In Portugal, research productivity is nowadays essential for the positive assessment of academics, research units and study programmes. Academic inbreeding has been highlighted in the literature as one of the factors influencing research productivity. This paper tests the hypothesis that inbreeding is detrimental for research productivity, measured through the number of publications listed in Scopus. The study resorts to a database provided by the national Agency for Assessment and Accreditation of Higher Education (A3ES), which comprises all academics teaching in all Portuguese institutions in the academic year 2015/2016. The sample selected for the analysis contains all academics with a PhD in Sociology (N=289). The study uses a special regression model for the analysis: the negative binomial logit hurdle. This was necessary given the large amount of academics with no publications or citations in Scopus, which were the dependent variables to assess research performance. The analysis provides separate results for the probability of inbred academics of having no papers/citations, and for the probability of producing more papers/citations than the non-inbred. Findings suggest that academic inbreeding, defined at the institutional level, has no negative effect on research productivity, contrary to what was expected. However, when defined at the national level, academic inbreeding is detrimental for the recognition and the impact of research: academics with a foreign PhD are more likely to have citations compared to academics who obtained their PhD in Portugal. A tendency was also noted that inbreeding might be more detrimental to research productivity in faculties of Economics than in Social Sciences and Humanities.

中文翻译：

---

葡萄牙社会学博士持有者的近亲繁殖和研究生产力

在葡萄牙，研究生产力如今对于学术界、研究单位和学习计划的积极评估至关重要。学术近亲繁殖在文献中被强调为影响研究生产力的因素之一。本文通过 Scopus 中列出的出版物数量来检验近亲繁殖不利于研究生产力的假设。该研究利用了国家高等教育评估和认证机构 (A3ES) 提供的数据库，其中包括 2015/2016 学年在所有葡萄牙机构教学的所有学者。为分析选择的样本包含所有拥有社会学博士学位的学者 (N=289)。该研究使用特殊的回归模型进行分析：负二项式 logit 障碍。考虑到大量学者在 Scopus 中没有发表或引用，这是评估研究绩效的因变量，这是必要的。该分析为近交学者没有论文/引用的概率以及产生比非近交的学者更多的论文/引用的概率提供了单独的结果。研究结果表明，与预期相反，在机构层面定义的学术近亲繁殖对研究生产力没有负面影响。然而，在国家层面上定义时，学术近亲繁殖不利于研究的认可和影响：与在葡萄牙获得博士学位的学者相比，拥有外国博士学位的学者更有可能获得引用。