Subscribe

Data analytics reveals SA's happiness level

Sibahle Malinga
By Sibahle Malinga, ITWeb senior news journalist.
Johannesburg, 15 May 2019
Professor Talita Greyling, researcher in the field of wellbeing economics and development economics at UJ.
Professor Talita Greyling, researcher in the field of wellbeing economics and development economics at UJ.

The University of Johannesburg (UJ) has partnered with Auckland University of Technology in New Zealand, to develop a World Happiness Index (WHI) that uses data analytics to track a country's happiness sentiment.

Professor Talita Greyling, a researcher in the field of wellbeing economics and development economics at UJ, along with Auckland University of Technology's Dr Stephanie Rossouw, are capitalising on the use of real-time data analytics and Twitter feeds to measure the gross national happiness (GNH) and real-time sentiment of SA, related to elections and economic activities.

The happiness level is measured on a scale from zero to 10, with zero being very unhappy, five neutral and 10 very happy.

According to the WHI, SA ranked 106th out of 156 countries for 2019, with a score of 4.7.

Governments are able to influence the GNH level through the institutional and policy framework they establish, notes UJ. The links between the government and GNH can operate in both directions: what governments do affects GNH, and in turn, the GNH of citizens in most countries determines what kind of governments they support.

"We have launched this index with the goal of measuring affect happiness on a daily basis, alongside other economic indicators, because we can see the importance of complementing 'hard' objective indicators with 'soft' subjective indicators," explains Greyling.

"Traditionally, economists measured the wellbeing of people or a nation by using objective economic indicators such as gross domestic product. These indicators do not measure wellbeing, but merely specific conditions, which are believed to lead to a good life."

This index was launched in SA in April, the week before the national election, and since then it has been launched in New Zealand and Australia, to monitor the sentiments around the elections in those countries.

The analysis software, provided by media monitoring and compliance company Afstereo, analyses each tweet to determine if the tweet is positive, negative or neutral.

"According to this outcome, we use these results in a sentiment balance algorithm to determine the happiness score. We do this at almost real-time, only with an hour delay, thus we can hourly report on the happiness levels of South Africans.

"We analyse between 45 000 to 100 000 tweets per day for SA, which currently has over 8.3 million active Twitter users (approximately 15% of the population), providing us with a much larger sample size than can be found in any other survey," adds Greyling.

The index reveals South Africans' emotions were on a roller-coaster from pre-elections to post-elections, with election day reaching a higher level of happiness of 6.48, but this was surpassed by Mother's Day, which reflected a very positive sentiment among South Africans.

The results from data collected during the elections and a few days later in SA.
The results from data collected during the elections and a few days later in SA.

The GNH is correlated to economic indicators, notes Greyling.

"When considering the All Share Index of the Johannesburg Stock Exchange, we saw that if the happiness level changes, the All Share Index moves in the same direction.

"With further testing, we found the relationship getting stronger with a time difference of three hours; thus three hours after a change in the level of happiness we observed a 50% correlation and the relationship is significant at a level of 99%, thus we can say we are 99% sure we can trust this result."

In future, the two universities intend to develop a Web portal which can be accessed by anyone to get a real-time happiness report.

Other international happiness tracking indexes include the Gallup World Happiness Index, the World Value Survey and the Happy Planet Index.

Share