The article by De Neve et al. (1) provides a useful basis for examining appropriate policy responses to Covid-19. One aspect, however, requires further elaboration.
The article implies that policymakers must address a trade-off between health benefits and wellbeing costs of lockdown. However, the presumption of a trade-off may be misplaced. Evidence from New Zealand (NZ) shows that lockdowns need not reduce wellbeing, even for disadvantaged groups.
Stats NZ has measured wellbeing before, during and after a Covid-19 lockdown. When implemented, NZ’s lockdown was rated as the most stringent of any OECD country (2); June quarter GDP fell 12.2%. Consequently, however, NZ has experienced a low Covid-19 death rate, ranked 174th globally (3).
NZ has a four-tier alert system. Level 4 involves almost complete lockdown; level 3 involves only a small relaxation. NZ moved to level 3 on 23 March 2020 and to level 4 on 25 March. It gradually returned to level 1 by 8 June. Thus much of the June quarter was spent in levels 2 to 4. A renewed community outbreak caused Auckland to revert to level 3 on 12 August, with the rest of NZ moving to level 2; these restrictions were unwound by 7 October.
Two Stats NZ adult population surveys, the General Social Survey (GSS) and the Household Labour Force Survey (HLFS), demonstrate the effect of these lockdowns on wellbeing. GSS data are available biennially (2014, 2016, 2018). Stats NZ introduced several GSS wellbeing questions into the (quarterly) HLFS for the June and September 2020 surveys (data is available from arthur.grimes@vuw.ac.nz).
We focus on responses to the evaluative subjective wellbeing (life satisfaction) question: “Where zero is completely dissatisfied, and ten is completely satisfied, how do you feel about your life as a whole?” We can compare the cumulative distribution function (CDF) of responses for the pre-lockdown (2018) and lockdown (June 2020) surveys for the total population (Stats NZ aggregates lower responses to a single group, 0-6). The June 2020 (lockdown) CDF lies everywhere below that of the 2018 CDF, so displaying first order welfare dominance (4) over the pre-lockdown position. We can conclude that population subjective wellbeing was unambiguously higher during lockdown than in the prior survey. Indeed, mean life satisfaction was higher during lockdown than it was in any of the three prior surveys.
Demographic decompositions are available for 32 sub-groups according to: sex, age-group, labour force status, migrant status, ethnicity, and region. Of these 32 demographic splits, mean life satisfaction rose for 30 and stayed constant for 2; none fell. Mean life satisfaction during lockdown for Māori and Pacific exceeded any of the prior three surveys; for the unemployed it equalled the previous peak; for sole parents it was bettered only once in the prior three surveys.
Life satisfaction in the post-lockdown period (Sept 2020) remained relatively high compared with the three pre-lockdown surveys, albeit lower than during lockdown.
Self-rated general health status in June 2020 also first order dominated that in 2018 as did self-rated financial wellbeing. The financial wellbeing result may reflect the wage guarantee programme which helped prevent large-scale job losses (unemployment fell from 4.2% to 4.0% between March and June 2020, albeit rising to 5.3% in September).
Another likely contributor to the rise in wellbeing was that the lockdown applied comprehensively to all (other than essential workers). The Prime Minister frequently referred to “the team of 5 million” (NZ’s population) which reinforced the already strong trust in institutions observed in NZ (see World Values Survey). General trust in others rose from 2018 to June 2020 as did institutional trust in each of the police, the media, the health system and parliament.
The NZ evidence indicates that an intense lockdown can improve both health and wellbeing outcomes, even for disadvantaged groups. The intense lockdown gave authorities the option (5) to achieve disease elimination, which proved successful and meant its duration could be short. The comprehensive coverage boosted community cohesion and trust in institutions from already high levels.
Improved wellbeing and health should therefore not be regarded as competing aims to be traded off: both can be achieved through an effective lockdown accompanied by other supportive policies.
Rapid Response:
Lockdowns need not reduce wellbeing
Dear Editor
The article by De Neve et al. (1) provides a useful basis for examining appropriate policy responses to Covid-19. One aspect, however, requires further elaboration.
The article implies that policymakers must address a trade-off between health benefits and wellbeing costs of lockdown. However, the presumption of a trade-off may be misplaced. Evidence from New Zealand (NZ) shows that lockdowns need not reduce wellbeing, even for disadvantaged groups.
Stats NZ has measured wellbeing before, during and after a Covid-19 lockdown. When implemented, NZ’s lockdown was rated as the most stringent of any OECD country (2); June quarter GDP fell 12.2%. Consequently, however, NZ has experienced a low Covid-19 death rate, ranked 174th globally (3).
NZ has a four-tier alert system. Level 4 involves almost complete lockdown; level 3 involves only a small relaxation. NZ moved to level 3 on 23 March 2020 and to level 4 on 25 March. It gradually returned to level 1 by 8 June. Thus much of the June quarter was spent in levels 2 to 4. A renewed community outbreak caused Auckland to revert to level 3 on 12 August, with the rest of NZ moving to level 2; these restrictions were unwound by 7 October.
Two Stats NZ adult population surveys, the General Social Survey (GSS) and the Household Labour Force Survey (HLFS), demonstrate the effect of these lockdowns on wellbeing. GSS data are available biennially (2014, 2016, 2018). Stats NZ introduced several GSS wellbeing questions into the (quarterly) HLFS for the June and September 2020 surveys (data is available from arthur.grimes@vuw.ac.nz).
We focus on responses to the evaluative subjective wellbeing (life satisfaction) question: “Where zero is completely dissatisfied, and ten is completely satisfied, how do you feel about your life as a whole?” We can compare the cumulative distribution function (CDF) of responses for the pre-lockdown (2018) and lockdown (June 2020) surveys for the total population (Stats NZ aggregates lower responses to a single group, 0-6). The June 2020 (lockdown) CDF lies everywhere below that of the 2018 CDF, so displaying first order welfare dominance (4) over the pre-lockdown position. We can conclude that population subjective wellbeing was unambiguously higher during lockdown than in the prior survey. Indeed, mean life satisfaction was higher during lockdown than it was in any of the three prior surveys.
Demographic decompositions are available for 32 sub-groups according to: sex, age-group, labour force status, migrant status, ethnicity, and region. Of these 32 demographic splits, mean life satisfaction rose for 30 and stayed constant for 2; none fell. Mean life satisfaction during lockdown for Māori and Pacific exceeded any of the prior three surveys; for the unemployed it equalled the previous peak; for sole parents it was bettered only once in the prior three surveys.
Life satisfaction in the post-lockdown period (Sept 2020) remained relatively high compared with the three pre-lockdown surveys, albeit lower than during lockdown.
Self-rated general health status in June 2020 also first order dominated that in 2018 as did self-rated financial wellbeing. The financial wellbeing result may reflect the wage guarantee programme which helped prevent large-scale job losses (unemployment fell from 4.2% to 4.0% between March and June 2020, albeit rising to 5.3% in September).
Another likely contributor to the rise in wellbeing was that the lockdown applied comprehensively to all (other than essential workers). The Prime Minister frequently referred to “the team of 5 million” (NZ’s population) which reinforced the already strong trust in institutions observed in NZ (see World Values Survey). General trust in others rose from 2018 to June 2020 as did institutional trust in each of the police, the media, the health system and parliament.
The NZ evidence indicates that an intense lockdown can improve both health and wellbeing outcomes, even for disadvantaged groups. The intense lockdown gave authorities the option (5) to achieve disease elimination, which proved successful and meant its duration could be short. The comprehensive coverage boosted community cohesion and trust in institutions from already high levels.
Improved wellbeing and health should therefore not be regarded as competing aims to be traded off: both can be achieved through an effective lockdown accompanied by other supportive policies.
(1) doi.org/10.1136/bmj.m3853
(2) https://ourworldindata.org/coronavirus
(3) https://www.worldometers.info/coronavirus/#countries
(4) doi/full/10.1080/00779954.2019.1697729
(5) doi.org/10.1080/00779954.2020.1806340
Competing interests: No competing interests