We model open access as facilitating full-text acquisition, which, while often increasing cites, can reduce cites from readers who refrain from citing superficially after realizing the article is not worth citing. We test the theory with data on over 200,000 science articles binned by cites in the pre-study period. Consistent with theory, we find that opening access to an article on the journal's website has a “Matthew effect” on citations: negative for the least-cited articles, positive for the most cited, and monotonic for quality levels in between. Estimates for broader open-access platforms and for cites coming from insiders versus outsiders also follow patterns consistent with theory.

中文翻译：

---

未见的引用：关于开放获取对跨质量范围的学术文章引用影响的理论和证据

我们将开放获取建模为促进全文获取，这虽然通常会增加引用次数，但可以减少读者的引用，他们在意识到文章不值得引用后避免表面引用。我们用在预研究期间按引用分类的超过 200,000 篇科学文章的数据来测试该理论。与理论一致，我们发现对期刊网站上的文章开放访问权对引用具有“马太效应”：引用最少的文章为负面，引用最多的文章为正面，两者之间的质量水平单调。对更广泛的开放获取平台和来自内部人员与外部人员的引用的估计也遵循与理论一致的模式。