Freedom, Happiness and IR: An analysis of the effect of freedom on happiness in countries

Introduction

At first glance, the importance of concepts as freedom and happiness may seem obvious. Something that people is accustomed to have, at least to some extent. However, when subjected to a deeper reflexion, these notions could have the potential to alter the foundations of the basic structures upon which the global system rests.

As political value, freedom is broadly referred to in speeches of Heads of State and arguably integrates one of the core values that conforms the International Society defined by Bull (2012, p. 13). As Human Right, freedom forms part of uncountable constitutions and international treaties. Concerning happiness, although it is also addressed by some domestic and international laws, it is yet to achieve its full potential as an important conductor of public policy. Additionally, as proposed by Penttinen, happiness could also constitute a novel methodology to understand International Relations (2013, pp. 116).

The main purpose of this work is to answer the question: how does freedom affects the level of happiness in countries? and reflect on the implications. Regarding its structure, this assignment will address, first, relevant literature on freedom, happiness, and gender equality to support their importance to IR. Second, this work is going to discuss the methodological issues surrounding this study, and finally, the results of several statistical tests will be presented.

Literature Review

Arguably, one of the most important steps when addressing a theoretical argument is to be able to conceptualize the main terminology involved. Therefore, to better understand the content yet to come within this work, it is essential to discuss what it is to be regarded as freedom, happiness and gender equality.

With reference to happiness, though there are extensive conceptions and research related to it dating back to ancient Greeks -each valid for its specific objective-, for practical reasons, this essay will attach its significance as a synonym of subjective well-being (SWB), in other words, an individual assessment of oneself personal life from a comprehensive perspective (Sarracino, 2012, p. 141). Additionally, the use of subjective well-being as the theoretical concept for happiness will directly add to the operationalization of the happiness variable that will be developed in the next section.

Furthermore, it would serve the case of this work to acknowledge that early significant global endeavours to include happiness as part of an action agenda have indeed been made. Therefore, it is important to mention the outcome of the UN resolution Happiness: towards a holistic approach to development. In this decision, the international community recognized “the importance of the pursuit of happiness and well-being in development with a view to guiding (the Member States’) public policies” (United Nations General Assembly, 2011, p. 1).

As for freedom, one of the possible conceptions could be simply summarized by the principle of having the “possibility to choose”, when there are reasons to suppose that that choice is not being constrained in anyway. However, as this is not possible when living in a structured (international or domestic) society, it would be appropriate to say that freedom is always subjected to degrees (Abdur-Rahman and Veenhoven, 2018, p. 437). In this sense, freedom relies on “the opportunity to choose, and the capability to choose”. The first condition refers to how much a specific context allows or constrains choices, and the second is related to the actual capacity to exercise one of those choices (Veenhoven, 2000, p. 259).

Concerning its importance to the realm of International Relations, Immanuel Kant’s theory of “democratic peace” is based in the concept of freedom. For Kant, the commitment of states to form a “free federation” of liberal democracies will, as a result of the extended use of (liberal) reason, will lead to a “perpetual peace” among states (Kant, 1990, p. 104), or at least among the members of such league. Therefore, in a sense, the hypothesis of Kant that liberal democracies do not go to war with one another contributed as a source of inspiration to the present work. At the end of the day, concepts such as freedom and happiness are intertwined at certain level.

Lastly, for reasons that will become more precise in the methodology section, it is necessary to understand the concept of gender equality as well. Therefore, according to Cornwall and Rivas, when referring to equality in gender relations there is an assumption that women occupy a lower place than men in the hierarchical social structure, and that their relationship is one of an adversarial nature (2015, p. 403). Hence, a gender equality agenda strives to overcome these injustices.

Regarding its applicability to International Relations, there is solid feminist scholarship that analyse how the field and practice of IR is gendered, and that the ongoing international system has relied on gender inequalities to operate (Enloe, 2000, p. 4). Nevertheless, it would be interesting to mention that the concept of gender equality is not well received by everyone. An important critic against it -held by non-Western feminists- is that it constitutes an imposition of “Western liberal societies” (Steans, 2013, p. 13). Thus, one of the theoretical weaknesses of this variable is that it is based upon Western scholarship, making it perhaps not necessarily applicable to the whole international community, whether because of ideology reasons or an actual gap in the realities of both political hemispheres.

Having defined all the pertinent substantive terminology for this essay, it would be paramount to ask ourselves what it is that freedom, happiness and gender equality have in common? Is there a point to hold them together in the same paper? Theoretically, several publications have proposed that there is an apparent strong causal relationship between the variable freedom and the variable happiness. Moreover, societies that value freedom above other important issues such as security, also present high levels of gender equality and tolerance in a broad sense (Inglehart and Welzel, 2005, p. 137). Following this logic, the objective of this assignment is to apply statistical analysis to test the correlation claim between freedom and happiness, and to examine gender equality as a possible intervening variable in that relationship.

Furthermore, considering the power imbalances and instability brought -in part- by the social and economic inequalities that permeate within the international community since economic neoliberalism entered scene (Bhutan, 2012, pp. 10-11), this analysis is particularly relevant for the field of International Relations in the sense that an approach that allocate country happiness at its centre is needed to tackle these concerns. Therefore, aiming to test some of the factors that are apparently related to happiness could develop into a useful tool -or at least a decent one- for the above purpose.

Consequently, to address the quantitative nature of the present work, and after considering the relevant literature above, it would be appropriate to conclude this section with the hypothesis put forward by this essay, namely: the more freedom there is in a country, the better its levels of happiness will be. This assumption is important for those who would want to contribute with the prelude of a tiered agenda focused on the problematic stated in the paragraph above, to attempt to improve it in the long term.

Methodology

Following the purpose of the present essay, it is essential to address the methodological issues that precede the statistical analyses of the subsequent section. Therefore, in terms of operationalization, i. e., the procedure by which theoretical concepts are translated into variables (Whittier, Wildhagen and Gold, 2020, p. 4), the notions of freedom, happiness, and gender equality will constitute variables by the following logic:

As independent variable (IV), that is, the one that provokes variation in another, this work is going to focus on the freedom status variable put forward by the Freedom in the World Report. With data of 2019, this variable is constructed by allocating its categories within a set of three ranges of ratings, which are based upon the general scores of the two main components of country freedom: “political rights and civil liberties”. Hence, ratings from 1 to 2.5 correspond to the “free” category; from 3 to 5 to the “partly free” category; and from 5.5 to 7 to the “not free” category (Freedom House, 2019, pp. 1-5, 18).

Regarding its level of measurement -the quantitative or qualitative form in which the values of a variable can be expressed-, the freedom status variable is nominal, i. e., the answers for each case correspond to categories (Whittier, Wildhagen and Gold, 2020, p. 9). However, it is important to mention that because of the multivariate analysis that will be conducted in the next section, the freedom status variable is going to be recoded -just for that part- into a “dummy” variable. The specifics of this decision are discussed in the relevant segment.

Furthermore, the gender equality variable -originally reported as gender inequality by The Quality of Government Institute-, will be eventually added to the analysis as an additional independent variable to test it as control variable (CV). In other words, it will be held constant to then observe the interaction of the main independent variable (freedom status) and the dependent one (happiness score) (Whittier, Wildhagen and Gold, 2020, pp. 11, 513-515, 520).

Using data of 2019, this (recoded) dichotomous variable -by considering reproductive health, empowerment, and economic factors- categorize countries by the following logic: nations that pose lesser levels of gender inequality are represented by a number 1, and nations where inequalities are extended and profound are expressed by a 0 (Dahlberg et al., 2020, p. 88).

As dependent variable (DV), i. e., the one that is altered by the independent variable (Whittier, Wildhagen and Gold, 2020, p. 23), this essay is going to adopt the happiness score variable disclosed by the World Happiness Report 2019. This variable, using the conception of subjective well-being as a synonym for happiness, is constructed by the national average responses to the Cantril life ladder (a scale of 0 to 10 answered by individuals to assess their life), that are in turn averaged from data between the 2016-2018 period (Helliwell, Layard and Sachs, 2019, p. 21). Consequently, in terms of its level of measurement, as it has numerical values that can be subdivided -due to the fact that they are national-level averages from the period mentioned above-, for the purpose of this work, this variable will be regarded as continuous.

Concerning the validity of the three variables, that is, “the extent to which variables actually measure what they claim to measure” (Whittier, Wildhagen and Gold, 2020, p. 24), it is appropriate to say that, taking into account the fact that freedom status (IV) was defined by the core principles of the Universal Declaration of Human Rights of 1948, this measure is conceptually and pragmatically valid in the sense that -to some extent- the values held by the United Nations (UN) are broadly accepted (Freedom House, 2019, p. 1).

By the same token, gender equality (CV) is based upon the gender inequality index of the UN Development Programme, which apply a gender approach to the three most relevant -and widely recognised- factors of human development (Dahlberg et al., 2020, p. 88). However, as already noted in the literature review section, Wester-centralism critiques may accurately affect the validity of this variable.

Furthermore, respecting happiness score (DV), there is a diverse catalogue of literature that supports its validity in different contexts (Kim-Prieto et al., 2005, p. 265). Additionally, it has the asset of being “democratic”, in the sense that it is each person, in their individuality, the ones who get to reflect and determine if their lives are “worthwhile” (Diener, 2000, p. 34).

In relation to reliability, i. e., “the extent to which responses are consistent” (Whittier, Wildhagen and Gold, 2020, p. 24), the entire (twice-reviewed) process that leads to the freedom status variable is done by a comprehensive team of experts (internal and external of Freedom House) in pertinent fields of knowledge, therefore, the level of the instrument’s complexity is not an issue that could affect the reliability of the measurement.

With respect to the gender equality variable, its reliability strengths and weaknesses have been assessed by several recognised international experts associated with the UN Human Development programme. Furthermore, for its construction, this variable relies on yearly data published by reputable global sources (Gaye et al., 2010, p. 10). Its decent reliability is accepted.

As for the happiness score variable (measured as SWB by the Cantril ladder), both its validity and reliability have been discussed and approved in multiple occasions for prestigious scholars, specifically regarding its use in aggregated levels of analysis (Veenhoven, 1984, cited in Abdur-Rahman and Veenhoven, 2018, p. 442), such as the national measurements used in this work.

To conclude this section, it would be of much value to address the method of analysis that will be executed in the following section. First, to independently understand better the variables included in a quantitative way, an univariate report will be made in order to address measures of central tendency (e. g. medians) were appropriate. Second, a bivariate evaluation will be conducted to analyse the relationship between freedom status (IV) and the happiness score (DV) of countries. Last, this work will direct a multivariate study with recoded “dummy” versions of the original independent variable (including gender equality as control variable), to assess their influence over the dependent variable (Whittier, Wildhagen and Gold, 2020, pp. 649, 651, 655).

Findings[1]

First of all, with the intention of providing a general insight of the independent variable (freedom status), the control variable (gender equality), and the dependent variable (happiness score), this work will effort to provide an univariate description of each. Therefore, from a universe of 143 cases[2] of study (countries), the categories are conformed as follows: free countries: 57 cases with a median happiness score of 6.223. Not free countries: 38 cases with a median happiness score of 4.8055. And partly free countries: 48 cases with a median happiness score of 5.013. The overall median happiness score for the 143 cases is 5.432.

Concerning gender equality, 107 cases are considered to have less gender inequality, with a median happiness score of 5.888, and with major gender inequality problems 36 cases, with a median happiness score of 4.382. Moreover, from the 107 cases that present relatively minor gender inequality, 53 are allocated in the free countries category (median happiness score of 6.293), 23 in the not free category (median happiness score of 5.323), and 31 in the partly free category (median happiness score of 5.386).

Having described each variable individually, before continuing with further analysis, it would be appropriate to visually observe if the expected relationship between the main variables freedom status (IV) and happiness score (DV) is indeed present. (See graph A)

Graph A

Source: Own making in STATA with information of the World Happiness Report,2019, and Freedom in the World, 2019.

From assessing the median happiness scores mentioned earlier along with graph A, it is possible to observe that the group of free countries[3] have significantly higher scores of happiness than the other two categories. Furthermore, it is interesting to notice that there appears to be practically no difference between the categories not free and partly free countries, however, by paying attention to the whiskers of the not free category, it is possible to distinguish more variance in the data of that category. Nevertheless, this first result is decidedly encouraging to continue testing the hypothesis put forward by this essay.

In terms of bivariate analysis, an Analysis of Variance (ANOVA) was selected as the best approach to further assess the relationship between freedom status (IV) and Happiness score (DV) due to the following logic: ANOVA is used for testing hypothesis between one categorical independent variable with 3 or more categories, and a dependent continuous variable (Whittier, Wildhagen and Gold, 2020, p. 440), which is the case in this work.

Consequently, the hypothesis set for this bivariate analysis, namely: there is difference between the happiness score means of the freedom status categories, is going to be tested against the null hypothesis which states that the happiness score means of the freedom status categories are not different. Additionally, the Bonferroni post-hoc test is going to be carried out to find where, within the freedom status categories, that difference is more meaningful (Whittier, Wildhagen and Gold, 2020, pp. 444-447), in order to strengthen the main hypothesis of this work: free countries are happier than not free countries. Alpha is going to be set at the standard value of 0.05.

Table 1

ANOVA

143 cases

Source: Own making in STATA with information of the World Happiness Report,2019, and Freedom in the World, 2019.

Table 2

Comparison of Happiness Score by Freedom Status

(Bonferroni)

143 cases

Source: Own making in STATA with information of the World Happiness Report,2019, and Freedom in the World, 2019.

As it is clear from analysing the p-value of the table 1 ANOVA (0.000 is less than 0.05), it is appropriate to say that there is statistical evidence to reject the null hypothesis. In other words, there is meaningful difference between the means of happiness score (DV) and the freedom status (IV) categories. Furthermore, by looking at the results of the post-hoc test (table 2), it is possible to see that there is no statistically significant difference between the not free and partly free categories (1 is higher than 0.05), but there is between the free countries category and the other two; being this the most relevant piece of information regarding the purpose of this essay.

Encouraged by these results, and considering the complexity of the present theoretical argument, a multivariate regression will be conducted to test whether the level of influence of freedom status (IV) over happiness score (DV) changes after including gender equality (CV) into the model. Moreover, freedom status (IV) and gender equality (CV) have a moderate association (Cramér’s V: 0.3425), therefore, it is possible to work with them.

In order to run the regression, freedom status (IV) had[4] to be recoded as a dichotomous variable, also known as “dummy”, which implicated the elaboration of 3 new variables taking the values 0 and 1, where 1 represent the status (free, not free, or partly free) in each, and 0 the absence of it (Whittier, Wildhagen and Gold, 2020, p. 559). Normally, the third category should be omitted from the recoding, but the present work considered interesting to leave it in. Furthermore, regarding gender equality (CV), it is important to mention that this variable had to be recoded from the original gender inequality variable stablished by The Quality of Government Institute, in the sense that the reference category was shifted from less inequality to more inequality. Therefore, the CV used in this work is focused on equality (countries with less inequality) rather than the original inequality measure.

Table 3

Multivariate Regression freedom status categories without gender equality (CV)

143 cases

Source: Own making in STATA with information of the World Happiness Report,2019, and Freedom in the World, 2019.

From the first set of models (without including the CV) it is possible to observe that, for every one unit change in the free countries category, there is a 1.30 change in happiness score (DV). Id est, there is a strong relationship between them and, furthermore, this relation is statistically significant (p-value of less than 0.05). On the other hand, not free countries have a negative association of -.11 with happiness score. However, both not free and partly free categories are not statistically significant (p-value over 0.05), and present collinearity problems. In terms of model fit, that is, the accurateness with which the independent variables are able to predict the dependent variable (Whittier, Wildhagen and Gold, 2020, p. 557), each report a modest R-squared result of .32. In other words, the models accounted for just 32% of the variation in happiness score.

In the light of the above results, table 4 will only focus on the free countries category and gender equality (CV), as these variables are the more relevant both theoretically and statistically; and in line with the main hypothesis of this work.

Table 4

Multivariate Regression free countries with gender equality (CV)

143 cases

Source: Own making in STATA with information of the World Happiness Report,2019, and Freedom in the World, 2019.

Overall, keeping in mind that this is not proof of causation, from the fourth model, it is pertinent to conclude that gender equality is a potential intervening variable due to the variation observed in the coefficient of free countries after controlling for it (coefficient decreased from 1.30 in model 1, to .93 units in model 4). Furthermore, including this variable improved the model fit (R-squared) from 30% to 51%. In this order of ideas, even though there is relevant statistical evidence that suggest that freedom status (IV) of countries does has an important positive effect over happiness score (DV), it is necessary to acknowledge that gender equality (CV) has an even stronger effect over it. Moreover, satisfactory diagnostic tests were conducted to discard problems of multicollinearity (VIF below 5: see tables 3 and 4) and heteroscedasticity (see graphs Band C in the Appendix section). No issues were found.

Conclusion

In conclusion, given the preliminary findings of the present essay, it is my belief that further attention should be paid to the design of domestic and foreign policy agendas, including freedom, gender equality and happiness as cross-cutting principles.

Considering the broad effects of happiness in the integral performance of nations, it would be of great benefit to construct a new discourse of happiness-based parameters to outline the knowledge, teaching, and practice of IR. Furthermore, considering the potential magnifying effect that freedom and gender equality have on happiness, and following the logic established by the Responsibility to Protect doctrine (United Nations General Assembly, 2009, p. 9), important efforts should be destined by the free and happy members of the international community to assist other nations in developing their gender equality and freedom components.

The findings of the present study should be approached with a healthy scepticism, for they consist only of an early stage of a preliminary analysis that is constrained to be incredibly complex (in theory, statistically and in practice). If I were to have more time, money and human resources, I would re-evaluate the instrument for data collection on the freedom variable for the categories Partly Free and Not Free, as they should reflect, at least according to literature, significant differences in relation to the happiness variable. Moreover, I would focus more on exploring the relation between freedom and gender equality from a historic perspective.

Reference List

Abdur-Rahman, A. and Veenhoven, R. (2018) ‘Freedom and Happiness in Nations: A Research Synthesis’, Applied Research Quality Life, vol. 13(June) pp. 435–456. doi: https://doi-org.ezproxy.is.ed.ac.uk/10.1007/s11482-017-9543-6

Bull, H. (2012) The Anarchical Society: A Study of Order in World Politics. 4th edn. New York: Columbia University Press.

Cornwall, A. and Rivas, A.M. (2015) ‘From ‘gender equality and ‘women’s empowerment’ to global justice: reclaiming a transformative agenda for gender and development’, Third World Quarterly, 36(2), pp. 396–415. DOI: 10.1080/01436597.2015.1013341

Dahlberg, S., Holmber, S., Rothstein, B., Alvarado-Pachon, N. and Axelssion, S (2020) The Quality of Government Basic Dataset. University of Gothenburg: The Quality of Government Institute. Available at:https://www.gu.se/en/quality-government/qog-data/data-downloads/basic-dataset (Accessed: 7 December 2020).

Diener, Ed. (2000) ‘Subjective Well-Being: The Science of Happiness and a Proposal for a National Index’, American Psychologist, vol. 55(1), pp. 34-43. doi: 10.1037//0003-066X.55.1.34

Enloe, C.H. (2000) Bananas, Beaches and Bases: Making Feminist Sense of International Politics. Rev. edn. Berkeley: University of California Press.

Freedom House (2019) Freedom in the World 2019 Methodology. Available at: https://freedomhouse.org/sites/default/files/2020-02/Methodology_FIW_2019_for_website.pdf (Accessed: 4 December 2020).

Gaye, A., Klugman, J., Kovacevic, M., Twigg, S., Zambrano, E. (2010) Measuring Key Disparities in Human Development: The Gender Inequality Index. United Nations Development Programme Research Paper 2010/46. Available at: http://hdr.undp.org/sites/default/files/hdrp_2010_46.pdf (Accessed: 6 December 2020).

Helliwell, J., Layard, R., and Sachs, J. (2019). World Happiness Report 2019. Available at: https://s3.amazonaws.com/happiness-report/2019/WHR19.pdf (Accessed: 4 December 2020).

Inglehart, R. and Welzel, C. (2005) Modernization, Cultural Change, and Democracy: The Human Development Sequence. Cambridge: Cambridge University Press.

Kant, I. (1990) ‘Perpetual Peace: A Philosophical Sketch’, in Reiss, H.S. (ed.) Kant: Political Writings. Translated from the German by Nisbet, H.B. 2nd edn. Cambridge: Cambridge University Press, pp. 93-130.

Kim-prieto, C., Diener, E., Tamir, M., Scollon, C. and Diener, M. (2005) ‘Integrating The Diverse Definitions of Happiness: A Time-Sequential Framework of Subjective Well-Being’, Journal of Happiness Studies, vol. 6(3), pp. 261-300. doi: 10.1007/s10902-005-7226-8

Penttinen, E. (2013) Joy and International Relations: A New Methodology. Oxfordshire, England; New York: Routledge.

Royal Government of Bhutan (2012) The Report of the High-Level Meeting on Wellbeing and Happiness: Defining a New Economic Paradigm. New York: The Permanent Mission of the Kingdom of Bhutan to the United Nations. Available at: https://sustainabledevelopment.un.org/content/documents/617BhutanReport_WEB_F.pdf (Accessed: 20 October 2020).

Sarracino, F. (2012) ‘Money, Sociability and Happiness: Are Developed Countries Doomed to Social Erosion and Unhappiness? Time-Series Analysis of Social Capital and Subjective Well-Being in Western Europe, Australia, Canada and Japan’, Social Indicators Research, vol. 109 (2), pp. 135–188. doi: 10.1007/sl 1205-011-9898-2

Steans, J. (2013) Gender and International Relations Theory, Practice, Policy. 3rd rev. edn. Oxford: Wiley.

United Nations General Assembly (2009) Report of the Secretary-General: Implementing the Responsibility to Protect. A/63/677. Available at: https://www.un.org/ruleoflaw/files/SG_reportA_63_677_en.pdf (Accessed: 6 December 2020).

United Nations General Assembly (2011) Happiness: towards a holistic approach to development. Resolution A/RES/65/309. Available at: https://digitallibrary.un.org/record/715187 (Accessed: 5 December 2020).

Veenhoven, R. (2000) ‘Freedom and Happiness: A Comparative Study in Forty-four Nations in the Early 1990s’, in Diener, E. and Suh, E.M. (eds.) Culture and Subjective Well-Being. Massachusetts: The MIT Press, pp. 257-288.

Whittier, N., Wildhagen, T. and Gold, H.J. (2020) Statistics for Social Understanding: With Stata and SPSS. Maryland: Rowman & Littlefield.

Appendix

Graph B

Heteroscedasticity Diagnostic

(Residuals)

Source: Own making in STATA with information of the World Happiness Report,2019, and Freedom in the World, 2019.

Graph C

Heteroscedasticity Diagnostic

(Residuals)

Source: Own making in STATA with information of the World Happiness Report,2019, and Freedom in the World, 2019.

[1] All the statistical analysis carried out in this section were made through Stata SE/16.1, except for the medians, which were calculated in Excel. [2] The analysis conducted in this section only used cases where data was available for all three variables. [3] The presence of two outliers, identified as India and Botswana, is noted. [4] Since regression only work with numeric values.