Measuring Equitable Prosperity in the EU-27: Introducing the IDDO, a Composite Index of Growth and Income Inequality (2005–2024)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is a very good and important paper which deserves to be published. The topic, the identification of a methodological gap, the approach taken to fill this gap and the clarity of the text all contribute. There is a very clear objective and focus of the paper, some very interesting results are presented and with potentially valuable conclusions. A few detailed comments are made in the following, but only as possible suggestions, except for comments 4 and 5 which do need to be addressed for the purpose of full methodological clarity, but should be easily done:

1) Lines 64-70: I noted some other relevant typologies and indices not mentioned, but then saw you have referenced them both in section 2.2 and in section 6. The only other relevant approach you do not mention might be Raworth’s doughnut which attempts to include environmental sustainability as well as economic and social issues. However, although this is a useful conceptual source it is more difficult to apply in practice to the contexts you are addressing, i.e. it needs detailed data collection and local analysis. Thus, my only suggestion for section 1’s introduction would be to signpost very briefly that there are limitations that which will be described later.

2) Section 2, before subsection 2.1, is also very good, so all I would suggest perhaps is to add the argument made in the “Spirit Level” book from Wilkinson and Pickett (2009), maybe also some of the later evidence building on their approach. Basically, this shows that societies with wider income inequalities experience worse outcomes across various indicators, including mental health, physical health, violence, and social mobility, and that – importantly – this also adversely impacts the wealthiest individuals so all lose. In other words, we need an approach (and indices) that could change the idea of a “tension between growth and equity” to something much closer to a “complementary relationship between growth and equity” in a sort of virtuous circle. (Maybe also include impacts on nature as a third element – see comment 6 below).

3) Lines 99-101: is there a source for the contention that “certain forms of inequality may actually stimulate investment and innovation—provided that strong institutions are in place”? I think this is correct and is a very important point, but it would be useful to give the source and, if possible, also specify in what circumstances, which type of inequalities and which type of institutions.

4) Line 169, and later: apart from the fact that this is dictated by the data available, what are the pros and cons of excluding pensions from disposal income used in the Gini coefficient data? In principle, wouldn’t it be better to include pension income, given that this is a very important component, often the only component, of the income pensioners receive? Assuming pensioners are included in the survey data (as I presume they are?) but their pensions are excluded, doesn’t this skew the data and thus the results? Perhaps this is not an issue given that all countries' pension data is excluded and the purpose of your analysis is to compare countries with each other? It is noteworthy that there is a huge amount of pensioner poverty (or “risk of poverty”) in the EU and that this is, on average, higher than for the working age population.

5) Section 3.2.3: The logic of creating a typology with four types (categories) is sound, but it is not clear how the actual types and their ranges are derived and defined. With the two dimensions of growth and equity, it’s easy to create four types when each dimension can be said to be ‘high’ or ‘low’. But you have also included a ‘moderate’ category which, if used consistently would lead to nine categories. I wonder therefore whether this is deliberate or whether type III should be labelled “low growth & high equity” rather than “low growth & moderate equity”? The logic of what you present is not clear here. Perhaps it is based upon an examination of the sequence of data shown in Table 3, but again this is not explained. Neither is there an explanation for the IDDO ranges? This might be based on an examination of the sequence of Table 3’s data, but this is not clear? Indeed, looking at Figure 2, there are only very small jumps (0.02 and 0,.01) between the actual data of Types I, II and III, whilst the gap between Types III and IV is larger at 0.04 (i.e. between Poland and Estonia). Neither is this explained by an attempt to have a similar number of countries in each type. For example, Type II has six countries and Type III has eleven countries. Using the “eyeball” approach (of course only a very rough method!), it would seem more sensible to include Finland, Cyprus and maybe Denmark in Type II, thereby approximately standardising the number of countries between the two types, also taking account of the size of the data gap between the two types. Indeed, if this was the method, there should be about 7 countries in each type, though adjusted according to the sequence of data, in an attempt to take account of significant gaps in the data sequence. Summa summarum and notwithstanding the useful comments in lines 389-402, it would be very useful for the actual methods used to be explained and justified, i.e. how and why are the four types derived, what sort of ‘clustering’ technique has been used, etc.?

6) Both sections 5 and 6 are excellent and well justified based on the methodology and the results produced. It might be useful to re-emphasise, perhaps in one sentence, that the IDDO values for individual countries is just a gateway to an analysis of the balance between growth and equity for the purposes of policy development. It’s useful to analyse both how and why existing developments and policies have produced the IDDO results and what these mean for future policy design. In relation to lines 457-464, it would be great to see this approach, notwithstanding the caveats you mention, extended to include the third dimension of the environment. There are possible candidates, as you mention, and this could be extremely useful in the context of the three pillars of the SDGs and how/if these need updating for whatever follows the fifteen year 2015 Paris Agreement once 2030 is reached.  

Author Response

General Comment: This is a very good and important paper which deserves to be published. The topic, the identification of a methodological gap, the approach taken to fill this gap and the clarity of the text all contribute. There is a very clear objective and focus of the paper, some very interesting results are presented and with potentially valuable conclusions. A few detailed comments are made in the following, but only as possible suggestions, except for comments 4 and 5 which do need to be addressed for the purpose of full methodological clarity, but should be easily done.

Response:
We are deeply grateful for your generous and encouraging assessment. Your recognition of the paper’s objectives, methodological contribution, and clarity is truly appreciated. We have carefully addressed all the comments provided, giving particular attention to Comments 4 and 5 as requested. The suggestions have significantly contributed to refining the manuscript and improving its conceptual and empirical clarity. Thank you once again for your thoughtful review and support.

Comment 1: Lines 64–70: I noted some other relevant typologies and indices not mentioned, but then saw you have referenced them both in section 2.2 and in section 6. The only other relevant approach you do not mention might be Raworth’s doughnut which attempts to include environmental sustainability as well as economic and social issues. However, although this is a useful conceptual source it is more difficult to apply in practice to the contexts you are addressing, i.e. it needs detailed data collection and local analysis. Thus, my only suggestion for section 1’s introduction would be to signpost very briefly that there are limitations that which will be described later.

Response:
We sincerely thank the reviewer for this thoughtful suggestion. We fully agree that Raworth’s “doughnut economics” is a highly valuable conceptual framework that integrates environmental sustainability with economic and social concerns. However, as rightly noted, its application requires localized and multidimensional datasets that are not readily available across all EU member states in a harmonized format. To acknowledge this limitation while maintaining the scope of the current study, we have inserted the following clarifying sentence at the end of the relevant paragraph in the Introduction (lines 76–80): “While this study focuses on the economic and distributive dimensions of development, broader frameworks such as Raworth’s ‘doughnut economics’ also emphasize environmental sustainability (Raworth, 2017). These alternative models, however, require localized and multidimensional data that are not available for a consistent EU-wide analysis.”
We hope this addition sufficiently addresses the reviewer’s concern and clarifies the study’s intended focus.

Comment 2: Section 2, before subsection 2.1, is also very good, so all I would suggest perhaps is to add the argument made in the “Spirit Level” book from Wilkinson and Pickett (2009), maybe also some of the later evidence building on their approach. Basically, this shows that societies with wider income inequalities experience worse outcomes across various indicators, including mental health, physical health, violence, and social mobility, and that – importantly – this also adversely impacts the wealthiest individuals so all lose. In other words, we need an approach (and indices) that could change the idea of a “tension between growth and equity” to something much closer to a “complementary relationship between growth and equity” in a sort of virtuous circle.

Response:
We are grateful for this insightful recommendation, which highlights a critical dimension of the equity-growth relationship. To incorporate this valuable perspective, we have added a new sentence before subsection 2.1 (lines 112–118), summarizing key findings from The Spirit Level (Wilkinson & Pickett, 2010) and subsequent studies. This sentence reads: “Research by Wilkinson and Pickett (2010), and subsequent empirical studies, shows that wider income inequality correlates with worse outcomes across a range of indicators including physical and mental health, education, violence, and social mobility. Notably, such inequality affects all socioeconomic groups, including the wealthiest. These insights support a paradigm shift: from viewing growth and equity as being in tension to understanding them as mutually reinforcing dimensions of inclusive development.”

We believe this addition strengthens the theoretical motivation of the index and aligns well with the paper’s broader normative rationale.

Comment 3: Lines 99–101: is there a source for the contention that “certain forms of inequality may actually stimulate investment and innovation—provided that strong institutions are in place”? I think this is correct and is a very important point, but it would be useful to give the source and, if possible, also specify in what circumstances, which type of inequalities and which type of institutions.

Response:
We appreciate this thoughtful comment and fully agree that this is a nuanced and important point. In response, we have added relevant citations and clarified the nature of the claim by inserting the following sentence in lines 103–107: “Some researchers have argued that certain forms of inequality can stimulate investment and innovation, provided that inclusive and well-functioning institutions are in place (Galor & Moav, 2004; Acemoglu & Robinson, 2012). This argument highlights the complex interplay between distribution, incentives, and governance.”

We believe this addition strengthens the theoretical grounding of the statement and aligns it with the broader literature on institutions and development. 

Comment 4: Line 169, and later: apart from the fact that this is dictated by the data available, what are the pros and cons of excluding pensions from disposal income used in the Gini coefficient data? In principle, wouldn’t it be better to include pension income, given that this is a very important component, often the only component, of the income pensioners receive? Assuming pensioners are included in the survey data (as I presume they are?) but their pensions are excluded, doesn’t this skew the data and thus the results? Perhaps this is not an issue given that all countries' pension data is excluded and the purpose of your analysis is to compare countries with each other? It is noteworthy that there is a huge amount of pensioner poverty (or “risk of poverty”) in the EU and that this is, on average, higher than for the working age population.

Response:
We are especially grateful for this important methodological observation. In response, we have added a new clarifying paragraph in lines 191–198 to address the rationale behind using the “Gini before social transfers” measure, which excludes pensions. The following sentence was introduced: “Regarding the Gini coefficient used in this analysis, we rely on the Eurostat (2021) measure ‘before social transfers,’ which explicitly excludes pensions by treating them as part of income rather than social benefits. While this selection omits a major income component—especially for pensioners—it allows us to focus on pre-redistribution inequality and better assess structural income disparities. Pensioners are included in the survey population, but pensions are considered part of personal income, not transfers. This approach upholds cross-country comparability, aligning with our objective of analyzing fundamental inequalities before welfare mechanisms intervene.”

We hope this clarification demonstrates our awareness of the issue and justifies the analytical choice within the study’s comparative framework.

Comment 5: Section 3.2.3: The logic of creating a typology with four types (categories) is sound, but it is not clear how the actual types and their ranges are derived and defined. With the two dimensions of growth and equity, it’s easy to create four types when each dimension can be said to be ‘high’ or ‘low’. But you have also included a ‘moderate’ category which, if used consistently would lead to nine categories. I wonder therefore whether this is deliberate or whether type III should be labelled “low growth & high equity” rather than “low growth & moderate equity”? The logic of what you present is not clear here. Perhaps it is based upon an examination of the sequence of data shown in Table 3, but again this is not explained. Neither is there an explanation for the IDDO ranges? This might be based on an examination of the sequence of Table 3’s data, but this is not clear? Indeed, looking at Figure 2, there are only very small jumps (0.02 and 0,.01) between the actual data of Types I, II and III, whilst the gap between Types III and IV is larger at 0.04 (i.e. between Poland and Estonia). Neither is this explained by an attempt to have a similar number of countries in each type. For example, Type II has six countries and Type III has eleven countries. Using the “eyeball” approach (of course only a very rough method!), it would seem more sensible to include Finland, Cyprus and maybe Denmark in Type II, thereby approximately standardising the number of countries between the two types, also taking account of the size of the data gap between the two types. Indeed, if this was the method, there should be about 7 countries in each type, though adjusted according to the sequence of data, in an attempt to take account of significant gaps in the data sequence. Summa summarum and notwithstanding the useful comments in lines 389-402, it would be very useful for the actual methods used to be explained and justified, i.e. how and why are the four types derived, what sort of ‘clustering’ technique has been used, etc.?

Response:
We are sincerely grateful for this extensive and thoughtful comment. In response, we have made two clarifications:
(1) We revised the text in Section 3.2.3 (lines 316–322) to explain the rationale behind the IDDO thresholds and the use of a heuristic, non-cluster-based typology. The following passage was added: “The four types used for interpretation (Types I–IV) were constructed based on an empirical reading of the IDDO score distribution over time and the need to balance clarity with analytical utility. Thresholds were chosen to reflect natural gaps in the data and to ensure internal consistency with the observed trajectories of GDP_norm and Gini_norm across countries. We did not apply clustering techniques such as k-means due to the limited number of cases (27) and the preference for an interpretive typology, consistent with practices in composite index design (OECD, 2008).”

(2) Additionally, Figure 2 was revised to better illustrate the IDDO distribution and to reflect these groupings more transparently.

We hope these revisions fully clarify the method behind the typology and the underlying rationale.

Comment 6: Both sections 5 and 6 are excellent and well justified based on the methodology and the results produced. It might be useful to re-emphasise, perhaps in one sentence, that the IDDO values for individual countries is just a gateway to an analysis of the balance between growth and equity for the purposes of policy development. It’s useful to analyse both how and why existing developments and policies have produced the IDDO results and what these mean for future policy design. In relation to lines 457-464, it would be great to see this approach, notwithstanding the caveats you mention, extended to include the third dimension of the environment. There are possible candidates, as you mention, and this could be extremely useful in the context of the three pillars of the SDGs and how/if this need updating for whatever follows the fifteen-year 2015 Paris Agreement once 2030 is reached.  

Response:
We greatly appreciate the reviewer’s positive assessment of Sections 5 and 6. In response to the suggestion, we have added a clarifying sentence at the end of the Conclusion (lines 525–528), emphasizing that the IDDO should be seen as an entry point for broader policy evaluation. The inserted sentence reads as follows: “In this sense, IDDO serves not only as an indicator of developmental balance but as a starting point for policy analysis and design. Future iterations could include environmental sustainability as a third dimension, aligning with the tripartite framework of the SDGs.”

We believe this addition strengthens the policy relevance of our conclusions and acknowledges the importance of integrating environmental aspects in future research.

 

We once again thank the reviewer for their thoughtful and constructive feedback. In addition to the revisions prompted directly by your suggestions, we have implemented further refinements based on the comments received from the other reviewers. These include clarifications of methodological aspects, refinements to the typological framework, and the integration of additional bibliographic and informational sources. These references were selected to support both the conceptual points raised in your review and the broader enhancements made throughout the manuscript. We sincerely hope that these efforts have resulted in a stronger and more coherent version of the paper.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This paper constructs an “IDDO” index based on GDP per capita and Gini coefficient. I think this paper can be improved in several aspects. I state my detailed comments and suggestions below.

 

Comments and Suggestions:

 

1. Based on the results presented in Table 3, we can see that when calculating the GDP_norm indicator, the GDP per capita of each country is expressed as a relative value compared to that of the EU27 in the same year (with the EU27 value set as 100 for each year). However, this approach has a significant problem: the GDP_norm values are only meaningful for comparisons within the same year. In other words, GDP_norm values cannot be used for cross-year comparisons. For example, Table 3 shows that Bulgaria’s GDP per capita is the lowest among all sample countries in every year, resulting in a GDP_norm value of 0 across all years. Yet, Bulgaria’s actual GDP per capita may have undergone substantial changes from 2005 to 2024. The GDP_norm values (consistently 0 for all years) fail to reflect these changes in GDP per capita over time. I think it would be more reasonable to calculate GDP_norm using actual GDP per capita adjusted for inflation and purchasing power parity (PPP). This adjustment would allow for meaningful comparisons both across regions and across different time periods. I suggest that the authors carefully reconsider how to compute GDP_norm in a more appropriate manner.

 

2. Line 225 on Page 6: “reprezintă” should be “represents”.

 

3. Line 268 on Page 7: What is the meaning of “și”? Please check.

 

4. Figure 1 on Page 9: The numbers on the vertical axis are written as “0,00”, “0,10”, “0,20” ... Those should be written as “0.00”, “0.10”, “0.20” ...

 

5. Section 4.2 on Page 10: Based on the IDDO score, different countries are classified into four types: Type I – High Balanced Development, Type II – Moderate and Stable Progress, Type III – Convergence with Inequality, and Type IV – High Distributive Imbalance. There are two problems. (1) The names of four types here are completely different from the names used in Table 2. (2) IDDO score is the average of GDP_norm and Gini_norm. Based on the IDDO score, we cannot infer the specific values of GDP_norm and Gini_norm. Therefore, based on the IDDO score, it is impossible to determine whether the growth is high or low, or whether the equity level is high or low. For example, consider a country with an IDDO value of 0.55. It is possible that its GDP_norm is 1 while its Gini_norm is 0.1. This country has a high growth but low equality, indicating a not balanced development scenario. In this case, the country should neither be labeled as “High Equity & High Growth” (the name given to Type I in Table 2) nor classified as “High Balanced Development” (the name given to Type I in Section 4.2). Fundamentally, I cannot understand the rationale behind these four typological classifications.

 

6. Figure 2 on Page 11: The meaning of these rectangles in the figure is unclear. For example, the authors should explain to readers how the length and width of these rectangles are determined and what they represent.

Author Response

Your observations have significantly contributed to enhancing the methodological clarity, analytical transparency, and overall quality of our work. We greatly appreciate both your critical insights and your constructive suggestions. Below, we present our point-by-point responses to each of your comments. All revisions are clearly marked in red in the revised version of the manuscript.

Comment 1: Based on the results presented in Table 3, we can see that when calculating the GDP_norm indicator, the GDP per capita of each country is expressed as a relative value compared to that of the EU27 in the same year (with the EU27 value set as 100 for each year). However, this approach has a significant problem: the GDP_norm values are only meaningful for comparisons within the same year. In other words, GDP_norm values cannot be used for cross-year comparisons. For example, Table 3 shows that Bulgaria’s GDP per capita is the lowest among all sample countries in every year, resulting in a GDP_norm value of 0 across all years. Yet, Bulgaria’s actual GDP per capita may have undergone substantial changes from 2005 to 2024. The GDP_norm values (consistently 0 for all years) fail to reflect these changes in GDP per capita over time. I think it would be more reasonable to calculate GDP_norm using actual GDP per capita adjusted for inflation and purchasing power parity (PPP). This adjustment would allow for meaningful comparisons both across regions and across different time periods. I suggest that the authors carefully reconsider how to compute GDP_norm in a more appropriate manner.

Response:
We thank the reviewer for this insightful observation. We agree that annual normalization of GDP per capita relative to the EU27 average limits the comparability of GDP_norm values across different years. In response, we have added a clarifying paragraph (lines 241–263) explicitly acknowledging this limitation and justifying the rationale behind this methodological choice. Specifically, we highlight that yearly normalization controls for EU-wide shocks and emphasizes relative positioning, which aligns with the purpose of IDDO as a distributive benchmarking tool rather than a temporal growth measure. Moreover, to address concerns about underlying growth dynamics, we refer readers to the raw GDP per capita values already included in Table 3, which complement the normalized scores. We hope this dual perspective enhances transparency and analytical utility.

Comment 2: Line 225 on Page 6: “reprezintă” should be “represents”.

Response:
We thank the reviewer for catching this language oversight. The term “reprezintă” was indeed a residue from the original language draft. It has now been correctly revised to “represents” in the English version of the manuscript.

Comment 3: Line 268 on Page 7: What is the meaning of “și”? Please check.

Response:
We appreciate the reviewer’s careful reading. Similar to the previous point, “și” was a remnant from the original language draft and has been corrected to the appropriate English conjunction (“and”). This has been verified throughout the manuscript to prevent recurrence.

Comment 4: Figure 1 on Page 9: The numbers on the vertical axis are written as “0,00”, “0,10”, “0,20” ... Those should be written as “0.00”, “0.10”, “0.20” ...

Response:
Thank you for drawing our attention to this formatting issue. We have corrected the decimal separators in Figure 1 to conform with international (English-language) conventions, replacing commas with dots (e.g., “0.00”, “0.10”, “0.20”, etc.).

Comment 5: Section 4.2 on Page 10: Based on the IDDO score, different countries are classified into four types: Type I – High Balanced Development, Type II – Moderate and Stable Progress, Type III – Convergence with Inequality, and Type IV – High Distributive Imbalance. There are two problems. (1) The names of four types here are completely different from the names used in Table 2. (2) IDDO score is the average of GDP_norm and Gini_norm. Based on the IDDO score, we cannot infer the specific values of GDP_norm and Gini_norm. Therefore, based on the IDDO score, it is impossible to determine whether the growth is high or low, or whether the equity level is high or low. For example, consider a country with an IDDO value of 0.55. It is possible that its GDP_norm is 1 while its Gini_norm is 0.1. This country has a high growth but low equality, indicating a not balanced development scenario. In this case, the country should neither be labeled as “High Equity & High Growth” (the name given to Type I in Table 2) nor classified as “High Balanced Development” (the name given to Type I in Section 4.2). Fundamentally, I cannot understand the rationale behind these four typological classifications.

Response:
We are sincerely grateful to the reviewer for highlighting two key issues related to the typological classification of countries in Section 4.2.

First, regarding the inconsistency in type names, we fully agree that alignment is essential for clarity. We have accordingly revised the terminology in Section 4.2 to match exactly the labels used in Table 2, ensuring coherence throughout the manuscript.

Second, we recognize the importance of clarifying the interpretive logic of the typology. As suggested, we have added a dedicated paragraph (lines 373–382) explaining that the four types are heuristic categories derived from an empirical reading of the IDDO distribution. The classification does not imply symmetric strength in both components (GDP_norm and Gini_norm), but rather reflects the overall balance captured by the composite index. We also acknowledge that the IDDO, being an average, may obscure extreme values in one dimension, which is why we caution readers to treat the types as interpretable tendencies, not as deterministic categories.

Furthermore, to visually support these clarifications, we have revised Figure 2, simplified its structure and made its construction and typological thresholds more transparent.

Comment 6: Figure 2 on Page 11: The meaning of these rectangles in the figure is unclear. For example, the authors should explain to readers how the length and width of these rectangles are determined and what they represent.

Response:
We thank the reviewer for the thoughtful observation regarding the interpretation of Figure 2.

In the revised version, we have replaced the previous visual format with a simplified representation based on the average IDDO scores across three reference years: 2005, 2014, and 2024. This average provides a synthetic overview of long-term developmental patterns and allows for a stable basis of typological comparison.

We have eliminated the original rectangular shapes to avoid ambiguity, and the current figure now displays a single horizontal line with countries grouped by type according to their average IDDO score. We also added a clarification in the caption and in Section 4.2, explicitly stating that the figure is meant to illustrate the empirical thresholds used in the typology, not to convey additional quantitative information through area or shape.

We hope these revisions improve both the readability and interpretive clarity of the classification system.

In addition to the revisions made in direct response to your thoughtful comments, we have also undertaken further clarifications throughout the manuscript to address observations raised by the other reviewers. Several new bibliographic sources have been integrated to support the methodological and conceptual refinements introduced.

We thank you once again for your valuable contribution to improving our manuscript. We have also incorporated additional refinements and relevant literature, including bibliographic support for the updated sections, as part of the broader revision process inspired by the feedback from both reviewers. We hope that these changes fully address your concerns and suggestions.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have revised the paper. I have no further comments.