Gini's Odyssey in Greece: Econometric analysis of income in-equality, GDP, and unemployment, through economic phases (pre crisis, crisis, memoranda and post memoranda)
Description
This study investigates the relationship between income inequality, economic growth, and unemployment in Greece over the period 2003–2020, encompassing distinct economic phases: pre-crisis (2003–2008), crisis and memoranda (2009–2014), and post-memoranda (2015–2020). The research focuses on two primary hypotheses: H1: There is a positive relationship between GDP (measured logarithmically as LOGGDP) and income inequality (Gini coefficient), suggesting that economic growth, when not accompanied by redistributive mechanisms, exacerbates income disparities. H2: There is a positive relationship between unemployment and income inequality, reflecting the disproportionate impact of labor market disruptions on lower-income groups. Description of the Data The analysis uses annual data for Greece over 18 years (2003–2020). The variables analyzed include: Dependent Variable: Gini Coefficient: A measure of income inequality where 0 represents perfect equality and 1 represents maximum inequality. Independent Variables: LOGGDP: The natural logarithm of Gross Domestic Product (GDP), reflecting economic output while accounting for non-linear effects of growth. Unemployment Rate (% of total labor force): Represents the share of the labor force actively seeking employment but unable to find work. Data Sources and Collection Gini Coefficient: Extracted from World Bank Open Data, calculated using annual household income surveys. GDP: Sourced from the World Bank and measured in constant US dollars to adjust for inflation and ensure comparability over time. Unemployment Rate: Sourced from the World Bank Analytical Methods Correlation Analysis: Pearson correlation coefficients were calculated to explore the strength and direction of relationships between variables. Linear Regression Model: The model estimated the impact of LOGGDP and unemployment on the Gini coefficient: Ginit=α+β1LOGGDPt+β2Unemploymentt+ϵt Ginit=α+β1LOGGDPt+β2Unemploymentt+ϵt α: Intercept. β1 and β2: Coefficients for LOGGDP and unemployment, respectively, representing their influence on the Gini coefficient. ε: Error term. Diagnostic Tests: Multicollinearity was assessed using Variance Inflation Factors (VIF). Model robustness was checked with Durbin-Watson statistics to evaluate autocorrelation. Residual analysis ensured compliance with normality assumptions. Hypothesis Validation Economic Growth and Inequality (H1): The positive relationship between LOGGDP and the Gini coefficient underscores that growth during the analyzed periods did not equitably benefit all segments of society. Unemployment and Inequality (H2): The stronger positive effect of unemployment on inequality highlights the critical role of labor market disruptions in intensifying disparities.
Files
Steps to reproduce
Data Retrieval Process Accessing Data Sources: The World Bank Open Data and ILO platforms were accessed online using publicly available APIs and search tools. The specific indicators used were: Gini Coefficient: SI.POV.GINI GDP: NY.GDP.MKTP.KD Unemployment Rate: SL.UEM.TOTL.ZS Downloading Data: Data was downloaded in CSV format for consistency across sources. Metadata associated with each dataset was reviewed to ensure alignment with the research period and context. Data Verification: Cross-referenced the primary data with Eurostat and OECD databases for consistency. Verified that units, timeframes, and definitions matched across all sources. Preprocessing Steps Handling Missing Data: Identified missing entries in the datasets. In cases where up to two consecutive years of data were missing, linear interpolation was applied to estimate the values. Any data gaps exceeding two years were excluded to maintain data integrity. Data Transformation: Converted GDP into its natural logarithmic form (LOGGDP) to better model its impact on income inequality and reflect diminishing marginal effects of economic growth. Normalization and Alignment: Standardized all variables to the same timeframe and ensured consistency in the frequency of observations (annual data). Analytical Tools and Workflows Software Used: IBM SPSS Statistics 23: For regression analysis, diagnostic tests, and residual analysis. Microsoft Excel: For initial data cleaning, preprocessing, and visualization of trends. Statistical Analysis: Conducted Pearson correlation analysis to determine initial relationships between variables. Used linear regression to quantify the effects of LOGGDP and unemployment on the Gini coefficient. Evaluated model fit using metrics such as R-squared, adjusted R-squared, and the F-statistic. Diagnostics and Robustness Checks: Variance Inflation Factor (VIF) analysis was conducted to assess multicollinearity between LOGGDP and unemployment. Durbin-Watson statistic checked for autocorrelation in the regression residuals. Residual plots ensured compliance with linearity and homoscedasticity assumptions. Reproducibility Protocol Preprocessing Steps: Load data into a spreadsheet or statistical software (e.g., SPSS, Excel). Address missing data points using interpolation, ensuring transparency in the methods applied. Analysis Workflow: Import the processed datasets into statistical software. Replicate correlation and regression analyses using the provided model specifications: Ginit=α+β1LOGGDPt+β2Unemploymentt+ϵt Ensure diagnostic tests (e.g., VIF, Durbin-Watson)