While there is a consensus on high polarization in Georgia as confirmed by the public perception of increasing polarization on one hand and the call of the country's development partners towards depolarization, there has not been any tool available to measure and monitor the polarization dynamics. On this basis, the ISET Policy Institute developed a media polarization index to explore and measure the dynamics of media polarization in Georgia. It looks at media polarization as a proxy for political polarization. The conducted research is based on the use of Natural Language Processing (NLP) techniques. Machine learning application makes the tool unbiased and independent of subjective interpretations. The note presents the Media Polarization Index and analyzes how media polarization relates to various factors, including political events, public perception, opinion polarization, political party ratings, and consumer confidence.

The research reveals that there are two distinct media clusters in the country. The Index shows an increase in media polarization since 2020, particularly acute since early 2022. While the Index captures significant polarization around specific events like elections, its response varies across different events and developments. Notably, public perception of polarization doesn't directly correlate with the media polarization index. Opinion polarization shows an upward trend, but its dynamics diverge from media polarization after March 2022. Political party ratings initially mirrored media polarization patterns, but this link weakened post-2022, possibly due to the impacts on political rhetoric around the Russian war in Ukraine. Analysis of party rating relevance to media polarization shows that the higher the polarization, the higher are ratings of the two largest political parties and the 'middle' (which mainly consists of smaller opposition parties) shrinks. Finally, the media polarization index exhibits a similar upward trend to the consumer confidence index, potentially reflecting the similar influence of events on both metrics. Based on the conducted research/Index the note offers three recommendations for the desired depolarization path, particularly in the context of the European Commission’s recommended nine steps, one of which is depolarization.