Comparing Wellbeing Economy with other OECD Nations: A Predictive Analysis of Subjective Well-Being Score using SDGs

doi:10.21203/rs.3.rs-5294452/v1

Download PDF

Article

Comparing Wellbeing Economy with other OECD Nations: A Predictive Analysis of Subjective Well-Being Score using SDGs

https://doi.org/10.21203/rs.3.rs-5294452/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Focusing on the difference among Wellbeing Economy governments (WEGo), WEGo Hub and non-WEGo OECD countries, this study investigates the relationship between Sustainable Development Goals (SDGs) and subjective well-being score (Gallup's World Poll data on life evaluations used for annual happiness rankings for countries). First, after confirming normality, Analysis of Variance (ANOVA) was conducted to compare the means of the groups and with a p-value of 0.00108, a highly statistically significant difference between the groups was confirmed; the variation in subjective well-being scores is unlikely to have occurred by chance and WEGo members had a higher mean.

Next, based on the Shapiro-Wilk normality test, Kruskal-Wallis rank sum test was conducted instead of ANOVA to compare median scores of the overall SDG goal score for the three groups: with a p-value (0.4492), it failed to reject the null hypothesis: There was no statistically significant difference between the groups being compared. However, WEGo countries still had higher minimum, mean, median, and maximum scores than the other two groups in 2024. WEGo countries had 11 out of 17 SDG goal mean and median scores higher than non-WEGO OECD countries, out of which 3 were statistically significant: Goal 3 (Good Health and Well-Being), Goal 5 (Gender Equality), and Goal 11 (Sustainable Cities and Communities).

WEGo countries have statistically significant differences from non-WEGO countries in both subjective well-being scores and Sustainable Development Goal 3 (Good Health and Well-Being) scores. Would they also have different results when it comes to predicting the subjective well-being of their citizens? Predictive models were trained and tested to extract and assess variable importance results using data from 2018–2024 for WEGo, WEGo Hub, and non-WEGo OECD countries. The Extreme Gradient Boosting model results showcased that only 3 SDG indicators were present in all three models: ‘n_sdg8_unemp’ (Unemployment rate), ‘n_sdg1_lmicpov’ (Poverty headcount ratio at $3.65/day), and ‘n_sdg16_rsf’ (Press Freedom Index).

For WEGo countries, SDG indicator ‘n_sdg2_obesity’ (The percentage of the adult population that has a body mass index of 30kg/m² or higher) emerged as the strongest predictor, followed by SDG indicator ‘n_sdg3_matmort’ (The estimated number of women who die from pregnancy-related causes) and ‘n_sdg7_renewcon’ (The share of renewable energy such as wind/solar in the total final energy consumption). For WEGo Hub countries in the OECD, ‘n_sdg8_unemp’ (Unemployment rate % of total labor force), ‘n_sdg12_explastic’ (The average annual amount of plastic waste exported), and ‘n_sdg16_admin’ (Timeliness of administrative proceedings) had significant coefficients. For Non-WEGo OECD countries, SDG indicators ‘n_sdg8_rights’ (Rating whether fundamental labor rights are effectively guaranteed), ‘n_sdg12_pollimp’ (Air pollution associated with imports), and ‘n_sdg3_traffic’ (Traffic deaths per 100,000 population) stood out.

While WEGo country model had 5 SDG indicators for Goal 3 (Good Health and Well-being), WEGo hub country model and Non-WEGo country model had 2 each. Findings contribute to understanding how SDG performance correlates to subjective well-being with variable importance results varying depending on the OECD country’s association to the Wellbeing Economy initiative.

The 2030 Agenda for Sustainable Development - “A shared blueprint for peace and prosperity for people and the planet, now and into the future” - was adopted by all 195 United Nations (UN) member states in 2015, but according to the UN, “the promises enshrined in the SDGs are in peril” (Geldard and Ellerbeck, 2023). In 2023, more than 50% of SDG targets had weak, insufficient progress and 30% had stalled or reversed (Butler, 2023).

5% of countries are estimated to achieve SDG goals by 2030 on only 11 of the 17 indicators (Weeks et al.,2023), but only 41 countries registered national commitments to SDG transformation at the SDG Summit (UN, 2023). This is problematic since the UN Sustainable Development Goals are a “vital component of the happiness” and development of countries around the world (World Happiness Report 2023). However, it is hard to shift from a Gross Domestic Product (GDP) growth-driven economy to a beyond-GDP paradigm even with criticisms raised by the Nobel Prize winner Joe Stiglitz and other scientists who have warned that GDP does not equate to success (Hoekstra, 2020).

Due to its prevalence today, it is easy to forget that GDP was invented out of necessity after the Wall Street crash of 1929 by the economist Simon Kuznets. Policymakers needed a reference to grasp the economic strength of the nation during the national emergency, and during World War 2, the same need surfaced across the globe - resulting in its confirmation as the main tool for measuring global economies in 1944 at the Bretton Woods conference. “For many markets and policymakers, GDP growth has become an all-consuming metric (Vanham, 2021)”, yet GDP is not a measure of the well-being of a country (Callen). “GDP’s shortcomings have become especially obvious in its failure to account for inequality (Coyle, 2017).” According to the United Nations, income inequality between countries has improved, yet income inequality within countries has become worse with 71% of the world’s population living in countries where inequality has grown (UN75), generating 85% of global GDP such as the United States, China, and India (Qureshi, 2023).

Launched in 2018, the Wellbeing Economy Governments partnership (WEGo) involving five governments are committed to progress toward the UN Sustainable Development Goals as well as designing and fostering economic systems that deliver human and ecological wellbeing. Its purpose is not to dismiss GDP: Economic growth is pursued not as an end in itself but as a means to achieving wellbeing (EuroHealthNet and Institute of Public Health, 2024). In fact, GDP growth remains important to all wellbeing economy countries (Hayden & Dasilva, 2022).

This paper aims to explore whether there are any statistical differences to be part of the Wellbeing Economy Governments partnership by comparing WEGo countries with other OECD members. Three hypotheses will be tested:

Subjective Well-Being Score

H₀: There is no significant difference for subjective well-being score between WEGo and non-WEGo OECD countries.

H₁: There is a significant difference for subjective well-being score between WEGo and non-WEGo OECD countries.

Correlation Between SDG and Subjective Well-Being Scores

H₀: There is no correlation between the SDG goal scores and subjective well-being scores across countries.

H₁: There is a positive correlation between the SDG goal scores and subjective well-being scores across countries.

Variable Importance

H₀: There is no difference in the variable importance output for predicting subjective well-being scores using SDG indicators among WEGo and non-WEGo OECD countries.

H₁: There is a difference in the variable importance output for predicting subjective well-being scores using SDG indicators among WEGo and non-WEGo OECD countries.

The Commission on the Measurement of Economic Performance and Social Progress created by the French Government (2008) explored existing measures of economic performance based on GDP and its key issues, stating how “time has come to adapt our system of measurement of economic activity to better reflect the structural changes which have characterized the evolution of modern economies.” Rather than dismissing GDP, the report stressed the importance of emphasizing well-being and working towards a system that captures data not found in the current system. While there have been concerns about GDP solely being used for economic performance, authors pointed out that GDP is inadequate when it comes to gauging well-being over time, particularly in sustainability.

Paper by Hayden and Dasilva (2022) did a case study of wellbeing economy governments. By reviewing New Zealand, Scotland, and Iceland’s policy changes since their commitment to beyond-GDP measures, the paper identified the countries’ efforts as “weak post-growth” approach and suggested further disintegration of dependence on economic growth to achieve a “strong post-growth” progress. While there was much qualitative analysis of the countries’ policies, quantitative analysis especially about the difference between the wellbeing economy and other OECD countries was not available.

Research by Neve and Sachs (2020) investigated correlations between SDGs and wellbeing and reported that in most cases showed a strong positive correlation with the exception of Goal 12 (responsible production and consumption) and Goal 13 (climate action); they both negatively correlated with well-being. However, this paper did not explore the concept of the wellbeing economy. It did highlight how there is a growing sentiment that subjective well-being has an increasing significance in policy-making and also more policy efforts are necessary to achieve sustainable development for well-being optimization.

Paper by Jannani, Sael, and Benabbou (2021) presented at the 2021 4th International Symposium on Advanced Electrical and Communication Technologies focused on predicting the World Happiness Index scores for the year 2021 using 3 linear models, 4 ensemble models, and 2 deep learning algorithm models. It concluded that the best performance was achieved using Lasso Regression. Another research by Kubiszewski et al. (2021) identified 8 of the 232 SDG indicators that explained 84% of the variation in life satisfaction using a LASSO model. World Happiness Index scores used by Jannani et al. and life satisfaction scores used by Kubiszewski et al. were based on Gallup’s World Poll data. While both research used machine learning models to predict subjective well-being scores, they did not provide the important variables for the outcome.

Data

For the analysis, the online database for the Sustainable Development Report 2024 was used, which included (1) the overall SDG score, (2) 17 SDG goal scores, and (3) 99 SDG indicator scores covering 192 countries. Rows for the 38 countries in the Organisation for Economic Co-operation and Development (OECD) were selected for further analysis. A new column was created to group countries based on their association with WEGo:

4 nations officially involved in WEGo
- Finland, Iceland, New Zealand, United Kingdom (Scotland and Wales combined)
9 OECD countries with established WEGo hubs
- Australia, Canada, Costa Rica, Denmark, Ireland, Netherlands, Spain, Sweden, United States
25 OECD countries with no WEGo associations
- Austria, Belgium, Chile, Colombia, Czechia, Estonia, France, Germany, Greece, Hungary, Israel, Italy, Japan, Korea, Latvia, Lithuania, Luxembourg, Mexico, Norway, Poland, Portugal, Slovak Republic, Slovenia, Switzerland, Türkiye

The 2024 data was looked at first and then combined with data on the ‘Backdated SDG Index’ tab of the excel sheet. Once again, OECD countries were filtered and grouped based on WEGO associations. Because the Wellbeing Economy Governments partnership (WEGo) launched in 2018, data prior were dropped. The combined data date ranged from 2018–2024.

It is important to note that subjective well-being (sdg3_swb) is not formally adopted as a specific indicator within the official SDG framework but included as a SDG indicator in the 2024 Sustainable Development Report. Subjective well-being is sourced from Gallup’s World Poll data for the World Happiness Report: Respondents are asked to think of their lives on a ladder, with the best possible life for them being a 10, and the worst possible life being a 0. Also known as the Cantril ladder, this survey uses responses from people in more than 140 nations (Gallup, 2024).

Methods

Columns with more than 10% missing values were removed. Then missing value was imputed using the Multivariate Imputation by Chained Equations (MICE) package in R. Data was already normalized so no scaling was done additionally.

For the 2024 SDG data, exploratory data analysis was conducted. First, normality was tested among the different WEGO association groups by observing the QQ plots and p-values from the Shapiro-Wilk normality test. When normally distributed, one-way ANOVA test was used to compare the mean among WEGO members, WEGO hubs in the OECD, and the rest of the OECD countries. When normal distribution could not be assumed, the Kruskal-Wallis Test was used to compare the median. SDG overall goal, 17 SDG goal scores, and the subjective well-being score were all compared along with the correlation analysis. Lastly, a hierarchical clustering dendrogram was used to illustrate natural groupings and optimal number of clusters based on the subjective well-being score among OECD countries.

To gain insights into the relative importance of the features in our dataset, predictive modeling was chosen for variable importance extraction. Columns with more than 10% missing values were removed before imputing. Subjective well-being was used as the target variable (y) and 89 SDG indicators as features (x). 5 machine learning models were used on datasets separated by their WEGO associations to compare the variable importance output to test our hypothesis that different SDG indicators will be considered among WEGO and non-WEGO countries.

Lasso Regression: Due to multicollinearity arising from having 98 features, lasso regression was chosen from the regression models as a regularization technique - reducing the coefficients of less significant features to zero.
Random Forest (RF): RF is a bagging technique that uses multiple decision trees and determines the final output via a majority vote; Compared to decision trees, RF model is robust to noise and less prone to overfitting.
Extreme Gradient Boosting (XGBoost): A popular choice for predictive analytics, XGBoost is a boosting technique that sequentially creates decision trees that improves upon the mistakes of the previous one and finalizes with the sum of all outputs.
Support Vector Machine (SVM): Known to be effective on high-dimensional data, SVM finds an optimal line/hyperplane that maximizes the distance between each class in an N-dimensional space.
K-Nearest Neighbor (KNN): Using the assumption that similar points can be found near one another, the average of the nearest neighbors is taken to make a prediction.

Since WEGo was launched in 2018, the models were trained on data found between 2018–2022 and tested on data between 2023–2024. RMSE, Rsquared, and MAE results for each WEGO group model training data results were compared. Test data results were used to validate the selection of the best model. With caret package’s VarImp() function, top 20 SDG indicators were showcased to provide insights into the relative importance of each feature in predicting the subjective well-being score.

Subjective Well-Being Score

Do WEGo countries have higher subjective well-being scores? To determine if there are statistically significant differences, the Shapiro-Wilk normality test was conducted on three groups to first test whether they are normally distributed: (1) WEGo members, (2) WEGo hubs where collaboration towards building a Wellbeing Economy take place and (3) the rest of the OECD countries that have no activity related to Wellbeing Economy. Normality needs to be assumed first to conduct ANOVA, otherwise Kruskal-Wallis Test is conducted as it is a non-parametric method.

The null hypothesis for a Shapiro Wilk test is that there is no difference between your distribution and a normal distribution. All three groups had p-value greater than 0.05, indicating that the data sets are normally distributed. Analysis of Variance (ANOVA) was then conducted to compare the means of the groups and with a p-value of 0.00108, a highly statistically significant difference between the groups was confirmed. In other words, the variation in subjective well-being scores is unlikely to have occurred by chance and WEGo members have a higher mean. Minimum, median, and maximum scores were also higher as showcased in the boxplot below.

One of the unsupervised learning methods for machine learning is hierarchical clustering, and when used for subjective well-being, 3 clusters were identified by cutting the tree at the height of 2. Except for the United Kingdom, all of the WEGo countries were grouped together, demonstrating a high degree of similarity. Out of the 9 WEGo Hub OECD countries, 5 were in cluster 1 with WEGo countries and 4 were in cluster 2 with the United Kingdom. Cluster 3 had entirely non-WEGo OECD members. Since clustering works by finding the most similar objects, the outcome highlights a potential divergence in subjective well-being scores between WEGO and non-WEGO countries and potential members to join the wellbeing economy movement.

Sustainable Development Goals

Do WEGo countries have higher Sustainable Development Goal scores? Based on the Shapiro-Wilk normality test, group 3 (rest of OECD) did not have normally distributed data. Kruskal-Wallis rank sum test was conducted instead of ANOVA to compare median scores of the three groups: with a p-value (0.4492) greater than the conventional threshold of 0.05, it failed to reject the null hypothesis. In other words, there was no statistically significant difference between the groups being compared. However, it is still important to note that WEGo countries did have higher minimum, mean, median, and maximum scores than the other two groups.

Next, are any of the 17 SDG Goal score differences statistically significant for WEGO countries compared to the rest of the OECD countries?

Goal 3 (p-value: 0.04619), Goal 5 (p-value: 0.00375), and Goal 11 (p-value: 0.02534) showed statistically significant differences among WEGo, WEGo hub, and the rest of the OECD countries. Higher mean and median scores were found in Goal 1, Goal 3, Goal 5, Goal 6, Goal 7, Goal 9, Goal 10, Goal 11, Goal 14, Goal 16, and Goal 17. In other words, WEGo countries had 11 out of 17 SDG goal mean and median scores higher than non-WEGO OECD countries, out of which 3 were statistically significant: Goal 3 (Good Health and Well-Being), Goal 5 (Gender Equality), and Goal 11 (Sustainable Cities and Communities).

It is important to note that Goal 14 should be viewed with caution as there were more than 10% of missing data originally. However, the three statistically significant results highlight WEGo potential. For instance, even though Norway (non-WEGO) is the only country to achieve Goal 3 (Sustainable Development Report), the Kruskal-Wallis rank sum test suggested that there is strong evidence that WEGo countries perform differently from non-WEGo OECD members with a higher median score. New Zealand is the only country in the world to achieve SDG Goal 11 Sustainable Cities and Communities (Sustainable Development Report), which may have contributed to the results seen in the analysis.

Correlation Between SDG and Subjective Well-Being Scores

Are there any correlations between SDGs and subjective well-being scores for data reported in 2024?

There is a strong positive correlation between subjective well-being score and Goal 5 (Gender Equality) with a correlation coefficient of 0.7428. There is a moderate positive correlation between subjective well-being score and the following SDG goal scores:

Goal 1 (No Poverty)
Goal 3 (Good Health and Well-being)
Goal 9 (Industry, Innovation and Infrastructure)
Goal 10 (Reduced Inequalities)
Goal 11 (Sustainable Cities and Communities)
Goal 16 (Peace, Justice and Strong Institutions)
Goal 17 (Partnerships for the Goals)

Correlation between subjective well-being score and Goal 3 should be dismissed since the former is used as an indicator calculated for Goal 3; in total, there are 7 SDG goals positively (strong and moderate) correlated. There is a moderate negative correlation between subjective well-being and Goal 12 and Goal 13. This aligns with Jan-Emmanuel De Neve and Jeffrey D. Sachs' observation in 2020 when they also found SDGs 12 (Responsible Consumption and Production) and 13 (Climate Action) being negatively correlated with well-being. It is also important to note that in the correlation plot above, SDG Goal 3 (Good Health and Well-being) is positively correlated to other SDG goals except Goals 12, 13, 14 (Life Below Water), and 15 (Life on Land).

Variable Importance

If we build separate machine learning models based on WEGo associations to predict subjective well-being, will the variable importance results show different SDG indicators? The Lasso regression model, random forest, extreme gradient boosting, support vector machine, and K-nearest neighbor models were used to assess the relative importance of SDG indicators in predicting subjective well-being scores. Extreme Gradient Boosting had the lowest RMSE and MAE for WEGo, WEGo Hubs, and Non-WEGo OECD country groups for both training and test data.

Three variable importance plots based on WEGo associations were generated based on the best performing model, Extreme Gradient Boosting. The model’s coefficients offer insights into which SDG indicators are most influential in predicting subjective well-being. For WEGo countries, SDG indicator ‘n_sdg2_obesity’ (The percentage of the adult population that has a body mass index of 30kg/m² or higher) emerged as the strongest predictor, followed by SDG indicator ‘n_sdg3_matmort’ (The estimated number of women, between the age of 15 and 49, who die from pregnancy-related causes while pregnant or within 42 days of termination of pregnancy, per 100,000 live births) and ‘n_sdg7_renewcon’ (The share of renewable energy such as wind/solar in the total final energy consumption). For WEGo Hub countries in the OECD, ‘n_sdg8_unemp’ (Unemployment rate % of total labor force, ages 15+), ‘n_sdg12_explastic’ (The average annual amount of plastic waste exported over the last 5 years expressed per capita), and ‘n_sdg16_admin’ (Timeliness of administrative proceedings) had significant coefficients. For Non-WEGo OECD countries, SDG indicators ‘n_sdg8_rights’ (Rating whether fundamental labor rights are effectively guaranteed, worst 0–1 best), ‘n_sdg12_pollimp’ (Air pollution associated with imports), and ‘n_sdg3_traffic’ (Traffic deaths per 100,000 population) stood out.

Only 3 SDG indicators were present in all three models: ‘n_sdg8_unemp’ (Unemployment rate), ‘n_sdg1_lmicpov’ (Poverty headcount ratio at $3.65/day), and ‘n_sdg16_rsf’ (Press Freedom Index). While WEGo country model had 5 SDG indicators for Goal 3 (Good Health and Well-being), WEGo hub country model and Non-WEGo country model had 2 each.

Table 1

SDG Goal 3 Indicators Found in Predictive Models’ Variable Importance Results
WEGo Model Result	Indicator Description	WEGo Hub Model	Indicator Description
n_sdg3_matmort	Maternal mortality rate	n_sdg3_fertility	Adolescent fertility rate
n_sdg3_neonat	Neonatal mortality rate	n_sdg3_u5mort	Mortality rate, under-5
n_sdg3_vac	Surviving infants who received 2 WHO-recommended vaccines	Non-WEGo Model	Indicator Description
n_sdg3_u5mort	Mortality rate, under-5	n_sdg3_traffic	Traffic deaths
n_sdg3_tb	Incidence of tuberculosis	n_sdg3_uhc	Universal health coverage index of service coverage

The results may be surprising, given that the WEGo country model encompasses 4 indicators related to mortality rate of women and children specifically and the WEGo hub country model also focuses on children. While there is a positive correlation between maternal mortality rate and average income across the world, not every country that achieved economic growth also reduced mortality rates: WEGo member states like Finland had a maternal mortality rate of 8.3 with a GDP of $47,371 while the United States (non-WEGO) had a maternity mortality rate of 21.1 with a GDP of $60,159 (Roser & Ritchie, 2024). Similarly, according to Emily Oster, an economist at Brown University, the infant mortality rate in the US is comparably higher than Finland, but "the difference is well-off women in Finland and less-educated women in Finland have very similar infant mortality profiles. Whereas that is not true in the US (Cassin, 2017)”. In fact, the United States has the highest infant and maternal mortality rates out of all OECD countries even though it spends the most on health care (Petrullo, 2023). The Finnish social insurance company KELA has distributed the baby box since 1937, gradually developing what initially started as a program for socially needy women to all expectant mothers (Neuezeit.At, 2022) - accompanied by national prenatal care and education for parents (Cassin, 2017). 87 years later, Finland has one of the lowest mortality rates for both women and infants globally.

From the Extreme Gradient Boosting model, different SDG indicators show more strongly associated relationships with subjective well-being scores based on whether the data is from a WEGo member state, WEGo hub in OECD, or no association with WEGo in OECD countries.

The global indicator framework includes 231 unique indicators, yet there are only 99 indicators in the official Sustainable Development Report database. Out of the 99, 11 indicators had missing values greater than 10%. In other words, the predictive models utilized only 38% of what was meant to be measured as a whole. This highlights the importance of the recent launch of the High Impact Initiative on the Power of Data at the SDG Summit in September 2023, which aims to ensure that every government has the data it needs to both track and drive SDG progress (General Assembly Economic and Social Council, 2024).

Also, overfitting happens when the model becomes too specialized to the training data and performs less effectively on unseen data. The likelihood increases when working with a smaller dataset, and that was the case for the models built. Despite normalization, regularization, and cross-validation, all of the models showed an R-squared value close to 1 for WEGo countries training data - meaning that the models explain 100% of the variation in the dependent variable. Test data confirmed concerns with lower Rsquared results. However, the Extreme Gradient Boosting (XGB) model showed robust performance with an R-squared of 0.98, an RMSE of 1.749, and an MAE of 1.5 for the test data, indicating that the model captured most of the variance in unseen data. Still, the results should be viewed with caution for the WEGo XGB model.

Table 1

Predictive Modeling Results - Training Data
Model	RMSE	Rsquared	MAE
Lasso Regression Model
- WEGo	2.214789	1	1.988371
- WEGo Hubs	3.689061	0.8614868	3.078462
- Rest OECD	5.657909	0.8347622	4.335634
Random Forest
- WEGo	2.527964	1	2.253735
- WEGo Hubs	2.859674	0.9028292	2.239574
- Rest OECD	5.591740	0.8617620	4.261445
Extreme Gradient Boosting
- WEGo	1.423883	1	1.242724
- WEGo Hubs	2.223763	0.9357628	1.783446
- Rest OECD	5.435988	0.8633881	4.213677
Support Vector Machine
- WEGo	3.223162	1	2.86126
- WEGo Hubs	3.564035	0.8557617	2.790985
- Rest OECD	7.348234	0.7894207	5.334445
K-nearest neighbor
- WEGo	4.035504	1	3.573560
- WEGo Hubs	3.908273	0.8021630	2.963432
- Rest OECD	6.228537	0.8242790	4.702998

Table 2

Predictive Modeling Results - Test Data
Model	RMSE	Rsquared	MAE
Lasso Regression Model
- WEGo	5.798824	0.633204	4.257352
- WEGo Hubs	5.563125	0.5849962	3.95862
- Rest OECD	6.369544	0.7699718	4.642443
Random Forest
- WEGo	2.8881	0.9890264	2.42319
- WEGo Hubs	4.697774	0.6997002	3.724698
- Rest OECD	5.785349	0.8052059	4.334191
Extreme Gradient Boosting
- WEGo	1.749412	0.9828592	1.533455
- WEGo Hubs	3.814036	0.8099891	3.146581
- Rest OECD	7.614009	0.7293595	5.806492
Support Vector Machine
- WEGo	4.223474	0.8038967	3.365675
- WEGo Hubs	4.883372	0.6887828	3.836774
- Rest OECD	7.53928	0.6860905	5.939502
K-nearest neighbor
- WEGo	4.280777	0.9926082	3.2128
- WEGo Hubs	6.294021	0.5014563	4.945622
- Rest OECD	6.629291	0.7863847	4.944696

According to the U.N. Secretary-General Antonio Guterres, “GDP is not a way to measure richness in the present situation in the world (Meredith, 2022).” If an unprecedented stock market crash in 1929 led to the creation of GDP and the deadliest and most destructive war in human history to systematize it globally in the 1940s, it makes perfect sense why the wellbeing economy has been introduced. “We are on the brink of an irreversible climate disaster (Ripple, 2024)”, with summer of 2024 being the Earth's warmest on record (King, 2024) and the 10 warmest years in the 174 years all occurring within the last decade (United Nations, 2024). Nearly three-quarters of the SDG targets (especially poverty, education, and decent work) are directly or significantly reliant on gender equality, but not a single country is on track to achieve it by 2030 (Gates, 2024). Global hunger levels have plateaued for three consecutive years, with 2.33 billion people globally facing moderate or severe food insecurity in 2023 (WHO, 2024).

Let us not wait for a tragedy that took 50 million lives to formally embrace the wellbeing economy as a global metric like how it was for GDP. Wellbeing economy is not a rejection of economic growth but a clear evolution needed for the sustainability of society and the planet. With the adoption of Pact for the Future in at the Summit of the Future on 22 September 2024 - where global leaders of the UN member states recommitted to accelerating SDGs to meet the 2030 Agenda (Whiting, 2024) - now is the right time to discuss about the wellbeing economy globally and actively.

Author Contribution

Jiwon Kim prepared and wrote the manuscript.

Data Availability

Data for Sustainable Development Goals is available for free and can be downloaded from the Sustainable Development Report database: https://dashboards.sdgindex.org/downloads.

Butler, Creon (2023). The sdgs are not on track: Here is what the world should do. Council on Councils https://www.cfr.org/councilofcouncils/global-memos/sdgs-are-not-track-here-what-world-should-do
Callen, Tim (n.d.). Gross Domestic Product: An Economy’s All. International Monetary Fund Finance & Development Magazine https://www.imf.org/en/Publications/fandd/issues/Series/Back-to-Basics/gross-domestic-product-GDP
Cassin, Elizabeth (2017). Do baby boxes really save lives? BBC https://www.bbc.com/news/magazine-39366596
Costanza, Robert et al. (2014) Development: Time to leave GDP behind. Nature 505, 283-285 https://doi.org/10.1038/505283a
Coyle, Diane (2017). Rethinking GDP. IMF Finance & Development https://www.imf.org/external/pubs/ft/fandd/2017/03/pdf/coyle.pdf
EuroHealthNet and Institute of Public Health (2024). Creating an impactful and sustainable Wellbeing Economy for better public health. Institute of Public Health. https://www.publichealth.ie/sites/default/files/resources/Sustainable%20Wellbeing%20Economy%20report%20final.pdf
Gates, Bill (2024). The Race to Nourish a Warming World. Gates Foundation https://www.gatesfoundation.org/goalkeepers/report/2024-report/#Introduction
Geldard, Rebecca & Ellerbeck, Stefan (2023). How much progress is being made on the UN’s sustainable development goals? World Economic Forum https://www.weforum.org/agenda/2023/09/un-sustainable-development-goals-progress-report/
General Assembly Economic and Social Council (2024). United Nations https://unstats.un.org/sdgs/files/report/2024/SG-SDG-Progress-Report-2024-advanced-unedited-version.pdf
Hayden, Anders & Dasilva, Clayton (2022) “The wellbeing economy: Possibilities and limits in bringing sufficiency from the margins into the mainstream.” Frontiers in Sustainability https://doi.org/10.3389/frsus.2022.966876
Hoekstra, Rutger (2020) Measuring the Wellbeing Economy: How to Go Beyond-GDP. Wellbeing Economy Alliance https://wellbeingeconomy.org/wp-content/uploads/WeAll-BRIEFINGS-Measuring-the-Wellbeing-economy-v6.pdf
Jannani, Ayoub, et al. (2021) Predicting Quality of Life using Machine Learning: case of World Happiness Index. IEEE https://ieeexplore.ieee.org/document/9668429
King, Simon (2024) Summer 2024 was world's hottest on record. BBC https://www.bbc.com/weather/articles/c93p5kz9elro#:~:text=Temperatures%20continue%20to%20rise
&text=So%20far%20this%20year%20the
,Copernicus%20%2C%20the%20EU's%
20climate%20service.
Kubiszewski, Ida et al. (2022). Toward better measurement of sustainable development and wellbeing: A small number of SDG indicators reliably predict life satisfaction. Sustainable Development, 30(1), 139–148. https://onlinelibrary.wiley.com/doi/full/10.1002/sd.2234
Meredith, Sam (2022) The world’s in a ‘polycrisis’ — and these countries want to quash it by looking beyond GDP. CNBC https://www.cnbc.com/2022/12/26/well-being-these-countries-are-looking-beyond-gdp-and-economic-growth.html
Neuezeit.at (2022). Simple: How Finland decreased maternal mortality rates with a cardboard box. The Better https://thebetter.news/finland-babybox/
Neve, JE. & Sachs, Jeffrey (2020). The SDGs and human well-being: a global analysis of synergies, trade-offs, and regional differences. Nature Scientific Reports 10, 15113 https://doi.org/10.1038/s41598-020-71916-9
Petrullo, Justina (2023). US Has Highest Infant, Maternal Mortality Rates Despite the Most Health Care Spending. AJMC https://www.ajmc.com/view/us-has-highest-infant-maternal-mortality-rates-despite-the-most-health-care-spending
Ripple, William, et al. (2024). The 2024 state of the climate report: Perilous times on planet Earth. BioScience https://doi.org/10.1093/biosci/biae087
Roser, Max and Ritchie, Hannah (2024). Maternal Mortality. Our World in Data. https://ourworldindata.org/maternal-mortality
Stiglitz, Joseph et al. (2008). Report by the Commission on the Measurement of Economic Performance and Social Progress. https://ec.europa.eu/eurostat/documents/8131721/8131772/Stiglitz-Sen-Fitoussi-Commission-report.pdf
Sustainable Development Report (2024). https://dashboards.sdgindex.org/map
UN75 (n.d.) Inequality – Bridging the Divide. United Nations https://www.un.org/en/un75/inequality-bridging-divide
United Nations (2024). The Sustainable Development Goals Report 2024. https://unstats.un.org/sdgs/report/2024/The-Sustainable-Development-Goals-Report-2024.pdf
Vanham, Peter (2021) A brief history of GDP - and what could come next. World Economic Forum https://www.weforum.org/agenda/2021/12/stakeholder-capitalism-episode-1-a-brief-history-of-gdp/
Weeks W, Weinstein J, Lavista J (2023) All Sustainable Development Goals Support Good Health and Well-Being. International Journal of Public Health 68:https://doi.org/10.3389/ijph.2023.1606901
Whiting, Kate (2024). What is the UN's Summit of the Future in 2024 and why is it important? World Economic Forum https://www.weforum.org/agenda/2024/09/un-summit-future-sdgs/
WHO (2024). Hunger numbers stubbornly high for three consecutive years as global crises deepen: UN report. World Health Organization https://www.who.int/news/item/24-07-2024-hunger-numbers-stubbornly-high-for-three-consecutive-years-as-global-crises-deepen--un-report

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Comparing Wellbeing Economy with other OECD Nations: A Predictive Analysis of Subjective Well-Being Score using SDGs

Status:

Version 1

Abstract

Figures

Introduction

Subjective Well-Being Score

Correlation Between SDG and Subjective Well-Being Scores

Variable Importance

Literature Review

Data and Methodology

Data

Methods

Results

Subjective Well-Being Score

Sustainable Development Goals

Correlation Between SDG and Subjective Well-Being Scores

Variable Importance

Discussion

Conclusion

Declarations

Author Contribution

Data Availability

References

Additional Declarations

Status:

Version 1