SLIDE 1
The Sustainable Development Oxymoron: Quantifying and Modelling the - - PowerPoint PPT Presentation
The Sustainable Development Oxymoron: Quantifying and Modelling the - - PowerPoint PPT Presentation
The Sustainable Development Oxymoron: Quantifying and Modelling the Incompatibility of Sustainable Development Goals Viktoria Spaiser Shyam Ranganathan Ranjula Bali Swain David J.T. Sumpter Sustainable Development Goals (SDG) Data
SLIDE 2
SLIDE 3
Data (RevoluBon) for SDGs
*
* World Bank Data API
1432 economic, social, poliBcal and environmental indicators, 217 countries, years 1980-2013, finally used 233 indicators
SLIDE 4
Are SDGs consistent?
CO2.emissions Air.Pollution Protected.Land Education Women.Parliament Child.Mortality Water GINI Hunger Violence Internet Protected.Sea Sanitation Poverty Alternative.Energy Unemployment.Youth
−1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0
Factor Dimension 1 Factor Dimension 2
20 40 60 80
A
SLIDE 5
Are SDGs consistent?
EFA-suggested model, CFA confirmed:
Development (Factor Dimension 1) CM Pov Hun Wat* San* Ed* Int* CO2 AP Vio .90 .91 .73 .90 .92 .93 .60 .30 .78 .68
- .53
Inequality & Violence (Factor Dimension 2) GINI
B
SLIDE 6
Model Latent Variable (L), by CFA
CFA factor scores for L used to create Model Latent Variable L (Latent Variable 1) SDL CM Edu CO2 .821 .981
- .531
ProporBon Variance: .73
R2=.673 R2=.961 R2=.282
Model Fits CFA: CFI: .973; TLI: .931; RMSEA: .031; SRMR: .063
CO2.emissions E d u c a t i
- n
Child.Mortality −2 2 4 −2 2
Factor Dimension 1 Factor Dimension 2
SLIDE 7
Finding predicBve models
Method
- 1. Feature SelecBon:
Variable EliminaBon Algorithm, using Ensemble PLS, accounBng for nonlineariBes èbest predictors
- 2. Data-driven Dynamical Systems Modeling:
Model CombinaBon Approach & Bayesian Model SelecBon with best predictors selected by Feature SelecBon Algorithm, iteraBve
SLIDE 8
Data-driven Dynamical Systems Modeling
Model Selec+on:
- 1. Log Likelihood (pre-selecBon):
- 2. Bayes Factor (final selecBon):
at higher order iteraBon steps
- nly Bayes Factor
Model Combina+on Approach: CombinaBon of increasing complexity (number of terms)
- f polynomial terms
SLIDE 9
Net foreign assets GDP per capita
2 4 6 6 8 10 −0.5 0.5
Net foreign assets Fertility rate
2 4 6 2 4 6 8 −0.5 0.5
Womens economic rights Independent Judicary
1 2 3 0.5 1 1.5 2 −0.5 0.5
Natural Depletion SDL
2 4 5 10 15 −0.3 −0.28 −0.26 −0.24 −0.22
Best-fit Models for change of L
D: Net foreign assets (indebtedness) G: GDP per capita Fr: Fertility rate Rf: Women’s economic rights J: Independence of Judicary Nd: Natural depletion costs
SLIDE 10
Factors contribuBng to incompaBbility Factors showing a way out
G: GDP per capita C: Final consumption expenditure Fr: Fertility rate Er: Renewable energy production M: Measles immunication Nd: Natural depletion costs Wg: Government spending Em: Particulate emission damage
Dynamical Systems Models for the three pillars
1. End Poverty (Model for changes in Child Mortality): 2. Socio-economic inclusion (Model for changes in EducaBon): 3. Environment (Model for changes in CO2 emissions):
SLIDE 11
SDG index 1
−0.6 1
Monitoring Sustainable Development
Based on L model
SLIDE 12
The SDG index vs. HDI and GDP
SDG index 1 predicts HDI predicts GDP per capita predicts 54% of changes in child mortality 41% of changes in child mortality 17% of changes in child mortality 6% of changes in educaBon 4% of changes in educaBon 2% of changes in educaBon 21% of changes in CO2 emissions 0.7% of changes in CO2 emissions 0.4 % of changes in CO2 emissions 16% of changes in L 7% of changes in L 4% of changes in L
SLIDE 13