Is the research performance of Chinese returnees better than that of their local counterparts?

Over the past few decades, the Chinese government has issued several policies to attract overseas elites to return to China and has given them privileges in academic resource allocation. This particular treatment has provoked complaints from domestic scientists who have received no additional resources. Therefore, the research performance of returnees and their local counterparts has been widely debated. We selected 4,770 returnees and their local counterparts in the same departments of the same institutions in the same year from 1984 to 2017, and compared their research performance using regression analysis. The results show that returnees have no significant advantage in publishing more papers or first-tier journal papers, but more returnees are corresponding authors in Natural Sciences. In addition, in Arts & Humanities and Social Sciences, returnees still have an advantage in publishing more papers and in more first-tier journal papers. This study has implications for the employment policies of Chinese universities/research institutions.


Introduction
Chinese academia has experienced a structural shift from a locally dominated system to a local-returnee hybrid system. Since the economic opening and reforms in 1978, the Chinese government has introduced several policies related to talented people. In 1978, the Ministry of Education issued the "Notice on Additional Selection of Foreign Students," specifying that sending students to study abroad increased to 3,000 that year. In 1982, allowing self-funded study abroad provided more opportunities for Chinese students to study abroad. Since then, China has witnessed a boom in studying abroad. Immediately after the economic system reform in 1984, the Chinese government launched institutional reform of the science and technology (S&T) system in 1985. The establishment of the post-doctoral system and appointment system significantly influenced Chinese academics, which supported the mobility of Chinese scientists (Yan et al., 2015). In 1993, the government adopted the policy to "support study overseas, encourage returns and guarantee 1 3 freedom of (international) mobility" (Cao et al., 2020). They also established the National Scholarship Council, supporting the outflow of students and scholars. There have also been several policies to attract overseas elites back to China. For example, the "Hundred Talent Program" (1993) provided 2 million yuan, and the "National Science Fund for Distinguished Young Scholars" (1994) offered 0.8 to 1 million yuan to each returning scientist to encourage outstanding scholars to return to China. In 1995, the Chinese leadership deepened the S&T reform by adopting a "Revitalizing the Nation through Science and Education Strategy." Programs such as the "Chang Jiang Scholars Program" (1998) have further supported excellent scientists. However, the previous policies did not attract enough elites to return home. By the end of 2002, the return rate of scholars who studied abroad and returned to China was only 14.5% (National Bureau of Statistics, 2022), indicating a noticeable brain drain trend in China.
To reverse the brain drain in China, the Chinese government added attraction policies as China's economy boomed (Jonkers & Tijssen, 2008). In 2008, the government announced the "Thousand Talents Plan", and in the following 2 years, the scheme grew to talents under 40 (Jia, 2018). All successful applicants were offered a one-million yuan (equivalent to $144,000 USD) starting bonus and had the chance to apply for research funds up to 3-5 million yuan (equivalent to $432,000-720,000 USD). The policies successfully attracted 1,510 overseas graduates by the end of 2011. Consequently, over 3.1 million, or 83.7% of Chinese overseas students who had completed their tertiary or post-graduate studies, returned to China by 2017 (Cao et al., 2020), which marked brain gain in China.
The policies for attracting returnees have given returnees privileges, such as enough money for resettling, more research funding, and priority promotion opportunities. However, these advantages have created a gap in career development between returnees and local scientists. They have also resulted in Chinese students and local scientists increasingly questioning the value and benefits of an international higher education degree and complaints about the privileges . Interviews have indicated that 21% of local scientists feel that returnees have excessive privileges in obtaining national academic resources, promotion, and fringe benefits such as housing. 1 For example, some PhDs and postdocs from first-class universities overseas were immediately promoted to full professorships, which upset their local counterparts because the returnees had yet to prove that they were capable of independent research (Zweig, 2006). Furthermore, the requirement of "overseas returnee only" in the recruitment notices of universities and research institutions upset local PhD students who could not apply irrespective of their academic achievements (Chen, 2022).
This study aims to determine if returnees still have an advantage over their local counterparts in terms of research publications. Hence, the research question we address in this study is: Is the research performance of returnees better than that of their local counterparts?

Literature review
The internationalization of science has allowed scientists to become internationally mobile during their careers, making international mobility a common phenomenon in academia. (Equeter & Hellemans, 2016;Netz et al., 2020). In Europe, 57% of university and 65% of institute respondents have at least one international mobility experience in their research careers (Boring et al., 2015). In the United States, one-third of the most frequently cited researchers were born abroad, similar to Western European countries (Ioannidis, 2004). Thus, scholars have shown a growing interest in the impact of mobility. International mobility is essential in promoting science (Morano-Foadi, 2005). It contributes to the research productivity of both destination and home countries (Jonkers & Cruz-Castro, 2013;Kahn & MacGarvie, 2016;Saxenian, 2002), promotes the transmission of knowledge and expands the international collaboration networks of countries, institutions and scientists (Barjak & Robinson, 2007;Baruffaldi & Landoni, 2012;Zhao et al., 2022), and boost the career success of scientists by providing cross-culture research creativity (Huisman, 2022).
Returnees may have privileges due to their social capital and innovations compared with their local counterparts (Bess, 1978;Enders, 2005;Gaule, 2014;Song & Gan, 2022;Zhang & Lucey, 2019). Therefore, they hold a distinct advantage in the domestic job market (Hao & Welch, 2012). Also, they are likely to be in a favourable position in international collaborations due to higher likelihood of global connectivity and thinking from a different perspective (Marini & Xu, 2023). For example, returnees in Turkey and Argentina (Jonkers & Cruz-Castro, 2013;Karakas, 2020) seem to have gained professional network ties, technical knowledge and skills, a broader worldview, and more resources (Jiang & Shen, 2019;Kahn & MacGarvie, 2016). However, other studies have shown that foreign degree holders are not always more productive than their colleagues with domestic degrees (Shin et al., 2014). The performance has been different in disciplines depending on the scientific environment of the home country. For example, in South Korea, returnees are less productive than domestic scientists in soft disciplines (Arts & Humanities, and Social Sciences).
The Chinese returnees have chances to win academic titles/programs which guarantee higher salaries, more funding, and better other academic resources (Cao, 2004), and these privileges lead to a positive effect on publications and promotion (Decramer et al., 2013;Zweig, 2006). Zhao et al. (2020) reported that Chinese returnees published more papers and more corresponding-author papers than their local counterparts. Cao et al. (2020) also showed that returnees tend to publish more and make outstanding contributions to the link between China and global networks. Shi et al. (2023) pointed out that the returning scientists attracted by China's young thousand talents (YTT) program are generally of high caliber in research. The assessments of more significant funding and larger research teams improve the research performance quality.
Returnees face challenges such as identity conflicts, different assessment policies, and a lack of domestic social networks (Ai, 2019;Lu & McInerney, 2016). On the one hand, returnees may find it difficult to (re)construct an academic identity in a familiar context that has become strange and lack of the opportunities to refresh their local knowledge because of their years of absence (Zhang et al., 2022). The increasing number of returnees and the "publish or perish" academic culture have also seemed to lead to a priority on quantity instead of the quality of the output, which weakens the returnees' advantages (Marini & Yang, 2021;Quan et al., 2017;Zhao et al., 2020). The policies issued in 2020 indicate that scholars should publish at least one-third of their research in Chinese journals but also continue engaging with the international research community (Liang et al., 2022), which puts more pressure on returnees. On the other hand, opportunities for postdoc research abroad have improved domestic PhDs' competence in publishing in top journals (Corley & Sabharwal, 2007;Horta, 2009;Melin, 2004) and narrowed the gap between returnees and their local counterparts. As a result, the challenges for returnees are even more evident in obtaining national academic titles, such as "Chang Jiang Scholars", a leading project for the construction of high-level talents in Chinese universities, because it typically takes returnees longer to earn academic titles than their local counterparts (Li & Tang, 2019).

Data
We collected data from the ORCID website and the web of science (WoS) database. The former is a unique 16-digit author identifier randomly assigned to each author, which helps disambiguate author names and contributes to the research of scholarly activities (Akers et al., 2016).
Using the profiles of scientists on the ORCID website, we first selected ~ 20,000 scientists who are native Chinese, earned PhD degrees, and had at least one employment experience. We presumed the nationalities of scientists based on the first degrees they obtained according to the records in ORCID (e.g., an undergraduate degree in China). We then identified returnees who met the following criteria: (a) Obtained a PhD degree from a country/territory other than mainland China. In addition, the overseas experience lasted at least 2 years. If a scientist had more than one PhD, we used the country for the PhD that took more years. (b) Returned to China and was employed by a university/research institute, excluding employment as a postdoc researcher.
We defined the local counterparts of a returnee as follows: (a) Obtained a PhD from mainland China (b) Employed by the same departments of the same institutions in the same year as the returnee.
As a result, our dataset comprised 637 returnees and 4,133 local counterparts to be compared after matching their experiences. Next, we searched for their publications in the WoS Core Collection. Our study followed the GIPP classification scheme system to identify scientists' disciplines. The global institutional profile project (GIPP) classification scheme is from the InCites database and has been widely used in literature. The GIPP discipline classification model aggregated all disciplines into six fields on the basis of the discipline classification of WoS, which includes Clinical, Pre-Clinical & Health, Engineering & Technology, Life Sciences, Physical Sciences, Social Sciences, and Arts & Humanities. Specifically, the discipline of each scientist was determined by the institutions with which they were affiliated. For example, a scientist was designated as belonging to economics if the affiliation was the Department of Economics, and belonging to Social Sciences according to the GIPP disciplines. We further grouped the six GIPP disciplines into two: Arts & Humanities and Social Sciences (AHSS), and Natural Sciences (NC, including the other four GIPP disciplines), which more most mentioned in China's research resources allocation, such as research grants and awards.

Methods
In this study, the research performance of a scientist was measured as follows: (a) the average number of papers the scientist published in a 5 year period since the year the scientist returned to China (b) the average number of corresponding-author papers the scientist published in a 5 year period since returning (c) the average number of first-tier journal papers the scientist published in a 5 year period.
The dependent, independent, and control variables are defined in Table 1. Fixed effect regression models are commonly used to reduce selection bias in the estimation. It controls the unobserved heterogeneity in regression (Mummolo & Peterson 2018). In our study, we utilized a fixed-effect model to control the institution-, departmentand time-effect, which makes the returnees and their local counterparts in our study comparable. The regression model takes the following form, where 0 ,..., 6 are unknown coefficients, and e 1 represents the error term. Table 2 shows the descriptive statistics of the variables in our dataset, consisting of 4,770 returnees and corresponding local scientists. As shown in Table 2, the statistics also demonstrate the dependent variables in the view of disciplines. Figure 1 displays

Regression results
In Table 3, Is_overseas is associated with P c , but has no significant correlation with P n and P ft . In other words, there is no significant difference between returnees and their local counterparts in terms of the number of publications and first-tier journal papers, but returnees published significantly more overall 0.126 than their local counterparts (p < 0.05). As (1)

Continuous
The annual number of papers in the continuous 5 year period before the year when a returnee and his/her local counterparts were employed by an institution in Mainland China

Continuous
The annual number of JCR first-tier journal papers in the continuous 5 year period before the year when a returnee and his/her local counterparts were employed by an institution in Mainland China

Dummy
The years between the year a scientist earned his/her PhD and the year he/she was employed by a Chinese institution

Start_Year
Dummy The year when a returnee and his/her local counterparts were employed by an institution in Mainland China

Dummy
The institution a scientist belongs to. The Chinese Government allocates more funding to elite universities, like Project 985 and Project 211 (Quan et al., 2017). Thus, the fixed effect of the institution is necessary when observing two groups of scientists

Dummy
The department/school a scientist belongs to. It matters what department/school scientists belong to as it represents the academic discipline. The development of different subjects is asymmetrical in China (Ai, 2019;Zhou & Leydesdorff, 2006), with strengths in Physics and Chemistry and weaknesses in Social Sciences. Scientists in the same discipline are comparable   Table 4. We found that the main effects have no statistical significance for both P n , and P ft , and the significant correlation with P c starts with the 5 year average publications. Interestingly, the corresponding-author advantage of returnees did not appear in the first 4 years after returning, but did appear in a longer-term period. We selected 2008 as the cutoff year and divided the samples into two groups. 2008 marks the first year China benefitted from brain gain due to the Thousand Talents Plan, and we observe that 80% of the returnees in our dataset returned to China after 2008. We found no significant difference between returnees and their local counterparts before 2008 in terms of the number of publications for returnees and corresponding-author publications and first-tier journal publications per year. However, after 2008, returnees published 0.125 (p < 0.1) more corresponding-author papers per year than their local counterparts, as shown in Table 5. We also grouped 4,770 scientists into AHSS and NS by discipline and found that scientists in AHSS have better research performances in publication numbers and first-tier papers, while scientists in NS publish more corresponding-author papers. Table 6 shows 1 3 that returnees in AHSS had significant advantages in the number of publications. In other words, returnees published 0.811 more papers and 0.506 more first-tier papers per year than their local counterparts, respectively. However, there was no significant difference between the two groups of scientists regarding the number of corresponding authors. The returnees in NS had significant advantages over corresponding-author publications, as shown in Table 7. The author's position in the publications reflects the role of a research team, and the corresponding author stands for the principal investigator (PI) and/or senior author in most cases (Shi et al., 2023;Tang, 2013;Zhao et al., 2020). In other words, returnees had more opportunities to be principal investigators (PI) than their local counterparts. However, returnees had no advantage in the total number of publications and first-tier publications.
Furthermore, we analyzed the changes in author positions in publications of returnees and their local counterparts before and after their return. As shown in Table 8, the position of both groups of scientists ascended along with their careers which is consistent with the increasing trend of the proportion of corresponding authors. For example, the returnees' author ordinal nearly increased from 3 rd to 5 th before and after returning. Comparatively, their local counterparts' author ordinal nearly increased from 2 nd to 3 rd , as shown in Table 8.

Discussion
This study found no significant difference in publishing papers between returnees and their local counterparts employed by the same departments of the same institutions in the same year. This finding challenges the argument that returnees have privileges compared to their local counterparts and are expected to be more productive. The advantages of returnees include English language skills, international connections and PhD training from developed countries/territories, all of which are significant for publishing in WoS-indexed journals, especially the first-tier journals.
Confounding factors influencing scientists' academic research performance are classified into two groups: individual-and institution-level factors. The former includes gender, age, academic training, degree, research collaboration, and discipline. The latter includes research policy, institution rank, research infrastructure and funding availability, access to scholarly resources, research rewards and culture. In our study, we utilized a fixed-effect model to address our research questions, including institution, department, and time fixed-effect. However, there are still confounding factors which were not controlled in this study, such as the gender of the sampled scientists, due to the data limitations.
There was a gap between returnees and their local counterparts decades ago regarding the academic resources they obtained. In the 1990s, there was a preference for returnees in talent recruitment. For example, Shanghai Jiao Tong University, one of the top nine universities receiving the ministry of education (MOE) funding, introduced a new hiring and promotion system that made the overseas experience a key criterion in recruitment and promotion (Zweig, 2006). This trend continued to grow in the following years. A survey of the top universities revealed that they recently hired faculty employees with 100% overseas backgrounds since 2018.
The gap in research performance between returnees and local hires has narrowed. To survive competition with returnees, local scientists and PhD students are financially supported by the Chinese government to visit foreign universities/institutions as visiting scholars/students for 6 months to 2 years to help them build international collaborations and networks. Many local PhD students also go overseas for postdoc positions to obtain overseas experience immediately after graduation. In addition, local scientists perform better in the number and quality of articles than before. In 2020, the State Council of the PRC issued a policy, "General Program for Deepening Educational Evaluation Reform in the New Era," to balance returnees and their local counterparts. This program states that universities should not make the overseas experience a requirement of employment or promotion. Thus, recruitment in two top universities, Tsinghua University and Peking University, have adopted a non-preferential stance toward applicants' academic credentials (Chen, 2022).
Returnees still have a clear publication advantage in AHSS, while local scientists perform better in NS as PIs. One explanation is that the difference is due to the unbalanced development of disciplines. In NS, China is witnessing a continued and significant increase in scientific production (Jin & Rousseau, 2005;Zhou & Leydesdorff, 2006) and has taken a world-leading position in several subfields of Chemistry and Physics (Malarvizhi et al., 2010;Zhou et al., 2009). However, the low priority of Social Sciences in the early period has caused slower development in China, which is yet to take off on the international stage (Ai, 2019). China's AHSS has experienced several stages, including the "preparing phase" , "introducing in" (1978the 1990s), "domestic practice" (1990-2000s), and "going global" (after 2000) stages during the development (Li & Li, 2015). Although China's AHSS is in line with international academic norms and becoming innovative, it is still not a major player in most fields.
Another explanation is that in AHSS, returnees are more familiar with international journals, and their previous works are internationally oriented. At the same time, local counterparts publish more papers in national and local journals that are not covered in the WoS collection (Wu et al., 2015). Moreover, there is an argument that the author's number of papers in AHSS possibly affects the performance of corresponding authors. In our dataset, the number of single-authored papers accounts for 3.39% of the total in AHSS, much higher than 0.82% in NS. However, there is no evidence that this explains the no-significance in the regression. The two explanations show that china continues to face problems with limited regional authorship, lack of individuality, and low international visibility and citations of publications in AHSS (Liu et al., 2015;Wang, 2008;Zhou et al., 2009).
There are a few limitations to this study. First, the findings in this study are generated from ORCID, which is not a random sample of the world's scientists (Bohannon & Doran, 2017). There are 15.7 million CVs on the ORCID website by 2022, and the number shrinks to 3.91 million after eliminating duplications webpages with no education and qualification because one registrant may register multi-ORCIDs (ORCID, 2022). China has a relatively low registration rate for ORCID, and most of the Chinese registrants are young, outstanding in English publications or active in international academia (Zhao et al., 2020). Second, the distribution of disciplines based on the ORCID dataset is uneven, with a large percentage of scientists in NS and a low percentage in AHSS.

Conclusion
In this study, we first explored the differences in the research performance between Chinese returnees and their local counterparts by strictly limiting the comparisons to those employed by the same departments of the same institutions in the same year. Based on a sample of 637 returnees who returned to China from overseas in 1984-2017 and their 4,133 local counterparts, we found no significant differences between returnees and their local counterparts in terms of productivity, although returnees had more opportunities to be PIs in research teams. However, returnees still have an advantage in AHSS.
This study has implications for employment and promotion policies regarding the balance between returnees and local scientists. There are almost equal opportunities to obtain overseas experience for local scientists in NS. In addition, employment and promotion policies are now not supposed to treat overseas and local doctoral degrees differently. However, institutions may treat scientists' research potential differently. Last year, the Natural Science Foundation of China (NSFC) set up a new program, "Overseas Distinguished Young Scholars of NSFC", targeting young scientists who studied overseas. Research funding, salary, and fringe benefits provided to the finalists are very competitive. However, current policies state that they should provide equal opportunities to local scientists with the same research potential. Nevertheless, in AHSS, there is still a gap between returnees and local scientists. Far fewer overseas postdoc positions are available for Chinese PhD students or scholars in AHSS compared to researchers in the NS. Policies should encourage local PhD students and scholars to obtain equivalent overseas research training, such as postdoc research.