effect of lockdown during covid 19 essay

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

The good, the bad and the ugly of lockdowns during Covid-19

Contributed equally to this work with: Talita Greyling, Stephanie Rossouw, Tamanna Adhikari

Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

Affiliation School of Economics, University of Johannesburg, Johannesburg, South Africa

Roles Conceptualization, Data curation, Investigation, Methodology, Project administration, Software, Supervision, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

Affiliation School of Economics, Auckland University of Technology, Auckland, New Zealand

Roles Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing

• Talita Greyling, 
• Stephanie Rossouw, 
• Tamanna Adhikari

• Published: January 22, 2021
• https://doi.org/10.1371/journal.pone.0245546
• Peer Review
• Reader Comments

Amidst the rapid global spread of Covid-19, many governments enforced country-wide lockdowns, with likely severe well-being consequences. In this regard, South Africa is an extreme case suffering from low levels of well-being, but at the same time enforcing very strict lockdown regulations. In this study, we analyse the causal effect of a lockdown and consequently, the determinants of happiness during the aforementioned. A difference-in-difference approach is used to make causal inferences on the lockdown effect on happiness, and an OLS estimation investigates the determinants of happiness after lockdown. The results show that the lockdown had a significant and negative impact on happiness. In analysing the determinants of happiness after lockdown, we found that stay-at-home orders have positively impacted happiness during this period. On the other hand, other lockdown regulations such as a ban on alcohol sales, a fear of becoming unemployed and a greater reliance on social media have negative effects, culminating in a net loss in happiness. Interestingly, Covid-19, proxied by new deaths per day, had an inverted U-shape relationship with happiness. Seemingly people were, at the onset of Covid-19 positive and optimistic about the low fatality rates and the high recovery rates. However, as the pandemic progressed, they became more concerned, and this relationship changed and became negative, with peoples' happiness decreasing as the number of new deaths increased.

Citation: Greyling T, Rossouw S, Adhikari T (2021) The good, the bad and the ugly of lockdowns during Covid-19. PLoS ONE 16(1): e0245546. https://doi.org/10.1371/journal.pone.0245546

Editor: Francesco Di Gennaro, National Institute for Infectious Diseases Lazzaro Spallanzani-IRCCS, ITALY

Received: July 29, 2020; Accepted: December 30, 2020; Published: January 22, 2021

Copyright: © 2021 Greyling et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: The following authors received salaries from their institutions, whom were also the funders of the research. 1. Prof T Greyling: University of Johannesburg via the University Research Fund. 2. Dr Stephanie Rossouw: Auckland University of Technology via the Faculty of Business, Economics and Law. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

1. Introduction

In an attempt to curb the spread of Covid-19 and minimise the loss of life, governments around the world have imposed their version of mandatory self-isolation through implementing lockdown regulations. Unfortunately, restricting people's mobility and depriving them of what matters most might intensify the negative effect on happiness levels.

In an extreme country case, this might be amplified. In this study, we treat an extreme country as a country with very strict lockdown regulations, with likely high infection rates, amidst low levels of well-being. We define well-being as those aspects of life that society collectively agrees are important for a person's quality of life, happiness and welfare. One of the dimensions of well-being, material (income) hinges on a bleak economic outlook.

To this end, our primary aim in this study is to use the Gross National Happiness Index (GNH), a real-time measure of well-being, derived from Big Data, to investigate if lockdown regulations in itself caused a decrease in happiness. Secondly, we determine which factors matter most (factors significantly influencing happiness) to happiness under these changed circumstances. We accomplish these aims by using two econometric techniques: difference-in-difference (DiD) and ordinary least squares (OLS).

Against this backdrop, the current study makes several contributions to the literature:

• Determining whether lockdown regulations cause a decline in happiness –in an extreme country case scenario.
• Investigating specifically the determinants of happiness during a lockdown, whereas other studies have focused on mental well-being and related matters (see section 2).
• Being one of the few studies (see also Rossouw, Greyling and Adhikari; Greyling, Rossouw and Adhikari [ 1 , 2 ]) that investigates the effect of lockdown on happiness making use of real-time Big Data . Other studies such as Hamermesh [ 3 ] and Brodeur et al. [ 4 ] also use Big Data, though limited to Google Trends (see section 2).

These results give policymakers the necessary information to take action in increasing the happiness of the nation and set the scene for increased economic, social and political well-being. It also allows them to reflect on happiness outcomes due to their policy actions. An additional benefit of the current study is that policymakers do not need to wait for extended periods to see the consequences of their policies, as we are making use of real-time data, with immediate information. Usually, policymakers can only evaluate their own decision making, with significant time-lags, prolonging the implementation of corrective actions.

Our results indicate lockdown itself causes a decrease in happiness. Furthermore, in an extreme country case (a country under stringent lockdown regulations coupled with low levels of well-being) what matters most to happiness under lockdown is the factors directly linked to the regulations that were implemented. These factors can be classified as (i) social capital issues; lack of access to alcohol (and tobacco), increased social media usage, and more time to spend at home, of which all are negatively related to happiness except the stay-at-home factor, and (ii) economic issues; concerns over jobs and the threat of retrenchments, which are negatively related to happiness. The finding on the stay-at-home order is interesting as even though lockdown itself caused a decline in happiness, it seems that people adjust and over time begin to appreciate the benefits of staying at home.

Noteworthy is that Covid-19, proxied by new deaths per day, had an inverted U-shape relationship to happiness. Seemingly people were, at the onset of Covid-19, positive and optimistic as the fatality rate was relatively low and recovery rates high. However, as the pandemic progressed, they became more concerned, and this relationship changed and became negative, with peoples' happiness decreasing as the number of new Covid-19 deaths increased.

The rest of the paper is structured as follows. The next section contains a brief background on South Africa and briefly discusses literature about happiness and studies conducted on the impacts of pandemics and consequently lockdown regulations. Section 3 describes the data, the selected variables and outlines the methodology used. The results follow in section 4, while the paper concludes in section 5.

2. Background and literature review

2.1 south africa.

In this study, we focus on South Africa because it presents us with a unique case to investigate the effect of a lockdown on happiness when levels of well-being are already low. Health and income, two dimensions of well-being, was significantly affected, although in opposite directions. Health was positively affected by the lockdown since it limited the spread of Covid-19. At the stage of writing the paper (3 June 2020), the number of new Covid-19 cases were nearly 120,000 (John Hopkins University [ 5 ]). On the other hand, the economic outlook of the country, and therefore peoples' incomes, was negatively affected. This opposite effect has led to significant debates on the value of the implementation of the lockdown.

Furthermore, South Africa implemented one of the most stringent lockdown regulations (comparable to the Philippines and Jordan), which exacerbated the costs to well-being and the economy while already experiencing a severe economic downturn. Therefore, South Africa is an example of an extreme country case which unfortunately amplifies the effects of the difficult choices made by policymakers. Therefore, we take advantage of this unique country case and determine how stringent lockdown regulations impact happiness during a one in 100-year event.

In South Africa, there are five levels of differing lockdown regulations, with alert level 5 being the most stringent and alert level 1 the most relaxed. The idea behind these levels is to curb the spread of Covid-19 and give time to South Africa's health system to prepare itself. Additionally, as they move down in levels, South Africans receive increasingly more of their previous liberties back. During level 5, which was announced 23 March 2020 and implemented on 27 March 2020, South Africans were only allowed to leave their homes to purchase or produce essential goods. All South Africans were instructed to work from home, there was no travel allowed, the sale of alcohol and tobacco were banned, people were not allowed to exercise outside their homes, and the police and defence force ensured compliance to the restrictions. South Africa moved to level 4 lockdown on 01 May 2020. With this move, they received back the ability to exercise outside from 6 am—9 am, purchase more than just essential goods, including food deliveries as long as it was within curfew.

Interestingly, the sale of alcohol and tobacco was still banned. On 01 June they moved to level 3, allowing restricted sales of alcohol (Mondays to Thursdays) and the re-opening of certain businesses. However, the services industry, especially beauty and tourism, remained closed. At the time of writing this paper, South Africa was still under level 3 lockdown.

Whereas everybody understands that the Covid-19 infections curve needs to be flattened, there are grave concerns that these stringent lockdown regulations will also flatline South Africa's well-being and economy. Before the Covid-19 lockdown, South Africa's average happiness levels were 6.32 compared to an average of 7.23 and 7.16 in Australia and New Zealand, respectively (Greyling et al. [ 2 ]). Furthermore, South Africa had a 29 per cent unemployment rate, and the gross domestic product (GDP) has been estimated to shrink by 7 per cent in 2020 (Bureau of Economic Analysis [ 6 ]). According to the South African Reserve Bank [ 7 ], an additional 3 to 7 million people can potentially become unemployed as a direct consequence of the pandemic, thereby increasing unemployment rates to approximately 50 per cent. The country's sovereign credit rating was downgraded to junk status in March 2020, which impacted on political stability, the level of the national debt and debt interest payments. Add to this already grim situation, the fact that consumption of South Africans has been declining in 2020, with a significant decrease seen after lockdown, then one can very easily see how the well-being and happiness levels in South Africa can plummet.

2.2 Happiness

Why should we care whether people's happiness is adversely impacted by not only a global pandemic but also by the response from the government? The studies of Helliwell, Layard, Stiglitz et al., Veenhoven, Diener and Seligman and others [ 8 – 12 ], have shown beyond a shadow of a doubt that if policymakers want to maximise the quality of life of their citizens, they need to consider subjective measures of well-being. Piekalkiewicz [ 13 ] states that happiness may act as a determinant of economic outcomes: it increases productivity, predicts one's future income and affects labour market performance. By measuring happiness, individuals themselves reveal their preference and assigned priority to various domains, which cannot be identified by a measure such as GDP. As was pointed out by Layard [ 9 ], while economists use exactly the right framework for thinking about public policy, the accounts we use of what makes people happy are wrong. In layman's terms, we say that utility increases with the opportunities for voluntary exchange. However, Layard [ 14 ] argues that this overlooks the significance of involuntary interactions between people. Policymakers should formulate policy to maximise happiness or well-being, as is the main aim of many constitutions. This can be achieved by directing economic, social, political and environmental policy to maximise well-being while acknowledging that people's norms, aspirations, feelings and emotions are important. Thereby underscoring that understanding and measuring happiness should be an integral part of the efforts to maximise the quality of life.

On the other hand, if people's happiness is negatively affected, such as it was in the wake of the Covid-19 pandemic and the implementation of lockdown regulations, there are far-reaching consequences.

These consequences are as follows:

• Social capital: unhappier people display less altruistic behaviour in the long run (Dunn et al. [ 15 ]). They are also less active, less creative, poor problem solvers, less social, and display more anti-social behaviour (Lyubomirsky et al. [ 16 ]). If unhappier people display more anti-social behaviour, South Africa could see an increase in behaviour such as unrests, violent strikes and perhaps higher crime rates.
• Health care: unhappier people are less physically healthy and die sooner (Lyubomirsky et al. [ 16 ]). Additionally, unhappy people engage in riskier behaviour such as smoking and drinking, thereby placing unnecessary pressure on national health systems.
• Economic: unhappy workers are typically less productive, in particular in jobs that require sociability and problem solving (Bryson et al. [ 17 ]). If an economy can raise the rate of growth of productivity, by ensuring their workers are happier, then the trend growth of national output can pick up.

2.3 Literature on the determinants of happiness during a lockdown

Having established that policymakers should strive to maximise the happiness of their people, it is necessary to know what determines happiness. Previous studies have investigated, at a macro-level, what influences happiness and found that economic growth, unemployment and inflation play a significant role (Stevenson and Wolfers, Perović, Sacks et al. [ 18 – 20 ]). However, these studies were conducted during 'normal' periods and not under such conditions that are currently plaguing the world. The current study will have the opportunity to investigate this exact question, namely what determines happiness during a lockdown driven by a global health pandemic.

Naturally, the number of studies being conducted to examine the effect of Covid-19 is growing exponentially. This increasing interest in the effect of a global pandemic as well as the policies implemented by governments on peoples' well-being, come on the back of relatively few studies conducted during prior pandemics such as SARS and the H1N1. When SARS hit in 2002 and then again when H1N1 hit in 2009, scholars were only truly starting to understand that for governments to formulate policies to increase well-being, you needed to measure well-being. Of the current studies being conducted on the effect of Covid-19 or lockdown regulations on all affected domains, not many studies are in a position to use real-time Big Data, such as we do.

In layman's terms, Big Data is a phrase used to describe a massive volume of both structured (for example stock information) and unstructured data (for example tweets) generated through information and communication technologies such as the Internet (Rossouw and Greyling [ 1 ]). At the time of writing this paper, the following studies were closest aligned with our study and focused on:

• nationwide lockdown on institutional trust, attitudes to government, health and well-being, using survey data collected at two points in time (December 2019 and April 2020) (1003 respondents) (Sibley et al. [ 21 ]). Their preliminary results showed a small increase in people's sense of community and trust. However, they also found an increase in anxiety/depression post-lockdown and hinted at longer-term challenges to mental health.
• the happiness of married and single people while in government-imposed lockdown by running simulations to formulate predictions, using Google Trends data (Hamermesh [ 3 ]). Not surprisingly, married people were more satisfied with life than single people.
• the timing of decision-making by politicians to release lockdown based on a comparison of economic benefits with the social and psychological benefits versus the cost, increase in deaths if policymakers released lockdown too early (Layard et al. [ 22 ])
• the stages of GNH using a Markov switching model in New Zealand (Rossouw et al. [ 23 ]). They found that happiness was at a lower level and the unhappy state lasted longer than was expected. Furthermore, they found that the factors important for New Zealand's happiness post-Covid-19 were related to international travel, employment and mobility.
• exploring Covid-19 related determinants of life dissatisfaction and feelings of anxiety in a cross-country study using survey data collected between 23 March and 30 April (de Pedraza et al. [ 24 ]). They found that persons with poorer general health, without employment, living without a partner, not exercising daily and those actively seeking out loneliness report higher dissatisfaction and higher anxiety. Additionally, they found that the effect of Covid-19 on dissatisfaction and anxiety levels off with a higher number of cases.

2.4 Literature on the causal effect of a lockdown

To the knowledge of the authors, there are only two papers that investigated the causal effect between lockdowns and population well-being. Brodeur et al. [ 4 ] investigated the changes in well-being (and mental health) in the United States and Europe after a lockdown was implemented, using Google Trends data. They found an increase in searches for loneliness, worry and sadness, which indicated a negative effect on mental health. Greyling et al. [ 2 ] conducted a cross-country study investigating the effect of lockdown on happiness. They found that lockdown caused a negative effect on happiness, notwithstanding the different characteristics of the countries (South Africa, New Zealand and Australia), the duration and the type of lockdown regulations. When they compared the effect size of the lockdown regulations, they found that South Africa, with the most stringent lockdown regulations incurred the greatest happiness costs.

Brodeur et al. [ 4 ] study analysed data from one Big Data source, Google Trends and collected data for a short period between only 01 January 2019 and 10 April 2020 in countries that had introduced a full lockdown by the end of this period. Greyling et al. [ 2 ] study used both Google Trends and the GNH index but did not investigate the determinants of happiness after lockdown for an extreme country case.

In summary, taking all of the above into consideration, there is not one study which determines causality between lockdown and happiness and analyses the determinants of happiness in an extreme country case using real-time , Big Data . Therefore, our study is the first of its kind.

3. Data and methodology

To estimate the causal effects of a lockdown on happiness, we use a Difference-in-Difference (DiD) approach (see section 3.3.1). The technique compares happiness (dependent variable), before and after the treatment (the lockdown) to a counterfactual time period in the year before. For the control period, we select the same time period, with the same number of days in 2019, corresponding to the number of days in 2020, thus 152 days in each year (01 January 2020 to 03 June 2020, excluding 29 February 2020). Our results should thus be interpreted as the average impact of the lockdown on happiness, comparing pre and post-lockdown in 2020 to the same time period in 2019, which we assume had normal levels of Gross National Happiness (see a discussion on the GNH in section 3.2.1). In this manner, we also account for seasonal trends in happiness.

In the analyses, we make use of daily data for South Africa. As high-frequency data available at almost real-time, is scarce, we make use of novel Big Data methodologies to harvest data. Additionally, we use the Oxford Stringency dataset that was released in May 2020, which includes data related to lockdown regulations, such as time-series data on the stay-at-home index, Covid-19 cases and Covid-19 deaths (Hale et al., Roser et al. [ 25 , 26 ]).

3.2 Selection of variables

The selection of the variables included in our estimations is based on the reviewed literature, the contents of tweets related to the lockdown and data availability.

3.2.1 Gross National Happiness Index–the dependent variable.

To measure happiness (the dependent variable), we make use of the Gross National Happiness Index (GNH), which was launched in April 2019 (Greyling, Rossouw and Afstereo [ 27 ]). This project measures the happiness (mood) of a country's citizens during different economic, social and political events.

Since February 2020, the researchers extended the project that initially analysed the sentiment of tweets, to incorporate the analysis of the emotions underpinning tweets. The team did this to determine which emotions are most prominent on specific days or events.

To construct the GNH index, the researchers use Big Data methods and extract tweets from the voluntary information-sharing social media platform Twitter. Big Data, such as Twitter, provides real-time information for policymakers to assist them when facing short-term deadlines with imperfect information. Big Data also allows governments to 'listen' and capture those variables which their citizens deem to be important for their well-being, rather than relying on pre-defined economic utility theories. Big Data offers governments the opportunity to observe people's behaviour and not just their opinions. This approach of revealed preferences unveils a reflexive picture of society because it allows the main concerns of citizens (and the priority ranking of those concerns) to emerge spontaneously, and it complements as such the information captured by gross domestic product. Lastly, Big Data does not suffer from non-response bias (Callegaro and Yang [ 28 ]).

Greyling, Rossouw and Afstereo [ 27 ] apply sentiment analysis to a live Twitter-feed and label every tweet as having either a positive, neutral or negative sentiment. This sentiment classification is then applied to a sentiment-balance algorithm to derive a happiness score. The happiness scores range between 0 and 10, with five being neutral, thus neither happy nor unhappy.

All tweets per day are extracted, and a happiness score per hour is calculated. The index is available live on the GNH website (Greyling, Rossouw and Afstereo [ 27 ]). In South Africa, the average number of tweets extracted for 2020 is 68,524 per day. South Africa has approximately 11 million Twitter users, representing almost 18 per cent of the population (Omnicore [ 29 ]). Although the number of tweets is extensive and represents significant proportions of the populations of the countries, it is not representative. However, Twitter accommodates individuals, groups of individuals, organisations and media outlets, representing a kind of disaggregated sample, thus giving access to the moods of a vast blend of Twitter users, not found in survey data.

Furthermore, purely based on the vast numbers of the tweets, it seems that the GNH index gives a remarkably robust reflection of the evaluative mood of a nation. Also, we correlate the GNH index with 'depression' and 'anxiety', derived from the 'Global behaviors and perceptions at the onset of the Covid-19 Pandemic data ' survey, for the period from 01 March 2020 (OFS [ 30 ]). We find it negative and statistically significant related, therefore, it seems that the GNH index derived from Big Data gives similar trends to survey data. (We would have appreciated the opportunity to correlate the GNH to a happiness measure–but a happiness measure, as such, was not included in the survey).

Considering the GNH index over time, we found that the index accurately reflects a nation's emotions, for example, when South Africa won the Rugby World Cup on 02 November 2019, the happiness index accurately depicted the joy experienced by South Africans ( Fig 1 ). The hourly happiness score was 7.9 at 13:00, the highest score ever measured, at the exact time that the final whistle was blown to announce the victory of the Springboks over England.

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

Source: Authors' calculations using GNH dataset (Greyling et al. [ 27 ]).

https://doi.org/10.1371/journal.pone.0245546.g001

Also, when the famous American basketball player, Kobe Bryant and his daughter Gigi, tragically passed away on 27 January 2020, the happiness index once again captured the negative mood of the nation, and the happiness score decreased to 5.8, significantly below the mean (see Fig 2 ). The result of the GNH mirrors the one determined by the Hedonometer, which recorded an average happiness score of 5.89 on the day of Bryant's death. The top three words that made this day sadder than the previous seven were 'crash', 'died' and 'rip'.

https://doi.org/10.1371/journal.pone.0245546.g002

3.2.2 The selection of covariates included in estimations.

We found ourselves in uncharted territory, as there are limited studies estimating happiness functions during a lockdown (see Brodeur et al., Greyling et al., Rossouw et al. [ 2 , 4 , 23 ]). As a result, we considered these studies and the tweets to determine the factors to consider, which might influence happiness during a lockdown , as well as the most tweeted subjects. It was evident from the tweets that the main topics of discussion related to economic concerns, the prohibition of the sale of alcohol and tobacco, the stay-at-home orders and the Covid-19 pandemic itself.

To estimate our difference-in-difference model, we restricted our covariates to the lockdown variable, a year effect, the difference-in-difference estimator and controlled for new Covid-19 deaths, job searches and searches for alcohol. We were restricted in the number of covariates due to the limited observations and potentially encountering the issue of over-identification of the models. Therefore, we selected those variables which were available for both 2019 and 2020, and which were also trending subjects during the lockdown period. We were not able to add a stay-at-home variable which captures the lack of mobility, as the counterfactual time period is then not comparable to 2020.

Lockdown, our treatment variable, divides the sample into two distinct time periods: before the announcement of the lockdown, 23 March 2020 and thereafter. We make use of the date of the announcement of the lockdown rather than the date of the implementation, as this showed the severest effect on happiness (see Brodeur et al. [ 4 ]).

The Covid-19 pandemic and consequent spread of the virus is the reason for the lockdown. As such, we include the number of new daily Covid-19 deaths as well as its square. This will allow us to control for the likelihood of a U-shaped relationship between the number of Covid-19 deaths and happiness. Furthermore, there is likely a lagged effect on happiness due to Covid-19 deaths being reported in the media only the following day. Therefore, we lag these variables by one day. We derive the data from the Oxford Stringency data set (Hale et al. and Roser et al. [ 25 , 26 ]).

To measure jobs (a proxy for future job uncertainty) and the sale of alcohol and tobacco, we use the methodology as set out by Nuti et al. and Brodeur et al. [ 4 , 31 ] and use daily searches on Google Trends (see also Simionescu and Zimmermann [ 7 ]). We considered searches for both the alcohol and tobacco topic; however, the variables follow the same trends during the lockdown period and are highly correlated (r = 0.83). We are, furthermore, restricted in the number of covariates to include in the model and decided to include only 'alcohol' in the estimations. We justify this decision since the ban of alcohol affects a larger proportion of the population. It is estimated that 41 per cent of males and 17.1 per cent of females consume on average 9.3 litres of alcohol per capita annually whereas only 17.6 per cent smoke (Peltzer et al. and Reddy et al. [ 32 , 33 ]). However, as a robustness check, we also run all estimations using the searches for tobacco.

It should be noted that when we use Google Trends data, it is expressed as an index between 0 and 100 with 0 being the "least" interest and 100 being the "most" interest shown in the topic for the year. However, the series are not comparable across years as the underlying data is sourced from different search requests for each of the two years. To address this, we use a scaling procedure outlined in Brodeur et al. [ 4 ]. First, we generate "weekly" interest weights for each day by expressing the average weekly score that a particular daily score fell on, as a proportion of the average yearly score. Then, we multiply the daily scores with these weights to obtain weighted search trends. Finally, we normalise these weighted search trends to render us a score between 0 and 100, which is comparable across years.

Other topics that are trending are related to the 'stay-at-home' orders. The Oxford Stringency dataset includes a time series variable on the stay-at-home orders. It differs on a day to day basis according to its stringency. It is an ordinal variable plus binary of geographic scope. It takes the value 0 if there are no stay-at-home orders and 1 if the government recommends not leaving your house. Value 2 represents people not leaving their homes with the exceptions of daily exercise, grocery shopping, and 'essential' trips. Not leaving your home with minimal exceptions (e.g. allowed to leave only once a week, or only one person can leave at a time, etc.) takes the value 3 (Hale et al. [ 25 ]).

Furthermore, we include the number of tweets per day, as it forms part of the Twitter data extracted daily for South Africa (Greyling et al. [ 27 ]), which is a proxy for connectivity. It also gauges the opportunity cost of not being able to have face to face interactions, which seems to be negatively related to happiness (Chae and Wilson et al. [ 34 , 35 ]). Interestingly the number of tweets increased markedly during the lockdown period, from an average of 60,708 to 80,000 tweets per day. Table 1 provides descriptive statistics for the variables included in the estimations.

https://doi.org/10.1371/journal.pone.0245546.t001

3.3 Methodology

3.3.1 difference-in-difference..

Where GNH it is the daily happiness for South Africa at time t . The treatment variable lockdown takes the value of 0 pre-announcement day (23 March) and one post-announcement of lockdown in both the year of the actual lockdown (2020) as well as the year before the lockdown (2019). Year is a dummy variable where 1 is the year 2020. We control for new deaths per million with a one-day lag as well as the quadratic effect of new deaths per million on GNH. Additionally, we control for the effect of job and alcohol searches. As a robustness test, we use the number of new Covid-19 cases instead of new Covid-19 deaths (see Table 4 in S1 Appendix ).

Our main coefficient of interest is the interaction between the lockdown and the year variable. If it is found to be significant, it provides evidence of a causal effect of the lockdown on the dependent variable, in the current year, notwithstanding the trend in 2019.

3.3.2 OLS regression.

Here, y t refers to the Gross National Happiness Index (GNH) for each day and X t is a vector of several relevant covariates to account for the changes in the happiness levels during the lockdown period. μ t is the error term.

Due to the various factors that affect happiness, some of our independent variables may be correlated with the error term, leading to endogeneity concerns. Depending on the direction of the correlation between the error term and the X-variable, the coefficient could be biased upwards or downwards. For instance, the coefficient on the indicator for jobs is likely biased upwards as it, in all likelihood, shows the effect of concerns about jobs as well as some other negative economic shock on happiness. In the absence of panel data or an appropriate instrument, it is difficult to ascertain causality to Eq ( 2 ). However, simply correlating the covariates and the error term we find all levels of correlation to be less than 0.3, although a very basic test, this still indicates that the likelihood that endogeneity might bias estimations is relatively small. A natural extension of the work, as better data becomes available with time, would be to address these concerns.

We cannot rule out the probability of autocorrelation and heterogeneity in our data, especially due to its time-series nature. We use robust standard errors to account for this. The choice of our controls, however, comes with a caveat. Seeing as we only have 81 daily observations using a larger battery of covariates would lead to problems arising due to overfitting of the model. This issue is considered in Green [ 36 ], who suggests a minimum of 50 observations for any regressions as well as an additional eight observations per additional term. As a robustness test, we included tobacco rather than alcohol products (see Table 5 in S1 Appendix ).

4. Results and analysis

4.1 difference-in-difference estimation.

Fig 3 tracks the dependent variable (GNH) over the time period before and after the date of the lockdown (23 March) in the year of the lockdown (2020) and the year preceding it. On the day of the announcement of the lockdown and for a few successive days, we see a sharp downwards departure from the 2019 trend, assumed to be normal.

https://doi.org/10.1371/journal.pone.0245546.g003

Table 2 provides the results for the difference-in-difference specification, which helps us to make causal inferences on the effect of the lockdown on the GNH. At the outset, we notice a negative and significant effect of the 'year' variable (p<0.001), showing that on average the GNH was lower in 2020 than in 2019.

https://doi.org/10.1371/journal.pone.0245546.t002

We control for trends in job searches (a proxy for job uncertainty) and alcohol searches (a proxy for increased interest in alcohol-related topics in the specification). Both variables show a negative association with GNH, implying if there are more searches for jobs or alcohol, reflecting a scarcity in these items, GNH decreases. The negative effect of alcohol is statistically significant at the 1% level (p<0.001). We also control for lagged new Covid-19 related deaths and lagged new Covid-19 related deaths squared, both are significant (p<0.001). Interesting is the finding of the significant inverted U-shaped relationship between new Covid-19 deaths and happiness ( Fig 4 ). In the earlier stages of the pandemic, with very few new Covid-19 deaths, it appears that people were positive and optimistic as the fatality rates were very low and the recovery rates very high. However, as time progressed, the higher fatality rates turned the relationship around so that the number of new Covid-19 cases were negatively related to happiness.

https://doi.org/10.1371/journal.pone.0245546.g004

To determine if the decrease in GNH was due to the lockdown (the treatment) specifically and not just due to the year trend, we must consider the estimated coefficient of the interaction variable 'lockdown and year'. We report a negative and statistically significant coefficient (p-value 0.064) on the interaction variable, indicating that 'lockdown' caused, on average a decrease in GNH of 0.101 points when compared to its mean values for average 2019 values, controlling for the general trend in the two years. Thus, we can conclude that the lockdown caused a decline in GNH in 2020 compared to 2019. The decline of 0.101 may seem small at first but given the low general levels of happiness in South Africa compared to other countries (Greyling et al. [ 27 ]) the reduction is substantial.

4.2 Regression analysis

To address the second research question, namely, to determine the factors that are related to happiness after the lockdown was implemented, we consider the results of Table 3 .

https://doi.org/10.1371/journal.pone.0245546.t003

Table 3 shows that job searches (p-value 0.09), searches for alcohol (p-value<0.001) and the number of tweets is negatively related to happiness. In contrast, the stay-at-home index is positively related to happiness (p-value<0.001). The squared relationship between new Covid-19 deaths and happiness is negative and statistically significant (p-value<0.001), indicating that this relationship changed over time as was highlighted in section 4.1. Suppose we consider the relatively low mortality rate (0.02 per cent of confirmed cases in the early stages) compared to countries such as the USA (3.9 per cent), the U.K. (15.4 per cent) and Spain (9.4 per cent). In that case, it could explain the initial positive relationship between the number of new Covid-19 deaths and happiness. Although as time passed and the death rate increased (currently, the mortality rate is at 1.5 per cent of confirmed cases), this relationship became negative. As a robustness test, we used the number of new Covid-19 cases and its square instead of the new Covid-19 deaths and found similar results (see Table 4 in S1 Appendix ).

As expected, job searches, a proxy for uncertainty about the future job market is negatively related to happiness (p-value<0.001). Analysing the tweets, we realised that this is a major concern, which is closely related to economic concerns. The economic performance of South Africa in the last year has been weak with high levels of unemployment (increase to 50 per cent), low growth rates (GDP is expected to contract with 7 per cent in 2020) and high debt to income ratios (government debt as a percentage of GDP– 80 per cent). In a recent survey conducted by Statistics South Africa on behavioural and health impacts of the Covid-19 pandemic (Statistics South Africa [ 37 ]), it was found that 95 per cent of the respondents were very concerned about the economy. In contrast, only 60 per cent was concerned about the Covid-19 pandemic itself. This supports our current findings in that economic factors matter more to South Africans happiness levels than Covid-19 itself.

Alcohol-related searches are also found to be negatively related to happiness (p-value<0.001). Considering the close correlation between alcohol and tobacco products, we can assume that what holds for alcohol products, also holds for tobacco products. As a robustness test, we excluded the alcohol variable and included searches for tobacco variable and found very similar results (see Table 5 in S1 Appendix ). South Africa is one of the very few countries globally that have banned alcohol and tobacco sales during the Covid-19 pandemic. It is argued that these products contribute to the negative effects of the virus. The banning of these products had severe implications on different levels of society. Individuals see this as a major infringement of their human rights, negatively affecting their happiness. Furthermore, research done by Sommer et al. [ 38 ] proved that because of the presence of hordenine in beer, it significantly contributes to mood-elevation. In South Africa, which is well-known for its high per capita beer and alcohol consumption (Statistics South Africa [ 39 ]), also related to 'socialising', the ban on these products had a severe negative effect on happiness. Even in level 3, when the ban on alcohol sales was lifted, but still restricted, we found this negative relationship.

The restricted sale of alcohol and tobacco has indirect consequences for South Africans happiness via the economic impact since these industries are two of the largest industries in South Africa. They employ people across the whole supply chain from production to retail. Due to the ban on these industries, people can potentially lose their jobs. Lastly, the government sector forgoes all taxes on these products. This is against the backdrop of the recent downgrade of South Africa's debt rating to junk status in an already very uncertain fiscal environment. If all of these factors are aggregated, we can understand that the cumulative effect of the banning and restriction of sales of these products severely decreases the happiness levels. In Table 5 in S1 Appendix , we use tobacco searches instead of alcohol to estimate the determinants of happiness, which gives us results that are qualitatively similar to Table 3 .

The number of tweets is negatively related to happiness (p-value<0.001). Previous research has shown that increases in the use of social media are negatively related to happiness (Rolland et al., Chae and Wilson et al. [ 34 , 40 , 41 ]). Noteworthy is that the number of tweets during the lockdown period increased significantly from an average of 60,708 per day before the lockdown to 80,000 per day after the lockdown indicating that more people used social media during the lockdown period.

Interesting is the result of the stay-at-home orders being positively related to happiness (p-value<0.001). On analysing the contents of the tweets, we find the following. South Africans are wary of contracting Covid-19, and therefore, they abide by the stay-at-home orders and social distancing regulations to minimise the risk. That means that the stay-at-home orders in itself increase happiness; it is only once the other lockdown regulations are added that a total decrease in happiness levels are experienced.

Additional benefits revealed from analysing the tweets show that being at home provides a more peaceful and calmer environment compared to the rushed experience outside their homes. Also, people in the suburbs seem to be more convivial, with strangers greeting one another as people went for short walks around their neighbourhoods. In general, people have more time to spend with their loved ones. People earning salaries incur major savings, as there is less opportunity to spend money. People also save on commuting to and from workplaces and other destinations. One of the unexpected benefits of the stay-at-home orders is the much lower crime rates experienced in the country. Homes are constantly occupied, limiting the risk of residential robberies (-3.8 per cent). Other types of crimes such as murder (-72 per cent), rape (-87.2 per cent) and carjacking (-80.9 per cent) are much lower as well (Adapted from the speech of Police Minister Cele 2020 [ 42 ]).

In summary, what changed when the lockdown regulations to curb the spread of Covid-19 were implemented? It caused a significant decrease in happiness, and factors related to the lockdown regulations became relevant determinants of happiness.

5. Conclusions

In this paper, we used the Gross National Happiness Index (GNH), a real-time measure of well-being, derived from Big Data, to investigate whether lockdown regulations caused a decrease in happiness. Additionally, we determined which factors matter to happiness under these changed circumstances. We accomplished these aims by using two models: difference-in-difference and ordinary least squares.

We added to the current literature by determining causality between lockdown and happiness and analysing the determinants of happiness after a lockdown in an extreme country case using real-time , Big Data . Subsequently, this was the first study of its kind.

To determine if the decrease in GNH was due to the lockdown, specifically, we considered the estimated coefficient of the interaction variable 'lockdown and year'. We found a negative and statistically significant coefficient on the interaction variable, indicating that the lockdown caused a decline in GNH in 2020 compared to 2019.

As regards to the factors that are related to happiness after the lockdown was implemented, we found searches for alcohol (tobacco), the number of tweets and uncertainty about the future job market to be negatively related to happiness. In contrast, stay-at-home orders are positively related to happiness. Interesting is the negative and statistically significant squared relationship between new Covid-19 deaths and happiness, indicating that this relationship initially was positive but became negative over time.

Considering the results mentioned above, it ultimately means that if policymakers want to increase happiness levels and increase the probability to achieve the happiness levels of 2019, they must consider those factors that matter most to peoples' happiness. These factors include allowing creatures of habits some of their lost comforts by reinstating the sale of alcohol and tobacco. In saying that, we do advocate for responsible alcohol use by all South Africans and note that the significant effect of the ban on the sale of alcohol could be confounded by the restriction on social gatherings as well.

These results are important for countries in similar well-being situations, thus low levels of happiness, a diverse state of the economy and an increasing number of Covid-19 cases to evaluate what the effect of a strict lockdown is.

Additionally, policymakers should assure citizens that there is a credible plan to get the economy, which is currently in dire straits, back on track. Such an economic plan should stimulate growth, create job opportunities and increase employment rates, supply the necessary infrastructure and deal with curbing vast budget deficits and debt burdens. Hopefully, such policies will fuel the dying embers of a dying economy and increase well-being levels.

Lastly, it would be remiss of us not to note the limitations of the study. First, we were restricted in the number of covariates we could add to our difference-in-difference model due to the limited observations and therefore potentially overidentifying the models. Second, regarding the inverted U-shaped relationship between new Covid-19 deaths and happiness, we acknowledge that there might be confounding factors at play, initially seen as ‘positives' of the lockdown, but later turned into negatives. However, using alternative sets of covariates in the regression analyses, the inverted U-shape between new Covid-19 deaths and happiness remained. Therefore we trust that the revealed relationship is robust.

Supporting information

S1 appendix. robustness checks..

https://doi.org/10.1371/journal.pone.0245546.s001

https://doi.org/10.1371/journal.pone.0245546.s002

Acknowledgments

We would like to thank our colleagues Professor Emeritus John Knight from Oxford University and Dr Kelsey O'Connor from STATEC Luxembourg, for their generosity in providing feedback on the study.

• 1. Rossouw S. & Greyling T. (2020). Big Data and Happiness. Invited chapter for the Handbook of Labor, Human Resources and Population Economics. Edited by Zimmermann Klaus F.
• 2. Greyling T, Rossouw S, Adhikari T. A tale of three countries: How did Covid-19 lockdown impact happiness? GLO Discussion Paper; 2020.
• 3. Hamermesh DS. Lockdowns, loneliness and life satisfaction. National Bureau of Economic Research; 2020 Apr 16. https://doi.org/10.1007/s11150-020-09495-x pmid:32837455
• View Article
• Google Scholar
• 5. John Hopkins University. Coronavirus Resource Centre. (2020). Available at https://coronavirus.jhu.edu/map.html Accessed 08 May 2020.
• 6. Bureau of Economic Research (2020). Available from https://www.ber.ac.za/BER%20Documents/Economic-Prospects/?doctypeid=1059 Accessed 08 June 2020
• 7. Simionescu M, Zimmermann KF. Big data and unemployment analysis. GLO Discussion Paper; 2017.
• 8. Helliwell JF. The Economics of Happiness. The National Bureau of Economic Research. 2015. Reporter 2:14–17.
• 10. Stevenson B, Wolfers J. Economic growth and subjective well-being: Reassessing the Easterlin paradox. National Bureau of Economic Research; 2008 Aug 29.
• PubMed/NCBI
• 14. Layard R. Happiness: Lessons from a new science. Penguin UK; 2011 Apr 7.
• 18. Statistics South Africa. Behavioural and health impacts of the COVID-19 pandemic in South Africa. 2020. Available at http://www.statssa.gov.za/publications/Report-00-80-02/Report-00-80-022020.pdf Accessed 08 June 2020
• 20. Sacks DW, Stevenson B, Wolfers J. Subjective Well-Being, Income, Economic Development, and Growth. In Annual World Bank Conference on Development Economics 2011: Development Challenges in a Post-crisis World 2013 Oct 11 (p. 283). World Bank Publications.
• 22. Layard R, Clark AE, De Neve JE, Krekel C, Fancourt D, Hey N, et al. When to Release the Lockdown? A Wellbeing Framework for Analysing Costs and Benefits. IZA Discussion Papers; 2020.
• 23. Rossouw S, Greyling T, Adhikari T, Morrison PS. Markov switching models for happiness during a pandemic: The New-Zealand experience. GLO Discussion Paper; 2020.
• 24. de Pedraza P, Guzi M, Tijdens K. Life Dissatisfaction and Anxiety in COVID-19 pandemic. GLO Discussion Paper; 2020.
• 25. Hale T, Petherick A, Phillips T, Webster S. Variation in government responses to COVID-19. Blavatnik School of Government Working Paper. 2020 May 27;31
• 26. Roser, M., Ritchie, H., Ortiz-Ospina, E. & Hasell, J. Coronavirus Pandemic (COVID-19) 2020. Available at https://ourworldindata.org/coronavirus Accessed on 18 June 2020.
• 27. Greyling T, Rossouw S, Afstereo. Gross National Happiness Index. The University of Johannesburg and Afstereo [producers]. 2019. Available at http://gnh.today . Accessed 08 May 2020.
• 29. Omnicore. Omnicore Agency. Available at https://www.omnicoreagency.com/ Accessed on 26 February 2020.
• 30. OFS. Global Behaviors and Perceptions in the COVID-19 Pandemic. 2020. Available at https://osf.io/g7fn9/ Accessed on 08 May 2020.
• 37. Statistics South Africa. Beer is more popular than vegetables. 2017. http://www.statssa.gov.za/?p=10269 Accessed 04 May 2020.
• 38. South African Reserve Bank (2020). Available from https://www.resbank.co.za/Research/Statistics/Pages/Statistics-Home.aspx 08 June 2020
• 42. Crime levels before South Africa's lockdown–and what they look like now. Business Tech. 2020 April 23. Available from https://businesstech.co.za/news/lifestyle/391877/crime-levels-before-south-africas-lockdown-and-what-they-look-like-now/

Log in using your username and password

• Search More Search for this keyword Advanced search
• Latest content
• Current issue
• BMJ Journals

You are here

• Volume 76, Issue 2
• COVID-19 pandemic and its impact on social relationships and health
• Article Text
• Article info
• Citation Tools
• Rapid Responses
• Article metrics

• http://orcid.org/0000-0003-1512-4471 Emily Long 1 ,
• Susan Patterson 1 ,
• Karen Maxwell 1 ,
• Carolyn Blake 1 ,
• http://orcid.org/0000-0001-7342-4566 Raquel Bosó Pérez 1 ,
• Ruth Lewis 1 ,
• Mark McCann 1 ,
• Julie Riddell 1 ,
• Kathryn Skivington 1 ,
• Rachel Wilson-Lowe 1 ,
• http://orcid.org/0000-0002-4409-6601 Kirstin R Mitchell 2
• 1 MRC/CSO Social and Public Health Sciences Unit , University of Glasgow , Glasgow , UK
• 2 MRC/CSO Social and Public Health Sciences Unit, Institute of Health & Wellbeing , University of Glasgow , Glasgow , UK
• Correspondence to Dr Emily Long, MRC/CSO Social and Public Health Sciences Unit, University of Glasgow, Glasgow G3 7HR, UK; emily.long{at}glasgow.ac.uk

This essay examines key aspects of social relationships that were disrupted by the COVID-19 pandemic. It focuses explicitly on relational mechanisms of health and brings together theory and emerging evidence on the effects of the COVID-19 pandemic to make recommendations for future public health policy and recovery. We first provide an overview of the pandemic in the UK context, outlining the nature of the public health response. We then introduce four distinct domains of social relationships: social networks, social support, social interaction and intimacy, highlighting the mechanisms through which the pandemic and associated public health response drastically altered social interactions in each domain. Throughout the essay, the lens of health inequalities, and perspective of relationships as interconnecting elements in a broader system, is used to explore the varying impact of these disruptions. The essay concludes by providing recommendations for longer term recovery ensuring that the social relational cost of COVID-19 is adequately considered in efforts to rebuild.

• inequalities

Data availability statement

Data sharing not applicable as no data sets generated and/or analysed for this study. Data sharing not applicable as no data sets generated or analysed for this essay.

This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/ .

https://doi.org/10.1136/jech-2021-216690

Statistics from Altmetric.com

Request permissions.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Infectious disease pandemics, including SARS and COVID-19, demand intrapersonal behaviour change and present highly complex challenges for public health. 1 A pandemic of an airborne infection, spread easily through social contact, assails human relationships by drastically altering the ways through which humans interact. In this essay, we draw on theories of social relationships to examine specific ways in which relational mechanisms key to health and well-being were disrupted by the COVID-19 pandemic. Relational mechanisms refer to the processes between people that lead to change in health outcomes.

At the time of writing, the future surrounding COVID-19 was uncertain. Vaccine programmes were being rolled out in countries that could afford them, but new and more contagious variants of the virus were also being discovered. The recovery journey looked long, with continued disruption to social relationships. The social cost of COVID-19 was only just beginning to emerge, but the mental health impact was already considerable, 2 3 and the inequality of the health burden stark. 4 Knowledge of the epidemiology of COVID-19 accrued rapidly, but evidence of the most effective policy responses remained uncertain.

The initial response to COVID-19 in the UK was reactive and aimed at reducing mortality, with little time to consider the social implications, including for interpersonal and community relationships. The terminology of ‘social distancing’ quickly became entrenched both in public and policy discourse. This equation of physical distance with social distance was regrettable, since only physical proximity causes viral transmission, whereas many forms of social proximity (eg, conversations while walking outdoors) are minimal risk, and are crucial to maintaining relationships supportive of health and well-being.

The aim of this essay is to explore four key relational mechanisms that were impacted by the pandemic and associated restrictions: social networks, social support, social interaction and intimacy. We use relational theories and emerging research on the effects of the COVID-19 pandemic response to make three key recommendations: one regarding public health responses; and two regarding social recovery. Our understanding of these mechanisms stems from a ‘systems’ perspective which casts social relationships as interdependent elements within a connected whole. 5

Social networks

Social networks characterise the individuals and social connections that compose a system (such as a workplace, community or society). Social relationships range from spouses and partners, to coworkers, friends and acquaintances. They vary across many dimensions, including, for example, frequency of contact and emotional closeness. Social networks can be understood both in terms of the individuals and relationships that compose the network, as well as the overall network structure (eg, how many of your friends know each other).

Social networks show a tendency towards homophily, or a phenomenon of associating with individuals who are similar to self. 6 This is particularly true for ‘core’ network ties (eg, close friends), while more distant, sometimes called ‘weak’ ties tend to show more diversity. During the height of COVID-19 restrictions, face-to-face interactions were often reduced to core network members, such as partners, family members or, potentially, live-in roommates; some ‘weak’ ties were lost, and interactions became more limited to those closest. Given that peripheral, weaker social ties provide a diversity of resources, opinions and support, 7 COVID-19 likely resulted in networks that were smaller and more homogenous.

Such changes were not inevitable nor necessarily enduring, since social networks are also adaptive and responsive to change, in that a disruption to usual ways of interacting can be replaced by new ways of engaging (eg, Zoom). Yet, important inequalities exist, wherein networks and individual relationships within networks are not equally able to adapt to such changes. For example, individuals with a large number of newly established relationships (eg, university students) may have struggled to transfer these relationships online, resulting in lost contacts and a heightened risk of social isolation. This is consistent with research suggesting that young adults were the most likely to report a worsening of relationships during COVID-19, whereas older adults were the least likely to report a change. 8

Lastly, social connections give rise to emergent properties of social systems, 9 where a community-level phenomenon develops that cannot be attributed to any one member or portion of the network. For example, local area-based networks emerged due to geographic restrictions (eg, stay-at-home orders), resulting in increases in neighbourly support and local volunteering. 10 In fact, research suggests that relationships with neighbours displayed the largest net gain in ratings of relationship quality compared with a range of relationship types (eg, partner, colleague, friend). 8 Much of this was built from spontaneous individual interactions within local communities, which together contributed to the ‘community spirit’ that many experienced. 11 COVID-19 restrictions thus impacted the personal social networks and the structure of the larger networks within the society.

Social support

Social support, referring to the psychological and material resources provided through social interaction, is a critical mechanism through which social relationships benefit health. In fact, social support has been shown to be one of the most important resilience factors in the aftermath of stressful events. 12 In the context of COVID-19, the usual ways in which individuals interact and obtain social support have been severely disrupted.

One such disruption has been to opportunities for spontaneous social interactions. For example, conversations with colleagues in a break room offer an opportunity for socialising beyond one’s core social network, and these peripheral conversations can provide a form of social support. 13 14 A chance conversation may lead to advice helpful to coping with situations or seeking formal help. Thus, the absence of these spontaneous interactions may mean the reduction of indirect support-seeking opportunities. While direct support-seeking behaviour is more effective at eliciting support, it also requires significantly more effort and may be perceived as forceful and burdensome. 15 The shift to homeworking and closure of community venues reduced the number of opportunities for these spontaneous interactions to occur, and has, second, focused them locally. Consequently, individuals whose core networks are located elsewhere, or who live in communities where spontaneous interaction is less likely, have less opportunity to benefit from spontaneous in-person supportive interactions.

However, alongside this disruption, new opportunities to interact and obtain social support have arisen. The surge in community social support during the initial lockdown mirrored that often seen in response to adverse events (eg, natural disasters 16 ). COVID-19 restrictions that confined individuals to their local area also compelled them to focus their in-person efforts locally. Commentators on the initial lockdown in the UK remarked on extraordinary acts of generosity between individuals who belonged to the same community but were unknown to each other. However, research on adverse events also tells us that such community support is not necessarily maintained in the longer term. 16

Meanwhile, online forms of social support are not bound by geography, thus enabling interactions and social support to be received from a wider network of people. Formal online social support spaces (eg, support groups) existed well before COVID-19, but have vastly increased since. While online interactions can increase perceived social support, it is unclear whether remote communication technologies provide an effective substitute from in-person interaction during periods of social distancing. 17 18 It makes intuitive sense that the usefulness of online social support will vary by the type of support offered, degree of social interaction and ‘online communication skills’ of those taking part. Youth workers, for instance, have struggled to keep vulnerable youth engaged in online youth clubs, 19 despite others finding a positive association between amount of digital technology used by individuals during lockdown and perceived social support. 20 Other research has found that more frequent face-to-face contact and phone/video contact both related to lower levels of depression during the time period of March to August 2020, but the negative effect of a lack of contact was greater for those with higher levels of usual sociability. 21 Relatedly, important inequalities in social support exist, such that individuals who occupy more socially disadvantaged positions in society (eg, low socioeconomic status, older people) tend to have less access to social support, 22 potentially exacerbated by COVID-19.

Social and interactional norms

Interactional norms are key relational mechanisms which build trust, belonging and identity within and across groups in a system. Individuals in groups and societies apply meaning by ‘approving, arranging and redefining’ symbols of interaction. 23 A handshake, for instance, is a powerful symbol of trust and equality. Depending on context, not shaking hands may symbolise a failure to extend friendship, or a failure to reach agreement. The norms governing these symbols represent shared values and identity; and mutual understanding of these symbols enables individuals to achieve orderly interactions, establish supportive relationship accountability and connect socially. 24 25

Physical distancing measures to contain the spread of COVID-19 radically altered these norms of interaction, particularly those used to convey trust, affinity, empathy and respect (eg, hugging, physical comforting). 26 As epidemic waves rose and fell, the work to negotiate these norms required intense cognitive effort; previously taken-for-granted interactions were re-examined, factoring in current restriction levels, own and (assumed) others’ vulnerability and tolerance of risk. This created awkwardness, and uncertainty, for example, around how to bring closure to an in-person interaction or convey warmth. The instability in scripted ways of interacting created particular strain for individuals who already struggled to encode and decode interactions with others (eg, those who are deaf or have autism spectrum disorder); difficulties often intensified by mask wearing. 27

Large social gatherings—for example, weddings, school assemblies, sporting events—also present key opportunities for affirming and assimilating interactional norms, building cohesion and shared identity and facilitating cooperation across social groups. 28 Online ‘equivalents’ do not easily support ‘social-bonding’ activities such as singing and dancing, and rarely enable chance/spontaneous one-on-one conversations with peripheral/weaker network ties (see the Social networks section) which can help strengthen bonds across a larger network. The loss of large gatherings to celebrate rites of passage (eg, bar mitzvah, weddings) has additional relational costs since these events are performed by and for communities to reinforce belonging, and to assist in transitioning to new phases of life. 29 The loss of interaction with diverse others via community and large group gatherings also reduces intergroup contact, which may then tend towards more prejudiced outgroup attitudes. While online interaction can go some way to mimicking these interaction norms, there are key differences. A sense of anonymity, and lack of in-person emotional cues, tends to support norms of polarisation and aggression in expressing differences of opinion online. And while online platforms have potential to provide intergroup contact, the tendency of much social media to form homogeneous ‘echo chambers’ can serve to further reduce intergroup contact. 30 31

Intimacy relates to the feeling of emotional connection and closeness with other human beings. Emotional connection, through romantic, friendship or familial relationships, fulfils a basic human need 32 and strongly benefits health, including reduced stress levels, improved mental health, lowered blood pressure and reduced risk of heart disease. 32 33 Intimacy can be fostered through familiarity, feeling understood and feeling accepted by close others. 34

Intimacy via companionship and closeness is fundamental to mental well-being. Positively, the COVID-19 pandemic has offered opportunities for individuals to (re)connect and (re)strengthen close relationships within their household via quality time together, following closure of many usual external social activities. Research suggests that the first full UK lockdown period led to a net gain in the quality of steady relationships at a population level, 35 but amplified existing inequalities in relationship quality. 35 36 For some in single-person households, the absence of a companion became more conspicuous, leading to feelings of loneliness and lower mental well-being. 37 38 Additional pandemic-related relational strain 39 40 resulted, for some, in the initiation or intensification of domestic abuse. 41 42

Physical touch is another key aspect of intimacy, a fundamental human need crucial in maintaining and developing intimacy within close relationships. 34 Restrictions on social interactions severely restricted the number and range of people with whom physical affection was possible. The reduction in opportunity to give and receive affectionate physical touch was not experienced equally. Many of those living alone found themselves completely without physical contact for extended periods. The deprivation of physical touch is evidenced to take a heavy emotional toll. 43 Even in future, once physical expressions of affection can resume, new levels of anxiety over germs may introduce hesitancy into previously fluent blending of physical and verbal intimate social connections. 44

The pandemic also led to shifts in practices and norms around sexual relationship building and maintenance, as individuals adapted and sought alternative ways of enacting sexual intimacy. This too is important, given that intimate sexual activity has known benefits for health. 45 46 Given that social restrictions hinged on reducing household mixing, possibilities for partnered sexual activity were primarily guided by living arrangements. While those in cohabiting relationships could potentially continue as before, those who were single or in non-cohabiting relationships generally had restricted opportunities to maintain their sexual relationships. Pornography consumption and digital partners were reported to increase since lockdown. 47 However, online interactions are qualitatively different from in-person interactions and do not provide the same opportunities for physical intimacy.

Recommendations and conclusions

In the sections above we have outlined the ways in which COVID-19 has impacted social relationships, showing how relational mechanisms key to health have been undermined. While some of the damage might well self-repair after the pandemic, there are opportunities inherent in deliberative efforts to build back in ways that facilitate greater resilience in social and community relationships. We conclude by making three recommendations: one regarding public health responses to the pandemic; and two regarding social recovery.

Recommendation 1: explicitly count the relational cost of public health policies to control the pandemic

Effective handling of a pandemic recognises that social, economic and health concerns are intricately interwoven. It is clear that future research and policy attention must focus on the social consequences. As described above, policies which restrict physical mixing across households carry heavy and unequal relational costs. These include for individuals (eg, loss of intimate touch), dyads (eg, loss of warmth, comfort), networks (eg, restricted access to support) and communities (eg, loss of cohesion and identity). Such costs—and their unequal impact—should not be ignored in short-term efforts to control an epidemic. Some public health responses—restrictions on international holiday travel and highly efficient test and trace systems—have relatively small relational costs and should be prioritised. At a national level, an earlier move to proportionate restrictions, and investment in effective test and trace systems, may help prevent escalation of spread to the point where a national lockdown or tight restrictions became an inevitability. Where policies with relational costs are unavoidable, close attention should be paid to the unequal relational impact for those whose personal circumstances differ from normative assumptions of two adult families. This includes consideration of whether expectations are fair (eg, for those who live alone), whether restrictions on social events are equitable across age group, religious/ethnic groupings and social class, and also to ensure that the language promoted by such policies (eg, households; families) is not exclusionary. 48 49 Forethought to unequal impacts on social relationships should thus be integral to the work of epidemic preparedness teams.

Recommendation 2: intelligently balance online and offline ways of relating

A key ingredient for well-being is ‘getting together’ in a physical sense. This is fundamental to a human need for intimate touch, physical comfort, reinforcing interactional norms and providing practical support. Emerging evidence suggests that online ways of relating cannot simply replace physical interactions. But online interaction has many benefits and for some it offers connections that did not exist previously. In particular, online platforms provide new forms of support for those unable to access offline services because of mobility issues (eg, older people) or because they are geographically isolated from their support community (eg, lesbian, gay, bisexual, transgender and queer (LGBTQ) youth). Ultimately, multiple forms of online and offline social interactions are required to meet the needs of varying groups of people (eg, LGBTQ, older people). Future research and practice should aim to establish ways of using offline and online support in complementary and even synergistic ways, rather than veering between them as social restrictions expand and contract. Intelligent balancing of online and offline ways of relating also pertains to future policies on home and flexible working. A decision to switch to wholesale or obligatory homeworking should consider the risk to relational ‘group properties’ of the workplace community and their impact on employees’ well-being, focusing in particular on unequal impacts (eg, new vs established employees). Intelligent blending of online and in-person working is required to achieve flexibility while also nurturing supportive networks at work. Intelligent balance also implies strategies to build digital literacy and minimise digital exclusion, as well as coproducing solutions with intended beneficiaries.

Recommendation 3: build stronger and sustainable localised communities

In balancing offline and online ways of interacting, there is opportunity to capitalise on the potential for more localised, coherent communities due to scaled-down travel, homeworking and local focus that will ideally continue after restrictions end. There are potential economic benefits after the pandemic, such as increased trade as home workers use local resources (eg, coffee shops), but also relational benefits from stronger relationships around the orbit of the home and neighbourhood. Experience from previous crises shows that community volunteer efforts generated early on will wane over time in the absence of deliberate work to maintain them. Adequately funded partnerships between local government, third sector and community groups are required to sustain community assets that began as a direct response to the pandemic. Such partnerships could work to secure green spaces and indoor (non-commercial) meeting spaces that promote community interaction. Green spaces in particular provide a triple benefit in encouraging physical activity and mental health, as well as facilitating social bonding. 50 In building local communities, small community networks—that allow for diversity and break down ingroup/outgroup views—may be more helpful than the concept of ‘support bubbles’, which are exclusionary and less sustainable in the longer term. Rigorously designed intervention and evaluation—taking a systems approach—will be crucial in ensuring scale-up and sustainability.

The dramatic change to social interaction necessitated by efforts to control the spread of COVID-19 created stark challenges but also opportunities. Our essay highlights opportunities for learning, both to ensure the equity and humanity of physical restrictions, and to sustain the salutogenic effects of social relationships going forward. The starting point for capitalising on this learning is recognition of the disruption to relational mechanisms as a key part of the socioeconomic and health impact of the pandemic. In recovery planning, a general rule is that what is good for decreasing health inequalities (such as expanding social protection and public services and pursuing green inclusive growth strategies) 4 will also benefit relationships and safeguard relational mechanisms for future generations. Putting this into action will require political will.

Ethics statements

Patient consent for publication.

Not required.

• Office for National Statistics (ONS)
• Ford T , et al
• Riordan R ,
• Ford J , et al
• Glonti K , et al
• McPherson JM ,
• Smith-Lovin L
• Granovetter MS
• Fancourt D et al
• Stadtfeld C
• Office for Civil Society
• Cook J et al
• Rodriguez-Llanes JM ,
• Guha-Sapir D
• Patulny R et al
• Granovetter M
• Winkeler M ,
• Filipp S-H ,
• Kaniasty K ,
• de Terte I ,
• Guilaran J , et al
• Wright KB ,
• Martin J et al
• Gabbiadini A ,
• Baldissarri C ,
• Durante F , et al
• Sommerlad A ,
• Marston L ,
• Huntley J , et al
• Turner RJ ,
• Bicchieri C
• Brennan G et al
• Watson-Jones RE ,
• Amichai-Hamburger Y ,
• McKenna KYA
• Page-Gould E ,
• Aron A , et al
• Pietromonaco PR ,
• Timmerman GM
• Bradbury-Jones C ,
• Mikocka-Walus A ,
• Klas A , et al
• Marshall L ,
• Steptoe A ,
• Stanley SM ,
• Campbell AM
• ↵ (ONS), O.f.N.S., Domestic abuse during the coronavirus (COVID-19) pandemic, England and Wales . Available: https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/articles/domesticabuseduringthecoronaviruscovid19pandemicenglandandwales/november2020
• Rosenberg M ,
• Hensel D , et al
• Banerjee D ,
• Bruner DW , et al
• Bavel JJV ,
• Baicker K ,
• Boggio PS , et al
• van Barneveld K ,
• Quinlan M ,
• Kriesler P , et al
• Mitchell R ,
• de Vries S , et al

Twitter @karenmaxSPHSU, @Mark_McCann, @Rwilsonlowe, @KMitchinGlasgow

Contributors EL and KM led on the manuscript conceptualisation, review and editing. SP, KM, CB, RBP, RL, MM, JR, KS and RW-L contributed to drafting and revising the article. All authors assisted in revising the final draft.

Funding The research reported in this publication was supported by the Medical Research Council (MC_UU_00022/1, MC_UU_00022/3) and the Chief Scientist Office (SPHSU11, SPHSU14). EL is also supported by MRC Skills Development Fellowship Award (MR/S015078/1). KS and MM are also supported by a Medical Research Council Strategic Award (MC_PC_13027).

Competing interests None declared.

Provenance and peer review Not commissioned; externally peer reviewed.

Read the full text or download the PDF:

• Fact sheets
• Facts in pictures
• Publications
• Questions and answers
• Tools and toolkits
• Endometriosis
• Excessive heat
• Mental disorders
• Polycystic ovary syndrome
• All countries
• Eastern Mediterranean
• South-East Asia
• Western Pacific
• Data by country
• Country presence 
• Country strengthening 
• Country cooperation strategies 
• News releases
• Feature stories
• Press conferences
• Commentaries
• Photo library
• Afghanistan
• Cholera 
• Coronavirus disease (COVID-19)
• Greater Horn of Africa
• Israel and occupied Palestinian territory
• Disease Outbreak News
• Situation reports
• Weekly Epidemiological Record
• Surveillance
• Health emergency appeal
• International Health Regulations
• Independent Oversight and Advisory Committee
• Classifications
• Data collections
• Global Health Observatory
• Global Health Estimates
• Mortality Database
• Sustainable Development Goals
• Health Inequality Monitor
• Global Progress
• World Health Statistics
• Partnerships
• Committees and advisory groups
• Collaborating centres
• Technical teams
• Organizational structure
• Initiatives
• General Programme of Work
• WHO Academy
• Investment in WHO
• WHO Foundation
• External audit
• Financial statements
• Internal audit and investigations 
• Programme Budget
• Results reports
• Governing bodies
• World Health Assembly
• Executive Board
• Member States Portal

Impact of COVID-19 on people's livelihoods, their health and our food systems

Joint statement by ilo, fao, ifad and who.

The COVID-19 pandemic has led to a dramatic loss of human life worldwide and presents an unprecedented challenge to public health, food systems and the world of work. The economic and social disruption caused by the pandemic is devastating: tens of millions of people are at risk of falling into extreme poverty, while the number of undernourished people, currently estimated at nearly 690 million, could increase by up to 132 million by the end of the year.

Millions of enterprises face an existential threat. Nearly half of the world’s 3.3 billion global workforce are at risk of losing their livelihoods. Informal economy workers are particularly vulnerable because the majority lack social protection and access to quality health care and have lost access to productive assets. Without the means to earn an income during lockdowns, many are unable to feed themselves and their families. For most, no income means no food, or, at best, less food and less nutritious food. 

The pandemic has been affecting the entire food system and has laid bare its fragility. Border closures, trade restrictions and confinement measures have been preventing farmers from accessing markets, including for buying inputs and selling their produce, and agricultural workers from harvesting crops, thus disrupting domestic and international food supply chains and reducing access to healthy, safe and diverse diets. The pandemic has decimated jobs and placed millions of livelihoods at risk. As breadwinners lose jobs, fall ill and die, the food security and nutrition of millions of women and men are under threat, with those in low-income countries, particularly the most marginalized populations, which include small-scale farmers and indigenous peoples, being hardest hit.

Millions of agricultural workers – waged and self-employed – while feeding the world, regularly face high levels of working poverty, malnutrition and poor health, and suffer from a lack of safety and labour protection as well as other types of abuse. With low and irregular incomes and a lack of social support, many of them are spurred to continue working, often in unsafe conditions, thus exposing themselves and their families to additional risks. Further, when experiencing income losses, they may resort to negative coping strategies, such as distress sale of assets, predatory loans or child labour. Migrant agricultural workers are particularly vulnerable, because they face risks in their transport, working and living conditions and struggle to access support measures put in place by governments. Guaranteeing the safety and health of all agri-food workers – from primary producers to those involved in food processing, transport and retail, including street food vendors – as well as better incomes and protection, will be critical to saving lives and protecting public health, people’s livelihoods and food security.

In the COVID-19 crisis food security, public health, and employment and labour issues, in particular workers’ health and safety, converge. Adhering to workplace safety and health practices and ensuring access to decent work and the protection of labour rights in all industries will be crucial in addressing the human dimension of the crisis. Immediate and purposeful action to save lives and livelihoods should include extending social protection towards universal health coverage and income support for those most affected. These include workers in the informal economy and in poorly protected and low-paid jobs, including youth, older workers, and migrants. Particular attention must be paid to the situation of women, who are over-represented in low-paid jobs and care roles. Different forms of support are key, including cash transfers, child allowances and healthy school meals, shelter and food relief initiatives, support for employment retention and recovery, and financial relief for businesses, including micro, small and medium-sized enterprises. In designing and implementing such measures it is essential that governments work closely with employers and workers.

Countries dealing with existing humanitarian crises or emergencies are particularly exposed to the effects of COVID-19. Responding swiftly to the pandemic, while ensuring that humanitarian and recovery assistance reaches those most in need, is critical.

Now is the time for global solidarity and support, especially with the most vulnerable in our societies, particularly in the emerging and developing world. Only together can we overcome the intertwined health and social and economic impacts of the pandemic and prevent its escalation into a protracted humanitarian and food security catastrophe, with the potential loss of already achieved development gains.

We must recognize this opportunity to build back better, as noted in the Policy Brief issued by the United Nations Secretary-General. We are committed to pooling our expertise and experience to support countries in their crisis response measures and efforts to achieve the Sustainable Development Goals. We need to develop long-term sustainable strategies to address the challenges facing the health and agri-food sectors. Priority should be given to addressing underlying food security and malnutrition challenges, tackling rural poverty, in particular through more and better jobs in the rural economy, extending social protection to all, facilitating safe migration pathways and promoting the formalization of the informal economy.

We must rethink the future of our environment and tackle climate change and environmental degradation with ambition and urgency. Only then can we protect the health, livelihoods, food security and nutrition of all people, and ensure that our ‘new normal’ is a better one.

Media Contacts

Kimberly Chriscaden

Communications Officer World Health Organization

Nutrition and Food Safety (NFS) and COVID-19

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• BMJ Open Access

COVID-19 pandemic and its impact on social relationships and health

1 MRC/CSO Social and Public Health Sciences Unit, University of Glasgow, Glasgow, UK

Susan Patterson

Karen maxwell, carolyn blake, raquel bosó pérez, mark mccann, julie riddell, kathryn skivington, rachel wilson-lowe, kirstin r mitchell.

2 MRC/CSO Social and Public Health Sciences Unit, Institute of Health & Wellbeing, University of Glasgow, Glasgow, UK

Associated Data

Data sharing not applicable as no data sets generated and/or analysed for this study. Data sharing not applicable as no data sets generated or analysed for this essay.

This essay examines key aspects of social relationships that were disrupted by the COVID-19 pandemic. It focuses explicitly on relational mechanisms of health and brings together theory and emerging evidence on the effects of the COVID-19 pandemic to make recommendations for future public health policy and recovery. We first provide an overview of the pandemic in the UK context, outlining the nature of the public health response. We then introduce four distinct domains of social relationships: social networks, social support, social interaction and intimacy, highlighting the mechanisms through which the pandemic and associated public health response drastically altered social interactions in each domain. Throughout the essay, the lens of health inequalities, and perspective of relationships as interconnecting elements in a broader system, is used to explore the varying impact of these disruptions. The essay concludes by providing recommendations for longer term recovery ensuring that the social relational cost of COVID-19 is adequately considered in efforts to rebuild.

Introduction

Infectious disease pandemics, including SARS and COVID-19, demand intrapersonal behaviour change and present highly complex challenges for public health. 1 A pandemic of an airborne infection, spread easily through social contact, assails human relationships by drastically altering the ways through which humans interact. In this essay, we draw on theories of social relationships to examine specific ways in which relational mechanisms key to health and well-being were disrupted by the COVID-19 pandemic. Relational mechanisms refer to the processes between people that lead to change in health outcomes.

At the time of writing, the future surrounding COVID-19 was uncertain. Vaccine programmes were being rolled out in countries that could afford them, but new and more contagious variants of the virus were also being discovered. The recovery journey looked long, with continued disruption to social relationships. The social cost of COVID-19 was only just beginning to emerge, but the mental health impact was already considerable, 2 3 and the inequality of the health burden stark. 4 Knowledge of the epidemiology of COVID-19 accrued rapidly, but evidence of the most effective policy responses remained uncertain.

The initial response to COVID-19 in the UK was reactive and aimed at reducing mortality, with little time to consider the social implications, including for interpersonal and community relationships. The terminology of ‘social distancing’ quickly became entrenched both in public and policy discourse. This equation of physical distance with social distance was regrettable, since only physical proximity causes viral transmission, whereas many forms of social proximity (eg, conversations while walking outdoors) are minimal risk, and are crucial to maintaining relationships supportive of health and well-being.

The aim of this essay is to explore four key relational mechanisms that were impacted by the pandemic and associated restrictions: social networks, social support, social interaction and intimacy. We use relational theories and emerging research on the effects of the COVID-19 pandemic response to make three key recommendations: one regarding public health responses; and two regarding social recovery. Our understanding of these mechanisms stems from a ‘systems’ perspective which casts social relationships as interdependent elements within a connected whole. 5

Social networks

Social networks characterise the individuals and social connections that compose a system (such as a workplace, community or society). Social relationships range from spouses and partners, to coworkers, friends and acquaintances. They vary across many dimensions, including, for example, frequency of contact and emotional closeness. Social networks can be understood both in terms of the individuals and relationships that compose the network, as well as the overall network structure (eg, how many of your friends know each other).

Social networks show a tendency towards homophily, or a phenomenon of associating with individuals who are similar to self. 6 This is particularly true for ‘core’ network ties (eg, close friends), while more distant, sometimes called ‘weak’ ties tend to show more diversity. During the height of COVID-19 restrictions, face-to-face interactions were often reduced to core network members, such as partners, family members or, potentially, live-in roommates; some ‘weak’ ties were lost, and interactions became more limited to those closest. Given that peripheral, weaker social ties provide a diversity of resources, opinions and support, 7 COVID-19 likely resulted in networks that were smaller and more homogenous.

Such changes were not inevitable nor necessarily enduring, since social networks are also adaptive and responsive to change, in that a disruption to usual ways of interacting can be replaced by new ways of engaging (eg, Zoom). Yet, important inequalities exist, wherein networks and individual relationships within networks are not equally able to adapt to such changes. For example, individuals with a large number of newly established relationships (eg, university students) may have struggled to transfer these relationships online, resulting in lost contacts and a heightened risk of social isolation. This is consistent with research suggesting that young adults were the most likely to report a worsening of relationships during COVID-19, whereas older adults were the least likely to report a change. 8

Lastly, social connections give rise to emergent properties of social systems, 9 where a community-level phenomenon develops that cannot be attributed to any one member or portion of the network. For example, local area-based networks emerged due to geographic restrictions (eg, stay-at-home orders), resulting in increases in neighbourly support and local volunteering. 10 In fact, research suggests that relationships with neighbours displayed the largest net gain in ratings of relationship quality compared with a range of relationship types (eg, partner, colleague, friend). 8 Much of this was built from spontaneous individual interactions within local communities, which together contributed to the ‘community spirit’ that many experienced. 11 COVID-19 restrictions thus impacted the personal social networks and the structure of the larger networks within the society.

Social support

Social support, referring to the psychological and material resources provided through social interaction, is a critical mechanism through which social relationships benefit health. In fact, social support has been shown to be one of the most important resilience factors in the aftermath of stressful events. 12 In the context of COVID-19, the usual ways in which individuals interact and obtain social support have been severely disrupted.

One such disruption has been to opportunities for spontaneous social interactions. For example, conversations with colleagues in a break room offer an opportunity for socialising beyond one’s core social network, and these peripheral conversations can provide a form of social support. 13 14 A chance conversation may lead to advice helpful to coping with situations or seeking formal help. Thus, the absence of these spontaneous interactions may mean the reduction of indirect support-seeking opportunities. While direct support-seeking behaviour is more effective at eliciting support, it also requires significantly more effort and may be perceived as forceful and burdensome. 15 The shift to homeworking and closure of community venues reduced the number of opportunities for these spontaneous interactions to occur, and has, second, focused them locally. Consequently, individuals whose core networks are located elsewhere, or who live in communities where spontaneous interaction is less likely, have less opportunity to benefit from spontaneous in-person supportive interactions.

However, alongside this disruption, new opportunities to interact and obtain social support have arisen. The surge in community social support during the initial lockdown mirrored that often seen in response to adverse events (eg, natural disasters 16 ). COVID-19 restrictions that confined individuals to their local area also compelled them to focus their in-person efforts locally. Commentators on the initial lockdown in the UK remarked on extraordinary acts of generosity between individuals who belonged to the same community but were unknown to each other. However, research on adverse events also tells us that such community support is not necessarily maintained in the longer term. 16

Meanwhile, online forms of social support are not bound by geography, thus enabling interactions and social support to be received from a wider network of people. Formal online social support spaces (eg, support groups) existed well before COVID-19, but have vastly increased since. While online interactions can increase perceived social support, it is unclear whether remote communication technologies provide an effective substitute from in-person interaction during periods of social distancing. 17 18 It makes intuitive sense that the usefulness of online social support will vary by the type of support offered, degree of social interaction and ‘online communication skills’ of those taking part. Youth workers, for instance, have struggled to keep vulnerable youth engaged in online youth clubs, 19 despite others finding a positive association between amount of digital technology used by individuals during lockdown and perceived social support. 20 Other research has found that more frequent face-to-face contact and phone/video contact both related to lower levels of depression during the time period of March to August 2020, but the negative effect of a lack of contact was greater for those with higher levels of usual sociability. 21 Relatedly, important inequalities in social support exist, such that individuals who occupy more socially disadvantaged positions in society (eg, low socioeconomic status, older people) tend to have less access to social support, 22 potentially exacerbated by COVID-19.

Social and interactional norms

Interactional norms are key relational mechanisms which build trust, belonging and identity within and across groups in a system. Individuals in groups and societies apply meaning by ‘approving, arranging and redefining’ symbols of interaction. 23 A handshake, for instance, is a powerful symbol of trust and equality. Depending on context, not shaking hands may symbolise a failure to extend friendship, or a failure to reach agreement. The norms governing these symbols represent shared values and identity; and mutual understanding of these symbols enables individuals to achieve orderly interactions, establish supportive relationship accountability and connect socially. 24 25

Physical distancing measures to contain the spread of COVID-19 radically altered these norms of interaction, particularly those used to convey trust, affinity, empathy and respect (eg, hugging, physical comforting). 26 As epidemic waves rose and fell, the work to negotiate these norms required intense cognitive effort; previously taken-for-granted interactions were re-examined, factoring in current restriction levels, own and (assumed) others’ vulnerability and tolerance of risk. This created awkwardness, and uncertainty, for example, around how to bring closure to an in-person interaction or convey warmth. The instability in scripted ways of interacting created particular strain for individuals who already struggled to encode and decode interactions with others (eg, those who are deaf or have autism spectrum disorder); difficulties often intensified by mask wearing. 27

Large social gatherings—for example, weddings, school assemblies, sporting events—also present key opportunities for affirming and assimilating interactional norms, building cohesion and shared identity and facilitating cooperation across social groups. 28 Online ‘equivalents’ do not easily support ‘social-bonding’ activities such as singing and dancing, and rarely enable chance/spontaneous one-on-one conversations with peripheral/weaker network ties (see the Social networks section) which can help strengthen bonds across a larger network. The loss of large gatherings to celebrate rites of passage (eg, bar mitzvah, weddings) has additional relational costs since these events are performed by and for communities to reinforce belonging, and to assist in transitioning to new phases of life. 29 The loss of interaction with diverse others via community and large group gatherings also reduces intergroup contact, which may then tend towards more prejudiced outgroup attitudes. While online interaction can go some way to mimicking these interaction norms, there are key differences. A sense of anonymity, and lack of in-person emotional cues, tends to support norms of polarisation and aggression in expressing differences of opinion online. And while online platforms have potential to provide intergroup contact, the tendency of much social media to form homogeneous ‘echo chambers’ can serve to further reduce intergroup contact. 30 31

Intimacy relates to the feeling of emotional connection and closeness with other human beings. Emotional connection, through romantic, friendship or familial relationships, fulfils a basic human need 32 and strongly benefits health, including reduced stress levels, improved mental health, lowered blood pressure and reduced risk of heart disease. 32 33 Intimacy can be fostered through familiarity, feeling understood and feeling accepted by close others. 34

Intimacy via companionship and closeness is fundamental to mental well-being. Positively, the COVID-19 pandemic has offered opportunities for individuals to (re)connect and (re)strengthen close relationships within their household via quality time together, following closure of many usual external social activities. Research suggests that the first full UK lockdown period led to a net gain in the quality of steady relationships at a population level, 35 but amplified existing inequalities in relationship quality. 35 36 For some in single-person households, the absence of a companion became more conspicuous, leading to feelings of loneliness and lower mental well-being. 37 38 Additional pandemic-related relational strain 39 40 resulted, for some, in the initiation or intensification of domestic abuse. 41 42

Physical touch is another key aspect of intimacy, a fundamental human need crucial in maintaining and developing intimacy within close relationships. 34 Restrictions on social interactions severely restricted the number and range of people with whom physical affection was possible. The reduction in opportunity to give and receive affectionate physical touch was not experienced equally. Many of those living alone found themselves completely without physical contact for extended periods. The deprivation of physical touch is evidenced to take a heavy emotional toll. 43 Even in future, once physical expressions of affection can resume, new levels of anxiety over germs may introduce hesitancy into previously fluent blending of physical and verbal intimate social connections. 44

The pandemic also led to shifts in practices and norms around sexual relationship building and maintenance, as individuals adapted and sought alternative ways of enacting sexual intimacy. This too is important, given that intimate sexual activity has known benefits for health. 45 46 Given that social restrictions hinged on reducing household mixing, possibilities for partnered sexual activity were primarily guided by living arrangements. While those in cohabiting relationships could potentially continue as before, those who were single or in non-cohabiting relationships generally had restricted opportunities to maintain their sexual relationships. Pornography consumption and digital partners were reported to increase since lockdown. 47 However, online interactions are qualitatively different from in-person interactions and do not provide the same opportunities for physical intimacy.

Recommendations and conclusions

In the sections above we have outlined the ways in which COVID-19 has impacted social relationships, showing how relational mechanisms key to health have been undermined. While some of the damage might well self-repair after the pandemic, there are opportunities inherent in deliberative efforts to build back in ways that facilitate greater resilience in social and community relationships. We conclude by making three recommendations: one regarding public health responses to the pandemic; and two regarding social recovery.

Recommendation 1: explicitly count the relational cost of public health policies to control the pandemic

Effective handling of a pandemic recognises that social, economic and health concerns are intricately interwoven. It is clear that future research and policy attention must focus on the social consequences. As described above, policies which restrict physical mixing across households carry heavy and unequal relational costs. These include for individuals (eg, loss of intimate touch), dyads (eg, loss of warmth, comfort), networks (eg, restricted access to support) and communities (eg, loss of cohesion and identity). Such costs—and their unequal impact—should not be ignored in short-term efforts to control an epidemic. Some public health responses—restrictions on international holiday travel and highly efficient test and trace systems—have relatively small relational costs and should be prioritised. At a national level, an earlier move to proportionate restrictions, and investment in effective test and trace systems, may help prevent escalation of spread to the point where a national lockdown or tight restrictions became an inevitability. Where policies with relational costs are unavoidable, close attention should be paid to the unequal relational impact for those whose personal circumstances differ from normative assumptions of two adult families. This includes consideration of whether expectations are fair (eg, for those who live alone), whether restrictions on social events are equitable across age group, religious/ethnic groupings and social class, and also to ensure that the language promoted by such policies (eg, households; families) is not exclusionary. 48 49 Forethought to unequal impacts on social relationships should thus be integral to the work of epidemic preparedness teams.

Recommendation 2: intelligently balance online and offline ways of relating

A key ingredient for well-being is ‘getting together’ in a physical sense. This is fundamental to a human need for intimate touch, physical comfort, reinforcing interactional norms and providing practical support. Emerging evidence suggests that online ways of relating cannot simply replace physical interactions. But online interaction has many benefits and for some it offers connections that did not exist previously. In particular, online platforms provide new forms of support for those unable to access offline services because of mobility issues (eg, older people) or because they are geographically isolated from their support community (eg, lesbian, gay, bisexual, transgender and queer (LGBTQ) youth). Ultimately, multiple forms of online and offline social interactions are required to meet the needs of varying groups of people (eg, LGBTQ, older people). Future research and practice should aim to establish ways of using offline and online support in complementary and even synergistic ways, rather than veering between them as social restrictions expand and contract. Intelligent balancing of online and offline ways of relating also pertains to future policies on home and flexible working. A decision to switch to wholesale or obligatory homeworking should consider the risk to relational ‘group properties’ of the workplace community and their impact on employees’ well-being, focusing in particular on unequal impacts (eg, new vs established employees). Intelligent blending of online and in-person working is required to achieve flexibility while also nurturing supportive networks at work. Intelligent balance also implies strategies to build digital literacy and minimise digital exclusion, as well as coproducing solutions with intended beneficiaries.

Recommendation 3: build stronger and sustainable localised communities

In balancing offline and online ways of interacting, there is opportunity to capitalise on the potential for more localised, coherent communities due to scaled-down travel, homeworking and local focus that will ideally continue after restrictions end. There are potential economic benefits after the pandemic, such as increased trade as home workers use local resources (eg, coffee shops), but also relational benefits from stronger relationships around the orbit of the home and neighbourhood. Experience from previous crises shows that community volunteer efforts generated early on will wane over time in the absence of deliberate work to maintain them. Adequately funded partnerships between local government, third sector and community groups are required to sustain community assets that began as a direct response to the pandemic. Such partnerships could work to secure green spaces and indoor (non-commercial) meeting spaces that promote community interaction. Green spaces in particular provide a triple benefit in encouraging physical activity and mental health, as well as facilitating social bonding. 50 In building local communities, small community networks—that allow for diversity and break down ingroup/outgroup views—may be more helpful than the concept of ‘support bubbles’, which are exclusionary and less sustainable in the longer term. Rigorously designed intervention and evaluation—taking a systems approach—will be crucial in ensuring scale-up and sustainability.

The dramatic change to social interaction necessitated by efforts to control the spread of COVID-19 created stark challenges but also opportunities. Our essay highlights opportunities for learning, both to ensure the equity and humanity of physical restrictions, and to sustain the salutogenic effects of social relationships going forward. The starting point for capitalising on this learning is recognition of the disruption to relational mechanisms as a key part of the socioeconomic and health impact of the pandemic. In recovery planning, a general rule is that what is good for decreasing health inequalities (such as expanding social protection and public services and pursuing green inclusive growth strategies) 4 will also benefit relationships and safeguard relational mechanisms for future generations. Putting this into action will require political will.

Twitter: @karenmaxSPHSU, @Mark_McCann, @Rwilsonlowe, @KMitchinGlasgow

Contributors: EL and KM led on the manuscript conceptualisation, review and editing. SP, KM, CB, RBP, RL, MM, JR, KS and RW-L contributed to drafting and revising the article. All authors assisted in revising the final draft.

Funding: The research reported in this publication was supported by the Medical Research Council (MC_UU_00022/1, MC_UU_00022/3) and the Chief Scientist Office (SPHSU11, SPHSU14). EL is also supported by MRC Skills Development Fellowship Award (MR/S015078/1). KS and MM are also supported by a Medical Research Council Strategic Award (MC_PC_13027).

Competing interests: None declared.

Provenance and peer review: Not commissioned; externally peer reviewed.

Data availability statement

Ethics statements, patient consent for publication.

Not required.

Weighing the benefits and costs of COVID-19 restrictions

July 20, 2021 – Over the course of the COVID-19 pandemic, policymakers and the public have often vehemently disagreed about the pros and cons of restrictions such as lockdowns. Proponents of restrictions argue that they save lives; opponents say they destroy livelihoods. Amid these often nasty debates, researchers have been churning out benefit-cost analyses that aim to shed light on which restrictions are “worth it” and which aren’t.

“There are now a huge number of these analyses,” said Lisa Robinson , senior research scientist and deputy director of the Center for Health Decision Science at Harvard T.H. Chan School of Public Health. “It’s become an enormous undertaking.” For her part, Robinson has been exploring the complexities involved in valuing deaths averted by COVID-19 policies , using a well-established but widely-misunderstood metric called the “value per statistical life,” or VSL.

In the big picture, analyses of COVID-19 policies need to take into account how many lives they could potentially save, of course. But they also need to consider the potential downsides. For instance, how many people could lose their jobs if a state puts a lockdown into effect? What educational losses will occur among kids who miss months of in-person school? If restrictions on restaurants are lifted, how do we weigh the economic benefits against the potential increase in COVID-19 cases—and the deaths that might result?

“What benefit-cost analysis does is require people to carefully and rigorously explore the impacts of a policy,” Robinson said. “Something may sound like a great idea on the surface, but digging more deeply into its real-world effects often unearths unexpected consequences. These analyses also highlight the trade-offs implicit whenever we make a decision about how to allocate resources.”

Assigning a money value to life

Robinson has been exploring how best to estimate the VSL—a metric commonly used to evaluate lifesaving interventions—in analyzing the relative costs and benefits of COVID-19 policies. She tackled the issue in a study she co-authored last year, which a recent article in The Economist characterized as “the best attempt at weighing up these competing valuations.”

Although the phrase “value per statistical life” suggests that the government, or someone else, is somehow placing a value on someone’s life, Robinson emphasizes that this is not the case. Rather, economists start by investigating how much of their own income individuals are willing to exchange to reduce their own chance of dying by a small amount—such as by paying extra to buy a safer car or choosing a less risky job for lower pay. These estimates are then converted into estimates of value of reducing expected deaths—that is to say, into the VSL.

Policymakers often use VSL, for example, when determining what safety requirements to impose on automobiles or how low to set standards for pollution emissions. In looking at the population as a whole, U.S. regulatory agencies making benefit-cost calculations currently estimate the VSL as roughly $10 or $11 million. A $10 million VSL means that a typical individual is willing to pay $1,000 to reduce his or her chance of dying within a given year by 1 in 10,000, Robinson explained in a 2020 blog post that looked at COVID-19 benefit-cost analysis and the VSL.

She noted, however, that the $10 million or $11 million figure is for the average member of the population—for someone middle-aged. An individual’s willingness to pay to reduce mortality risk may not stay the same across their life course. For example, older people have fewer expected years of life remaining than the average member of the population, and less opportunity for future earnings, which could change the VSL calculation and therefore make a difference when COVID-19 policies are assessed.

Yet another consideration to take into account is the value people may place on avoiding the substantial pain and suffering caused by COVID-19—including struggling to breathe, or being put on a ventilator—which could increase the VSL. In an October 2020 paper , Robinson’s colleague and frequent co-author James Hammitt , professor of economics and decision sciences, also noted that people may be more likely to place a higher value on avoiding risks they view as “dreaded, uncertain, catastrophic, and ambiguous”—like COVID-19.

The overall point, according to Robinson, is that analysts and policymakers comparing the benefits and costs of particular COVID-19 policies should take care to examine uncertainty in the VSL estimates. VSL is likely to vary depending on who is affected by the policy and by how they view the risks that they experience.

She acknowledges the difficulty involved in deciding on COVID-19 restrictions. But she is also gratified to see that people are using benefit-cost analysis to carefully explore the implications of those tough decisions.

“When COVID hit, benefit-cost analysis really got a lot of attention in the mass media,” she said. “I was so excited that this field that I’ve been involved in for so long is now the subject of stories in major news outlets like the New York Times and the Wall Street Journal. People are really paying attention to its usefulness in policymaking.”

– Karen Feldscher

photo: Anthony Quintano/Wikimedia Commons

Read these 12 moving essays about life during coronavirus

Artists, novelists, critics, and essayists are writing the first draft of history.

by Alissa Wilkinson

The world is grappling with an invisible, deadly enemy, trying to understand how to live with the threat posed by a virus . For some writers, the only way forward is to put pen to paper, trying to conceptualize and document what it feels like to continue living as countries are under lockdown and regular life seems to have ground to a halt.

So as the coronavirus pandemic has stretched around the world, it’s sparked a crop of diary entries and essays that describe how life has changed. Novelists, critics, artists, and journalists have put words to the feelings many are experiencing. The result is a first draft of how we’ll someday remember this time, filled with uncertainty and pain and fear as well as small moments of hope and humanity.

• The Vox guide to navigating the coronavirus crisis

At the New York Review of Books, Ali Bhutto writes that in Karachi, Pakistan, the government-imposed curfew due to the virus is “eerily reminiscent of past military clampdowns”:

Beneath the quiet calm lies a sense that society has been unhinged and that the usual rules no longer apply. Small groups of pedestrians look on from the shadows, like an audience watching a spectacle slowly unfolding. People pause on street corners and in the shade of trees, under the watchful gaze of the paramilitary forces and the police.

His essay concludes with the sobering note that “in the minds of many, Covid-19 is just another life-threatening hazard in a city that stumbles from one crisis to another.”

Writing from Chattanooga, novelist Jamie Quatro documents the mixed ways her neighbors have been responding to the threat, and the frustration of conflicting direction, or no direction at all, from local, state, and federal leaders:

Whiplash, trying to keep up with who’s ordering what. We’re already experiencing enough chaos without this back-and-forth. Why didn’t the federal government issue a nationwide shelter-in-place at the get-go, the way other countries did? What happens when one state’s shelter-in-place ends, while others continue? Do states still under quarantine close their borders? We are still one nation, not fifty individual countries. Right?

• A syllabus for the end of the world

Award-winning photojournalist Alessio Mamo, quarantined with his partner Marta in Sicily after she tested positive for the virus, accompanies his photographs in the Guardian of their confinement with a reflection on being confined :

The doctors asked me to take a second test, but again I tested negative. Perhaps I’m immune? The days dragged on in my apartment, in black and white, like my photos. Sometimes we tried to smile, imagining that I was asymptomatic, because I was the virus. Our smiles seemed to bring good news. My mother left hospital, but I won’t be able to see her for weeks. Marta started breathing well again, and so did I. I would have liked to photograph my country in the midst of this emergency, the battles that the doctors wage on the frontline, the hospitals pushed to their limits, Italy on its knees fighting an invisible enemy. That enemy, a day in March, knocked on my door instead.

In the New York Times Magazine, deputy editor Jessica Lustig writes with devastating clarity about her family’s life in Brooklyn while her husband battled the virus, weeks before most people began taking the threat seriously:

At the door of the clinic, we stand looking out at two older women chatting outside the doorway, oblivious. Do I wave them away? Call out that they should get far away, go home, wash their hands, stay inside? Instead we just stand there, awkwardly, until they move on. Only then do we step outside to begin the long three-block walk home. I point out the early magnolia, the forsythia. T says he is cold. The untrimmed hairs on his neck, under his beard, are white. The few people walking past us on the sidewalk don’t know that we are visitors from the future. A vision, a premonition, a walking visitation. This will be them: Either T, in the mask, or — if they’re lucky — me, tending to him.

Essayist Leslie Jamison writes in the New York Review of Books about being shut away alone in her New York City apartment with her 2-year-old daughter since she became sick:

The virus. Its sinewy, intimate name. What does it feel like in my body today? Shivering under blankets. A hot itch behind the eyes. Three sweatshirts in the middle of the day. My daughter trying to pull another blanket over my body with her tiny arms. An ache in the muscles that somehow makes it hard to lie still. This loss of taste has become a kind of sensory quarantine. It’s as if the quarantine keeps inching closer and closer to my insides. First I lost the touch of other bodies; then I lost the air; now I’ve lost the taste of bananas. Nothing about any of these losses is particularly unique. I’ve made a schedule so I won’t go insane with the toddler. Five days ago, I wrote Walk/Adventure! on it, next to a cut-out illustration of a tiger—as if we’d see tigers on our walks. It was good to keep possibility alive.

At Literary Hub, novelist Heidi Pitlor writes about the elastic nature of time during her family’s quarantine in Massachusetts:

During a shutdown, the things that mark our days—commuting to work, sending our kids to school, having a drink with friends—vanish and time takes on a flat, seamless quality. Without some self-imposed structure, it’s easy to feel a little untethered. A friend recently posted on Facebook: “For those who have lost track, today is Blursday the fortyteenth of Maprilay.” ... Giving shape to time is especially important now, when the future is so shapeless. We do not know whether the virus will continue to rage for weeks or months or, lord help us, on and off for years. We do not know when we will feel safe again. And so many of us, minus those who are gifted at compartmentalization or denial, remain largely captive to fear. We may stay this way if we do not create at least the illusion of movement in our lives, our long days spent with ourselves or partners or families.

• What day is it today?

Novelist Lauren Groff writes at the New York Review of Books about trying to escape the prison of her fears while sequestered at home in Gainesville, Florida:

Some people have imaginations sparked only by what they can see; I blame this blinkered empiricism for the parks overwhelmed with people, the bars, until a few nights ago, thickly thronged. My imagination is the opposite. I fear everything invisible to me. From the enclosure of my house, I am afraid of the suffering that isn’t present before me, the people running out of money and food or drowning in the fluid in their lungs, the deaths of health-care workers now growing ill while performing their duties. I fear the federal government, which the right wing has so—intentionally—weakened that not only is it insufficient to help its people, it is actively standing in help’s way. I fear we won’t sufficiently punish the right. I fear leaving the house and spreading the disease. I fear what this time of fear is doing to my children, their imaginations, and their souls.

At ArtForum , Berlin-based critic and writer Kristian Vistrup Madsen reflects on martinis, melancholia, and Finnish artist Jaakko Pallasvuo’s 2018 graphic novel Retreat , in which three young people exile themselves in the woods:

In melancholia, the shape of what is ending, and its temporality, is sprawling and incomprehensible. The ambivalence makes it hard to bear. The world of Retreat is rendered in lush pink and purple watercolors, which dissolve into wild and messy abstractions. In apocalypse, the divisions established in genesis bleed back out. My own Corona-retreat is similarly soft, color-field like, each day a blurred succession of quarantinis, YouTube–yoga, and televized press conferences. As restrictions mount, so does abstraction. For now, I’m still rooting for love to save the world.

At the Paris Review , Matt Levin writes about reading Virginia Woolf’s novel The Waves during quarantine:

A retreat, a quarantine, a sickness—they simultaneously distort and clarify, curtail and expand. It is an ideal state in which to read literature with a reputation for difficulty and inaccessibility, those hermetic books shorn of the handholds of conventional plot or characterization or description. A novel like Virginia Woolf’s The Waves is perfect for the state of interiority induced by quarantine—a story of three men and three women, meeting after the death of a mutual friend, told entirely in the overlapping internal monologues of the six, interspersed only with sections of pure, achingly beautiful descriptions of the natural world, a day’s procession and recession of light and waves. The novel is, in my mind’s eye, a perfectly spherical object. It is translucent and shimmering and infinitely fragile, prone to shatter at the slightest disturbance. It is not a book that can be read in snatches on the subway—it demands total absorption. Though it revels in a stark emotional nakedness, the book remains aloof, remote in its own deep self-absorption.

• Vox is starting a book club. Come read with us!

In an essay for the Financial Times, novelist Arundhati Roy writes with anger about Indian Prime Minister Narendra Modi’s anemic response to the threat, but also offers a glimmer of hope for the future:

Historically, pandemics have forced humans to break with the past and imagine their world anew. This one is no different. It is a portal, a gateway between one world and the next. We can choose to walk through it, dragging the carcasses of our prejudice and hatred, our avarice, our data banks and dead ideas, our dead rivers and smoky skies behind us. Or we can walk through lightly, with little luggage, ready to imagine another world. And ready to fight for it.

From Boston, Nora Caplan-Bricker writes in The Point about the strange contraction of space under quarantine, in which a friend in Beirut is as close as the one around the corner in the same city:

It’s a nice illusion—nice to feel like we’re in it together, even if my real world has shrunk to one person, my husband, who sits with his laptop in the other room. It’s nice in the same way as reading those essays that reframe social distancing as solidarity. “We must begin to see the negative space as clearly as the positive, to know what we don’t do is also brilliant and full of love,” the poet Anne Boyer wrote on March 10th, the day that Massachusetts declared a state of emergency. If you squint, you could almost make sense of this quarantine as an effort to flatten, along with the curve, the distinctions we make between our bonds with others. Right now, I care for my neighbor in the same way I demonstrate love for my mother: in all instances, I stay away. And in moments this month, I have loved strangers with an intensity that is new to me. On March 14th, the Saturday night after the end of life as we knew it, I went out with my dog and found the street silent: no lines for restaurants, no children on bicycles, no couples strolling with little cups of ice cream. It had taken the combined will of thousands of people to deliver such a sudden and complete emptiness. I felt so grateful, and so bereft.

And on his own website, musician and artist David Byrne writes about rediscovering the value of working for collective good , saying that “what is happening now is an opportunity to learn how to change our behavior”:

In emergencies, citizens can suddenly cooperate and collaborate. Change can happen. We’re going to need to work together as the effects of climate change ramp up. In order for capitalism to survive in any form, we will have to be a little more socialist. Here is an opportunity for us to see things differently — to see that we really are all connected — and adjust our behavior accordingly. Are we willing to do this? Is this moment an opportunity to see how truly interdependent we all are? To live in a world that is different and better than the one we live in now? We might be too far down the road to test every asymptomatic person, but a change in our mindsets, in how we view our neighbors, could lay the groundwork for the collective action we’ll need to deal with other global crises. The time to see how connected we all are is now.

The portrait these writers paint of a world under quarantine is multifaceted. Our worlds have contracted to the confines of our homes, and yet in some ways we’re more connected than ever to one another. We feel fear and boredom, anger and gratitude, frustration and strange peace. Uncertainty drives us to find metaphors and images that will let us wrap our minds around what is happening.

Yet there’s no single “what” that is happening. Everyone is contending with the pandemic and its effects from different places and in different ways. Reading others’ experiences — even the most frightening ones — can help alleviate the loneliness and dread, a little, and remind us that what we’re going through is both unique and shared by all.

• Recommendations

Most Popular

• How Republicans became the party of raunch
• Laura Loomer, Trump’s new favorite conspiracy theorist, explained
• Sign up for Vox’s daily newsletter
• The impact of the Supreme Court’s reversal of affirmative action, explained in one chart
• Take a mental break with the newest Vox crossword

Today, Explained

Understand the world with a daily explainer plus the most compelling stories of the day.

This is the title for the native ad

More in Culture

The truth behind the ongoing controversy over the highly memeable dancer.

The highly coveted endorsement comes after a year of paranoid speculation.

From prebiotic sodas to collagen waters, beverages are trying to do the most. Consumers are drinking it up.

Why the new luxury is flip phones and vinyl LPs

Kids are still playing, but they need help.

From TikTok to Two Hot Takes, everyone sucks here.