Inequality and Life Satisfaction in Low- and Middle-Income Countries: The Role of Opportunity

Round 1

Reviewer 1 Report

The article addresses an important issue for developing countries: the role of opportunities and their moderating effect in the relationship between income inequality and life satisfaction. 

Over the last years, a flourishing literature has addressed the relationship between income inequality and SWB in different countries. I feel that the authors could benefit from reading 

Schneider, S. M. (2016). Income inequality and subjective wellbeing: Trends, challenges, and research directions. Journal of Happiness Studies, 17(4), 1719-1739.

for a review. The literature proposed at the beginning of the article is incomplete and not very precise (e.g. developing countries are not a new focus in the literature, see also 

Veenhoven, R. (1990). Inequality in happiness: inequality in countries compared across countries.). In many occasions, the references quoted do not correspond with the statements either because of the sample of countries or because of the strong positioning taken by the authors. Please downsize the innovative connotation given to this paper and be more precise in the references.

In line 43-44, the authors suggest that they will test the tunnel hypothesis. I expected this test in the empirical application and I was disappointed to see that they have a cross-sectional sample with no panel data. How can the authors test the rapid development of societies with static data? Please correct this by either using panel data (e.g. more waves of the WVS) or by changing the positioning of the paper. Please note also that all the papers in the literature that use panel fixed-effects and have a sample of developing countries implicitly test the tunnel hypothesis.

I do not understand why Ravazzini &Chavez-Juarez would have suggested to use objective measures of opportunities when their measure was already objective. With reference to line 74-75, I would say that the first attempt to test how income inequality and opportunities interfere in life satisfaction was Ravazzini &Chavez-Juarez. What the authors could say is that they particularly focus on developing countries, whereas Ravazzini & Chavez-Juarez focused on Europe.

In the literature review on page 3, I do not see a big difference between the relative deprivation argument and the influence of income on the relationship between income inequality and SWB (lines 141-145), please clarify this point.

Please restructure the arguments in the theoretical framework. The paragraph in lines 114-119 is very similar to the paragraph in lines 168-176. Please read also:

Schneider, S. M. (2019). Why Income Inequality Is Dissatisfying—Perceptions of Social Status and the Inequality-Satisfaction Link in Europe. European Sociological Review.

The interpretation given to the results of Ravazzini & Chavez-Juarez in lines 150-158 is wrong, please read the article again. The relationship between income inequality and the income quintiles is not considered by the authors as the risk of downward mobility, but the intensity of the possible loss of income that rich individuals would have. The authors propose a mix of arguments to explain their results and namely status anxiety, distaste for inequality and negative externalities linked to inequality, please clarify this point (line 152). It is true that the measure of inequality of opportunity is constructed on the basis of micro data, but the outcome is macro: 1 observation of inequality of opportunity per year per country, please be more precise.

I the theoretical framework, which is not a theoretical framework, but a literature review, the authors must explain the concept they want to use. I believe that many authors on social mobility would criticize the approach taken by the authors. Where is mobility? Mobility must have a starting and an ending point (e.g. inequality of opportunity is a measure that compares the characteristics of individuals at birth -including characteristics of their parents- with outcome measures of these individuals during their adult life). With a static cross-sectional analysis and the measure of opportunity used in the Social Progress Index, what the authors have in this paper is very far from the concept of social mobility. Please reposition all your statements linked to social mobility (see also line 267). In relation to this (Line 370), I’m not sure that the Opportunity index you use measures if individuals have the chance to reach better social position. The index is a measure of fairness on different spheres, but it contains no indicator of the fairness of the labour market.

Figure 1 and 2, please display only the 25 countries that you use later in your regressions.

Table 1, please use the same number of digits everywhere.

I have problems with the measures of social trust and social fairness used by the authors. First, they go in opposite directions, second, why are they at the macro level? Please give an explanation for this, as they are simple controls of perceptions not particularly explained in the results, why not leaving them at the micro level?

At the micro level, your variables are tailored based on developed countries. Do you have some indicators more specific to developing countries? Whether the person works in the formal or informal economy, whether the person has a legal contract, the right to social security or paid holidays would be for instance good indicators that you could add.

In table 1, please provide also a model with only the opportunity index.

I’m very sceptical of the results presented in Table 3. The models look overcrowded with triple interactions and only 25 countries. Figure 4 is also very suspicious with flat lines for the reference categories in countries with low opportunities. Moreover, the dichotomisation between low and high opportunity countries is arbitrary and not well explained (how many countries are in each group? Can we still run models on them?). I suggest the authors should drop this part.

What are the scores for each component of the opportunity index at the country level? Maybe your results could be explained by a higher variability across countries in inclusiveness and access to advanced education compared to personal rights and personal freedom.

Please correct or justify two statements in the conclusions:

-lines 469-471, your paper is not the first to use an objective measure of opportunity and the effect of inequality you find is positive and significant per se

- line 497-498, do you evidence to support the statement that international institutions to not target inclusiveness and access to advanced education? (please be more cautious and see the Sustainable Development Goals, target 4 is quality education, target 10 reduced inequality, target 11 makes cities human settlements inclusive, resilient and sustainable).

Author Response

Inequality and Life Satisfaction in Middle-Income Countries: The Role of Opportunity

Submitted to Societies (societies-489264)

"Response to Reviewers"

Answers to Reviewer #1

We would like to thank you for your comments and suggestions. In light of your comments we have made several changes, which in our opinion have notably contributed to improving the manuscript. Furthermore, in the revised manuscript we have also introduced some changes required by Referee #2 (in particular a section devoted to robustness checks).

We have also modified the title because our sample is not only related to Middle-Income countries. The new title is “Inequality and Life Satisfaction in Low- and Middle-Income Countries: The Role of Opportunity”.

The text has been revised by a professional proofreader/translator.

We have provided a much richer manuscript that hopefully has greatly improved our research and we are extremely grateful for your valuable comments.

In what follows, we address your specific comments, so that we hope to allay any doubts you may have concerning our research (the referee’s comments are in italics).

The article addresses an important issue for developing countries: the role of opportunities and their moderating effect in the relationship between income inequality and life satisfaction.

Thank you very much for your general appraisal of our analysis. We do believe that the paper tackles an interesting topic and could offer a contribution to the literature.

1 Over the last years, a flourishing literature has addressed the relationship between income inequality and SWB in different countries. I feel that the authors could benefit from reading

Schneider, S. M. (2016). Income inequality and subjective wellbeing: Trends, challenges, and research directions. Journal of Happiness Studies, 17(4), 1719-1739.

for a review. The literature proposed at the beginning of the article is incomplete and not very precise (e.g. developing countries are not a new focus in the literature, see also

Veenhoven, R. (1990). Inequality in happiness: inequality in countries compared across countries.). In many occasions, the references quoted do not correspond with the statements either because of the sample of countries or because of the strong positioning taken by the authors. Please downsize the innovative connotation given to this paper and be more precise in the references.

Thank you for this appreciation of our work. We have made an effort to highlight the contribution of this manuscript to the literature.

First, we have included both suggested references in the literature review. Second, we have carefully reviewed the statements. In the last paragraph of literature review, we have included a reference to Schneider (2016) and Schneider (2012) that advocate for studies that do not follow a strict economic rationale when studying the Inequality-SWB nexus. We also add a reference to Veenhoven (1990), who suggests that level of happiness and inequality in happiness may be explained by indicators of social inequality such as ‘socially produced difference in life chances’, which are broader than the mere inequality in income. We also included references to Haller and Hadler (2004 and 2006) who point out the influence of democracy and political freedom in explaining SWB.

Indeed, the literature review included in the Introduction was incomplete because it was further developed in Section 2. We have shortened and modified the introduction, taking into account the above comments and comments 2 and 8. Sentences in lines 36-39 have been replaced by more precise sentences taking into account comment 4.

2 In line 43-44, the authors suggest that they will test the tunnel hypothesis. I expected this test in the empirical application and I was disappointed to see that they have a cross-sectional sample with no panel data. How can the authors test the rapid development of societies with static data? Please correct this by either using panel data (e.g. more waves of the WVS) or by changing the positioning of the paper. Please note also that all the papers in the literature that use panel fixed-effects and have a sample of developing countries implicitly test the tunnel hypothesis.

Thank you for this comment. Indeed, our sample is not a panel dataset since we are skeptical about pooling different waves from the WVS that do not rely on same individuals. Therefore, as stated in your comments, we are not able to test the Tunnel hypothesis. Instead, we can only conjecture that individuals leaving in countries at different levels of development would have on average different tolerance towards inequality. To avoid confusion and make the overview of the literature included in the introduction clearer, we have removed sentences in lines 43-49 of the original manuscript because these studies are detailed in the literature review.

3 I do not understand why Ravazzini &Chavez-Juarez would have suggested to use objective measures of opportunities when their measure was already objective. With reference to line 74-75, I would say that the first attempt to test how income inequality and opportunities interfere in life satisfaction was Ravazzini &Chavez-Juarez. What the authors could say is that they particularly focus on developing countries, whereas Ravazzini & Chavez-Juarez focused on Europe.

Thank you for pointing out this issue. We have corrected the mistake and suppressed the sentence in lines 63-65 of the original manuscript. Instead, we have added two new sentences to justify the need to take into account social inequalities (lines 69-72 of the new manuscript). Following your suggestion we have also replaced sentence in lines 74-77 of the original manuscript with “This study tackles a similar issue to Ravazzini and Chávez-Juárez [20] for Europe but for a sample of developing countries. Here, we test how differences in social opportunities between countries explain why individuals living in more unequal countries would be more satisfied”.

4 In the literature review on page 3, I do not see a big difference between the relative deprivation argument and the influence of income on the relationship between income inequality and SWB (lines 141-145), please clarify this point.

Thank you for pointing out this issue. We think you are right. Both are two approaches of the same concept: one from economics (relative income hypothesis) and another from Sociology (relative deprivation effect). Accordingly, we have joined together the first and second paragraphs (respectively, lines 93-104 and 105-114 in the original manuscript) in a sole paragraph (90-112 in the new manuscript).

The following sentence (lines 141-145 in the original manuscript) and the references it included have been removed:  “The permissiveness towards inequality may not only depend on the macroeconomic contexts but also on individuals’ situation. Inequality aversion may be shaped by individuals’ characteristics regarding income, regions, ethnicity, and culture [41], and social preferences [42].”

5 Please restructure the arguments in the theoretical framework. The paragraph in lines 114-119 is very similar to the paragraph in lines 168-176. Please read also:

Schneider, S. M. (2019). Why Income Inequality Is Dissatisfying—Perceptions of Social Status and the Inequality-Satisfaction Link in Europe. European Sociological Review.

Thank you for this useful suggestion. Accordingly, we have combined both paragraphs (respectively, lines 115-124 and 170-177 in the original manuscript) in a sole paragraph (113-124 in the new manuscript). We have eliminated the references 31-40 (original list) and included the ones of Wilkinson and Pickett (2010) and Schneider (2019).

In order to introduce social mobility into the debate, we have substituted lines 140-144 with the paragraph (143-149 in the new manuscript)-

We have attempted to improve the writing in the rest of the paragraphs (151-161 of the new manuscript).

6 The interpretation given to the results of Ravazzini & Chavez-Juarez in lines 150-158 is wrong, please read the article again.

The relationship between income inequality and the income quintiles is not considered by the authors as the risk of downward mobility, but the intensity of the possible loss of income that rich individuals would have. The authors propose a mix of arguments to explain their results and namely status anxiety, distaste for inequality and negative externalities linked to inequality, please clarify this point (line 152). It is true that the measure of inequality of opportunity is constructed on the basis of micro data, but the outcome is macro: 1 observation of inequality of opportunity per year per country, please be more precise.

Thank you for drawing our attention to these aspects. We have made an effort to make a better description of the results of Ravazzini & Chavez-Juarez (lines 150-158 in the original manuscript). We have also decided that it would be better to move this reference to another paragraph devoted to the studies that directly examine the role of actual social mobility in the association between inequality and SWB (166-174 in the new manuscript).

To stick to the main topic of the paper, we have removed the references to Graafland and Lous [19] and Graafland and Compen [49], which are more related to the link between economic freedom and life satisfaction.

7 I the theoretical framework, which is not a theoretical framework, but a literature review, the authors must explain the concept they want to use.

Thank you for pointing this out. We have changed the title of section 2 (originally “2. Theoretical framework: Income inequality and subjective wellbeing” by “2. Literature review: Income inequality and subjective wellbeing” .

In this section, we focus on the main theoretical arguments for a negative relationship between Income inequality and subjective wellbeing and their empirical validations. Secondly we present studies that explain the circumstances under which this relationship may be negative: namely macro circumstances and individuals’ perceptions of the mobility of their society. We finish with a paragraph justifying the need for paying attention to the influence of a broader social context instead of taking into account the economic or labor market contexts. The literature tends to show that individuals’ perception of social mobility is not the whole story.

8 I believe that many authors on social mobility would criticize the approach taken by the authors. Where is mobility? Mobility must have a starting and an ending point (e.g. inequality of opportunity is a measure that compares the characteristics of individuals at birth -including characteristics of their parents- with outcome measures of these individuals during their adult life). With a static cross-sectional analysis and the measure of opportunity used in the Social Progress Index, what the authors have in this paper is very far from the concept of social mobility. Please reposition all your statements linked to social mobility (see also line 267). In relation to this (Line 370), I’m not sure that the Opportunity index you use measures if individuals have the chance to reach better social position. The index is a measure of fairness on different spheres, but it contains no indicator of the fairness of the labour market.

Thank you for this suggestion. The Opportunity index used in this study does not reflect social mobility. Social mobility and perceptions of social mobility may explain the link between income inequality and SWB but inequality of opportunities is a broader context. As stated by Ravazzini and Chávez-Juárez (2006) on p6, “a high level of inequality of opportunity is generally associated with low levels of social mobility”. As suggested by this reviewer and in line with Veenhoven (1990) among others, we justify the use of measure of opportunity used in the Social Progress Index as an indicator of social inequalities. We have tackled this issue in two ways:

1.     In the literature review (lines 180-201), we have justified the need for taking into account different facets of social inequalities to fully assess the relationship between income inequalities and SWB.

2.     In the section 3.2.3 Contextual variables, we have eliminated paragraph 256-269 in the original manuscript, and substitute it for lines 257-266 in the revised version in order to justify the use of the Opportunity Index.

9 Figure 1 and 2, please display only the 25 countries that you use later in your regressions.

Table 1, please use the same number of digits everywhere.

Both figures have been changed in this sense and, now we have used the same number of digits everywhere.

10 I have problems with the measures of social trust and social fairness used by the authors. First, they go in opposite directions, second, why are they at the macro level? Please give an explanation for this, as they are simple controls of perceptions not particularly explained in the results, why not leaving them at the micro level?

Thank you for raising this issue. The use of macro variables describing social trust and social fairness of countries is suggested by Bjornskov  et al. [11] and Grosfeld and Senik [15], and we have decided maintain them in the regressions. They provide an additional control of the macro contexts.  Since they are not significant, we haven’t comment on the signs but we expected individuals living in countries with higher levels of trust and fairness to be more satisfied.

However, following the reviewer’s suggestion, we have also included these variables at the micro level. The description of the new variables has been added to Table A.4 Definition of variables. This reduces the size of the sample due to missing data. Equally, we expected individuals displaying higher levels of trust and fairness to be more satisfied. The first variable fails to be significant and the second one displays the expected positive sign (line 318 of the new manuscript).

11 At the micro level, your variables are tailored based on developed countries. Do you have some indicators more specific to developing countries? Whether the person works in the formal or informal economy, whether the person has a legal contract, the right to social security or paid holidays would be for instance good indicators that you could add.

This is an interesting suggestion. Unfortunately, such indicators describing individual work situations are not available in the WVS. In turn, we have included three additional variables to account for possible extreme economic situations that may be common in middle and low-income countries: frequency without enough food, frequency without cash and possibility of saving money. Including these additional individuals’ characteristics do not change the results of the key macro variables. We comment on the results in lines 322-324. The description of the new variables has been added to Table A.4 Definition of variables.

12 In table 1, please provide also a model with only the opportunity index.

Thank you for commenting on this issue. Heeding to your suggestion, we now report a new model with only the opportunity index (model 3 in Table 1). The main results are robust to this specification.  

13 I’m very sceptical of the results presented in Table 3. The models look overcrowded with triple interactions and only 25 countries. Figure 4 is also very suspicious with flat lines for the reference categories in countries with low opportunities. Moreover, the dichotomisation between low and high opportunity countries is arbitrary and not well explained (how many countries are in each group? Can we still run models on them?). I suggest the authors should drop this part.

In this regard, and as suggested in your review, we have eliminated section 4.5. Heterogeneity among socioeconomic groups that made the reading of the paper more difficult and the objective more diffuse.

14 What are the scores for each component of the opportunity index at the country level? Maybe your results could be explained by a higher variability across countries in inclusiveness and access to advanced education compared to personal rights and personal freedom.

Thank you for your advice. The scores for each component of the opportunity index have been included in Table A.3. Components of the Opportunity index by countries (in Appendix). In addition, in the bottom rows of this table we have reported the mean and standard deviation of the macro variables. As can be observed, variabilities of each component of the opportunity index are not very different. This table is mentioned in lines 384-385.

15 Please correct or justify two statements in the conclusions:

-lines 469-471, your paper is not the first to use an objective measure of opportunity and the effect of inequality you find is positive and significant per se

Thank you for your correction. We have modified sentences 469-471 by “Our main contribution consists in corroborating that social opportunities interfere in this relationship” (lines 429-430), and lines 472-473 by “Opportunity and inequality exert significant effects per se on life satisfaction. These effects are positive when considered separately and negative when considered jointly. More interestingly,…”. Corresponding changes have been made in the abstract and introduction.

16 - line 497-498, do you evidence to support the statement that international institutions to not target inclusiveness and access to advanced education? (please be more cautious and see the Sustainable Development Goals, target 4 is quality education, target 10 reduced inequality, target 11 makes cities human settlements inclusive, resilient and sustainable).

Thank you for suggesting this reference. We have hedged our statements and substitute the sentence in lines 493-495 with lines 453-458.

Reviewer 2 Report

The paper is written well with a nice flow. The literature overview is very good.  Please check when you first introduce abbreviations if the full name is given in brackets. E.g. page 1, LS is introduced, but the full name is not given. I did not see checking of assumptions of the model (autocorrelation, etc.) - thus this should be added to the results wich comments - is this why you used robust errors? Which procedure was used for robust errors? Page 6 you cite in row 254 "several authors..." - please add literature here. Robustness of the results should be checked as well.

Author Response

Inequality and Life Satisfaction in Middle-Income Countries: The Role of Opportunity

Submitted to Societies (societies-489264)

"Response to Reviewers"

Answer to Reviewer #2

We would like to thank you for your comments and suggestions. In light of your comments we have made several changes, which in our opinion have notably contributed to improving the manuscript. Furthermore, in the revised manuscript we have also introduced some changes required by Referee #1, in particular we have worked on the justification of the use of the Opportunity Index, included other control variables, included model 2 in Table 1 and eliminated section 4.5. Heterogeneity among socioeconomic groups, which made the reading of the paper more difficult and the objective more diffuse.

We have also modified the title because our sample is not only related to Middle-Income countries. The new title is “Inequality and Life Satisfaction in Low- and Middle-Income Countries: The Role of Opportunity”.

The text has been revised by a professional proofreader/translator.

We have provided a much richer manuscript that hopefully has greatly improved our research and we are extremely grateful for your valuable comments.

In what follows, we address your specific comments, so that we hope to allay any doubts you may have concerning our research (the referee’s comments are in italics).

The paper is written well with a nice flow. The literature overview is very good

Thank you very much for your kind comments.

Please check when you first introduce abbreviations if the full name is given in brackets. E.g. page 1, LS is introduced, but the full name is not given.

Thank you for your advice. We have taken it into account.

I did not see checking of assumptions of the model (autocorrelation, etc.) - thus this should be added to the results wich comments –

is this why you used robust errors?

Which procedure was used for robust errors?

Thank you for commenting on this issue. We have used robust standard errors because variances of the residual errors were different among countries (Levene’s tests statistic=37.1, p=0.00). Robust standard errors have been calculated using the White estimate of variance. We have included a comment in the text (section 4) explaining it.

Page 6 you cite in row 254 "several authors..." - please add literature here.

Thank you for your correction. Indeed, in the section 3.2.3 Contextual variables, we have eliminated the paragraph 256-269 in the original manuscript, and substitute it for lines 257-264 in the revised version in order to justify the use of the Opportunity Index, which is a broader context than social mobility, more related with social inequalities. Likewise, in the literature review (lines 181-201), we have justified the need for taking into account different facets of social inequalities to fully assess the relationship between income inequalities and SWB.

Robustness of the results should be checked as well.

Thank you for this suggestion. We have included a new section with robustness checks (lines 412-424). To check the robutness of our results we have repeated the estimation of Model 4 (Table 1) splitting the sample successively into: people with low (no formal or primary) versus high (secondary or universitary) education, people with low (below median) versus high (above median) household income and people with low (lower/lower middle/working) versus high (upper/upper middle class) subjective social status. The results are presented in Table 3 and corroborate the relationship found between LS, inequality and opportunity.

Round 2

Reviewer 1 Report

This version has largely improved, and I would like to congratulate the authors for the big work they have done in such a short amount of time. The literature review is now more grounded, and the arguments are clearly explained. There are however still some few points that need to be tackled in the other sections of this paper:

Figure 1 and 2 display more countries than the ones used in the analysis. Please create also a set of figures with the same countries used in the regressions.

You mention that the countries have been surveyed in the years 2010-2013. This is a quite large timeframe considering also that there was a global crisis in the previous years. Please mention the precise year of the survey for each country of your study.  

Table 1: use “lower class” as reference category so that the signs follow the coefficients for income and education.

Please perform a robustness check where you exclude social trust and social fairness both at the micro and at the macro level. Do your results hold?

In the limitations of your study, please mention that you use only one year and that future research could study how the relationship between income inequality, opportunity and SWB changes in a dynamic setting and particularly when low and middle-income country improve their social conditions (which might mean an improvement in opportunities, but a polarisation of incomes and therefore an increase in income inequality).

Please correct the following statements:

Page 12: “Our finding corroborates a not-yet validated intuition formulated by other authors [i.e. 1, 19, 36, 50] according to which social mobility interferes in the relationship between inequality and LS.”

-as previously mentioned, with your data, you can talk about opportunities, but not directly about social mobility. The sentence is ambiguous, but I think you should state clearly that you are also unable to validate this intuition properly. Please be more precise.  

Page 14: “Finally, once opportunity at the macro level is accounted for, we do not find clear evidence that the relationship between income inequality and life satisfaction differs between income groups. People with no basic education would behave differently from people having reached at least the primary level who display a relatively more indifferent attitude towards inequality.”

-can you still claim this?

Page 15: “people are sensitive to social indicators such as inclusiveness and access to advanced education, aspects that are usually not in the priority of international institutions’ agenda.”

-you correctly mention the sustainable development goals, but you still claim that international institutions do not set the right priorities. I believe that critical statements should be justified so that the actual policy makers can do something about this. Can you justify this strong statement with some examples? Do you really want to address yourself to international organisations or more to national or local governments? Why should some goals be a priority for policy makers who live in countries where people are satisfied despite their poor outcomes?   

Author Response

Inequality and Life Satisfaction in Low- and Middle-Income Countries: The Role of Opportunity

Submitted to Societies (societies-489264)

"Response to Reviewers"

Answers to Reviewer #1

We would like to thank you for your quick revision.

In light of your comments, we have made several changes, which we hope have improved the article. Below, we address each of your suggestions (the referee’s comments are in italics). Changes are highlighted in the new manuscript using the track changes option.

This version has largely improved, and I would like to congratulate the authors for the big work they have done in such a short amount of time. The literature review is now more grounded, and the arguments are clearly explained. There are however still some few points that need to be tackled in the other sections of this paper:

We are extremely grateful for your valuable suggestions and your general appraisal of our revision. We have mentioned it in the acknowledgments.

Figure 1 and 2 display more countries than the ones used in the analysis. Please create also a set of figures with the same countries used in the regressions.

Both figures have been changed in this regard.

You mention that the countries have been surveyed in the years 2010-2013. This is a quite large timeframe considering also that there was a global crisis in the previous years. Please mention the precise year of the survey for each country of your study. 

Thank you for this comment. Following your suggestion, in Table A.2 we have included the year of the survey for each country of our study.

Table 1: use “lower class” as reference category so that the signs follow the coefficients for income and education.

We have made this change (see Table 1) and changed the comment accordingly in line 316.

Please perform a robustness check where you exclude social trust and social fairness both at the micro and at the macro level. Do your results hold?

In this regard, and as suggested in your review, we have incorporated a new table in the appendix (Table A.6), displaying the same models as Table 1 but excluding variables describing social trust and social fairness of countries and individuals. The main results hold. We have included a reference to these results in line 302.

In the limitations of your study, please mention that you use only one year and that future research could study how the relationship between income inequality, opportunity and SWB changes in a dynamic setting and particularly when low and middle-income country improve their social conditions (which might mean an improvement in opportunities, but a polarisation of incomes and therefore an increase in income inequality).

Thank you for your suggestion. At line 475 we have included the following sentences: “Further work needs to explore the soundness of our results in a more comprehensive panel. Indeed, dynamic patterns may be crucial in emerging countries where social conditions, including opportunities, evolve quickly while income inequality may temporarily increase.“

Please correct the following statements:

Page 12: “Our finding corroborates a not-yet validated intuition formulated by other authors [i.e. 1, 19, 36, 50] according to which social mobility interferes in the relationship between inequality and LS.”

-as previously mentioned, with your data, you can talk about opportunities, but not directly about social mobility. The sentence is ambiguous, but I think you should state clearly that you are also unable to validate this intuition properly. Please be more precise. 

Yes, this is a mistake. We have corrected two sentences in this section.

Line 339 “To test if respondents regard income inequality as a signal of (future) upward social mobility, we include the opportunity index in the model”, has been substituted by “To test if respondents are affected by social inequalities in a broader sense, we include the opportunity index in the model”.

Lines 356 “Our finding corroborates a not-yet validated intuition formulated by other authors [i.e. 1, 19, 36, 50] according to which social mobility interferes in the relationship between inequality and LS”. ”, has been substituted by “Our finding is in line with the intuition formulated by several authors [i.e. 2, 11. 19, 37, 39, 50] according to which SWB may be influenced by inequality in “life chances” and not only by income inequality.”

Page 14: “Finally, once opportunity at the macro level is accounted for, we do not find clear evidence that the relationship between income inequality and life satisfaction differs between income groups. People with no basic education would behave differently from people having reached at least the primary level who display a relatively more indifferent attitude towards inequality.”

-can you still claim this?

Based on the new robustness checks, we have corrected this statement. Lines 442-443 have been substituted by: Finally, once opportunity at the macro level is accounted for, we do not find clear evidence that the relationship between income inequality and life satisfaction differs between low- and high-income groups, low and high socio-economic status and educated and less educated people.

Page 15: “people are sensitive to social indicators such as inclusiveness and access to advanced education, aspects that are usually not in the priority of international institutions’ agenda.”

-you correctly mention the sustainable development goals, but you still claim that international institutions do not set the right priorities. I believe that critical statements should be justified so that the actual policy makers can do something about this. Can you justify this strong statement with some examples? Do you really want to address yourself to international organisations or more to national or local governments?  

Thank you for pointing out this issue. We agree that international institutions are more aware of the importance of other social inequalities while national governments in developing or emerging countries, which are often non-democratic and not always concerned by the life satisfaction of the population are, at best, merely concerned by economic outcomes.

Accordingly, we have substituted lines 464-463, “On the contrary, people are sensitive to social indicators such as inclusiveness and access to advanced education, aspects that are usually not in the priority of international institutions’ agenda.“ for: “On the contrary, people are sensitive to social indicators such as inclusiveness and access to advanced education, aspects in which low- and middle-income countries usually stand far behind the high-income countries. For instance, in 2014 the average opportunity index was 72.54, 51.83, 41.38 for high-, middle- and low-income countries, respectively.[1]

Why should some goals be a priority for policy makers who live in countries where people are satisfied despite their poor outcomes?

Based on our results, part of the answer is that poor outcomes in terms of inequality do not matter so much if the performance in social inequalities is able to compensate. However, according to the Tunnel hypothesis, this situation may only be transitory.

Line 468 onward, we have modified the sentence: “Nevertheless, when opportunities are low, more redistribution and improvement of life chances for all would definitively prove useful for increasing SWB.”

We have also added (line 471 et seq.):  “In countries with a medium and high level of opportunities, inequality of income seems relatively irrelevant to SWB, all else being constant. This result should be interpreted cautiously. According to the Tunnel effect [4, 15-17], this optimistic perception could vanish in a second step if, despite more opportunities, income inequalities do not subside.” This is in line with the already mentioned suggestion that further research is needed to consider a dynamic setting.  

We have also corrected two typesets in lines 166 and 168.

[1] These averages were computed on 18, 20 and 16 high-, middle- and low-income countries, respectively, included in the last wave of the WVS. Differences among groups are all significant at 1%.