Measuring and analyzing political polarization in the UK using Geopandas and PySAL.

Chinna Subbaraya Siddharth Ramavajjala ^a, Ramakrishna Raju Gangaraju ^a
a - Department of Geography, University of Wisconsin - Madison

Motivation and topic of Interest:

Political polarization is often considered a crucial parameter for the stability of a government, social tensions (Esteban et al., 2012), policy uncertainty (Funke et al., 2016), economic fluctuations, and low economic growth. However, quantifying political polarization is challenging and complicated as it depends on several factors like income, level of education, age of the population, variability of job status (percentage of employed), and a fraction of employed. Python analysis applied in our repository utilizes information from British Household Panel Survey (BHPS), which is a questionnaire under taken at a region level across different time periods. The crux of dataset is to ask the same survey question and analyze if the opinion is changed or deviation against previous responses. Therefore, the idea is to compute categorical data encoded into values, giving a scale for qualitative data (survey responses). Interestingly, most of research on political polarization is carried out in the United States, and the measurement polarization is not a common sight across developed economies despite the enormous power of data and information.

Data:

• British Household Panel Survey (BHPS) data: - https://beta.ukdataservice.ac.uk/datacatalogue/studies/study?id=5151#!/access-data
• UK Shapefiles - https://geoportal.statistics.gov.uk/datasets/ons::nuts1-jan-2018-super-generalised-clipped-boundaries-in-the-uk/explore?location=49.270533%2C12.096034%2C5.65

Spatial Bigdata Visualization Flowchart:

Methods:

As part of the project, we will implement methods and techniques used in analyzing BHPS titled "Political polarization in the UK: measures and socioeconomic correlates” by Daryna Grechyna (2022). Therefore, the project uses raw data, consisting of several folders in SPSS format, which would be read in pandas. Indirect calculations are performed based on variables of interest. The evaluation of political polarization will be performed on the following statements, which are available in BHPS data.




• Statement 1 (S1): “Private enterprise is the best way to solve Britain’s economic problems.”
• Statement 2 (S2): “Major public services and industries ought to be in state ownership.”
• Statement 3 (S3): “It is the government’s responsibility to provide a job for everyone who wants one.”



The political polarization measures implemented are:

• Lindqvist and Ostling (2010)
• Duca and Saving (2016)
• Abramowitz and Saunders (2008)



The extracted and analyzed data would then be combined with Geopandas and PySAL libraries to calculate hotspots and cold spots of polarization in the UK.

Results:



Local Moran's I of Fraction of Employed in the years 1995, 2000 and 2004



Correlation matrices between variables and political polarization measurements in the years 1995, 2000 and 2004



Opinion change for Statement 3 by each region in the years 1995, 2000 and 2004



Opinion distribution change for all Statements in the years 1995, 2000 and 2004