Fransuellen Santos, Luciana Luz (p)

Eda Yucesoy

New educational demands arise from changes in a country's productive structure, especially with the modernization of production processes. Economically dynamic regions tend to be more attractive and concentrate investments in higher education. Between 2003 and 2014, higher education expansion programs created new institutions, increasing access to higher education in various regions in Brazil. It is plausible to think that these programs may have promoted changes in terms of the prevailing regional hierarchy. This study analyzes whether municipalities that received new higher education institutions experienced changes in their centrality. Using a synthetic indicator to measure centrality based on the number of enrollments by field of study, along with spatial econometric techniques, we analyze differences in Brazil's urban hierarchies and discuss the factors associated with these changes. The results show that over the years, there has been an increase in the centrality of cities outside major urban areas, reflecting the effects of higher education interiorization policies and the decentralization of educational opportunities. The study sheds light on the mechanisms driving the change in hierarchies, and discusses the role of population size, and municipality's capacity to attract services and qualified jobs.

Eda Yucesoy (p), Ahmet Salih Odabaş, Alim Arlı

Barbara Martini

Higher education institutions are considered as foundations of economic growth for attracting dynamic labor, services and industries and grounds for innovative practices. The geography of higher education in Turkey has drastically changed with the spreading of higher education institutions from the center to periphery since 1990s. Large state investments for new public universities covering entire country has gone hand in hand with mushrooming private universities, especially in Istanbul. Today higher education in Turkey is formed by a complex system of 208 universities with approximately 20,000 departments, accommodating more than 8.2 million university students (of which 4.6 million are undergraduate and half a million are graduate students). This dispersion of higher education institutions all over the country generates a vibrant yet obscure landscape. While some areas / cities manage to respond to the presence of higher education institutions, others fail despite their advantageous political positions. The distinctive geography of higher education institutions in Turkey produces urban relational systems that call for new tools and analyses.
This paper aims at exploring the relationship between higher education institutions and social characteristics of Turkish cities. Understanding these unique patterns and formulating new perspectives to analyze the complexity of this landscape have become an interesting research topic for us. A novel database is compiled for the research funded by Tubitak (The Scientific and Technological Research Council of Türkiye, project number 119K766). Data covers the 2019 outcomes of university placement test, published by the Higher Education Council and TurkStat 2019 census. The analysis is conducted on a district level in order to locate clusters of specializations, densities and agglomerations that describe the distinctive characteristics of the higher education “field" and social “milieu”. Analyzing and representing placements of 825.000 students in 922 districts reveal unique spatial typologies through the lens of higher education institutions. The paper adopts a relational mapping perspective, i.e. quantitative inquiry for identifying clusters of higher education institutions and social characteristics of districts. The paper indicates an alternative formation of urban regions and hierarchies in Turkey, based on knowledge-related clusters, social capacities and their competitiveness.

Barbara Martini (p), Martina Dal Molin , Chiara Leggerini, Miron Tequame

Luciana Luz

The literature on the effect of diversity in scientific team and publication quality gap and citation impact is inconclusive depending on the dataset, field of study and country of analysis. This paper tries to answer part of this puzzle by studying the role of gender diversity on publication quality, proxied by the number of citation impact in the field of social sciences. We use a novel dataset extracted from Scopus and merged with the Scimago Journal Ranking to include information on journal metrics. Our dataset consists of around 145 000 peer reviewed articles from 155 journals in Social Sciences, published between 2010 and 2025 and combined with indicators that assess scientific domains, and hence allowing comparison of journals and different indicators of journal prestige. We implement a novel methodology that estimates the association between gender diversity and citation impact, based on a two-step procedure. We first predict citation by factors that do not depend on gender, regressing citation on journal metrics. In this step we use the residuals, interpreting them as factors that account for the part of citation that are not explained by journal metrics. In the second step, we use such quartile distribution of the residuals in a probabilistic/logistic model to study the relationship between residuals and gender team diversity. Our measure of gender diversity is based on five levels of diversity, as measured by 5th-quantile distribution of the percentage of men among the authors. We find significant difference across gender diversity on how such residuals are predicted. We find the reverse for residuals that are worse predicted by journal metrics: women have 0.24 marginal effect of predicting these residuals, that depend least on metrics, while men have 0.30. Our results suggest that women are set to a higher bar when it comes to publication, i.e. most of their citation is explained by factors that are related to journal metrics, indicating that women’s citations are mostly explained by journal metrics. We also find that the relationship is reversed when we increase the composition of men in the authors group, as the number of men increase in the group, their likelihood of predicting high levels of residuals increases.