73) 2.11** (0.86) Δ FUNDING –0.57* (0.32) –0.84 (0.63) ΔFUNDING(–1) –0.39 (0.36) 0.04 (0.39) LOG(PARTICIPANTS) 2.63** (1.26) 2.07 (1.52) 2.71* (1.27) 3** (1.33) LOG(SUPERVISOR’S SCIENTIFIC AGE) –1.67 (2.81) –2.96 (2.6) –5.34** (2.33) –8.43** (3.66) Δ R & D –0.017** (0.007) –0.013* (0.007) 0.08 (0.69) 0.08 (0.07) Constant term –3.1 (7.2) 0.23 (5.01) –6.95 (3.12) –4.75 (2.57) Panel data model with fixed effects Robust standard errors are shown in parentheses ** significant at the 5% significance level * significant at the 10% significance level https://doi.org/10.15826/recon.2022.8.2.012 R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 157 r-economy.com Online ISSN 2412-0731 Conclusion This paper contributes to the study of the factors of scientific productivity at the level of research groups. The econometric models based on the data of the Ural Federal University have brought to light the factors that affect the scien- tific performance of research groups. The main factor influencing the growth in the number of articles is the number of research group’s participants. The positive effect of this factor turned out to be statistically significant for most of the models. The influence of the next two factors was paradoxical. First, there is a negative influence of the supervisor’s academic age on the growth in the number of articles. Although the paper explains this result as well as cautions against its misinterpretation, the main recom- mendation is that more credit should be given to younger managers. Secondly, the negative impact of increased funding on the growth in the num- ber of articles of the research group. This result is explained by the specific motivation of research groups, but it should also be interpreted with great caution, especially because it can affect the university leadership’s decision-making regarding funding allocation. The age of the research group is also one of the factors that positively affects the growth in scientific performance, but only for social sci- ences and arts & humanities, and especially for high-impact articles. Perhaps this is because so- cial sciences and arts & humanities in Russia are younger, which is why the effect of the creation of such groups is stronger. In both models for other sciences, a negative relationship between the growth in articles and the growth of R&D income was detected. This suggests that a simultaneous growth in these indi- cators can be problematic. The value of these results may be influenced by the fact that only research groups of the Ural Federal University are included in the sample. This was a forced limitation caused by the fact that we had access only to one university’s data on in- dividual research groups while the corresponding data for other universities are closed. It is shown that the structure of the Ural Federal University’s publications to some extent reflects the specifics of Sverdlovsk region, and with a high degree of reliability, the conclusions can be applied only to this university, but the large sample size and variety of subject areas allow us to assess the pos- sibility of applying these conclusions to other uni- versities optimistically. It will be interesting to observe the changes in the performance of research groups in con- nection with the launch of the new PRIORITY 2030 federal program in Russia and changes in the target indicators in comparison with the previous program (Project 5-100). Due to the new emphasis on the number of articles in the first and second quartiles, we should expect an increase in the number of high-quality publica- tions. It is unlikely that this increase will lead to a decrease in the total number of publications (as quality and quantity usually go together), but the growth rate of the total number of publications of research groups is likely to decrease. References Acs, Z.J., Audretsch, D.B., & Feldman, M.P. (1994) ‘R&D spillovers and recipient firm size’. The Review of Economics and Statistics, 76, 336–340. Albert, C., Davia, M.A., & Legazpe, N. (2016). Determinants of research productivity in Spanish academia. European Journal of Education, 51(4), 535–549. Adams, J.D., Black, G.C., Clemmons, J.R., & Stephan, P.E. (2005) ‘Scientific teams and institu- tional collaborations: Evidence from U.S. universities, 1981–1999’. Research Policy, 34(3), 259–285. https://doi.org/10.1016/j.respol.2005.01.014 Aldieri, L., Guida, G., Kotsemir, M., & Vinci, C.P. (2019). An investigation of impact of research collaboration on academic performance in Italy. Quality & Quantity, 53(4), 2003–2040. Aldieri, L., Kotsemir, M., & Vinci, C.P. (2018). The impact of research collaboration on aca- demic performance: An empirical analysis for some European countries. Socio-Economic Planning Sciences, 62, 13–30. Beaudry, C., & Allaoui, S. (2012). Impact of public and private research funding on scientific production: The case of nanotechnology. Research Policy, 41(9), 1589–1606. https://doi.org/10.15826/recon.2022.8.2.012 https://doi.org/10.1016/j.respol.2005.01.014 158 r-economy.com R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 Online ISSN 2412-0731 De Witte, K., & Rogge, N. (2010). To publish or not to publish? On the aggregation and drivers of research performance. Scientometrics, 85(3), 657–680. Fischer, M. M. and Varga, A. (2003) ‘Spatial knowledge spillovers and university research: Evi- dence from Austria’. Annals of Regional Science, 37, 303–322. Garcia, R., Araújo, V., Mascarini, S., Santos, E.G., & Costa, A.R. (2020). How long-term uni- versity-industry collaboration shapes the academic productivity of research groups. Innovation, 22(1), 56–70. Geiger, R.L. (2004). Knowledge and money: Research universities and the paradox of the mar- ketplace. Stanford University Press. Griliches, Z. (1979) ‘Issues in assessing the contribution and development of research to produc- tivity growth’. The Bell Journal of Economics, 10(1), 92–116. https://doi.org/10.2307/3003321 Hall, B.H., Griliches, Z., & Hausman, J.A. (1986). Patents and R and D: Is There A Lag? Interna- tional Economic Review, 27, 265–283. Hayati, Z., & Ebrahimy, S. (2009). Correlation between quality and quantity in scientific produc- tion: A case study of Iranian organizations from 1997 to 2006. Scientometrics, 80(3), 625–636. Jaffe, A.B. (1989). Real effects of academic research. The American Economic Review, 79, 957–970. Jaffe, A.B., & Trajtenberg, M. (1996). Flows of knowledge from universities and federal labo- ratories: Modeling the flow of patent citations over time and across institutional and geographic boundaries. Proceedings of the National Academy of Science, 93, 12671–12677. Jung, H., Seo, I., Kim, J., & Kim, B.K. (2017). Factors affecting government-funded research quality. Asian Journal of Technology Innovation, 25(3), 447–469. Landry, R., Traore, N., & Godin, B. (1996). An econometric analysis of the effect of collaboration on academic research productivity. Higher Education, 32(3), 283–301. Lawani, S.M. (1986). Some bibliometric correlates of quality in scientific research. Sciento- metrics, 9(1-2), 13–25. Levin, S.G., & Stephan, P.E. (1989). Age and research productivity of academic scientists. Research in Higher Education, 30(5), 531–549. Martin, F. (1998). The economic impact of Canadian university R&D. Research Policy, 27, 677–687. Maslennikov, V.V. (2013). Project management of scientific activities of the university. Methodological tools. Moscow: Paleotype Michalska-Smith, M.J., & Allesina, S. (2017). And, not or: quality, quantity in scientific pub- lishing. PloS ONE, 12(6). Pakes, A.S. (1978). Economic incentives in the production and transmission of knowledge: an em- pirical analysis. Harvard University, Cambridge, MA. Pardey, P.G. (1989). The Agricultural Knowledge Production Function : An Empirical Look. The Review of Economics and Statistics, 71(3), 453–461. Perovic, S., Radovanovic, S., Sikimic, V. and Berber, A. (2016). Optimal research team composi- tion: data envelopment analysis of Fermilab experiments. Scientometrics, 108(1), 83–111. https://doi. org/10.1007/s11192-016-1947-9 Rezapour, A., Ebadifard Azar, F., Yousef Zadeh, N., Roumiani, Y.A., & Bagheri Faradonbeh, S. (2015). Technical efficiency and resources allocation in university hospitals in Tehran, 2009–2012. Medical Journal of the Islamic Republic of Iran, 29(1), 839–850. Riddel, M., & Schwer, R.K. (2003). Regional innovative capacity with endogenous employment: Empirical evidence from the U.S. Review of Regional Studies, 33, 73–84. Sandler, D.G., & Gladyrev, D.A. (2020). Construction of a cost-effective system of target indica- tors for the development of university research activities, taking into account correlation dependen- cies. Statistics and Economics, 17(4), 71–84 Varga, A. (1998). University research and regional innovation: A spatial econometric analysis of academic technology transfers. Boston: Kluwer. https://doi.org/10.15826/recon.2022.8.2.012 https://doi.org/10.2307/3003321 https://doi.org/10.1007/s11192-016-1947-9 https://doi.org/10.1007/s11192-016-1947-9 R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 159 r-economy.com Online ISSN 2412-0731 Varga, A. (2000). Local academic knowledge transfers and the concentration of economic activity. Journal of Regional Science, 40, 289–309. Varga, A. (2001). Universities and regional economic development: Does agglomeration matter? In Johansson, B., Karlsson, C., & Stough, R. (eds). Theories of endogenous regional growth: Lessons for regional policies. New York/Berlin: Springer-Verlag, pp. 345–367. Vlasova, N.Y., & Lyashenko, E.A. (2021). University-business-government relations in the de- velopment of the institutional environment of Russian regions. R-Economy, 7(4), 214–224. Zinchenko D.I., Egorov A.A. (2019). Modeling the effectiveness of Russian universities. Eco- nomic Journal of the Higher School of Economics, 23(1), 143–172. Information about the authors Daniil G. Sandler – PhD in Economics, Associate Professor at the Department of Interna- tional Economics and Management, Graduate School of Economics and Management; Leading Re- searcher, Research Laboratory for University Development Issues, First Vice-Rector (Economics and Strategy), Ural Federal University (19 Mira Str., 620002 Ekaterinburg, Russia); Scopus Author ID: 56581474400; ORCID: 0000-0002-5641-6596; e-mail: d.g.sandler@urfu.ru Dmitry A. Gladyrev – Senior Lecturer at Department of Economics, Graduate School of Eco- nomics and Management, Ural Federal University (19 Mira Str., 620002 Ekaterinburg, Russia); Sco- pus Author ID: 57208191401; ORCID: 0000-0001-5746-0495; e-mail: d.a.gladyrev@urfu.ru Dmitry M. Kochetkov – PhD in Economics, Senior Researcher at the Laboratory for University Development, Ural Federal University (19 Mira Str., 620002 Ekaterinburg, Russia); PhD Candidate at the Center for Science and Technology Studies, Leiden University (Willem Einthoven Building, Kolffpad 1, 2333 BN Leiden, Netherlands); Scopus Author ID: 57194605735; ORCID: 0000-0001- 7890-7532; e-mail: d.kochetkov@cwts.leidenuniv.nl Anna D. Zorina – Deputy of Head, Department of Strategic Development and Marketing, Ural Federal University (19 Mira Str., 620002 Ekaterinburg, Russia) ARTICLE INFO: received April 15, 2022; accepted June 2, 2022 Информация об авторах Сандлер Даниил Геннадьевич – кандидат экономических наук, доцент кафедры международной экономики и менеджмента, Институт экономики и управления; ведущий специалист научно-исследовательской лаборатории по проблемам университетского развития, первый проректор по экономике и стратегическому развитию, Уральский федеральный университет (620002, Россия, Екатеринбург, ул. Мира, 19); Scopus Author ID: 56581474400; ORCID: 0000-0002-5641-6596; e-mail: d.g.sandler@urfu.ru Гладырев Дмитрий Анатольевич – старший преподаватель кафедры экономики, Институт экономики управления, Уральский федеральный университет (620002, Россия, Екатеринбург, ул. Мира, 19); Scopus Author ID: 57208191401; ORCID: 0000-0001-5746-0495; e-mail: d.a.gladyrev@urfu.ru Кочетков Дмитрий Михайлович – кандидат экономических наук, старший научный сотрудник научно-исследовательской лаборатории по проблемам университетского развития, Уральский федеральный университет (620002, Россия, Екатеринбург, ул. Мира, 19); аспирант Центра исследований науки и технологий, Лейденский университет (2333 BN, Нидерланды, Лейден, Willem Einthoven Building, Kolffpad 1); Scopus Author ID: 57194605735; ORCID: 0000- 0001-7890-7532; e-mail: d.kochetkov@cwts.leidenuniv.nl Зорина Анна Дмитриевна – заместитель директора управления стратегического развития и маркетинга, Уральский федеральный университет (620002, Россия, Екатеринбург, ул. Мира, 19); e-mail: a.d.zorina@urfu.ru ИНФОРМАЦИЯ О СТАТЬЕ: дата поступления 15 апреля 2022 г.; дата принятия к печати 2 июня 2022 г. https://doi.org/10.15826/recon.2022.8.2.012 https://www.scopus.com/authid/detail.uri?authorId=56581474400 https://orcid.org/0000-0002-5641-6596 mailto:d.g.sandler@urfu.ru https://www.scopus.com/authid/detail.uri?authorId=57208191401 https://orcid.org/0000-0001-5746-0495 mailto:d.a.gladyrev@urfu.ru https://www.scopus.com/authid/detail.uri?authorId=57194605735 https://orcid.org/0000-0001-7890-7532 https://orcid.org/0000-0001-7890-7532 mailto:d.kochetkov@cwts.leidenuniv.nl https://www.scopus.com/authid/detail.uri?authorId=56581474400 https://orcid.org/0000-0002-5641-6596 mailto:d.g.sandler@urfu.ru https://www.scopus.com/authid/detail.uri?authorId=57208191401 https://orcid.org/0000-0001-5746-0495 mailto:d.a.gladyrev@urfu.ru https://www.scopus.com/authid/detail.uri?authorId=57194605735 https://orcid.org/0000-0001-7890-7532 https://orcid.org/0000-0001-7890-7532 mailto:d.kochetkov@cwts.leidenuniv.nl mailto:a.d.zorina@urfu.ru 160 r-economy.com R-ECONOMY, 2022, 8(2), 148–160 doi: 10.15826/recon.2022.8.2.012 Online ISSN 2412-0731 作者信息 桑德勒·丹尼尔·根纳季耶维奇——经济学博士，国际经济管理系副教授，经济管 理学院，大学发展研究实验室资深专家，经济与战略发展第一副校长，乌拉尔联邦大 学（邮编：620002，俄罗斯，叶卡捷琳堡，米拉大街19号）；Scopus Author ID: 56581474400; ORCID: 0000-0002-5641-6596; 邮箱：d.g.sandler@urfu.ru 格拉德列夫·德米特里·阿纳托利耶维奇——经济系高级讲师，经济管理学院，乌拉尔 联邦大学（邮编：620002，俄罗斯，叶卡捷琳堡，米拉大街19号）；Scopus Author ID: 57208191401; ORCID: 0000-0001-5746-0495; 邮箱：d.a.gladyrev@urfu.ru 科切特科夫·德米特里·米哈伊洛维奇——经济系博士，大学发展研究实验室高级研究 员，乌拉尔联邦大学（邮编：620002，俄罗斯，叶卡捷琳堡，米拉大街19号）；莱顿大 学科学技术研究中心博士在读（邮编：2333 BN，荷兰，莱顿，Willem Einthoven Build- ing, Kolffpad 1）；Scopus Author ID: 57194605735; ORCID: 0000-0001-7890-7532; 邮箱：d.kochetkov@cwts.leidenuniv.nl 佐丽娜·安娜·德米特里耶夫娜——战略发展与市场部副部长，乌拉尔联邦大学（邮 编：620002，俄罗斯，叶卡捷琳堡，米拉大街19号）；邮箱：a.d.zorina@urfu.ru https://doi.org/10.15826/recon.2022.8.2.012 https://www.scopus.com/authid/detail.uri?authorId=56581474400 https://orcid.org/0000-0002-5641-6596 mailto:d.g.sandler@urfu.ru https://www.scopus.com/authid/detail.uri?authorId=57208191401 https://orcid.org/0000-0001-5746-0495 mailto:d.a.gladyrev@urfu.ru https://www.scopus.com/authid/detail.uri?authorId=57194605735 https://orcid.org/0000-0001-7890-7532 mailto:d.kochetkov@cwts.leidenuniv.nl mailto:a.d.zorina@urfu.ru