Groundbreaking research by University of Johannesburg researchers

Please note: Embargo 31 January 2024:

In global news, groundbreaking research by University of Johannesburg researchers on the effects of macroeconomic shocks on happiness levels in countries applying Big Data and Machine Learning.

A manuscript titled “Reactions to macro-level shocks and re-examination of the adaptation theory using Big Data”, authored by Prof Greyling and Prof Rossouw, published in one of the top-ranked journals, PLOS One, has been selected by PLOS One’s press team to be sent to media houses globally and posted to EurekAlert. (EurekAlert! is one of the biggest online, global news services that brings groundbreaking research to the media)

The study investigated whether two different types of macro-level shocks (lockdown – endogenous and the invasion of Ukraine – exogenous) affected happiness differently. Additionally, they wanted to determine whether the “adaptation theory” (Diener et al., 2006), established using survey data on a micro-level, is upheld at a macro level using Big Data rather than survey data,

They used Difference-in-Difference and Event study models to estimate the effect of these shocks on happiness and to determine the time period that elapses before happiness returns to levels before the shock.

The results showed that happiness strongly decreased after the shocks and that happiness levels adapted quickly after the shocks, only one week after the Ukrainian invasion (remember, the study did not include Ukraine but neighbouring countries) and two to three weeks after the lockdown shock.  The study used data derived in real-time from social media.

These results are significant:

1. It defines shocks' effects on happiness in countries, which is very important because happiness affects economic outcomes, including productivity, future income, and labour market performance. Increased happiness also positively affects a nation's social and health sectors, fosters altruistic behaviour, and enhances various cognitive and social capabilities. Knowing how happiness reacts to shocks in advance informs policymakers to take corrective actions.
2. The study shows that Big Data and 4IR methodologies, such as machine learning, used to measure happiness in real-time, are essential as they provide policymakers with timely information to implement corrective policies sooner rather than later. 
3. The study proves that the adaptation theory (after a shock, happiness levels of individuals return to levels before the shock) also holds at a macro level. However, the time period of adaptation might be slightly longer, lasting up to three weeks.

Contact details: Prof T Greyling – talitag@uj.ac.za