Fear of violence and brain drain analysis among healthcare workers in Turkey

Abstract

Background

This study investigates the relationship between healthcare professionals' intention to emigrate and their exposure to violence in Turkey, using a quantile regression model. Through this approach, it aims to reveal how healthcare professionals' attitudes toward brain drain vary across different levels of fear of violence, considering factors such as professional experience and income.

Methods

A cross-sectional study design was employed, utilizing a quantile regression model to analyze the variation in brain drain attitudes across different percentiles. The model examines how fear of violence affects these attitudes at various levels.

Results

The analysis reveals that the intention to engage in brain drain increases with professional experience. Interestingly, a high fear of violence is associated with a reduced tendency to emigrate. Furthermore, the findings indicate that as income increases, attitude scores toward brain drain decrease, suggesting a complex interplay of factors in this phenomenon.

Conclusions

The study’s findings have significant implications for policymakers. By understanding the role of factors such as fear of violence, income level, and professional experience in healthcare professionals’ decisions to stay or leave, policymakers can develop targeted strategies to prevent or manage brain drain. Future research could further investigate these variables, providing valuable insights for policy development.

Key Messages

The increase in income level negatively affects the attitudes of health professionals towards brain drain; This shows that brain drain has an economic dynamic.

This study, which examines the relationship between brain drain attitude and fear of violence among health professionals, provides detailed findings with a quantile regression model.

Research results reveal that brain drain intention is positively affected as the duration of professional experience increases. It is observed that as healthcare professionals progress in their careers, they tend to find better opportunities abroad and therefore approach brain drain more positively.

It has been determined that fear of violence is a determining factor in individuals’ perceptions of security and therefore their brain drain tendencies. It appears that individuals with a high fear of violence are more likely to stay in countries where they feel safer.

Reflexivity statement

This study involves researchers of different seniority levels working collaboratively. There are three women in the team. The authors have different backgrounds and specializations (economics, medicine, health management, statistical analysis and econometric analysis). All authors have in-depth knowledge of low- and middle-income countries. Ethical approval was received from Istanbul Beykent University Scientific Research and Publication Ethics Board for Social and Human Sciences (date: 12 September 2023, no: 20238/8).

Background

At the level of threat or activity, physical force is deliberately directed at oneself, another person, a group or a community. Violence is defined as violence that results in or has a high probability of resulting in injury, death, psychological harm, atrophy or deprivation [1]. The phenomenon of violence comes to the fore in the health sector and causes social, mental and physical damage to health workers. In this context, violence in healthcare is defined as a set of physical, verbal and psychological attacks committed against any healthcare personnel by the patient himself, his relatives or third parties [2]. According to the World Health Organization, on a global basis, nearly 62% of healthcare workers are exposed to violence in the workplace, with verbal abuse (58%) being the most common form of violence, followed by threats (33%) and harassment (12%) [3]. According to the reports of the Health and Social Service Employees Union, in Turkey in 2021, there were 190 incidents of violence in healthcare, and in 2022, 316 healthcare workers were exposed to violence. It is stated that 422 healthcare workers were victims of violence in 249 cases of violence [4, 5]. It can also be noted that increasing violent incidents are a risk factor for the quality of health care. In the literature, it has also been determined that violence in health is a root cause of brain drain [6].

Brain drain is defined to move from one country to another to obtain a better quality of life and technological and political conditions [7]. Healthcare worker migration, called the "medical brain drain", refers to the mass migration of trained and skilled healthcare workers (doctors, nurses, midwives) from low-income countries to high-income countries [8]. Healthcare workers’ brain drain mainly occurs in underdeveloped or developing countries [9]. Brain drain continues to increase in Turkey. For example, The educated population migrating abroad from Turkey doubled from 2016 to 2019. Physicians have an important place in this number, increasing 27 times in the last decade [10]. The desire to go abroad among physicians has increased by 90 percent in the previous two years [11]. Another study among nurses showed that more than half wanted to migrate to other countries due to a lack of career opportunities, low wages and negative working conditions [12].

The literature states that, in parallel with the increase in the education level and professional experience of healthcare workers, their desire for an increase in their income level is a strong attraction factor for their migration tendencies [13, 14]. Studies indicate that as health professionals' education levels increase, their search for professional advancement, job security, and better working conditions also increases. While these factors strengthen brain drain tendencies, this tendency becomes more pronounced among highly educated professionals in regions with low-income levels or limited career advancement opportunities [15, 16]. Healthcare workers with high experience levels tend to emigrate due to job dissatisfaction, low wages, and limited development opportunities. Insufficient government support and heavy workload reinforce this tendency [17, 18]. Studies indicate that low-income levels are a determining factor in healthcare professionals' attitudes toward brain drain, increasing the tendency to migrate; higher incomes strengthen healthcare professionals' decisions to remain in their current countries [16, 17, 19, 20].

The desire to migrate is not only found in senior healthcare professionals but is also seen in young people at the beginning of their careers in healthcareIn various studies conducted with medical students and resident physicians in Turkey, participants stated that they wanted to work abroad due to issues in the healthcare system, as well as challenging living and working conditions. The main reasons for this desire were identified as high income, better employment and living conditions, violence in healthcare, and long working hours [21, 22]. A study on medical students and young doctors in Egypt determined that 89.4% of the participants wanted to emigrate. That salary, working hours, lack of appreciation, dissatisfaction with their relationships with patients and colleagues, and violence paved the way for brain drain. Moreover, the participants expressed their opinions on improvements made in the sector. They also said they would change it [6]. In this context, it can be stated that the medical brain drain caused by violence and living conditions in healthcare is a global problem. World Health Organization reports said there was a 60% increase in immigrant doctors and nurses working in OECD countries between 2010 and 2020. They pointed out an increasing mismatch between the supply and economic demand of healthcare workers and the ongoing acceleration in the international migration of healthcare workers [23]. In addition to its impact on the healthcare system, the brain drain of healthcare workers represents a financial loss for the country of origin due to the expenditures on health education [24]. The third of the Sustainable Development Goals, "Healthy and Quality Life", also aims to significantly increase the recruitment, development, training and maintenance of the health workforce, especially in underdeveloped and developing countries [25].

In light of this information, it can be stated that violence in healthcare is a significant risk factor for the sustainability and quality of the health system. This study specifically aims to examine the relationship between violence in healthcare and the brain drain intentions among healthcare professionals in Turkey, utilizing a quantile regression model. By focusing on the connection between exposure to violence and brain drain intentions, our study provides an original evaluation by reflecting the perspectives of healthcare professionals who experience violence directly in their work environments.

Through the quantile regression model, the analysis enables us to understand how brain drain intentions vary across different levels of fear of violence. This method allows for an examination of how brain drain attitudes change at various points within the distribution, offering insights that traditional mean-based models may overlook. By capturing the variations in brain drain attitudes across different percentiles, particularly at the lower and higher ends, our approach introduces an econometric perspective that enriches the literature and brings a methodological innovation to studies on violence in healthcare and its broader impacts on workforce retention.

Methods

This study includes the methods used to address the research questions and test the hypotheses, along with details on the data collection process and analysis approach.

Sampling and measurement tools

Based on the literature, this study used the Attitude Scale Towards Brain Drain for Nursing Students (16 items, 5-point Likert scale) developed by Öncü et al. [26] and the Fear of Violence Scale (20-point Likert scale) developed by Sağlam [27], along with demographic information. The study was conducted following the principles of the Declaration of Helsinki. Consent was obtained from the participants through an informed consent form for their participation in the survey. Ethical approval was received from Istanbul Beykent University Scientific Research and Publication Ethics Board for Social and Human Sciences (date: 12 September 2023, no: 20238/8).

The research population consists of individuals working in the health sector in 2023. The sample population represents the population consisting of 210 individuals using the convenience sampling method. Data were collected through an online survey, which enabled the participation of healthcare professionals from different regions of Turkey. This diversity ensured that the study included different regional experiences, increasing the general validity of the findings. In the study, the term “health sector workers” covers both health professionals who are directly involved in patient care (nurses, physicians, dentists, etc.) and a broader definition such as administrative and support roles in the health sector. In order to avoid the sampling error that would occur when the determined sample’s ability to represent the population is insufficient, in any multivariate study, especially in factor analysis applications, the requirement to take observations that are at least five times the number of items is taken into account for the adequacy of the sample size [28].

Research hypotheses

The following hypotheses were discussed to understand which factors affect healthcare professionals’ attitudes toward brain drain.

H ₁: Health workers’ fear of violence has a statistically significant effect on attitudes towards brain drain.

H ₂: Demographic characteristics of healthcare professionals have a statistically significant effect on attitudes toward brain drain.

Results

The data obtained within the research were analyzed using SPSS 25 and Stata 15 package programs. Table 1 presents the demographic characteristics of the study participants, along with the mean, standard deviation and frequency values of variables such as age, gender, education level, marital status, income level and occupational distribution.

Table 1.

Descriptive statistics

	Mean	Std. Dev.	Min	Max	25%	50%	75%
Gender	1.3523	0.4788	1	2	1	1	2
Age	39.6142	10.8895	20	70	31	39.5	46
Education Level	4.6523	1.0928	1	6	4	5	5
Marital Status	1.3904	0.4890	1	2	1	1	2
Income (TL)	51053.57	63567.7132	11000	700000	28000	39500	53500
Professional Groups	2.9666	1.4389	1	6	2	3	3

Education level (1: Primary School, 2: High School, 3: 2-Year University, 4: 4-Year University, 5: Master's, 6: Doctorate), years of professional experience (1: Less than 1 year, 2: 1–3 years, 3: 4–6 years, 4: 7–9 years, 5: 10 years or more), professional group (1: Nurse, 2: Manager, 3: Doctor, 4: Health Technician, 5: Dentist, 6: Other), gender (1: Female, 2: Male), marital status (1: Married, 2: Single) are coded as shown

64.76% of the survey participants are women, and 35.24% are men. Although the age range of the participants varies, the youngest participant’s age is 20, and the oldest participant’s age is 70. The participants' average age was 39.6 years (SD = 10.9). The youngest participant is 20 years old, while the oldest is 70. It was observed that 60% of the participant’s education level was mainly at the postgraduate level. Most participants had a high level of education, with an average education level corresponding to “undergraduate” to “postgraduate” (Mean = 4.65). The lowest participation is primary school graduates with 0.48%. It was observed that 60.95% of the participants were married, with the average marital status coded as 1.39 (where 1 = married, 2 = single). Their average monthly personal income was 51,053.57₺. When the professional experience of the participants was examined, it was determined that 62.38% of them were ten years or more; It was determined that 2.38% had less than one year of experience. When examined according to the profession, the highest participation rate was medical doctors with 45.71 percent; It was observed that dentists made the lowest participation with a rate of 2.38.

It was determined that 73.81% of the participants witnessed physical violence, while 79.52% were not directly exposed to physical violence. This shows that physical violence is widely observed in the environment where healthcare workers are present, but most of the participants do not directly experience it. In terms of verbal violence, 94.29% of the participants witnessed verbal violence and 83.33% directly experienced verbal violence. This shows that verbal violence is a more common problem than physical violence in the healthcare sector.

In this study, brain drain attitude was considered as the dependent variable. Factors such as income status, length of professional experience and fear of violence were evaluated as independent variables. Fear of violence was measured with scales based on participants' experiences of physical and verbal violence at work, and the survey questions were structured using Likert-type scales. This structure is a technique that develops the relationship between measurement and factors based on the relevant theoretical framework. Therefore, this technique tests existing theories rather than producing a new theory. Unlike explanatory factor analysis, only the relationships between the determined factors were examined from various perspectives. As a result, the established model [1] is based on a theoretical basis and comprehensively analyzes the effects of independent variables on brain drain attitude.

$$x=\mathbf{\Lambda }n+\epsilon$$ 1

$x$: $p\times 1$ dimensional vector,

$\mathrm{\Lambda }$: $p\times q$ dimensional factor loading matrix,

$n$: $q\times 1$ dimensional vector of latent factors,

$\epsilon$: $p\times 1$ it is the theoretical representation of factor analysis with dimensional error vector [29].

Based on this, the reliability of the scales used in the study was examined first. It was determined that the Attitude Scale toward Brain Drain $\left(\alpha ,=,0.9326\right)$ and Fear of Violence $\left(\alpha ,=,0.8705\right)$ were high. In addition, as a result of the Bartlett sphericity test $\left(0.000,<,0.05\right)$, it was observed that there was a high correlation between the variables ($H_0$: Correlation matrix is ​​the identity matrix (R = I)). In order to ensure the significance of the original factor matrix in the interpretation, the Varimax method recommended by Kaiser, one of the orthogonal axis rotation methods, was used. According to the varimax rotation result of the principal components analysis, the total explained variance was 67.43% in the Attitudes Towards Brain Drain Scale and 55.56% in the Fear of Violence Scale. Based on these findings, It is determined that the scales meet the necessary conditions for structural validity. Based on this, to determine whether the data distribution complies with the normal distribution, the skewness and kurtosis values must be between−1.5 and + 1.5 [30]. Considering the overall scale, it was determined that the distribution was normal, so it was decided that parametric methods could be used ${(\gamma }_1=-0.0001;{\gamma }_2=0.9949)$.

It was determined that the average score of the Attitude Scale towards Brain Drain, where a maximum score of 80 and a minimum of 16 points can be obtained and with no cut-off point, was 40.32 ± 9.656. An increase in the score indicates that the tendency to migrate increases. The scale items were classified as attractive and repulsive factors, and the first two items with the highest scores were determined and given in Table 2.

Table 2.

Push and pull factors in brain drain

Group	Ordinal	Scale item	$\overline{X}\pm SD$
Push Factor	1	I would like to live in another country where I can be away from political pressures	1.961905 ± 1.106084
	2	I would like to work in another country where I won’t have to worry about the future	1.919048 ± 1.061785
Pull Factor	1	I would like to work abroad because I can earn more money	3.714286 ± 1.073511
	2	I think that working abroad will increase my living standards	3.538095 ± 1.128566

When push and pull factors are examined, working and living conditions likely affect the participants’ attitudes toward brain drain. According to the Fear of Violence Scale, it was determined that the participants had a moderate level of fear of violence (3.28 ± 0.710). When the Pearson correlation coefficient was examined to determine the relationships between the variables, it was seen that there was a low degree (−0.2989), reverse and statistically significant (p-value: 0.000 ≤ 0.05) relationship. Based on the research hypotheses, a multiple regression model was estimated to determine the variables that statistically affect attitudes toward brain drain.

Multiple regression analysis prediction results are presented in Table 3. The multiple regression model is the final model established with the variables determined to be statistically significant on the dependent variable. It is understood that there is no autocorrelation problem since the regression model is statistically significant and the Durbin-Watson value is very close to 1.604 and 2 [31]. According to the White test, which was examined to determine whether there was heteroscedasticity in the regression model, it was seen that the basic hypothesis stating that the error terms had constant variance was not rejected. Therefore, it was found that there was no heteroscedasticity problem in the estimated model. When the independent variables did not affect the model examined in the diagnostic tests, it was determined that the average score of the attitude scale toward brain drain was 43.6007. According to the model, an increase in the fear of violence reduced the attitude towards brain drain by 0.2471 units. In addition, income status hurts attitude. A statistically significant positive effect of the duration of professional experience was detected.

Table 3.

Multiple regression analysis prediction results

Dependent Variable: Attitude Towards Brain Drain							VIF
Statistics	Coefficient	Std. Err.	t-statistic	p-value	Confidence interval (95%)
Variables					Lower limit	Upper limit
Fear of Violence	−0.2471	0.0580	−4.26	0.0000	−0.3615	−0.1328	1.02
Professional Experience Period	0.1569	0.0719	2.18	0.0300	0.0151	0.2988	0.88
Income status	−0.1711	0.0679	−2.52	0.0130	−0.3051	−0. 0372	0.80
Constant	43.6007	4.1239	10.57	0.0000	35.4698	51.7315
$R^2=0.5802$
${\overline{R}}^2=0.4211$
F(3,206) = 8.19; p = 0.0000
Durbin-Watson: 1.604
White Test: 10.67; p = 0.2990
Jarque–Bera Test: 12.05; p = 0.0024

However, the Jarque–Bera test based on the kurtosis and skewness of the sample determined that the distribution of errors was not normal. Normality expresses whether the observations are compatible with a certain distribution curve. The suitability of the variables for this distribution is a critical feature that should be considered before multivariate analyses. If the normality assumption is addressed, the analysis results may stay consistent with the true values, which may affect the reliability of the analysis [32]. Accordingly, since classical regression analysis assumes normally distributed error terms, it is appropriate to use alternative regression techniques when this assumption is violated. Therefore, we adopted quantile regression, a robust technique that captures effects at all points of the dependent variable’s conditional distribution and accounts for asymmetry in the tails. This flexible method does not rely on strict assumptions [33, 34].

Quantile regression model

It was developed as an alternative to the least squares estimator in linear models where the regression error term distribution is abnormal and has a long tail feature [35]. The model, which is based on the logic of modeling conditional quantiles as a function of the independent variables in the model, is aimed to explain the change in conditional quantiles. In the traditional regression model, changes in the conditional mean of the dependent variable are examined. It enables measuring variables in a distribution center and the area within the lower and upper tail [36].

$$\genfrac{}{}{0.0pt}{}{\mathit{min}}{\beta ϵR}\left[{\sum }_{tϵ\left\{t,:,y_t,\ge ,x_t,\beta \right\}},q,\left|y_t,-,x_t,\beta \right|,+,{\sum }_{tϵ\left\{t,:,y_t,<,x_t,\beta \right\}},(1-q),\left|y_t,-,x_t,\beta \right|\right]$$ 2

While any Y variable takes a value in the range $0<q<1$ with its ordinal quantile degree, the regression process occurs with the minimization problem [37].

Quantile regression analysis prediction results are presented in Table 4. Quantiles were created to investigate how the attitudes of individuals in the data set toward brain drain were distributed according to their fear of violence. These quantile ranges, for example, will help to understand how attitudes toward brain drain are distributed in different groups according to their fear of violence. The econometric approach used in this study is based on the quantile regression model to analyze the determinants of brain drain attitudes. Quantile regression allows estimating the relationship between different quantiles in the conditional distribution of the dependent variable and the independent variables. This approach allows evaluating the effects at different points of the distribution (e.g. low, medium, high quantiles), whereas the classical OLS (Least Squares) method only considers the average effects. The use of the quantile regression model in the study allows us to examine how the effects of variables such as income level, professional experience and fear of violence on brain drain attitudes change at different points of the distribution. This provides a more comprehensive and detailed analysis by capturing the heterogeneous effects at the extreme values ​​of the dependent variable. Thus, this approach, which clarifies the potential heterogeneities that classical average models may miss, strengthens the methodological contribution of the research.

Table 4.

Quantile regression analysis prediction results

Quantile	q = 0,10		q = 0,50		q = 0,90
	Coefficient	t-value	Coefficient	t-value	Coefficient	t-value
Variable
Fear of Violence	−0.0207a	−0.95	−1.5333a	−2.08	−1.5926a	−2.07
Professional Experience Period	0.1282a	2.53	0.3206a	0.50	0.8661a	8.20
Income status	−2.1513a	−2.27	−3.2323a	−3.30	−0.5218	−0.64
Fixed Term	31.7921a	9.54	46.9959a	10.48	61.5414a	6.48
${\mathit{PseudoR}}^2$	0.2573		0.2899		0.3817

The ^a symbol represents statistical significance at the 5% level

A quantile regression model was estimated to determine whether the relationship between factors affecting attitudes towards brain drain changes at different attitude scores. Model results are examined in percentiles between the 10% with the lowest attitude toward brain drain and the 90% with the highest attitude score.

It is seen that attitude scores towards brain drain decrease as the fear of violence increases. In other words, as the fear of violence increases, individuals’ attitudes towards brain drain become more negative. It shows that low fear of violence has a limited effect on attitudes towards brain drain. However, it shows that individuals with a high fear of violence have a lower desire to engage in brain drain. Especially at highlevels of fear, attitudes towards brain drain are negatively affected, and the willingness of t, these individuals to engage in brain drain decreases. The effect of the fear of violence variable on attitude continues to increase negatively at all quantile levels.

It is seen that attitude scores towards brain drain decrease as income increases. However, it is seen that it is statistically significant at all quantile levels except 0.90. In the 0.10 quantile, a 1-unit increase in income increases the attitude score to 2.15 units on average. It was determined that it reduced attitude scores by 3.23 units on average at the 0.50 quantile. Therefore, the increase in income generally reduces the positive attitude towards brain drain. This suggests that individuals with higher income levels have a more negative attitude towards brain drain.

It is seen that attitude scores towards brain drain increase with the increase in professional experience. In the 0.10 quantile, a 1 unit increase in the duration of professional experience increases the attitude score by 0.12 units on average; at the 0.50 quantile, attitude scores are on average 0.32 units; and it was determined that it increased attitude scores by 0.86 units on average at the 0.90 quantile. Therefore, the increase in the duration of professional experience generally increases the positive attitude towards brain drain.

Discussion

Violence and brain drain in healthcare are among modern healthcare systems’ most serious problems. Incidents of violence against healthcare workers reduce the quality of healthcare services and negatively affect the morale and motivation of healthcare professionals. Increasing cases of violence around the world reduce the satisfaction of healthcare professionals with their profession, and this causes talented individuals to migrate abroad. Especially in developing countries, the prevalence and effects of violence in healthcare stand out as an important factor contributing to the brain drain of healthcare workers. Brain drain refers to the migration of highly educated and skilled professionals, especially from developing countries to developed countries. This means the transfer of qualified human capital [38, 39]. Individuals who are well-educated and specialized in their fields migrate to countries that offer opportunities such as better living and working conditions, success in their careers and higher salaries. Developed countries aim to become more attractive to educated and promising young talents by allocating large shares of their budgets to offer these opportunities. As a natural result of this situation, brain drain has increased greatly in recent years [39]. When the main reasons healthcare professionals turn to brain drain are examined within the scope of the literature, It seems that better living conditions, higher income, career and education opportunities and better working opportunities are effective. In particular, employees’ perception of satisfactory wages increases their motivation and performance. On the contrary, wages perceived as insufficient lead to negative consequences such as job dissatisfaction, intention to quit, and life dissatisfaction [12, 40, 41].

Our study was discussed from the perspective of healthcare professionals who are actively working in the field, who are intertwined with the concept of violence in their work lives, and who have experienced it in determining the root causes of violence and brain drain in health and offer suggestions. Demographic characteristics of the participants play a determining role in attitudes towards brain drain. The fact that 64.76% of female participants provide important data to examine the effects of gender on brain drain. The findings of the study indicate that there may be differences in attitudes towards brain drain between male and female participants. In the literature, no conclusion has been reached that reveals a causal relationship between two variables within the scope of the health sector. However, there have been studies showing that female health workers are more exposed to violent incidents [42–44]. In addition, studies have shown that the migration of highly skilled women is higher than that of men. The health sector is female-dominated, especially since nurses are predominantly women [45, 46]. Based on the findings and literature, it is recommended that the role of gender in brain drain (sector, social structure, working and living conditions, etc.) be examined in more detail in future studies.

The fact that 60% of the participant’s education level is at the postgraduate level shows how a high education level impacts brain drain [47]. It has also been determined that the brain drain tendency of qualified and competent academic staff has increased in developing countries [48]. In this context, our findings are compatible with the literature. Based on this, As mentioned before, individuals’ self-efficacy and academic self-efficacy based on self-efficacy theory may be an important factor in brain drain. Academic self-efficacy refers to individuals’ beliefs that they can successfully fulfill their academic duties, and this belief significantly affects individuals’ motivation in education, performance and career planning in general. The relationship between academic self-efficacy and brain drain can be explained by individuals’ search for opportunities to maximize their talents and potential. Individuals with high academic self-efficacy can contribute to brain drain by seeking better conditions in countries where they can find more suitable career and education opportunities. In this context, increasing academic self-efficacy can be critical in reducing brain drain and enabling individuals to find better opportunities in their home countries. Therefore, it is an important factor that can be addressed in future studies examining the attitudes of healthcare professionals toward brain drain.

The findings showed that attitude scores towards brain drain increased with increased professional experience. Regression analyses determined that increasing the duration of professional experience positively affected the brain drain intention. In the quantile regression analysis findings, it was observed that the attitude scores towards brain drain increased with the increase in the duration of professional experience. An increase in the length of professional experience reflects that individuals are more advanced in their careers and are more likely to find better opportunities abroad. This shows that more experienced individuals have more positive attitudes towards brain drain. Especially at the 0.90 quantile, it was determined that a 1-unit increase in the duration of professional experience increased the attitude score by 0.86 units on average. This finding expresses the positive effect of professional expertise on brain drain decisions. In the literature, it is stated that the work experience of healthcare professionals and the problems they experience in advancing in the profession are reasons for brain drain [49]. It is noted that the medical brain drain is affected by immigration policies aimed at attracting highly skilled workers such as doctors, and immigration is encouraged by visa restrictions, recognition of diplomas, tax deductions for immigrants and options to obtain permanent residence status [50]. In this context, our findings are compatible with the literature and that high education level is a factor that enables participants to be more competitive in the global labor market and have a higher tendency to brain drain. More experienced individuals are likelier to find better career opportunities abroad and view brain drain more positively.

Fear of violence stands out as an important determining factor in attitudes towards brain drain. While 73.81% of the participants had witnessed physical violence before, 79.52% had not been exposed to physical violence. On the other hand, the rates of witnessing and being exposed to verbal violence are quite high. It is seen that attitude scores towards brain drain decrease as the fear of violence increases. This finding shows that individuals with a high fear of violence have a reduced tendency to go abroad to feel safer. This result was critical for understanding how the fear of violence shapes individuals’ plans for the future and affects their brain drain decisions. In the regression analysis findings, it was determined that the increased fear of violence negatively affected the attitude toward brain drain. Quantile regression analysis findings show that this effect increases negatively at all quantile levels. It has been determined that individuals with a particularly high fear of violence have a lower desire to engage in brain drain. This finding reveals that the fear of violence significantly affects individuals’ perceptions of security and, therefore, their tendency to brain drain. In a study conducted with medical students in Turkey, it was determined that violence in healthcare was a reason for brain drain [51]. In a study conducted on health professionals in Egypt, it was stated that violence in health is a root cause of brain drain [52]. In this context, our findings differ from the literature. The reason for this may be that the experiences of the people participating in the study, their regions, and their working conditions are different from those in other studies. Looking at the findings show that the majority of employees in the health sector have negative experiences with violence in the workplace. Verbal violence is a type of violence that is more commonly both witnessed and experienced than physical violence.

On the other hand, since it was determined that the individuals in the sample were not exposed to high rates of physical violence, it is thought to be related to the findings. This result can also be examined in depth as a topic of future research. Because at this point; It is thought that healthcare professionals’ professional and personal competencies, especially their self-efficacy, will be important factors in combating violence.

Individuals with high self-efficacy demonstrate high performance in problem-solving and success indicators and effectively solve the problems they encounter in healthcare delivery. However, no studies directly examine the relationship between self-efficacy and brain drain. Considering the positive effects of self-efficacy on healthcare workers’ job satisfaction, stress management and career development, it is clear that it may also be an important factor in brain drain. It is possible to state that the fear of violence hurts individuals’ self-efficacy. Still, there is no consensus on this issue yet in the literature [27]. Since healthcare professionals with high self-efficacy may be more successful in overcoming the difficulties they face and maintaining their professional development, the brain drain tendencies of these individuals may also differ. Increasing the self-efficacy levels of healthcare professionals can both strengthen their capacity to cope with violence and reduce brain drain rates. Therefore, it is thought that addressing self-efficacy and academic self-efficacy factors in future research is important.

Another important finding is how the income levels of the participants affect their attitudes toward brain drain. It has been observed that attitude scores towards brain drain decrease as income increases. According to the regression analysis findings, income level increases hurt brain drain intention, but this effect varies at different income levels. This shows that individuals with higher income levels are more satisfied with their current living conditions and, therefore, have less desire to undergo brain drain. However, quantile regression analyses determined that this relationship was statistically significant at all quantile levels except the 90% quantile. This finding shows that income increases generally increase negative attitudes towards brain drain and that individuals’ satisfaction levels with their current economic situation affect brain drain decisions. The fact that this relationship varies according to quantile levels indicates that the effect of income increase on brain drain is heterogeneous. This shows that brain drain trends according to income level in Turkey should be examined carefully. When studies are investigated, it is stated that there is a causal relationship between income, education and life expectancy indices and that income inequality is a cause of brain drain [53, 54].

Additionally, studies indicate that receiving a high salary triggers brain drain [55]. For example, higher-skilled employees are affected by higher wages, and salary levels may cause healthcare professionals to migrate to other countries [56]. Our findings are consistent with the literature, showing that individuals with higher income levels are more satisfied with their current living conditions, do not want to change their routine, and therefore have less desire to migrate.

Conclusions

The quantile regression model provided a valuable tool to understand how attitudes towards brain drain differ according to factors such as fear of violence, income level and length of professional experience. Model results revealed significant differences in the percentiles between the 10% with the lowest attitude score towards brain drain and the 90% with the highest. These findings show that attitudes towards brain drain do not depend on a single variable but are shaped by complex interactions of multiple factors. The fact that income status generally hurts brain drain is closely related to the economic conditions in Turkey. The decrease in attitude scores towards brain drain as income increases indicates that individuals are more satisfied with their living conditions at higher income levels.

Based on the study findings, the fear of violence may also increase individuals’ tendency to stay in the country because the migration process also involves uncertainties and risks. While security problems may be temporary, and individuals can often develop more adaptive behaviors against these problems, economic difficulties have a long-term and ongoing impact. Although security concerns are important, financial concerns greatly impact individuals’ brain-drain decisions. Therefore, Turkey must first improve economic conditions and increase employment and professional development opportunities to reduce the brain drain. This study makes important contributions to understanding the factors affecting attitudes toward brain drain. Variables such as fear of violence, income level and length of professional experience play an important role in shaping individuals’ brain drain decisions. By examining the effects of these variables in more detail, future research may help develop policies to prevent or manage brain drain. In particular, understanding how employment and security policies affect individuals’ attitudes toward brain drain will provide important information to decision-makers.

Authors’ contributions

Conception and design of the work: G.O., H.M. and G.B. Survey design and implementation: G.O., H.M. and G.B. Data analysis: G.B. Interpretation of data for the work: H.M. and G.B.Drafting the work: G.O. and H.M. Revising it critically: all authors. Final approval of the version to be published: all authors.

Funding

There is no financial support from any institution.

Data availability

The data that support the findings of this study are available from the authors.

Declarations

Ethics approval and consent to participate

Ethical approval was received from Istanbul Beykent University Scientific Research and Publication Ethics Board for Social and Human Sciences (date: 12 September 2023, no: 20238/8). Informed consent to participate in the study was obtained from all participants. All procedures contributing to this work comply with the Helsinki Declaration.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.