The Determinants of Brain Drain and the Role of Citizenship in Skilled Migration

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This article offers an examination of the determinants of brain drain on a global scale. The authors test six broad factors that might either increase or decrease brain drain--or the out-migration of skilled migrants to other countries. 

I think topically this study is very novel and important to the study of migration. However, I found this manuscript was very difficult to read, and empirically the ms was difficult to gauge its contributions. I therefore have suggestions largely to improve readability and empirical contributions:

1. The ms was hard to read, so I'd invest resources into an editor (see comments below on language)

2. The manuscript dove too quickly (in my opinion) into a discussion of variables, without first motivating and justifying the research question in full. There are other theoretical accounts of in- and out-migration generally speaking, that don't necessarily get at determinants, that could motivate and justify the study. I think the study would flow better with an introduction to the overall question (within sociological literatures aside from migration, for instance), before getting to the variables.

3. As it stands, the paper reads as if the authors cherry-picked their literature review/hypotheses from the variables they knew they could test--rather than the literature more generally. I think the motivation of determinants could be more broad and less of a laundry-list of the variables that then happen ot perfectly match up with what the authors test empirically.

4. There is no discussion of the data, which makes it hard tot gauge the contribution of the empirical section. This should be several paragraphs. I suggest elaborating this section and defending it against other potential sources (as well as moving some of the comparison of other empirical studies currently in the discussion so that readers can see how this study extends on other empirical contributions so far).

Comments on the Quality of English Language

I thought the quality of the English language was high, but the readability of the manuscript was low. There were many instances where sentences were just dependent clauses, and there were lots of phrases/jargon without any explanation of what those phrases meant. I think the authors should consider hiring an editor to help them with readability, as I personally found the manuscript very hard to read and understand.

Author Response

Dear Reviewer 1, we appreciate your recommendations, which have greatly helped us to improve our manuscript. We have highlighted in green color all the changes requested by the team of four reviewers.

1. The ms was hard to read, so I'd invest resources into an editor (see comments below on language).

Response: We have re-edited much of the manuscript, making the text more fluent and adapting it to a more general readership.

2. The manuscript dove too quickly (in my opinion) into a discussion of variables, without first motivating and justifying the research question in full. There are other theoretical accounts of in- and out-migration generally speaking, that don't necessarily get at determinants, that could motivate and justify the study. I think the study would flow better with an introduction to the overall question (within sociological literatures aside from migration, for instance), before getting to the variables.

Response: We have added an introduction between lines 33 and 133, allowing a general understanding of the study and its overall objective, before giving way to the theoretical framework and explanation of the variables studied, which has also been expanded.

3. As it stands, the paper reads as if the authors cherry-picked their literature review/hypotheses from the variables they knew they could test--rather than the literature more generally. I think the motivation of determinants could be more broad and less of a laundry-list of the variables that then happen ot perfectly match up with what the authors test empirically.

Response: An introduction section has been added (lines 33-133) and the literature review has been expanded with two subsections (lines 142-295). In addition, a figure 1 has been incorporated to show the two main factors studied as determinants of brain drain, which have been disaggregated into three categories and the eleven variables studied.

4. There is no discussion of the data, which makes it hard tot gauge the contribution of the empirical section. This should be several paragraphs. I suggest elaborating this section and defending it against other potential sources (as well as moving some of the comparison of other empirical studies currently in the discussion so that readers can see how this study extends on other empirical contributions so far).

Response: Retaining the article structure of the journal. More details have been incorporated in the results section and the discussion section (lines 372-448) has been rewritten for better understanding. Additionally, improvements have also been introduced in the conclusion (highlighted in light green).

Reviewer 2 Report

Comments and Suggestions for Authors

The introduction covers the state-of-the-art comprehensively. The empirical findings on the determinants of brain drain are manifold, hence complex relationships occur. The paper summarizes findings from different fields of research and perspectives, but the structure could be improved contentwise. I would recommend to group the findings based on the most relevant dimensions (and possibly add subtitles). This would improve comprehensibility.

The methodological approach is describe clearly, but the data used for analysis should be described at lease with a few sentences.

After the regression equation, the same paragraph is added twice. Please remove one.

The discussion and conclusions refer very much to the results in methodological terms. The interpretation respectively the description of the content-related meaning of the results could be improved.

Author Response

The introduction covers the state-of-the-art comprehensively. The empirical findings on the determinants of brain drain are manifold, hence complex relationships occur.

1. The paper summarizes findings from different fields of research and perspectives, but the structure could be improved contentwise. I would recommend to group the findings based on the most relevant dimensions (and possibly add subtitles). This would improve comprehensibility.

Response: Dear Reviewer 2, we greatly appreciate your corrections, which have been very helpful in improving our manuscript.
In response to the team of four reviewers, we have highlighted in light green all the changes made to the document.
We have added an introductory section and separated the literature review into two sections. In addition, a figure 1 has been incorporated to show the two main factors studied as determinants of brain drainage, which have been disaggregated into three categories and the eleven variables studied. This has allowed us to give a greater order to the conclusions, allowing a better understanding of the manuscript.

2. The methodological approach is describe clearly, but the data used for analysis should be described at lease with a few sentences.

Response: We have improved the description of the data in Table 1, and have added an appendix A, where the 11 independent variables and the dependent variable are explained, based on the data article we have previously published.

3. After the regression equation, the same paragraph is added twice. Please remove one.

Response: We have removed the duplicate paragraph.

4. The discussion and conclusions refer very much to the results in methodological terms. The interpretation respectively the description of the content-related meaning of the results could be improved.

Response: We have rewritten the discussion (lines 372-448), as well as added a paragraph to the conclusion (470-478). In both texts we have emphasized the significance of the result in relation to the variables, categories and factors under study.

Reviewer 3 Report

Comments and Suggestions for Authors

Interesting paper which deserves publication but should be improved.

The structure should be improved. The introduction already lists potential drivers of brain drain migration. This should be moved to one theoretical section where a synthesis of the potential drivers of brain drain is made (potentially with a figure).

The paper seems to take for granted that brain drain is bad. I agree but highly qualified emigration is not necessarily bad. There is a scholarly debate that should at least be mentioned.

Reference should be made to the theories of migration (especially to the New Economics of Migration)…

Massey, D. S., J. Arango, G. Hugo, A. Kouaouci, A. Pellegrino & J. E. Taylor (1993) Theories of international migration: A review and appraisal. Population and Development Review, 19, 431-466.

Stark, O. & J. E. Taylor (1989) Relative Deprivation and International Migration. Demography, 26, 1-14.

An in-depth discussion of the dependent variable is needed. How is brain drain measured what are the possible bias in the measurement.

The same for each potential explanatory variable: in-depth critical discussion is needed.

The results should be explained with much more details. The signs of the coefficient should be discussed (why a + for “public services” and a – for “rule of law” for ex.).

The whole “Discussion” pat should be rewritten streamlined and clarified.

Part of the conclusion should be in the discussion part. The conclusion should mainly show the added value of the research in comparison with the existing literature.

Other points

l. 32 Not clear why poverty has to do with brain drain / highly qualified emigration.

l. 119 explain “triple helix”

l. 135 typo.

p.4 why do you start again to cite authors for each potential drivers ? This should be done in one single theoretical/literature review section.

l. 171-177 Duplicated § !

p. 7 the technicalities as well as Figure 2 could be in Annex to make space for a real discussion of the results.

 

Author Response

0. Interesting paper which deserves publication but should be improved.

Response: Dear Reviewer 3 we appreciate your comments, which have been invaluable in improving our manuscript. Changes in response to the team of 4 reviewers have been highlighted in light green throughout the manuscript.

1. The structure should be improved. The introduction already lists potential drivers of brain drain migration. This should be moved to one theoretical section where a synthesis of the potential drivers of brain drain is made (potentially with a figure).

Response: We have completely edited the introduction, adding subtitles. In addition we have generated a literature review section, giving a more orderly structure to the theoretical determinants of brain drain, at the end of this new section the requested figure is incorporated (see figure 1).

2. The paper seems to take for granted that brain drain is bad. I agree but highly qualified emigration is not necessarily bad. There is a scholarly debate that should at least be mentioned.

Response: Thank you for this correction and opportunity to incorporate the broader debate, accounting for brain drain, brain gain and brain circulation. We have added this in the new version of the introduction.

3. Reference should be made to the theories of migration (especially to the New Economics of Migration)…

Massey, D. S., J. Arango, G. Hugo, A. Kouaouci, A. Pellegrino & J. E. Taylor (1993) Theories of international migration: A review and appraisal. Population and Development Review, 19, 431-466.

Stark, O. & J. E. Taylor (1989) Relative Deprivation and International Migration. Demography, 26, 1-14.

Response: Thank you very much for the references you have indicated, both have been included in the improvement of the introduction, together with several articles that we have added as references to the manuscript.

4. An in-depth discussion of the dependent variable is needed. How is brain drain measured what are the possible bias in the measurement.

Response: Thank you very much for the references you have indicated, both have been included in the improvement of the introduction, together with several articles that we have added as references to the manuscript.

5. The same for each potential explanatory variable: in-depth critical discussion is needed.

Response: We have added a literature review section (lines 134-297) and further discussion in the new presentation of the discussion (lines 371-448). The central ideas have also been highlighted at the end of the conclusion. In addition, Table 1 and Appendix A provide more detail on each variable. We have also incorporated an appendix A detailing this and other variables.

6. The results should be explained with much more details. The signs of the coefficient should be discussed (why a + for “public services” and a – for “rule of law” for ex.).

Response: We greatly appreciate this comment. Accordingly, we have adjusted the methodology and limited the dozens of models explored (see table B.1, in appendix B) to those with coefficient signs (betas) in accordance with the identified theoretical causality (tables 2 and 3).

7. The whole “Discussion” pat should be rewritten streamlined and clarified.

Response: We have completely reformulated the discussion, respecting the journal's instructions.

8. Part of the conclusion should be in the discussion part. The conclusion should mainly show the added value of the research in comparison with the existing literature.

Response: We have moved the parts you point out to the discussion and have rewritten part of the conclusion.

9. Other points

9.1. 32 Not clear why poverty has to do with brain drain / highly qualified emigration.

Response: Regarding Poverty, we have specified in line 216, that the literature mentions poverty in the countries of origin.

9.2. 119 explain “triple helix”.

Response: We have incorporated a brief explanation regarding the meaning of the triple helix, in addition to adding a fundamental reference on that subject (165-168).

9.3. 135 typo.

Response: We have completely rewritten the introduction.

9.4. p.4 why do you start again to cite authors for each potential drivers ? This should be done in one single theoretical/literature review section.

Response: We have added a literature review section, where all these references have been incorporated.

9.5. 171-177 Duplicated § !

Response: We have removed the duplicate paragraph.

9.6. 7 the technicalities as well as Figure 2 could be in Annex to make space for a real discussion of the results.

Response: We have added the interpretation of the histogram of errors and also added a paragraph on the importance of this technicality in terms of providing validity to the conclusions.

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript “Determinants of Brain Drain and the Role of Citizenship in the Integration of Skilled Migrants” is a well-written research article dealing with an important and contemporary subject of brain drain almost on a global scale. However, the second part of the title is missing from the analyses and discussion. The article mainly analyses and discusses determinants of the brain drain. The role of citizenship in the integration of skilled migrants is not mentioned either in the theoretical framework or in the analyses (and consecutive discussion). Since the manuscript is quite short, I recommend the authors to address this aspect as well.

The theoretical framework of the brain drain is well-developed and includes key references.

In the Materials and Methods section, more information on predictor variables should be provided. There is no information on how these variables/indices are constructed, what is the range of presented values, and what is a description of high and low scores on each independent construct. This information could be introduced in Table 1. In this segment author(s) repeat twice the same paragraph – p.4, l.164-170 and l.171-177.

The results are clearly presented. Is there any chance that there would be differences in conducting the regression analysis if the predictor variables were inserted in a different order? For speaking about model 8, the first with the unfavourable multicollinearity indicators – what would happen if the predictor G6 was included in the model instead of E1? Or G3, or S1. Is there any possibility the model fit would change? In other words, does the sequence of including predictor variables play a significant role in the outcome of the model?

Finally, in the light of the incomplete description of predictor variables, the discussion fails to provide deeper insights into possible interpretation of the obtained results. The authors merely repeat the obtained correlations and link them to other research without trying to explain what it means and possibly how to deal with the societal-level shortcomings of the significant determinants of brain drain.

 

If the authors decide not to develop the idea of citizenship and integration of skilled migrants, then the presented analysis could be further elaborated by comparing and analysing possible differences between the countries of the Global South and Global North.

Author Response

0. The manuscript “Determinants of Brain Drain and the Role of Citizenship in the Integration of Skilled Migrants” is a well-written research article dealing with an important and contemporary subject of brain drain almost on a global scale. However, the second part of the title is missing from the analyses and discussion. The article mainly analyses and discusses determinants of the brain drain.

Response: Dear Reviewer 3 we appreciate your comments, which have been invaluable in improving our manuscript. Changes in response to the team of 4 reviewers have been highlighted in light green throughout the manuscript.

1. The role of citizenship in the integration of skilled migrants is not mentioned either in the theoretical framework or in the analyses (and consecutive discussion). Since the manuscript is quite short, I recommend the authors to address this aspect as well.

Response: We have addressed the citizenship aspect more broadly, incorporating it in the introduction, literature review, discussion and conclusions.

2. The theoretical framework of the brain drain is well-developed and includes key references.

Response: We appreciate this comment, however we have improved the introduction and literature review, giving greater order and classification to the variables under study.

3. In the Materials and Methods section, more information on predictor variables should be provided. There is no information on how these variables/indices are constructed, what is the range of presented values, and what is a description of high and low scores on each independent construct. This information could be introduced in Table 1. In this segment author(s) repeat twice the same paragraph – p.4, l.164-170 and l.171-177.

Response: In the methods section we have incorporated more information of the variables (with origin in the FSI and WGI), in table 1 and appendix A details and definitions of all the variables under study are added. Regarding the repeated paragraph, it has been eliminated.

4. The results are clearly presented. Is there any chance that there would be differences in conducting the regression analysis if the predictor variables were inserted in a different order? For speaking about model 8, the first with the unfavourable multicollinearity indicators – what would happen if the predictor G6 was included in the model instead of E1? Or G3, or S1. Is there any possibility the model fit would change? In other words, does the sequence of including predictor variables play a significant role in the outcome of the model?

Response: In response to the review team, we have modified the selection methods, excluding proposed models that present inconsistencies between proportionality (sign of their betas coefficients) and theoretical causality related to brain drain. In Appendix B, we have reported the dozens of proposed models (see Table B.1), which are summarized in Table 2. Among these models stated in appendix B, the requirements outlined in the table are covered.

5. Finally, in the light of the incomplete description of predictor variables, the discussion fails to provide deeper insights into possible interpretation of the obtained results. The authors merely repeat the obtained correlations and link them to other research without trying to explain what it means and possibly how to deal with the societal-level shortcomings of the significant determinants of brain drain.

Response: Based on the description of the variables that we have made, we have given a better coverage to the interpretation of the results, discussing these results with the pre-existing literature (discussion section) and giving a better interpretation in the conclusions, indicating the lines of action to address the phenomenon of brain drain.

6. If the authors decide not to develop the idea of citizenship and integration of skilled migrants, then the presented analysis could be further elaborated by comparing and analysing possible differences between the countries of the Global South and Global North.

Response: We have addressed the aspect of citizenship more broadly, incorporating it in the introduction, literature review, discussion and conclusions. We have also pointed out that in future lines of research we have proposed to establish new models and comparisons along the lines indicated.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

Clearly improved manuscript.

One important comment was not adressed (the answer of the authors is actually a copy-past of an answer to another comment). I checked in the manuscript and the comment is not adressed either. 

4. An in-depth discussion of the dependent variable is needed. How is brain drain measured what are the possible bias in the measurement.

This comment should be adressed before publication.

Author Response

C.1: Clearly improved manuscript.

R.1: Dear Reviewer 3, we appreciate your overall assessment of the improvements incorporated in our manuscript.

C.2: One important comment was not adressed (the answer of the authors is actually a copy-past of an answer to another comment). I checked in the manuscript and the comment is not adressed either. 

4. An in-depth discussion of the dependent variable is needed. How is brain drain measured what are the possible bias in the measurement.

This comment should be adressed before publication.

R.2: Dear Reviewer 3, in this new version of the manuscript we have incorporated methodological details regarding the construction of the E3 variable (line 309-314), as well as how this variable incorporates the concept of brain drain (line 316-322). Finally, details on possible biases in the measurement of E3 have been discussed in lines 444-453. All changes have been highlighted in light green.

Reviewer 4 Report

Comments and Suggestions for Authors

The authors addressed all the remarks and provided a more elaborate paper on determinants of brain drain and citizenship – the concept which was practically missing in the previous version. The introduction, theoretical framework, and all other segments of the paper were revised in a more precise manner and contribute to a greater extent to the studied field of migration studies. I suggest the paper is accepted in the present form.

Author Response

C1: The authors addressed all the remarks and provided a more elaborate paper on determinants of brain drain and citizenship – the concept which was practically missing in the previous version. The introduction, theoretical framework, and all other segments of the paper were revised in a more precise manner and contribute to a greater extent to the studied field of migration studies. I suggest the paper is accepted in the present form.

R1: Dear Reviewer 4, we greatly appreciate your appraisal of our manuscript and its contribution to migration studies. Also, we appreciate your suggestion to accept this article in its present form.