Inter- and Intra-Regional Disparities in Russia: Factors of Uneven Economic Growth

Round 1

Reviewer 1 Report

The topic of the research is interesting. I found the manuscript well-structured. The research questions and also contribution of the paper are clear. In my opinion, this research can contribute to the literature by providing insightful information regarding the uneven economic growth with a focus on the municipalities level. Just a minor comment regarding the limitation as well as future directions for research. I recommend the authors move the last paragraph of the discussion section (limitations) to the end of the conclusion section. Moreover, please provide a better clarification of potential directions for research in the future based on your key findings, which have not been properly covered in the current version of the manuscript.

Author Response

Thank you very much for your positive assessment of the study. We would like to provide some clarification on the comments.

I recommend the authors move the last paragraph of the discussion section (limitations) to the end of the conclusion section. Moreover, please provide a better clarification of potential directions for research in the future based on your key findings, which have not been properly covered in the current version of the manuscript.

Dear reviewer, we agree with your comment. In the revised manuscript we provide a better clarification of potential directions for research in the future. We removed the last paragraph of the discussion section (limitations) to the end of the conclusion section (lines 1000-1004).

Our paper is a pilot study for such a big country as Russia, it lays as an initial playground due to few studies on uneven development. Future research should be aimed at identifying the most significant factors in reducing inter-municipal inequality at the regional level to develop a set of policy recommendations provided with financial resources of municipalities, regions, and the federal center (lines 996-1000).

In the future, researchers should expand the list of indicators with spatial lags at the municipal level. In our study, an obvious drawback that prevents the detailed disclosure of spatial patterns is the construction of a model based on a spatial matrix at the level of the region, not municipalities, as well as the inclusion of only spatial errors. The revealed dependencies made it possible to explain part of the variation at the municipal level and, to a lesser extent, the variation attributed to the regional level, which requires expanding the list of factors taken into account at the regional level (lines 1007-1013).

In addition, we would like to note that the new edition of the article also reflects the changes taking into account the comments made by other reviewers. Models 3,4,5,8,9,10 were recalculated according to reviewers comments (tables 3, 4) . We exclude temperature variable and replaced variables of shares import and export in GRD by net export variable.

We hope that the perception of the article has improved in the new edition.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors took up an interesting and important topic related to the causes of uneven economic growth. The research is carried out on the example of the Russian Federation. Research assumptions, the section on previous research, and the section describing Russia's economy and its spatial context should be assessed positively. However, the connection between the performed statistical surveys and the interpretation and discussion of the results, as well as the entire section of the final conclusions, raise great doubts in my opinion and generate a number of questions that have not been answered in this paper. In my opinion, the adopted objectives of the research have not been achieved and I do not see clearly defined by the authors "key factors of uneven economic growth" at the regional level in Russia, which would bring something new and allow to "rethink" the causes of these inequalities and indicate what to do to change this state.

It seems to me that there are difficulties in the proper translation of the text into English. This impression is especially true for the discussion of the results and in the conclusions section. The sentences are often long, the phrases used can be interpreted in various ways and it is not entirely clear what exactly the author is talking about? Also, some of the explanations of an economic nature to the results obtained raise my doubts, because in my opinion they are a simplification or do not take into account various other factors that may determine the obtained results and indicate possible reasons influencing these results. 

Therefore, I think that this version of the paper should be rejected. Below are detailed examples of doubts and errors that determine such an assessment. 

Examples of doubts and errors:

1. Lines 788-791 "We revealed 20% of growth rate variance for social payments and taxable income and 6% for production volume in municipalities attributed to the region. Thus, region contribution depends on variables capturing economic growth, whereas factors that influence variation at various scales are different." In my opinion, this is a description of statistical results, and not an explanation of the relationship between the size of a region's contribution to its economic growth. If we talk about the level of "contribution" of the region, it should be understood how much in the total economic growth depends on regional factors and how much on external factors. This would mean that if a region pays more for social transfers, then its contribution to the economic growth of that region is higher, and vice versa. Unfortunately, in the conducted research we do not know whether social transfers depend on the region at all, and if so, what is the structure of these payments between the region and the central budget. In these studies, we only have an indicator of the share of social payments in the income of the population, and not what is the region's contribution in this regard.
2. Lines 818-826. I agree that the structure of the economy influences economic growth. But... How do I understand the statements "a weaker negative impact" or "a less positive impact"? This style of the results description is for me hard to understand and raises additional questions.  We are scientists and we should prove our theses with facts and figures where possible. The use of phrases like "a weaker negative impact" or "a less positive effect" is in opposition to this, and the impact can be positive, negative or nonexistent. If, in regions with a high share of extractive industries and small businesses, the initial income level has "a weaker" negative impact on economic growth, is there "a stronger" negative impact in other regions, or are we to understand that the impact is positive in regions with different structure? Where are the numbers that prove something is "a weaker" or "a stronger"? In section "Disscussion" are more places with this style of results description.
3. Lines 818-826 - continuation. What does it mean? Do lower incomes have a weaker negative impact on later growth? Or vice versa - higher incomes have a "weaker" negative impact? Do the authors mean that the rate of economic growth is slowing down, or that in these regions some initial income level in general is influencing the real decline of the Gross Regional Product? Or maybe the authors write about diversification inside both groups of regions - with a large share of the mining industry and with a small share of this industry? This style of results description raises a lot more questions for me, and I don't see the answer here. It also says that in regions with a high share of agriculture, investment has less impact on production growth. Should the regional authorities conclude from these results that where there is a lot of agriculture, we should give up investments because they are not very productive? Perhaps the problem is elsewhere - wrong investment direction or investing in ineffective solutions instead of using the best solutions available on the market? Besides, how are we going to change the economic structure of the region if we don't invest in its development? The fundamentals of macroeconomics tell us that investment, along with consumption and public spending, are the main drivers of economic growth, so the question is how to interpret your results in regions with a high share of agriculture? This is beyond the scope of this paper, but the results described in this way cause controversy and further questions. 
4. Lines 827-833. Again, it is not clear what the authors want to say in the text: "Our calculations show that in regions with a higher ratio of total R&D expenditures in GRP in municipalities, the negative impact of the initial production condition on its growth is weaker than in regions with a low ratio of total R&D expenditures in GRP..." Does this mean that the level of regional production grows depending on the increase in R&D spending? Or do the results always indicate a negative impact of R&D on economic growth, and the difference between regions is only whether this negative impact is "weaker" in regions with a higher ratio of total GRP R&D expenditure or "stronger" - in regions with a low share of R&D in GRP? And of course this is a completely opposite conclusion to what is in Sterlacchinni's work. The main conclusion from the article by A. Sterlacchini says (twice quoted in this paragraph - position 36 in bibliography) that the most effective factors influencing the growth of GDP per capita are the percentage of adults with higher education and the intensity of spending on R&D in added value. So it is a positive effect, and the differences between the regions studied by him concern whether this impact is statistically significant or not. Reading the discussion of the results proposed by the authors here, one can read it differently.
5. Lines 875-881 and related with that comments to the methodology. There is text about influnce of export and import to economy growth. In my opinion better factor for such research is taking a value of Net Export (NX).  We know from macroeconomic theory that if NX is positive, it affects economic growth, and if NX is negative, it reduces economic growth in a given area. Researching the relationship between the value of exports alone and economic growth in the regions is not a very good idea, because there may be regions with a relatively low value of exports, but at the same time with lower imports in terms of value - and there you will probably see a positive impact of NX on growth. There may also be regions with a high export value, but an even higher import value - there NX will be negative and its impact will probably limit economic growth in the region. If your research shows that a higher value of exports has a positive impact on the growth of the region, it is probably due to the positive value of NX, and not the value of this export itself. For this reason, I believe it is wrong to accept in your research the relationship between export and import values ​​separately. 
6. Lines 882-897 (about climate and temperature). In my opinion, you misinterpret the results or you should change the methodological assumptions when examining this relationship. From what I have read, the reason for the differences in industrialization and economic growth between the northern and southern regions of Russia does not depend on the average temperature in the region. It depends on the location of the natural resources. In Russia, there are more of these resources in the north and less in the south. That is why economic growth is not due to the temperature in these regions, but to the location of natural resources - especially oil, gas, and coal. The climate and temperature are important for the structure of the region's economy, because it affects the conditions of agricultural production. In the north it is cold and the conditions for agricultural production are worse, and in the south it is warmer and therefore the share of agriculture in the structure of the economy (plus fewer other natural resources than in the north) is greater.
7. Lines 911-916. The same sentence is used twice.
8. Section "Conclusions". In my opinion, your conclusions are not helpful in understanding what is included in the paper title and is described in goals. The wording used are very general and does not answer what decisions to make on the basis of your research results and their interpretation. It is true that the factors that are listed as key can be considered correct, but I do not know whether innovation, human capital, openness of the economy (etc.) should be at a higher level or at a lower level in a given type of region. In the last paragraph, you rightly say that these issues should be taken into account in development strategies and decisions in the regions to reduce inequalities in economic growth. Unfortunately, there are no clear guidelines resulting from research on what decisions are to be made by the regional authorities - whether and when to increase the level of investment and innovation expenditure, or to abandon them, because you claim that in some cases it negatively affects economic growth, etc., etc. 

Author Response

Thank you very much for your kind attitude to our research and the insightful comments to improve it. We want to provide some clarification on the comments.

 The connection between the performed statistical surveys and the interpretation and discussion of the results, as well as the entire section of the final conclusions, raise great doubts in my opinion and generate a number of questions that have not been answered in this paper. In my opinion, the adopted objectives of the research have not been achieved and I do not see clearly defined by the authors "key factors of uneven economic growth" at the regional level in Russia, which would bring something new and allow to "rethink" the causes of these inequalities and indicate what to do to change this state. It seems to me that there are difficulties in the proper translation of the text into English. This impression is especially true for the discussion of the results and in the conclusions section. The sentences are often long, the phrases used can be interpreted in various ways and it is not entirely clear what exactly the author is talking about? 

Dear reviewer, we agree with your comment. We improved our discussion and results sections according to your detailed comments. We should confess that the initial title was too ambiguous, it included "key factors of uneven economic growth" and «rethinking». We have not just identified regional and municipal factors influencing uneven development but have shown their complex interaction. With this in mind, we have changed the title of the article.

Also, some of the explanations of an economic nature to the results obtained raise my doubts, because in my opinion they are a simplification or do not take into account various other factors that may determine the obtained results and indicate possible reasons influencing these results. 

Dear reviewer, we agree with your comment. We tried to make our discussion section clearer by adding explanations to phrases that could be interpreted in various ways. We tried to exclude explanations that contain simplifications or can be read differently. We tried to justify factors used in the model according to the existing body of literature on uneven development. We revealed complex relationships between municipal and regional contexts, which suggested the need for more complex approaches to alleviate the uneven development of territories.

 

Lines 788-791 "We revealed 20% of growth rate variance for social payments and taxable income and 6% for production volume in municipalities attributed to the region. Thus, region contribution depends on variables capturing economic growth, whereas factors that influence variation at various scales are different." In my opinion, this is a description of statistical results, and not an explanation of the relationship between the size of a region's contribution to its economic growth. If we talk about the level of "contribution" of the region, it should be understood how much in the total economic growth depends on regional factors and how much on external factors. This would mean that if a region pays more for social transfers, then its contribution to the economic growth of that region is higher, and vice versa. Unfortunately, in the conducted research we do not know whether social transfers depend on the region at all, and if so, what is the structure of these payments between the region and the central budget. In these studies, we only have an indicator of the share of social payments in the income of the population, and not what is the region's contribution in this regard.

We agree with this comment. It relates to the statistical description. Region contribution is a complex category and requires many factors to be considered. We replace "region contribution" with a more accurate statistical term "variance attributed (related) to regional level". (lines 638, 639, 643)

 

Lines 818-826. I agree that the structure of the economy influences economic growth. But... How do I understand the statements "a weaker negative impact" or "a less positive impact"? This style of the results description is for me hard to understand and raises additional questions.  We are scientists and we should prove our theses with facts and figures where possible. The use of phrases like "a weaker negative impact" or "a less positive effect" is in opposition to this, and the impact can be positive, negative or nonexistent. If, in regions with a high share of extractive industries and small businesses, the initial income level has "a weaker" negative impact on economic growth, is there "a stronger" negative impact in other regions, or are we to understand that the impact is positive in regions with different structure? Where are the numbers that prove something is "a weaker" or "a stronger"? In section "Discussion" are more places with this style of results description.

We agree that our style of results description was not successful for our models including cross-level interaction. We tried to improve it. We precisely denoted slopes of regression line and showed it as dependence of municipal factors’ relationship from regional determinants. We detailed conditions upon which slopes of regression lines could be different. (lines 683-689, 726-727, 811-815, 817-823, 874-882, 889-900, 918-922, 924-927)

 

Lines 818-826 - continuation. What does it mean? Do lower incomes have a weaker negative impact on later growth? Or vice versa - higher incomes have a "weaker" negative impact? Do the authors mean that the rate of economic growth is slowing down, or that in these regions some initial income level in general is influencing the real decline of the Gross Regional Product? Or maybe the authors write about diversification inside both groups of regions - with a large share of the mining industry and with a small share of this industry? This style of results description raises a lot more questions for me, and I don't see the answer here. It also says that in regions with a high share of agriculture, investment has less impact on production growth. Should the regional authorities conclude from these results that where there is a lot of agriculture, we should give up investments because they are not very productive? Perhaps the problem is elsewhere - wrong investment direction or investing in ineffective solutions instead of using the best solutions available on the market? Besides, how are we going to change the economic structure of the region if we don't invest in its development? The fundamentals of macroeconomics tell us that investment, along with consumption and public spending, are the main drivers of economic growth, so the question is how to interpret your results in regions with a high share of agriculture? This is beyond the scope of this paper, but the results described in this way cause controversy and further questions. 

All mentioned questions are associated with unsuccessful results description. We don’t concern matters related to diversification inside groups. We should not give up investments from regions with high agriculture. In our article, we focus on the complex dependencies of factors at the municipal and regional levels that determine the unevenness of development. We showed that slope of regression lines could be different. Hierarchical models describe the relationship between factors and a dependent variable, within which, in addition, each coefficient can be considered as a function of other factors. As a result, the slope of the regression line depends on the sets of factors that vary for different regions; it is stronger for some groups of regions, while for others, it is weaker. We are talking about the fact that the slope of regression line is not static, but changes in a certain range: for some regions it is steeper, for others it is flatter. (lines 764, 766-770, 811-815)

As for recommendations for agriculture, we should be cautious. Rural areas feature specifics not only in Russia. Kilroy A. & Ganau R. (2020) underlined that their results on agriculture "could not be perceived as strongly advised policy recommendations but allowed to know what typically characterized regional economic growth in each type of region." Policymakers should not refuse investments in agricultural regions because of their low productivity. The development of agriculture is the pivotal food security issue, and the primary federal center deals with it, but not regions. But we should take into account this finding and create recommendations to increase the productivity in agriculture. Separately, we note that the slope of the regression line showed that territories with low values of investments located in regions with a high share of agriculture have higher growth rates than those that are located in regions with a low share of agriculture. Accordingly, within certain volumes, investments in regions with a high share of agriculture in GRP contribute to production growth. (lines 827-846)

Lines 827-833. Again, it is not clear what the authors want to say in the text: "Our calculations show that in regions with a higher ratio of total R&D expenditures in GRP in municipalities, the negative impact of the initial production condition on its growth is weaker than in regions with a low ratio of total R&D expenditures in GRP..." Does this mean that the level of regional production grows depending on the increase in R&D spending? Or do the results always indicate a negative impact of R&D on economic growth, and the difference between regions is only whether this negative impact is "weaker" in regions with a higher ratio of total GRP R&D expenditure or "stronger" - in regions with a low share of R&D in GRP? And of course this is a completely opposite conclusion to what is in Sterlacchinni's work. The main conclusion from the article by A. Sterlacchini says (twice quoted in this paragraph - position 36 in bibliography) that the most effective factors influencing the growth of GDP per capita are the percentage of adults with higher education and the intensity of spending on R&D in added value. So it is a positive effect, and the differences between the regions studied by him concern whether this impact is statistically significant or not. Reading the discussion of the results proposed by the authors here, one can read it differently.

 

In this case, we explore the relationship between initial production and its growth. From the macroeconomy, we know that the higher the initial conditions (production), the lower the economic growth. However, we found that the slope angles are different for regions with a high and low share of R&D expenditures in GRP. Moreover, they have different slopes, so the lines describe the relationship between initial production and growth for the regions with a high and low share of R&D expenditures in GRP crossing. They show different effects not only for various levels of initial production but for the different ratios of total R&D expenditures in GRP with the same level of initial production. (lines 848-856)

We refer to Sterlacchini A. and Cappelen A. et al. because they suggested that in diverse groups of regions, the relationship between R&D expenditures in GRP and economic growth manifests itself in different ways.

Sterlacchini A. revealed that for some regions, the relationship is statistically significant, but not for others. Cappelen A. et al. find that the R&D variable could be both negative or positive in different groups. In the first edition, we have combined their findings. In the revised manuscript, we have made an explanation so that the result of each of them is more visible. (lines 859-863)

  

Lines 875-881 and related with that comments to the methodology. There is text about influence of export and import to economy growth. In my opinion better factor for such research is taking a value of Net Export (NX).  We know from macroeconomic theory that if NX is positive, it affects economic growth, and if NX is negative, it reduces economic growth in a given area. Researching the relationship between the value of exports alone and economic growth in the regions is not a very good idea, because there may be regions with a relatively low value of exports, but at the same time with lower imports in terms of value - and there you will probably see a positive impact of NX on growth. There may also be regions with a high export value, but an even higher import value - there NX will be negative and its impact will probably limit economic growth in the region. If your research shows that a higher value of exports has a positive impact on the growth of the region, it is probably due to the positive value of NX, and not the value of this export itself. For this reason, I believe it is wrong to accept in your research the relationship between export and import values ​​separately. 

 

We agree with the comment. Including exports and imports per GRP, we took into account only their volume relative to the regional economies but did not consider the relationship between the volumes of imports and exports. Therefore, we removed them. The peculiarities of our function forced us to add Net Export (NX) as a dummy variable (1 is a positive Net Export value, 0 is a negative Net Export value), because in a number of regions the net export value was negative, which causes a logarithm problem. In this model specification, the Net Export influence is accounted for through dummy.  Dummy is measured through the exp (Koeff * dummy) multiplier for the entire model equation. There are questions about the coefficient at dummy obtained in the model, but since it differs insignificantly from zero, and therefore, the multiplier is equal to one, we decided not to focus on this and left it. (lines 354-356, 932-935)

 

Lines 882-897 (about climate and temperature). In my opinion, you misinterpret the results or you should change the methodological assumptions when examining this relationship. From what I have read, the reason for the differences in industrialization and economic growth between the northern and southern regions of Russia does not depend on the average temperature in the region. It depends on the location of the natural resources. In Russia, there are more of these resources in the north and less in the south. That is why economic growth is not due to the temperature in these regions, but to the location of natural resources - especially oil, gas, and coal. The climate and temperature are important for the structure of the region's economy, because it affects the conditions of agricultural production. In the north it is cold and the conditions for agricultural production are worse, and in the south it is warmer and therefore the share of agriculture in the structure of the economy (plus fewer other natural resources than in the north) is greater.

 

We agree with this comment. We deleted the temperature variable from models. Models 3,4,5,8,9,10 were recalculated (tables 3, 4).

 

Lines 911-916. The same sentence is used twice.

We agree with this comment. We deleted the excessive sentence.

 

Section "Conclusions". In my opinion, your conclusions are not helpful in understanding what is included in the paper title and is described in goals. The wording used are very general and does not answer what decisions to make on the basis of your research results and their interpretation. It is true that the factors that are listed as key can be considered correct, but I do not know whether innovation, human capital, openness of the economy (etc.) should be at a higher level or at a lower level in a given type of region. In the last paragraph, you rightly say that these issues should be taken into account in development strategies and decisions in the regions to reduce inequalities in economic growth. Unfortunately, there are no clear guidelines resulting from research on what decisions are to be made by the regional authorities - whether and when to increase the level of investment and innovation expenditure, or to abandon them, because you claim that in some cases it negatively affects economic growth, etc., etc. 

Our study is a pilot for analysis at this level of aggregation in Russia. The paper shows that only 20% and 6% of the variation is attributed to the region, respectively, for income and production growth. Therefore, the key highlight is the variation at the municipal level. The interaction of regional and municipal factors affects this variation.  Yes, we found that the slopes in the regions are different, but returning to politics, we draw attention to the fact that the article does not consider the groups of regions (for example, development clubs in Europe, coastal and central part in China). We analyze 85 regions, each of which has its combination of factors at the regional level that complicates the formation of common policy recommendations. We model the growth of 2239 municipalities, and within the framework of the article; it is rather difficult to show specific solutions for each of them. This study shows how complex is the existing uneven development in terms of scales, factors, and spatial effects. Based on the graphs and coefficients for each municipality as a whole, it is possible to designate a list of recommendations. But the publication format is limited and does not allow this. At the same time, we hope that the new version of the description of the results will make it clear to the reader what decisions the government to make, whether and when to increase factors that in some cases it negatively affects economic growth.

In addition, we would like to note that the new edition of the article also reflects the changes taking into account the comments made by other reviewers. We hope that the perception of the article has improved in the new edition.

Author Response File: Author Response.pdf

Reviewer 3 Report

Dear authors,

First of all, let me congratulate you for the theme you have chosen and for your research!

I suggest you the following:

1. In the Results section you should have only the analysis results, with no connections and personal interpretation. Please remove in the Discussion section all the consequences of the results (for instance, lines 695-697).

2. I would like to suggest a more accurate writing. For instance, lines107-112 should be compressed in one section - which should present the article structure. Also, the title of 2.2. should be Factors influencing the economic growth.

 

Author Response

Thank you very much for your kind attitude to our research and the insightful comments to improve it. We want to provide some clarification on the comments.

In the Results section you should have only the analysis results, with no connections and personal interpretation. Please remove in the Discussion section all the consequences of the results (for instance, lines 695-697).

Dear reviewer, we agree with your comment. We deleted lines 695-697 in the manuscript. But another reviewer asked to describe slopes in details that is why it may seem like discussion in Results section. But, please, kindly note it is not like this.  We just tried to show the variety and complexity of these relationships. The Discussion section encapsulates the interpretation of results.

I would like to suggest a more accurate writing. For instance, lines107-112 should be compressed in one section - which should present the article structure.

Dear reviewer, we agree with your comment. We compressed lines107-112 in one section (lines 105-110).

Also, the title of 2.2. should be Factors influencing the economic growth.

Dear reviewer, we agree with your comment. We improved the title of the Section 2.2. (lines 186)

In addition, we would like to note that the new edition of the article also reflects the changes taking into account the comments made by other reviewers. Models 3,4,5,8,9,10 were recalculated according to reviewers comments (tables 3, 4). We exclude temperature variable and replaced variables of shares import and export in GRP by net export variable.

We hope that the perception of the article has improved in the new edition.

Author Response File: Author Response.pdf

Reviewer 4 Report

The research presented in the article is interesting and very relevant. The results are clear and practical. The test method used is appropriate. The results obtained are widely described. Great article. Success in further research.

Author Response

Thank you very much for your positive assessment of the study. But we would like to note that the article was changed taking into account the comments made by other reviewers. Models 3,4,5,8,9,10 were recalculated  (tables 3, 4) . We exclude temperature variable and replaced variables of shares import and export in GRD by net export variable.
We hope that the perception of the article has improved in the new edition.

Round 2

Reviewer 2 Report

Dear Authors!
Thank you for your comments and discussion with my reviews. I am glad about revised version of your paper and accept it in a present form. I hope that our common work help to better understand  your results and conclusions of the research for all potential readers.

There is only one very small thing to improve. In line 935 they are used 2 words in Russian cirilical form ("Таким образом"). I think you should change it to English to mainten equal style of a entire paper:)

I wish you further successes in building your scientific career and kindly regards. 

Author Response

Dear reviewer! Thank you very much for reaching out to improve our paper. Your comments are invaluable and insightful! Our joint work, undoubtedly, provides a better understanding of our results and conclusion. We appreciate your time and efforts invested in our study. We translated "Таким образом" in line 921. 

Warm and kind regards from the Russian authors)

Author Response File: Author Response.docx