The effects of shocks, tobacco sales bans, and COVID-19 restrictions on mental health and wellbeing: evidence from Tanzania, South Africa, and Australia

Perry, Samuel ORCID: 0000-0002-6098-4369 (2024). The effects of shocks, tobacco sales bans, and COVID-19 restrictions on mental health and wellbeing: evidence from Tanzania, South Africa, and Australia. University of Birmingham. Ph.D.

Preview

Perry2024PhD.pdf
Text - Accepted Version
Available under License All rights reserved.
Download (3MB) | Preview

Abstract

This thesis comprises three empirical studies that delved into the impact of various shocks on mental health and wellbeing. The first study investigated the relationships between self-reported household shocks, conflict events, and food insecurity, on the one hand, and subjective wellbeing (SWB), on the other, in Tanzania. The second study examined the effect of the 2020 tobacco sales ban on the depressive feelings of South African smokers. The third chapter estimated the influence of COVID-19 lockdowns on mental health in Australia, with a specific focus on how personality traits may moderate this effect. All these studies employed individual-level panel survey data.

Using two waves of the Tanzanian National Panel Survey (TZNPS), we investigated the impact of self-reported household shocks, conflict, and food insecurity on SWB. The data was supplemented by data from the Armed Conflict Location and Event Data (ACLED) Project to match conflict events to households based on their geographical proximity. We firstly estimated the direct impact of the shocks, conflict, and food insecurity on life satisfaction using ordered Probit and linear fixed effects models. The findings suggested that only certain self-reported were negatively associated with SWB. These tended to be more idiosyncratic to the household. Additionally, we found increases in conflict events were negatively associated with SWB, albeit the magnitude of this relationship was relatively small. We further found a significant negative relationship between self-reported food insecurity and SWB. We subsequently tested for factors which moderate or mediate these relationships. We found that labour market shocks had a significantly greater impact on SWB for individuals from food insecure households, whilst receipt of financial assistance mitigated this effect. We further found the effect of food insecurity on SWB was heightened for households in the upper consumption quartile, we which postulated may reflect a greater stigma within their socioeconomic status group. Finally, our analysis largely ruled out significant mediating roles for food insecurity and consumption in the effects of the shocks and conflict on SWB.

Next, we used five waves of the South African National Income Dynamics Survey (NIDS) and three waves of the NIDS Coronavirus Rapid Response Mobile Survey (NIDS-CRAM) to estimate the impact of the tobacco sales ban introduced by the South African government in response to the COVID-19 pandemic in March 2020. Using a difference-in-differences (DD) approach, we found that the ban significantly increased frequency of depressive feelings amongst smokers who continued to smoke during the ban, but not amongst those who quit. We attributed the ‘struggling quitters hypothesis’ as a potential driver behind this finding. We subsequently conducted a gender-based analysis which indicated that female smokers were most susceptible to poorer mental health during the ban. We postulated this finding may reflect evidence that women are more likely to smoke to reduce negative emotion and less likely to be successful in cessation attempts than men. Our findings therefore indicated the imposition of the ban may have had an unintended negative mental health effect on smokers and emphasise the need for appropriate cessation support to be provided to smokers.

Lastly, we exploited state-level heterogeneity in COVID-19 policies in Australia to estimate the effect of extended lockdowns on mental health, and further investigated how the ‘Big Five’ personality traits may moderate this effect. To this end, we utilised over 180,000 observations on individuals across 17 waves of the Household, Income, and Labour Dynamics Australia (HILDA) survey covering the period 2005 to 2021. Using an ‘extended’ two-way fixed effects (ETWFE) difference-in-differences design which accounted for staggered treatment timing, we firstly found that extended lockdowns significantly adversely affected the mental health of individuals in affected states, with some variations across cohorts. Moreover, extroversion, emotional stability, and openness to experience traits were found to moderate these effects to varying degrees. Highly extroverted individuals were notably more susceptible to deteriorating mental health during extended lockdowns than highly introverted individuals, although this moderating effect somewhat diminished between the 2020 and 2021 extended lockdowns. Additionally, individuals with high levels of emotional stability were largely protected from the mental health effects of extended lockdowns, as were, to a lesser extent, individuals who scored higher on openness to experience, with this latter effect dissipating in 2021. Our findings add to the growing literature documenting the unintended consequences of lockdown policies, and indicate policymakers need to seriously consider the pros and cons such polices if they were ever again required.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):