Skip to main content
NCBI home page
Elsevier - PMC COVID-19 Collection logoLink to Elsevier - PMC COVID-19 Collection
. 2022 Aug 21;83:101867. doi: 10.1016/j.jenvp.2022.101867

Household crowding during the COVID-19 lockdown fosters anti-democracy even after 17 months: A 5-wave latent growth curve study

Silvia Russo a, Pasquale Colloca b, Nicoletta Cavazza c, Michele Roccato a,∗
PMCID: PMC9392657  PMID: 36034614

Abstract

In an earlier cross-sectional study, Roccato et al. (2021) showed that household crowding during the COVID-19 lockdown was positively related to support for anti-democratic political systems. However, little is known about the persistence of these effect over time. In this study, we examined its duration in a longitudinal study structured in five waves, the first in May–June 2019 (before the COVID-19 outbreak, N = 1504) and the others during the pandemic, in April 2020 (during the lockdown, N = 1199), October 2020 (N = 1156), April 2021 (N = 1148), and October 2021 (N = 1151). The increase in support for anti-democratic systems associated with household overcrowding in the initial phase of the lockdown (Wave 2) did not change over the subsequent 17 months. Moreover, the effect was stronger among those who had high (compared with low) trust in democratic political institutions before the pandemic. Strengths, limitations, and potential developments of the study are discussed.

Keywords: Household crowding, Lockdown, COVID-19, Authoritarianism, Latent growth analysis, Longitudinal research

1. Introduction

The lockdown measures taken by many governments in the Spring of 2020 to prevent the spread of the COVID-19 pandemic were an almost unprecedented event in modern history. In Italy, where we conducted this study, people were prohibited from leaving their homes from March 11 to May 3, 2020, except to buy food and medicines at the nearest shop and to go to work if they were key workers. These unique living conditions prompted researchers to conduct several studies on the social-psychological effects of being forced to live with others 24 h a day, 7 days a week. Cross-sectional studies have shown that locked down people report increases in their stress and anxiety and even their need for mental health care in many countries around the world, including the United Kingdom (Bu et al., 2021), Greece (Fountolakis et al., 2021), India (Pal & Danda, 2021), Singapore (Olszewska-Guizzo, Fogel, Escoffier, & Ho, 2021), Ecuador (Mautong et al., 2021) and Italy (Somma et al., 2021). These findings were confirmed by Henssler et al.’s (2021) meta-analysis and by some longitudinal studies with at least one pre- and one post-COVID-19 wave (e.g., Probst et al., 2021). However, little is known about the persistence of these effects over time. Are these a trivial shift that is destined to be reabsorbed in a few weeks after the end of the lockdown or do they represent a stable psychological change?

In this study we aimed to examine whether the political effects of restriction observed in previous studies can be long-lasting, especially when certain contextual characteristics such as household overcrowding or personal predisposition are present. Three longitudinal studies offer suggestions for our objectives. First, Andersen et al. (2021) showed that pre-COVID-19 affective disorders amplified the negative effects of lockdown on people's well-being, demonstrating the importance in this area of study of focusing on the interaction between individual predispositions and the environment in which people live. Second, these negative effects were still evident two or three weeks after the end of the lockdown (Probst et al., 2021). In a study conducted in South Korea, where the government responded to the 2015 MERS pandemic with a two-week lockdown, the impact on participants' mental health declined significantly after 4–6 months (Jeong et al., 2016).

Most studies on the impact of COVID-19 lockdown have focused on the stressful role of factors such as length of confinement, fear of infection, frustration, boredom and financial loss (e.g., Brooks et al., 2020). However, household overcrowding is another critical variable that could play a role when people are confined. Household overcrowding is known to have negative effects on health and well-being even in normal times. Studies conducted in animals have shown that living in a crowded environment promotes some stress indicators, such as increased blood pressure and decreased reproductive capacity, and even decreases life expectancy (e.g., Calhoun, 1962; Christian, 1963). The few studies conducted on human samples, initially in offices, correctional facilities and university dormitories (Baum et al., 1981; Evans, 2003; Veitch, 2012; Wener, 2012), confirmed these findings and showed that living in crowded environments promotes psychological distress, anxiety and even mental illness, both in noncontact and contact cultures (Evans et al., 2000).

The lockdown measure, which forces people to spend their entire day together, restricts individual privacy and exposes them to undesirable social interactions, likely exacerbated these effects. Indeed, the study conducted by Amerio et al. (2020) in the context of the COVID-19 pandemic found that poor housing was associated with an increased risk of depressive symptoms. Interestingly, one study has documented that household crowding can have an impact that goes beyond individual wellbeing and even affects individual social and political orientations. Indeed, Cavazza et al. (2021) showed that household overcrowding promoted participants’ support for anti-democratic political systems during the lockdown, via the partial mediation of their perceptions of the relative impact of COVID-19 on their family and their expectation of future lifestyle restrictions, which these authors considered a measure of perceived social competition. However, the study by Cavazza et al. (2021) left two important questions unanswered.

First, we do not know whether the lockdown had a general anti-democratic effect or whether it prompted anti-democracy mainly (or even only) in certain subgroups. The main point of reference in this case is the social-psychological literature on right-wing authoritarianism, which shows that societal (e.g., Mirisola et al., 2014) and environmental (e.g., Russo et al., 2020) threats promote right-wing authoritarianism primarily among low-authoritarian people, who are pushed to increase their authoritarianism to cope with the loss of perceived control over their lives. In this light, the threat makes low- and high-authoritarians similar in terms their socio-political attitudes. Based on this literature, we wondered whether the anti-democratic shift fostered by housing overcrowding identified by Cavazza et al. (2021) is a general process or whether it is limited to people who were particularly trustful in democratic political institutions before the pandemic.

Second, we do not know how long this anti-democratic change will last. Is it a temporary change, like the “rally effects” studied in public opinion research (Mueller, 1970), which lead to temporary (typically, lasting a few months) increases in institutional trust in response to exogenous chocks such as wars and terrorist attacks (Parker, 1995)? Or is it a more stable change that permanently threatens democracy by strengthening potential anti-democracy in the population? In this longitudinal study, we explored these research questions using a broad quota sample of the adult Italian population.

2. Material and methods

2.1. Participants and procedure

We sought to answer our research questions using the database of the COnsequences of COvid-19 project (COCO), a 28-month longitudinal study conducted on a broad quota sample of the Italian adult population, stratified with respect to sex, age, educational level and geographic area of residence. The study consists of 5 waves, the first of which was conducted in May–June 2019 (before the COVID-19 outbreak, N = 1504) and the others during the pandemic in April 2020 (during the lockdown, N = 1199), October 2020 (N = 1156), April 2021 (N = 1148) and October 2021 (N = 1151). A total of 990 participants (51.92% women, M age = 49.32, SD = 14.34) were included in the analyses. The research protocol was approved by the Bio-Ethical Committee of the University of Turin. The dataset we used is available at https://osf.io/j2pyr/?view_only=f8e815f92a9a45388d8cdea2bda2a3d3.

2.2. Measures

2.2.1. Support for anti-democratic political systems

In Waves 2, 3, 4 and 5, participants were asked to respond to two items from the European Values Study (https://europeanvaluesstudy.eu/), asking them to indicate the extent to which ‘Having a strong leader who does not have to bother with parliament and elections’ and ‘Having the army rule the country’ would be good or bad forms of government for Italy. Response categories ranged from 1 (‘It would be a very bad system’) to 4 (‘It would be a very good system’). Based on rs ranging from 0.47, p < .001 (Wave 5) to 0.51, p < .001 (Wave 2), we calculated four indices of support for anti-democratic political systems.

2.2.2. Household crowding during the lockdown

Household crowding was measured in Wave 2 using the American Crowding Index (Torshizian & Grimes, 2020), i.e., the ratio of the number of people living in a house (including the respondent) to the number of available rooms.

2.2.3. Pre-pandemic trust in democratic political institutions

In Wave 1, we measured pre-pandemic trust in three democratic political institutions (the political parties, the President of the Republic and the Italian Parliament), using – like in the European Social Survey (www.europeansocialsurvey.org) – an 11-category format (α = 0.83). We estimated participants’ trust in democratic political institutions as a latent variable, by resorting to confirmatory factor analysis.

2.2.4. Control variables

In our analyses, we partialled out participants' gender (0 = man), age, education, size of the place of residence and perceived economic situation, using the following item from the European Social Survey: ‘Your present income (or your household's income if you don't live alone) allows you to live … ‘. The response categories were four: ‘Finding it very difficult on present income’ (= 1), ‘Finding it difficult on present income’ (= 2), ‘Coping on present income’ (= 3) and ‘Living comfortably on present income’ (= 4). All these variables were measured in Wave 1.

3. Results

As a first step, we used univariate latent growth curve models to determine the rate of change in people's support for anti-democratic political systems over time. These models allow us to estimate the initial level of support for anti-democratic political systems and the linear and quadratic change in that support over time. Because the initial level and change over time are represented as latent factors, it is also possible to estimate the variances of the latent intercept and slope, which represent the extent of interindividual differences in the mean level of anti-democratic attitudes in Wave 2 and change over time (Hertzog & Nesselroade, 2003). Because four time points were available, we were able to construct a baseline model with one linear slope and one quadratic curve. Although the model had good fit indices (χ2(2) = 4.26, p = .12; CFI = 0.999, TLI = 0.996, RMSEA = 0.026), neither the linear (coeff. = 0.06, p = .85) nor the quadratic slope (coeff. = −0.03, p = .28) indicated significant change over time. The variances also showed no significant individual variation around these flat trends (coeff. = 0.00, p = .95 for the linear slope and coeff. = 0.00, p = .85 for the quadratic curve). Overall, this model shows that support for anti-democratic political systems has not changed from Wave 2 and beyond.

In a second step, we added time-invariant covariates to our model. Because we found no change in support for anti-democratic systems over time, we limited our investigation to the effects of these covariates on the latent intercept, i.e., the level of support for anti-democratic political systems in Wave 2. This model included the pre-pandemic trust in democratic political institutions and household crowding (both mean-centred), as well as their latent interaction and the control variables.

Table 1 shows the results of this model (χ2(27) = 29.97, p = .32; CFI = 0.998, TLI = 0.997, RMSEA = 0.011). Age and education showed a negative association with support for anti-democratic political systems in Wave 2, while the other control variables were not significantly associated with the dependent variable. More interestingly as concerns our research goals, living in a crowded household was positively associated with support for anti-democratic political systems in Wave 2, while pre-pandemic trust in democratic political institution was not associated with the dependent variable. Moreover, the interaction between these two covariates reached statistical significance. A simple slope analysis revealed that the effect of household overcrowding on support for anti-democratic political systems in Wave 2 was stronger (coeff. = 0.25, SE = 0.04, p < .001) among those who had high trust in democratic political institutions (−1 SD) than among those who had a low trust in them before the pandemic (+1 SD, coeff. = 0.18, SE = 0.03, p < .001).

Table 1.

Time-invariant covariates of the latent intercept of the support for anti-democratic political systems.

Covariate coeff. SE p 95% CIs
Woman −.02 .04 .64 −.09/.09
Age −.01 .00 .008 −.01/-.00
Size of area of residence −.01 .02 .51 −.06/.02
Education −.10 .04 <.001 −.17/-.06
Perceived economic situation .00 .04 .97 −.09/.06
Pre-pandemic trust in democratic political institutions .02 .02 .12 −.02/.05
Household crowding .22 .03 <.001 .13/.27
Pre-pandemic trust in democratic political institutions*Household crowding .04 .02 .04 .01/.07
Open in a new tab

4. Discussion

The lockdown measures used to combat the pandemic COVID-19 promoted short-term and time-limited negative psychological effects, such as anxiety and stress (Andersen et al., 2021; Probst et al., 2021). Moreover, they also had political effects and promoted people's support for anti-democratic political systems (Cavazza et al., 2021). However, while there is evidence of the short-term duration of the psychological effects of lockdowns (e.g., Jeong et al., 2016), it is less clear what the nature of these socio-political orientations is and how long they last, as they might not disappear overnight. Moreover, it is not known whether these effects are pervasive or – consistent with findings in the literature on authoritarianism (e.g., Mirisola et al., 2014; Russo et al., 2020) – occur only among people who, under ‘normal’ conditions, are trustful in democratic political institutions. Our analysis sought to address this knowledge gap by examining longitudinally whether the effects of lockdown can be long-lasting, particularly when environmental characteristics interact with individual socio-political predispositions.

Our results show that individual support for anti-democratic political systems fostered by the lockdown did not change in the subsequent phases of the COVID-19 pandemic; in particular, our models showed no significant inter- and intra-individual variability over time. It is plausible that the COVID-19 crisis structurally increased the salience of perceived threat above a certain threshold, thereby triggering more extremist political orientations. These dispositions are potentially more stable and long-term compared to psychological effects on well-being, as they involve value orientations and worldviews that are psychologically costlier to change in the short term. A follow-up of this study, conducted after the end of the COVID-19 pandemic, may be interesting to understand if and when the threats to our democracies caused by the lockdown will subside.

We also showed that household overcrowding in the initial phase of the lockdown (Wave 2) had strong anti-democratic effects, in terms of support for anti-democratic systems, especially among those who express high trust in democratic political institutions before the pandemic. This finding is consistent with the social-psychological literature on right-wing authoritarianism (e.g., Mirisola et al., 2014; Russo et al., 2020), which explains the rise in authoritarian orientation as a coping mechanism triggered by perceived threats. Moreover, the study is particularly interesting because it extends the literature on the consequences of household crowding by showing that the negative effects of environmental discomfort are particularly relevant among citizens that, in ‘normal times’, are the core of the democratic system.

The current study has some limitations that should be acknowledged and that depend mainly on the use of secondary data. First, in this study we measured household crowding with an objective index (number of people per room). This measure is widely used (e.g., Torshizian & Grimes, 2020). However, some recent studies have shown that subjective indices related to feeling too close to others (e.g., Fornara et al., 2022; Thornock et al., 2019) may have even higher predictive power than the objective measures. Therefore, our results on the relationship between household crowding and support for anti-democratic political systems may actually underestimate the true relationship between the two variables. Future studies that consider both indices and examine the interplay between subjective and objective dimensions of crowding could be interesting. Second, we have examined the role of housing crowding, implicitly emphasising the role of housing size. However, as recently stressed, the spatial configuration of the house could also be considered when assessing the housing distress (Campagna, 2016). It might be interesting to extend this study by considering features such as the presence of partitions that promote residents’ privacy and protect them from unwanted stimuli and intruders.

These limitations should be weighed against the strengths of the study. Because of the longitudinal nature of the dataset, our approach allowed us to capture a trend over a long period of time. Our longitudinal analysis can be considered the more accurate test for assessing changes in anti-democratic orientation induced by housing condition during the lockdown. Moreover, by including variables measured before the pandemic outbreak, our model can rule out the possibility that observed outcomes depend on post-pandemic variables. In sum, we believe that these findings may provide the impetus for more careful evaluation of the socio-political effects of household crowding in future research, possibly examining the conditions under which living in a crowded environment influences political orientations.

Author statement

Silvia Russo: Conceptualization, Data curation, Formal analysis, Writing - original draft and review & editing, Funding acquisition.

Pasquale Colloca: Writing – original draft and review & editing, Funding acquisition.

Nicoletta Cavazza: Conceptualization, Writing - review & editing, Funding acquisition.

Michele Roccato: Conceptualization, Formal analysis, Writing - original draft and review & editing, Funding acquisition.

Declaration of competing interest

None.

Handling Editor: L. McCunn

References

  1. Amerio A., Brambilla A., Morganti A., Aguglia A., Bianchi D., Santi F., Costantini L., Odone A., Costanza A., Signorelli C., Serafini G.H., Amore M., Capolongo S. COVID-19 lockdown: Housing built environments' effects on mental health. International Journal of Environmental Research and Public Health. 2020;17:5973. doi: 10.3390/ijerph17165973. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Andersen A.J., Mary-Krause M., Bustamante J.J.H., Héron M., El Aarbaoui T., Melchior M. Symptoms of anxiety/depression during the COVID-19 pandemic and associated lockdown in the community: Longitudinal data from the TEMPO cohort in France. BMC Psychiatry. 2021;21:381. doi: 10.1186/s12888-021-03383-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Baum A., Singer J.E., Baum C.S. Stress and the environment. Journal of Social Issues. 1981;37:4–35. doi: 10.1111/j.1540-4560.1981.tb01056.x. [DOI] [Google Scholar]
  4. Brooks S.K., Webster R.K., Smith L.E., Woodland L., Wessely S., Greenberg N., Rubin G.J. The psychological impact of quarantine and how to reduce it: Rapid review of the evidence. Lancet. 2020;395:912–920. doi: 10.1016/S0140-6736(20)30460-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bu F., Mak H.W., Fancourt D. Rates and predictors of uptake of mental health support during the COVID-19 pandemic: An analysis of 26,720 adults in the UK in lockdown. Social Psychiatry and Psychiatric Epidemiology. 2021;56:2287–2297. doi: 10.1007/s00127-021-02105-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Calhoun J.B. Population density and social pathology. Scientific American. 1962;206:139–149. doi: 10.1038/scientificamerican0262-139. [DOI] [PubMed] [Google Scholar]
  7. Campagna G. Linking crowding, housing inadequacy, and perceived housing stress. Journal of Environmental Psychology. 2016;45:252–266. doi: 10.1016/j.jenvp.2016.01.002. [DOI] [Google Scholar]
  8. Cavazza N., Russo S., Colloca P., Roccato M. Household crowding can have political effects: An empirical study on support for anti-democratic political systems during the COVID-19 lockdown in Italy. Journal of Environmental Psychology. 2021;76 doi: 10.1016/j.jenvp.2021.101628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Christian J.J. The pathology of overpopulation. Military Medicine. 1963;128:571–603. doi: 10.1093/milmed/128.7.571. [DOI] [PubMed] [Google Scholar]
  10. Evans G.W. The built environment and mental health. Journal of Urban Health. 2003;80:536–555. doi: 10.1093/jurban/jtg063. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Evans G.W., Lepore S.J., Allen K.M. Cross-cultural differences in tolerance for crowding: Fact or fiction? Journal of Personality and Social Psychology. 2000;79:204–210. doi: 10.1037/0022-3514.79.2.204. [DOI] [PubMed] [Google Scholar]
  12. Fornara F., Mosca O., Bosco A., Caffo A.O., Lopez A., Iachini T., Ruggiero G., Ruotolo F., Sbordone F.L., Ferrara A., Cattaneo Z., Arioli M., Frassinetti F., Candini M., Miola L., Pazzaglia F. Space at home and psychological distress during the COVID-19 lockdown in Italy. Journal of Environmental Psychology. 2022;79 doi: 10.1016/j.jenvp.2021.101747. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fountolakis K.N., Apostolidou M.K., Atsiova M.B., Filippou A.K., Florou A.K., Gousiou D.S., Katsata A.R., Mantzari S.N., Padouva-Marakoulaki M., Papatriamtafyllou E.I., Sacharidi P.I., Tonia A.I., Tsagalidou E.G., Zymara V.P., Prezakerso P.E., Koupidis S.A., Fountoulakis N.K., Chrousos G.P. Self-reported changes in anxiety, depression and suicidality during the COVID-19 lockdown in Greece. Journal of Affective Disorders. 2021;279:624–629. doi: 10.1016/j.jad.2020.10.061. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Henssler J., Stock F., van Bohemen J., Walter H., Heinz A., Brandt L. Mental health effects of infection containment strategies: Quarantine and isolation—a systematic review and meta-analysis. European Archives of Psychiatry and Clinical Neuroscience. 2021;271:223–234. doi: 10.1007/s00406-020-01196-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hertzog C., Nesselroade J.R. Assessing psychological change in adulthood: An overview of methodological issues. Psychology and Aging. 2003;18(4):639–657. doi: 10.1037/0882-7974.18.4.639. [DOI] [PubMed] [Google Scholar]
  16. Jeong H., Yim H.W., Song Y.-J., Ki M., Min J.-A., Cho J., Chae J.-H. Mental health status of people isolated due to Middle East respiratory syndrome. Epidemiology and Health. 2016;38 doi: 10.4178/epih.e2016048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Mautong H., Gallardo-Rumbea J., Alvarado-Villa G.E., Fernández-Cadena J.C., Andrade-Molina D., Orellana-Romain C.E., Cherrez-Ojeda I. Assessment of depression, anxiety and stress levels in the Ecuadorian general population during social isolation due to the COVID-19 outbreak: A cross-sectional study. BMC Psychiatry. 2021;21:212. doi: 10.1186/s12888-021-03214-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Mirisola A., Roccato M., Russo S., Spagna G., Vieno A. Societal threat to safety, compensatory control, and right-wing authoritarianism. Political Psychology. 2014;35:795–812. doi: 10.1111/pops.12048. [DOI] [Google Scholar]
  19. Mueller J.E. Presidential popularity from Truman to Johnson. American Political Science Review. 1970;64:18–33. doi: 10.2307/1955610. [DOI] [Google Scholar]
  20. Olszewska-Guizzo A., Fogel A., Escoffier N., Ho R. Effects of COVID-19-related stay-at-home order on neuropsychophyisiological response to urban spaces: Beneficial role of exposure to nature? Journal of Environmental Psychology. 2021;75 doi: 10.1016/j.jenvp.2021.101590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Pal K., Danda S. Stress, anxiety triggers and mental health care needs among general public under lockdown during COVID-19 pandemic: A cross-sectional study in India. International Journal of Mental Health and Addiction. 2021 doi: 10.1007/s11469-021-00596-x. Advance online publication. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Parker S.L. Toward an understanding of ‘rally’ effects: Public opinion in the Persian Gulf war. Public Opinion Quarterly. 1995;59:526–546. https://doi-org.eres.qnl.qa/10.1086/269492 [Google Scholar]
  23. Probst T., Budimir S., Pieh C. Depression in and after COVID-19 lockdown in Austria and the role of stress and loneliness in lockdown: A longitudinal study. Journal of Affective Disorders. 2021;277:962–963. doi: 10.1016/j.jad.2020.09.047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Russo S., Mirisola A., Dallago F., Roccato M. Facing natural disasters through the endorsement of authoritarian attitudes. Journal of Environmental Psychology. 2020;68 doi: 10.1016/j.jenvp.2020.101412. [DOI] [Google Scholar]
  25. Somma A., Krueger R.F., Markon K.E., Gialdi G., Colanino M., Ferlito D., Liotta C., Frau C., Fossati A. A longitudinal study on clinically relevant self-reported depression, anxiety and acute stress features among Italian community-dwelling adults during the COVID-19 related lockdown: Evidence of a predictive role for baseline dysfunctional personality dimensions. Journal of Affective Disorders. 2021;282:364–371. doi: 10.1016/j.jad.2020.12.165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Thornock C.M., Nelson L.J., Porter C.L., Evans-Stout C.A. There's no place like home: The associations between residential attributes and family functioning. Journal of Environmental Psychology. 2019;64:39–47. doi: 10.1016/j.jenvp.2019.04.011. [DOI] [Google Scholar]
  27. Torshizian E., Grimes A. Household crowding measures: A comparison and external test of validity. Journal of Happiness Studies. 2020;22:1925–1951. doi: 10.1007/s10902-020-00302-z. [DOI] [Google Scholar]
  28. Veitch J.A. In: The Oxford handbook of environmental and conservation psychology. Clayton S.D., editor. Oxford University Press; Oxford: 2012. Work environments; pp. 248–275. [Google Scholar]
  29. Wener R. Cambridge University Press; New York, NY: 2012. The environmental psychology of prisons and jails: Creating humane spaces in secure settings. [Google Scholar]

Articles from Journal of Environmental Psychology are provided here courtesy of Elsevier

RESOURCES