Does the Number of Publications Matter for Academic Promotion in Higher Education? Evidence from Lebanon

Abstract

This paper uses unique data from Lebanon to explore the factors affecting promotion in academia. We use the Ordered Probit model to answer the research question empirically. The results indicate that the number of publications is an essential factor affecting promotion decisions. Moreover, our findings indicate a quadratic relationship between age and promotion, reaching a peak at 62 years. After this turning point, age and promotion become negatively correlated. When dividing our sample by academic generations, we find that the number of publications is an important determinant for promotion only for the cohort who graduated after the year 2000. Finally, after dividing by gender, the results suggest that males and females who publish more have equal chances of earning a promotion.

1. Introduction

Academic advancement is still one of the most evident signs of an academic’s standing. The hiring, training, and advancement of scholars into leadership positions foster a favorable academic atmosphere and a positive departmental image. Academic promotion is a process of rank progression in which a university acknowledges academics for their achievements, translated into increased salaries and greater responsibilities and influence (Hardré and Cox 2009). However, in some countries such as Lebanon, departments in different universities sets their standards for academic advancement.1 The standards established by one higher institute may differ from those defined by others. According to Demirdover (2020), the following two categories may be used to summarize these criteria.

The first one is research activity and academic publication. Rospigliosi and Bourner (2019) describe it as a very recent phenomenon. According to Google’s Ngram viewer, the term “research development” began to take place during the beginning of the twenty-first century (Google Books Ngram 2019). Perhaps the easiest way for the committee that checks for promotion standards is to look into the total number of publications. However, these requirements might be fulfilled by predatory publishing2. Therefore, academic councils must look into other indices, such as the number of citations and the journal’s quality.

In many universities, having teaching experience is a must for academic promotion. In turn, the teaching activity would be the second most sought after criterion by academic councils. However, a study by Graham (2015) argues that around 75 percent of the surveyed lecturers and professors considered teaching a “not very important” criterion, while 70 percent of the department heads and academic councils find it essential. As mentioned earlier, enormous attention has been attributed to research outcomes and productivity for the past 20 years. Therefore, fresh graduates now see teaching contribution as a poor indicator of achievement (Schimanski and Alperin 2018).

However, Lebanon has suffered from the issue of political instability (Al-Hajj et al. 2021). According to Jaspal et al. (2020), the Lebanese economic crisis depends on many factors that are mainly wars and civil disorganization, including the long civil war that was between 1975 and 1990, in addition to various Israeli strikes and the regional political and economic instability, such as the Syrian war that has forced around 1.5 million people to migrate as refugees, and finally the new revolution of 17 October 2019. Similarly, Hazbun (2016) considers that these combined circumstances have led to persistent political insecurity and weakening political and institutional assurance. Sadly, since 2016, Lebanon has been experiencing its most significant economic crisis and political upheaval in decades. Salem (2020) asserts that the country’s researchers have probably been the ones most impacted by the country’s current economic and political unrest. Numerous domestic researchers have reduced or completely stopped their work due to Lebanon’s ongoing challenges, particularly in light of the severe lack of US dollars (Salem 2020). El Achi et al. (2020) also noted that the biggest issue facing researchers in Lebanon was the absence of sustainable and sufficient funding for research. Particularly during disasters, securing appropriate funding for research is substantially more challenging and, occasionally, nearly impossible.

In the past, most studies tackling this kind of research used information from more extensive bibliometric databases, such as the Web of Science. The Web of Science does not index a sufficient number of pertinent articles or books, which is a problem in several subjects, including the social sciences and the arts and humanities. Thus, this could give a false sense of research activity (Hicks 2004). However, the social sciences, the humanities, and the arts can be thoroughly covered in our paper. Additionally, most previous research has looked at the volume of publications produced by scientists over a specific period. As a result, we add to the literature by identifying the factors influencing promotion decisions for our respondents throughout their careers. Furthermore, in developing nations, this kind of study is still scarce. Hence, this research uses unique data from the Lebanese American University (LAU) to study the determinants of academic promotion. LAU is a leading private and nonsectarian higher education provider. It has campuses in Beirut and Byblos, Lebanon. It is accredited by the Commission on Institutions of Higher Education of the New England Association of Schools and Colleges and is run under a license from the Board of Regents of the University of the State of New York. LAU is dedicated to academic achievement and gives its students access to all of the tools they need for success in the future. The university’s objective is to put students first, encourage civic involvement and variety in society, and support a variety of values it hopes to instill in its students, such as openness, integrity, and quality.

The rest of the paper is structured as follows. Section 2 offers a survey of the related literature, while Section 3 outlines the data and research methodology utilized. Section 4 describes the empirical findings and insights. Finally, Section 5 provides concluding remarks and a set of policy recommendations.

2. Literature Review

2.1. Number of Publications

Many studies indicate that the number of publications is important for academic promotion. Faria et al. (2013) examined the impact of different factors affecting faculty academic promotion and tenure across 175 economic departments in the United States (U.S.). They argued that the number of academic publications was associated with a higher chance of achieving academic promotion across all institutional settings. Zinovyeva and Bagues (2010) reached the same conclusion; members who publish more have a higher tendency to earn a promotion. They also suggested that academic promotion practices differ between academic generations, particularly those with a PhD earned before 2000 and those earned after 1999. Furthermore, in their study, Ginther and Hayes (2003) found that the number of publications had no significant effect on promotion decisions for those who obtained their PhD between 1975 and 1979. Moreover, J. S. Long et al. (1993) sought to determine whether the quantity of publications or the quality, as indicated by the number of citations or the journal rank in this study, drove promotion among American biochemists. Based on the results of their study, they were able to support their argument. Their findings showed that an increase in the number of articles published while an assistant professor increased the likelihood of obtaining a promotion to associate professor by 49 percent, while an increase in the number of citations had no effect. In addition, Tien (2007) conducted a study to determine whether Taiwan’s promotion system adequately rewards faculty research productivity. The results revealed that publication number is an essential predictor of promotion prospects. Another finding of this study was that the publication type and date affected promotion. For example, publishing books had a greater influence on one’s possibility of promotion than publishing articles, as reported in the study. In terms of timing of publications, the findings indicated that current-year publications were more important for determining promotion chances than prior publications. We thus propose our hypothesis stating that publishing more increases the probability of academic promotion.

2.2. Socioeconomic Characteristics

In spite of rapid advances in female education attainment in recent decades, women continue to experience gender inequality in the workplace. There is still a substantial wage and seniority gap between men and women in almost all fields (Bosquet et al. 2019). The academic field is one of those fields subject to this discrimination. Academic wage gaps are more widely studied than differences across genders in academic promotion, but a good number of papers still examine gender’s influence on academic promotion. Many of these studies showed that a male faculty was more likely to be promoted compared to their female colleagues. For example, Faria et al. (2013) found that the probability of male faculty becoming associate professors was 13 percentage points higher than that of female faculty. Furthermore, Marini and Meschitti (2018) aimed in their paper to investigate whether gender affected the probability of promotion to full professor in Italian universities. They reached the conclusion that male faculty members had a higher likelihood of becoming full professors than their female colleagues by around 24 percent. In addition, Tien (2007) asserted that females had lower chances of earning academic promotions. Many authors (J. S. Long et al. 1993; Ornstein et al. 2007) found that, generally, men were promoted to associate and full professor faster than women. J. S. Long et al. (1993) also claimed that promotions to the rank of associate professor were expected to be based on higher standards for women than for men. In addition, Vithal et al. (2013) argued that male faculty members were more likely to apply for higher ranks, such as associate professors and professors. Even though more men applied to become associate professors and professors, a higher percentage of women actually achieved these positions. The findings also indicated that women were more likely than men to apply for promotions at the lower rank levels of lecturer and senior lecturer, but both men and women had equal chances of being promoted at these levels. Moreover, Kahn (2012), who conducted a study at the University of New South Wales (UNSW), claimed that female lecturers were less likely to be promoted than males. However, Kahn (2012) reached a surprising result. Women associate professors were more likely to be promoted to professor rank than men. The intuition behind this finding is that only a limited number of highly skilled female academics were promoted to associate professors (Kahn 2012). Yet, a study by Smart (2009) demonstrated no statistically significant difference between men’s and women’s chances of obtaining a promotion.

The faculty member’s age and nationality may also affect academic promotion. According to Tien (2007), the probability of promotion was significantly influenced by one’s age. This study shows that the chance of being promoted increased by 3 percent with every additional year of age. In addition, this result was significant among all the models. Additionally, Smart (2009) included in his paper the variable age and the square of age to capture the non-linear relationship between age and the probability of promotion. Specifically, the probability of promotion may increase as faculty members grow up until they reach a threshold age where the probability of promotion might be lower. He suggested that there is a significant relationship between age and being promoted. Particularly, in three different schools, the faculty member had a higher probability of being promoted till a certain age, while this probability decreased after reaching a specific age. Another study completed among French and Italian physicists demonstrated that both French and Italian physicists had higher chances of being promoted as they increased in age until they turned sixty (Lissoni et al. 2011). Furthermore, there were also differences in gender among age groups. Accordingly, males were more likely to obtain college positions at a younger age than females (Zinovyeva and Bagues 2010). They further argued that male members were more likely to obtain associate professor positions at older ages than females (Zinovyeva and Bagues 2010). Faria et al. (2013) included this variable in their study regarding the nationality of a faculty member. They found that faculty of American nationality were more likely to obtain a promotion by 15 percentage points compared to non-American faculty members. Finally, Caucasian faculty have a higher probability of promotion by 11.5 percentage points compared to non-Caucasian faculty members.

2.3. University Characteristics

Another factor that might affect the academic promotion of a faculty member is experience. Many studies have agreed that the more experienced a staff member is, the more he/she is likely to be promoted. According to a study by Ginther and Hayes (2003), being more experienced positively and significantly affected promotion odds among humanities faculty. Additionally, the authors divided the sample into two cohorts, those who earned their PhD between 1975 and 1979 and those who obtained it between 1980 and 1989. The results demonstrated that experience affected promotion positively in the most recent cohorts, whereas it had a negative and insignificant effect on the previous cohorts. Moving to Perna (2001), her study’s results showed a positive correlation between experience and promotion to full professor among tenured faculty members. Furthermore, Smart (2009), in his paper, claimed that inexperienced staff had fewer opportunities for promotion.

The size of the department might also influence academic promotion. Faria et al. (2013) used the Probit model and found that faculty size did not affect the academic promotion of staff members in the economics’ departments. Even after dividing the sample size into two different academic generations, the first including individuals who received their doctoral degree before 1999 and the latter including those who were awarded their doctoral degree after 1999, the results showed that the size of the economics department still did not have any significant relationship with the likelihood of obtaining promotion among the two generations.

Individuals’ academic promotion may be affected by how many citations their publications receive and what journal rank it is published in. Starting with the number of citations, Hargens and Farr (1973) showed in a study completed with scientists such as mathematicians, physicists, and chemists that there was a positive correlation between rapid promotion and the number of citations received by a scientist. However, J. S. Long et al. (1993) found that the number of citations of articles that belong to an author before starting his/her assistant professor position had no significant effect on the promotion’s likelihood. Another study by Abbott et al. (2010) stated that the respondents believed that the number of citations had a significant role in determining their promotion; however, the study showed that the administrators did not rely directly on quantitative metrics such as the number of citations while evaluating or promoting. On the other side of the argument, Bagues and Zinovyeva (2012) found that academic promotion depended on the network, which could be enhanced as the number of citations increased. Moving to the journal rank, Faria et al. (2013) included this factor while examining the determinants of academic promotion. They found that publishing in top journals had a significant positive impact on academic promotion at public universities.

3. Research Methodology

3.1. Setting the Context

We chose LAU as our study’s focus for several reasons. The main reason behind choosing LAU as our case study is that this university does not offer any PhD programs. Hence, the faculty members are heavily responsible for the university’s research output. Moreover, since LAU is a medium-sized academic institution, faculty specialties rarely overlap based on their PhD theses. Furthermore, LAU is a leading non-sectarian higher education organization built on strong values and rooted in an authentic sense of mission. In 1924, LAU was founded in the Ottoman Empire as the first college for women.3 With Lebanon’s rising educational demands, LAU continuously grew throughout the years. The university is now divided into two campuses in Lebanon, one in Beirut and one in Byblos. Currently, LAU offers 60 different significant fields distributed over 65 distinct programs through 7 schools: the Adnan Kassar School of Business, the Alice Ramez Chagoury School of Nursing, the School of Architecture and Design, the School of Arts and Sciences, the School of Engineering, the Gilbert and Rose-Marie Chagoury School of Medicine, and the School of Pharmacy.4 In the fall of 2021, LAU had 8012 students, 750 staff, and 1140 faculty members.5

Over the past few years, the university’s research outcome has improved significantly. According to the QS World University Rankings 2023, LAU is ranked #601-650.6 In fact, the Times Higher Education Subject Rankings place the School of Business as the top business school in the country.7 In addition, the tables in Appendix A show that none of the schools at LAU are publishing in Q4 journals,8 and very few are publishing in Q3 journals, namely the School of Pharmacy with two papers in their record, and the School of Arts and Sciences with only one paper published in a Q3 journal. These results confirm that LAU is dedicated to maintain a high reputation which further asserts its global ranking. Another observation that could be made by looking at these tables is that the School of Medicine has the highest number of papers published in Q1 journals (22 papers), followed by the School of Arts and Sciences with 20 papers in their record. The School of Business lies in third place with 19 papers published in Q1 journals. The Schools of Engineering, Pharmacy, and Nursing come after that with 17, 16, and 15 papers in their record, respectively.

When it comes to Q2 journals, the School of Nursing takes the lead with nine papers, followed by the Schools of Engineering and Pharmacy with seven papers, and the School of Business with six papers. The schools with the least papers published in Q2 journals are the School of Medicine and the School of Arts and Sciences, with only three papers in their record.

3.2. Data and Variables

Our study investigated the determinants of academic promotion in Lebanon. The database we used contained all of the varieties of scholarly and scientific works in all of the areas of study from this top university in Lebanon. As a result, as of fall 2021, there were 1996 articles published overall. The documentation system, a routine practice utilized by academic institutions, was applied to gather the data. As a result, accurate, trustworthy, and well-structured data were produced (Sivertsen 2010). Additionally, we ensured that all of the publications were approved in accordance with national standards and regulations. Furthermore, LAU constantly logs all of the publication outcomes. As a result, missing papers did not affect our study. After that, we integrated this data with another database that records socioeconomic information on professors, department size, faculty level, and the overall number of publications that faculty members have produced as of fall 2021. The data were collected upon our request from the Department of Institutional Research and Assessment (DIRA) at LAU. We were left with 242 observations after data cleaning.

Table 1. Summary statistics.

	Mean	Std. Dev.	Min.	Max.
Number of Publications	23.207	35.442	1	370
Age	47.538	9.768	29.460	70.180
Gender	0.616	0.487	0	1
Nationality	0.574	0.495	0	1
Years of Experience	10.724	8.039	0.170	35.230
Tenured	0.694	0.462	0	1
Department Size	33.636	32.500	5	90
Number of Citations	429.062	994.342	0	11867
Top University	0.492	0.501	0	1
Q1	10.612	16.771	0	185
Q2	5.583	10.175	0	111
Q3	1.975	3.694	0	40
Q4	0.574	1.711	0	23
Number of Observations	242

presents the summary statistics of the variables.

To investigate the determinants of academic promotion in Lebanon, we resorted to a categorical variable that captured individuals’ academic advancement. Hence, our dependent variable took the value of one if the respondent was an assistant professor, two if he was an associate professor, and three if he was a full professor. As for the explanatory variables, our study included one key independent variable, a continuous variable that captured the total number of publications. The other independent variables were divided into the following categories: socioeconomic and university characteristics. First, the socioeconomic factors included age and age squared9 (continuous variables), gender (1 for male and 0 for female), and nationality (1 if the respondent is Lebanese and has single citizenship and 0 if not).

Second, there were seven university characteristics. They were defined as follows: experience (continuous variable), which represented the years of professional experience of the faculty members; tenured (1 if the respondent was tenured and 0 if not); faculty size (continuous variable), which was defined as the total number of faculty members in a department; the total number of citations (continuous variable), a variable that captured the total number of citations to date; top university (1 if the faculty had graduated from a top university10 and 0 otherwise); journal quality (four dummy variables), that captured the rank of the journals the research work was published in. They corresponded to the quartiles of the journals in Q1 (reference group), Q2, Q3, and Q4. Finally, we included a variable called PhD graduation year, a dummy variable used to divide our sample into two: faculty who earned their PhD degree before the year 2000. In that case, the variable took the value of 1, and it took the value of 0 if the respondents received their Ph.D. degree after 2000.

3.3. Collinearity Test

When suspecting the presence of correlation among the explanatory variables, testing for it is crucial since having a collinearity problem leads to an increase in the variance of the regression coefficients, making our statistical significance inaccurate. We used one of the most common diagnostic tests, the Variance Inflator Factor (VIF), where a value of 10 or more for the VIF is considered problematic (Hair et al. 1995).

Table 2. Collinearity Diagnostics.

Variable	VIF	Tolerance	R-Squared
Number of Publications	8.01	0.1248	0.8752
Age	2.93	0.3407	0.6593
Gender	1.08	0.9239	0.0761
Nationality	1.14	0.8797	0.1203
Experience	2.45	0.4079	0.5921
Tenured	2.64	0.3786	0.6214
Faculty Size	2.93	0.3414	0.6586
Citations	7.74	0.1293	0.8707
Top University	1.13	0.8857	0.1143
Q2	6.41	0.1561	0.8439
Q3	5.14	0.1945	0.8055
Q4	3.39	0.2946	0.7054
Graduation Year	2.62	0.3823	0.6177
Mean VIF	3.66

shows a mean value of 3.66 for the VIF, indicating the absence of a multicollinearity problem between our independent variables.

3.4. Modeling Approach

To identify the factors affecting promotion decisions, we used an Ordered Probit model as our dependent variable was ordinal and categorical. Our model could be represented as follows:

$$P{r}_j^{*}=\theta +\alpha X_j+\beta Y_j+\delta Z_j+{\epsilon }_j$$

Let $P{r}_j^{*}$ denote the promotion granted to a faculty member j (j = 1, 2, …, J). $X_j$ refers to our main independent variable, the number of publications. $Y_j$ and $Z_j$ represents the socioeconomic characteristics variables and the university characteristics variables, respectively. α, β, and $\delta$ are the coefficients to be estimated, $\theta$ is the constant term, and ${\epsilon }_j$ is the random error term. However, the variable $P{r}_j^{*}$ was not observed in the dataset as it is a latent variable. As an alternative, we can observe $P{r}_j$ by using the Ordered Probit rule as follows:

$$P{r}_i=\{\begin{array}{l}1 \left(\mathrm{Assistant} \mathrm{Professor}\right) \mathrm{if} P{r}_i^{*}\le {\mu }_1\\ 2 \left(\mathrm{Associate}\mathrm{Professor}\right) \mathrm{if} {\mu }_1<P{r}_i^{*}\le {\mu }_2\\ 3 \left(\mathrm{Full} \mathrm{Professor}\right) \mathrm{if} P{r}_i^{*}>{\mu }_2 \end{array}$$

where $P{r}_i$ is the academic position granted to a faculty member and ${\mu }_j$ is the introduced threshold of the unobserved concern level.

4. Results

4.1. Benchmark Results

Table 3. Results of the Ordered Probit model, benchmark results.

	(1)	(2)	(3)	(4)
	Promotion	Promotion	Promotion	Promotion
Number of Publications	0.019 ***	0.020 ***	0.022 ***	0.022 ***
	(0.003)	(0.003)	(0.006)	(0.009)
Age		0.370 ***	0.251 **	0.243 **
		(0.093)	(0.100)	(0.101)
Age Squared		−0.003 ***	−0.002 **	−0.002 *
		(0.001)	(0.001)	(0.001)
Gender (=1 if male)		0.036	0.060	0.029
		(0.166)	(0.179)	(0.181)
Nationality (=1 if Lebanese)		−0.135	−0.212	−0.253
		(0.161)	(0.175)	(0.178)
Experience			0.074 ***	0.066 ***
			(0.016)	(0.019)
Tenured (=1 if tenured)			0.981 ***	0.912 ***
			(0.328)	(0.333)
Faculty Size			0.005	0.005
			(0.005)	(0.005)
Number of Citations			0.002	0.001
			(0.001)	(0.001)
Top University			−0.140	−0.108
			(0.184)	(0.187)
Q2				0.018
				(0.026)
Q3				−0.122 **
				(0.051)
Q4				−0.024
				(0.121)
Experience X Number of Publications				−0.025
				(0.001)
Number of Observations	242	242	242	242
Pseudo R2	0.096	0.236	0.368	0.381
Log Likelihood	−219.510	−185.685	−153.497	−150.406

Robust standard errors are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.10.

shows the coefficients of the independent variables on the probability of a faculty member being promoted. Column 1 only includes our key independent variable, the number of publications. Column 2 adds the socioeconomic factors of the respondents to column 1. Regarding column 3, we added the university characteristics variables, excluding the journal rank. Finally, the last column adds to the previous model of the journal rank dummies.

The findings suggested that as the number of publications of a faculty member increased, the probability of obtaining a promotion became higher. This result was significant among all of the models. Hence, our hypothesis, stating that publishing more increases the likelihood of being promoted, was supported. This finding was suspected and corresponds with the results of Zinovyeva and Bagues (2010) and Faria et al. (2013), which suggested that the number of academic publications was associated with a higher chance of earning a promotion. It was argued that the number of publications determined the faculty member’s productivity and success, affecting the promotion decisions. Regarding age, the coefficient of the age variable was positive, while the coefficient of the age squared variable was negative. All the coefficients were statistically significant. This implied a quadratic relationship between promotion and the faculty member’s age that reached a peak at approximately the age of 62.11 After this age, the promotion rate started to increasingly decrease. Smart (2009) and Lissoni et al. (2011) reached the same conclusion of a non-linear relationship between age and obtaining a promotion. Regarding the turning point, our results were close to those of Lissoni et al. (2011), who found that academics were more likely to obtain promotions as they became older until they reached the age of 60. That is, the young faculty members were more active and energetic, which may attract promotion faster than older faculty members. Moreover, the older faculty members have reached the top of the hierarchy and cannot be promoted further as they were already full professors. When it comes to gender, we found that there was no significant relationship between being male and academic promotion. This finding is consistent with Smart (2009), who demonstrated no statistically significant relationship between gender and obtaining a promotion. However, our result contradicts (J. S. Long et al. 1993; Ornstein et al. 2007; Tien 2007; Faria et al. 2013), who suggested that males have higher chances of earning academic promotion. Concerning nationality, being Lebanese was not significantly associated with being promoted. Nevertheless, Faria et al. (2013) reached a different conclusion when they found that American and Caucasian faculty members had a higher probability of obtaining a promotion in the U.S. Our results indicated no discrimination regarding gender and nationality of the faculty member at LAU.

Concerning experience, our results revealed that having more years of experience in the academic field significantly increased the likelihood of earning a promotion. This result was consistent with what was found by Perna (2001) and Ginther and Hayes (2003) who argued that there was a positive correlation between experience and promotion. That is, the decision to promote a faculty member highly relied on mastering the skills gained from experience. Regarding tenure, tenured people had a higher likelihood of being promoted. The intuition behind this is that being tenured is a requirement at LAU to become a full professor.12 When it comes to faculty size, we did not find any significant relationship between this variable and promotion. This corresponds with the findings of Faria et al. (2013). In addition, we found that the number of citations did not significantly impact on promotion odds. This result was similar to that of J. S. Long et al. (1993) who suggested that the number of citations that belonged to the author’s articles before obtaining the assistant professor position had an insignificant impact on the probability of obtaining a promotion. Additionally, an insignificant relationship existed between graduating from a top university and promotion. Moving to our journal rank dummies, our finding suggested that those who published in low rank journals were less likely to be promoted compared to those who published in Q1 journals. Faria et al. (2013) reached a similar conclusion where their findings showed that publishing in a top journal positively and significantly affected academic promotion at public universities.

4.2. Results by Academic Generation

Table 4. Results of the Ordered Probit model, results by faculty cohort.

	(1)	(2)
	PhD Received before 2000	PhD Received after 2000
Number of Publications	−0.001	0.053 ***
	(0.018)	(0.015)
Age	1.903 ***	0.531 ***
	(0.636)	(0.169)
Age Squared	−0.016 ***	−0.005 ***
	(0.005)	(0.002)
Gender (=1 if male)	0.270	−0.170
	(0.399)	(0.248)
Nationality (=1 if Lebanese)	−0.530	−0.144
	(0.334)	(0.254)
Experience	0.065 *	0.133 ***
	(0.034)	(0.027)
Tenured (=1 if tenured)	0.071	1.145 **
	(0.662)	(0.449)
Faculty Size	0.001	0.001
	(0.010)	(0.007)
Number of Citations	0.001 **	−0.001 *
	(0.001)	(0.001)
Top University	0.228	−0.118
	(0.374)	(0.259)
Q2	−0.001	0.055
	(0.035)	(0.040)
Q3	0.092	−0.285 ***
	(0.091)	(0.089)
Q4	−0.122	0.020
	(0.171)	(0.213)
Number of Observations	71	171
Pseudo R2	0.339	0.529
Log Likelihood	−51.172	−71.422

Robust standard errors are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.10.

presents the coefficients when dividing our sample by academic generation: those who received their PhDs before 2000 and those who earned them after 2000. Starting with our main explanatory variable, our results suggest that the number of publications of a faculty member has an insignificant impact among those who obtained their PhDs before 2000. This result corresponds with what was found by Ginther and Hayes (2003), who argued that publishing more did not significantly affect promotion decisions among the old cohort. However, for those who graduated after 2000, our results were similar to the benchmark results. The number of publications of this cohort had a positive and significant effect on obtaining a promotion. This could be explained by the fact that the old faculty members concentrated more on teaching productivity while the recent faculty members focused on research productivity (Graham 2015). This supported the suggestion of Zinovyeva and Bagues (2010) that there was a difference between the academic generations regarding promotion practices.

We obtained the same findings as the benchmark results for both of the generations regarding age and age squared. The age and age squared coefficients were positive and negative; they were all statistically significant. Moreover, we noticed that both of the variables, gender and nationality, did not significantly impact on academic promotion for both of the cohorts. Consistent with our benchmark results, those who are more experienced, whether from old or recent cohorts, were more likely to earn a promotion. In addition, being tenured only had a significant positive impact on promotion among those who graduated after 2000. That is, more challenges and new requirements are presented today in order to obtain a promotion, including tenure. According to Gardner and Veliz (2014), most universities only wanted to see excellence in at least one out of teaching, service, or research back in the 80s. Nowadays, excellence is expected in all three of them, while the focus is on research (Youn and Price 2009). Concerning the faculty size, our results suggested that there was no significant relationship between faculty size and obtaining promotions among both of the academic generations. This corresponds with the findings of Faria et al. (2013), who claimed that even after dividing the sample into two cohorts, faculty size did not significantly affect the promotion chances of a faculty member. Furthermore, after dividing our sample, our results regarding the number of citations became significant. It was found that the number of citations was positively related to promotion among the old cohort, while they were negatively related to promotion among the recent one. When it comes to a top university, the finding was similar to the benchmark result where there was no significant correlation between promotion and the rank of the university that the faculty member graduated from. Additionally, publishing in low rank journals remained significant only among the recent academic generation. In addition, publishing in low rank journals negatively affected promotion chances compared to those who published in Q1 journals.

4.3. Results by Gender

Table 5. Results of the Ordered Probit model, results by gender.

	(1)	(2)
	Male	Female
Number of Publications	0.026 ***	0.073 **
	(0.010)	(0.034)
Age	0.202	0.639 ***
	(0.124)	(0.219)
Age Squared	−0.001	−0.006 ***
	(0.001)	(0.002)
Nationality (=1 if Lebanese)	−0.201	−0.608 *
	(0.230)	(0.318)
Experience	0.078 ***	0.064 **
	(0.022)	(0.029)
Tenured (=1 if tenured)	0.962 **	1.296 **
	(0.439)	(0.619)
Faculty Size	0.003	0.014
	(0.006)	(0.009)
Number of Citations	0.001	−0.002 *
	(0.001)	(0.001)
Top University	−0.190	−0.016
	(0.244)	(0.321)
Q2	0.011	−0.002
	(0.026)	(0.071)
Q3	−0.063	−0.219 **
	(0.062)	(0.103)
Q4	−0.113	0.364
	(0.125)	(0.283)
Number of Observations	149	93
Pseudo R2	0.381	0.455
Log Likelihood	−95.007	−47.838

Robust standard errors are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.10.

shows the marginal effect by gender. We found that an increase in the number of publications increased the likelihood for both males and females to obtain a promotion. This is consistent with our benchmark result. For female faculty members, the coefficients for both variables, age, and age squared were significant with positive and negative signs, respectively. However, we lost significance for male faculty members. Moreover, we noticed that the effect of being Lebanese became significant for females after dividing our sample by gender. It was found that the female Lebanese faculty members were less likely to be promoted. Accordingly, it seemed that there is sex discrimination about nationality. Lebanese female faculty members were more likely to face promotion challenges than non-Lebanese female faculty members. Only a few papers studied the impact of nationality on academic promotion. Medgyes (1992), for instance, argued that in the United States, non-native professors were less likely to obtain a promotion since their English was not as good as the native ones. Floris and Renandya (2020) concluded that discriminatory practices towards non-native speaking teachers still existed to date. Consistent with our previous results, male and female faculty members with more experience had higher chances of promotion. Moving to tenure, our findings suggested a positive and significant relationship between promotion and male and female tenured faculty members. In addition, faculty size was still insignificant after dividing by gender. Furthermore, it was found that the number of citations for female faculty members became significant and negative. Concerning the top university variable, the finding revealed that the rank of the university from which the individual graduated, whether male or female, had an insignificant impact on academic promotion. Lastly, female faculty members who published in low-rank journals had a lower probability of obtaining a promotion than those who published in Q1 journals. However, this result for male individuals was insignificant.

5. Conclusions

In this study, we examined the factors affecting academic promotion in Lebanon. In order to conduct this empirical study, we collected data from LAU, one of the country’s leading universities in Lebanon. The findings show that the number of publications significantly determines the likelihood of obtaining a promotion. In other words, the number of publications indicates one’s hard work and productivity. After dividing by gender, it is shown that the number of publications of a faculty member, whether male or female, increases the probability of earning academic promotion. Nevertheless, when dividing our sample into two academic generations, old and recent cohorts, the finding remains only significant among those who graduated after 2000. That is because, by the 2000s, excellence in teaching was no longer enough to obtain a promotion as research had taken center stage (Graham 2015). The preference for research over teaching is widespread among academics, and they tend to perceive that promotion comes more from research than from teaching.13

Moreover, we found a non-linear quadratic relationship between a faculty member’s age and academic promotion, where it reaches a turning point at the age of 62. Additionally, the findings reveal that gender and nationality are not significantly related to promotion. This implies no discrimination among faculty members in terms of gender and nationality at LAU. In addition, more years of experience are associated with a higher probability of earning an academic promotion. That is due to the skills mastered during these years, which play a significant role in the promotion decision. Finally, the findings reveal that faculty members who publish in low rank journals are less likely to obtain academic promotion compared to those who publish in Q1 journals.

This paper offers evidence that might be useful for administrators in the university who are responsible for the promotion decision as it helps them to make better decisions. For instance, the positive effect of the faculty member’s number of publications suggests that this factor should be a priority for administrators to look at when making decisions. Furthermore, academic councils should not overlook the faculty members’ years of experience or whether he/she is tenured or not, since they are important determinants for academic promotion. On the other side, faculty members who seek academic promotion also benefit from this empirical study. For instance, faculty members should be eager to apply for promotion before reaching 62, as age and promotion becomes negatively correlated after this age. Lastly, tenured faculty members should take advantage of that and apply for a promotion. In short, universities should make the promotion process more transparent so faculty members have a clearer vision of what they should work on to achieve promotion.

Finally, more independent variables could be included in the study. For instance, one could distinguish between publications (i.e., articles, books, and reports), which could be regarded as a limitation. Hence, as more data become available, researchers could include more independent variables. In terms of possible directions for future research, this study could be further extended by collecting data from other universities in Lebanon or the region to increase the sample size, and hence study more in depth the gender discrimination present in higher education in the region.