key: cord-0828415-w0v37m4h
authors: Hananel, Ravit; Fishman, Ram; Malovicki-Yaffe, Nechumi
title: Urban diversity and epidemic resilience: The case of the COVID-19
date: 2021-12-10
journal: Cities
DOI: 10.1016/j.cities.2021.103526
sha: bd5b7899f87b3ca16399ed1dc7f5a9cc60310969
doc_id: 828415
cord_uid: w0v37m4h

The spread of the coronavirus pandemic offers a unique opportunity to improve our understanding of the role of urban planning strategies in the resilience of urban communities confronting a pandemic. This study examines the relationship between urban diversity and epidemiological resilience by empirically assessing the relation between the level of neighborhood homogeneity and the probability of being infected by the coronavirus. We focus on the ultra-Orthodox Jewish community in Israel, a relatively closed community that was disproportionately and severely affected by the pandemic. The findings indicate a monotonic but nonlinear relationship between the level of ultra-Orthodox prevalence in a neighborhood and a resident's probability of contracting COVID-19. As the fraction of ultra-Orthodox individuals in the neighborhood decreases, the fraction of infected population decreases significantly and more strongly that can be explained without recourse to urban diversity considerations. This relationship is found to be significant and strong, even when other variables are accounted for that had hitherto been perceived as central to coronavirus distribution, such as housing density, socioeconomic level of the neighborhood, and number of people per household. The findings are important and relevant to many societies around the globe in which a variety of populations have a separatist lifestyle.

J o u r n a l P r e -p r o o f

In recent months the world has experienced a severe pandemic caused by the novel coronavirus. According to the World Health Organization, as of July 13, 2021, more than 186 million people had been diagnosed with COVID-19, and more that 4 million had died, in 216 countries (WHO, 2021) . While morbidity rates vary both across and within countries, some of the hardest hit are closed, homogeneous communities, of which the ultra-Orthodox Jewish communities in Israel, New York, and other countries offer a stark example (Gilman, 2021; Halbfinger & Kershner, 2020; Heilman, 2020; Holmes 2020; Schattner & Klepfish 2020; Stack, 2020) .

The global spread of the Covid-19 pandemic has spurred a large body of research aimed at understanding the factors that influence variation in infection and mortality rates across countries, communities and individuals. Much of this research has focused on demographic (Venkastesan, 2020; Volk et al., 2021; Vanhamel, 2021) , economic (Baldwin & Weder di Mauro, 2021; Nicola et all, 2020; Omer et al., 2020; Saban, Myers, Shachar, Miron, & Wilf-Miron, 2021) , socioeconomic and ethnic (Saban et al., 2021 (Saban et al., 2021 , health (Peters, et al. 2020; Pirutinsky et all, 2020; Wang & Tang, 2020) and environmental circumstances (Eroğlu, 2021; Gautam, 2020; Gautam & Trivedi, 2020) . However relatively little attention has been given to the impact of Florida, 2020; Sokol, 2020; Bogost, 2020; Stack, 2020) . However, as far as we are aware, no empirical research has thus far focused on this question at the neighborhood level.

In contemporary urban studies, urban diversity is considered a planning strategy that helps produce better, more just, and more livable cities. The premise of this approach is that mixing diverse populations with different backgrounds and income levels and mixing land use and various forms of housing can produce neighborhoods whose residents are better off both economically and socially (Grant, 2002; Fainstein, 2016; Hananel, 2017; Talen, 2005) . Consequently, many scholars since the 1960s have supported urban-redevelopment strategies that stimulate physical and social heterogeneity (Fainstein, 2005a; Talen, 2012; Shamai & Hananel, 2020) .

This study examines the relationship between urban diversity and epidemiological resilience by empirically assessing the relation between the level of neighborhood homogeneity and the probability of being infected by the coronavirus.

We focus on the homogeneity of the neighborhood in terms of populations that are relatively closed and isolated, and which have been widely blamed for the spread of the pandemic. Our case study is concerned with the ultra-Orthodox (haredi in Hebrew; henceforth, UO) community, a relatively closed community (Malovicki-Yaffe et al., 2018; Cahaner & Shilhav, 2012) that was disproportionately and severely affected by the pandemic in both Israel and the United States, especially in its early stages (Bateman, 2020; Halbfinger, 2020; Sales, 2020) . In recent years, the UO community has undergone major changes in its degree of closure and isolation. Some members of the community continue to live in closed neighborhoods, while others have moved to more mixed neighborhoods (Alfasi et al., 2013; Malach & Cahaner 2017 ). This enables us to examine a wide range of neighborhood homogeneity values, especially as we are able to utilize unique, detailed data spanning all neighborhoods in Israel.

The structure of this paper is as follows: The next section briefly presents the theoretical research framework, namely, the urban diversity approach. The third section introduces the UO community, its unique characteristics, and its spatial dispersal. The fourth section describes the research methodology. The last two sections present the findings and the conclusions of the research and discuss their broader relevance for future decision making.

Our results indicate a monotonic but nonlinear relationship between the level of UO prevalence in a neighborhood and a resident's probability of contracting COVID-19. As the fraction of UO individuals in the neighborhood decreases, the fraction of infected population decreases significantly and more strongly that can be explained without recourse to urban diversity considerations. This relationship is found to be significant and strong, even when other variables are accounted for that had hitherto been perceived as central to coronavirus distribution, such as housing density, socioeconomic level of the neighborhood, and number of people in the household.

Although the findings focus on the UO community in Israel, they are important and relevant at various levels to many societies around the globe in which a variety of populations have a separatist lifestyle. On a practical level, the findings emphasize the central role of urban planning strategies in maintaining public health and their ability to affect the resilience of the public and reduce its likelihood of contracting epidemic diseases.

Despite the seeming unanimity of urban theorists on the merits of urban diversity, in recent years there has been a growing opposition that has put a spotlight on some negative social effects of urban diversity (Fainstein, 2005) . Social diversity usually means the integration and assimilation of population from various income level. However, studies show that social mix not always contribute to improving the socioeconomic status of disadvantaged local residents, but also exacerbates economic inequalities, displacement, segregation and isolation. In other cases, disadvantaged tenants are not readily accepted into more affluent communities, and they experience oppression, stigmatization, exclusion, and even hostility (August, 2016; Biddulph, 2011; Ruming et al., 2004; Teernstra, 2015; Ye, 2017) .

Resilience is defined as "the capacity of a system to absorb disturbance and reorganize, while undergoing change so as to still retain essentially the same function structure, identity and feedbacks" (Walker, Holling, Carpenter, & Kinzig, 2004, p. 6 ).

Originally, the term was used by physical scientists to denote the characteristics of a spring and to describe the stability of materials and their resistance regarding external shocks (Davoudi, 2012) . In recent decades, resilience has become an increasingly popular concept and has begun to be used also by social scientists, economists, and urban planners and in relation to government policy and strategies (Feinstein, 2015;

Porter & Davoudi, 2012; Wilkinson, 2012) .

In urban studies the concept in general is aimed to describe the qualities that help communities and individuals overcome natural crises, such as earthquakes and floods, or sociopolitical crises, such as financial calamities, mass immigration, war, peace agreements, and social protests (Barrios, 2014; Davoudi, J o u r n a l P r e -p r o o f 2012; Fitzgibbons & Mitchell, 2019; Hananel, 2019; S. Meerow et al., 2016; S. Meerow & Newell, 2016; Simone el al., 2021) .

Definitions for urban resilience vary, but all the definitions and approaches, explain Albers & Deppisch (2013), refer to diversity as a key principle for urban and regional resilience. The principle of "diversity", meaning that a city or a region has a number of functionally different components that exist side by side, help to protect a city or a region system against various threats and to reduce vulnerability (Wardekker et al., 2010; Godschalk, 2003) . In this context, diversity refers not only to ecological, but also to social, cultural or economic diversity (Albers & Deppisch, 2013; Kumagai et al., 2010; Wardekker et al., 2010) .

At the neighborhood level, studies emphasize the need for diversity, and explore how the lack of diversity (within both neighborhoods and surrounding contexts) affects the neighborhoods' ability to withstand a shock such as the Great Recession It is often argued that cultural diversity plays a major role in creating mechanisms for innovation, providing new ways to adapt to change, and generating knowledge and institutions to deal with the challenges, opportunities and threats generated by change (UNESCO, 2008) . According to Coldin & Barthel (2013) , the term cultural diversity encompasses a diversity of social relations among people of different ethnic background, age, or gender. Friedman, 1989; Stadler, 2002) . In the urban context, the ultra-Orthodox community can be considered as a gated community, in which forming an environment that protects unique cultural values, lifestyles, and social cohesion serves as a major motive for enclosure (Rosen and Razin, 2008) .

Communal life is a central characteristic of the UO lifestyle. The community is the basic unit because it correlates with religious needs, many of which are met only in communal life (Friedman, 1991; Stadler, 2002; Stadler, at all, 2008) . Such needs include prayer quarters, religious facilities such as ritual baths, and shops selling kosher products (Shoshana, 2014) .

There are UO communities all over the world. Currently, this population is estimated to number some 1.8 million people who share many characteristics. Israel has the biggest community, with some 1 million members (CBS, 2019). i Next is the United States, with about 500,000, concentrated mainly in the area of New York City and New Jersey (Pew Research Center, 2013; Wisse, 2007; Brown, 2000) . The remainder live in England, France, and the rest of Europe and in other small communities globally.

The UO society established in Israel continued this spirit of isolation-"walls of holiness" between them and the rest of the Israeli population-and is considered the most segregated and religiously stringent of all UO communities (Brown, 2000 (Brown, , 2014 .

A key feature of the UO community in Israel is its being a "society of learners": Most men in the community invest many years studying sacred Jewish texts, and their wives are the main breadwinners. This phenomenon emerged in the J o u r n a l P r e -p r o o f late 1970s because of historical and political factors in Israeli society and affected the geographical boundaries of the community, leading to a unique social and spatial structure. Because UO men do not leave their neighborhoods, where they have their study centers and community facilities, they could stay isolated within the "walls of holiness" (Malovicki-Yaffe, et al., 2018a Stadler, 2009 ).

Until the mid-1980s, the geographical area of the UO community in Israel was known as the "Haredi Triangle," referring to an area encompassing Jerusalem, Bnei

Brak, and Ashdod, three main cities in central Israel. In those cities, the UO were segregated in well-defined neighborhoods. Most of those neighborhoods were established on the margins of the cities and were designed for the UO only (Cahaner, & Shalav, 2012; Zicherman, 2016) In the beginning of the 1990s, demographic growth and a housing shortage led to significant spatial changes. (Malach & Cahaner, 2017; Regev, 2019) . Often, the UO lifestyle leads to friction and conflict with the local residents, who wish to preserve the secular character of their "gentrified" neighborhoods; in many places it has led to the exodus of the secular population from the neighborhood. (Cahaner & Shilhav, 2012; Mansfeld, & Cahaner, 2013; Zicherman, 2016) . iv

The COVID-19 outbreak in Israel occurred when relations between secular and UO Israelis were already strained (Friedman, 2020) . When a general lockdown was declared on March 16, the leaders of the UO community responded with J o u r n a l P r e -p r o o f suspicion and refused to cooperate ( Malchi, Malach & Friedman, 2020 ; Kingsley 2021 ). The traditional OU lifestyle involves high levels of daily interaction between community members, such as prayer in a quorum of ten men, thrice a day, and ritual bathing. There was particularly sensitivity with regard to educational institutions, as

Torah study is considered an essential religious activity, and any attempt to interrupt or interfere with the education of the community's children and youth is viewed as an attack on its "holy of holies." Consequently, some leading rabbis initially opposed the instructions to close schools and Torah study halls, and instructed their followers to maintain group study and prayer. Overall, the public health restrictions were seen as unnecessary and as a continuation of the state's supposed religious discrimination and harassment vis-à-vis the UO community (Malach & Freidman 2020; Stern 2020; ) However, rates of infection and morbidity rose rapidly within the community, and 42% of all COVID cases diagnosed in Israel in the first wave of the outbreak of the epidemic were in the UO community. By October the rate rose to 50% of all cases in Israel while ultra-Orthodox society forms only 12.5% of Israeli citizens. Moreover, 1 in 73 UO over the age of 65 has died during the outbreak. This is more than four times the number in the same cohort of the general population) Hanuau, 2020; Schwartz & Lieber, 2021) .

In order to stop the spread of infection the Israeli government imposed strict quarantine on some UO homogeneous neighborhoods as they were identified as epicenters of the outbreak. The army was called in to enforce curfews and to distribute food and other essentials to residents of Bnei Brak, a major UO city, in which as many as 38 percent of the 200,000 residents were infected. The town was declared a "restricted zone." (Holmes 2020; Saban, 2021; Guttentag, 2020) .

The high mortality rates were not unique to the Israeli UO community. Other UO communities around the world display similarly high rates of infaction. The UO comminty acoounted for 30% of all cases in New York city (Goldstein 2020; Maslin

Nir & Otterman 2020) and the UO community in London displayed infection rates of 64%, nine times higher than the UK average of 7% (Burgess, 2021; Gilman, 2020) .

Behavior such as ignoring mask mandates and social distancing requirements, and continuing to attend synagogues, Yeshivot (houses of study) and even mass funerals and weddings (Hanau, 2020; Kingsley 2021) were common across UO communities around the world. The autonomy and the insularity of UO neighborhoods allowed the communities to do as they pleased and ignore government regulation.

This paper seeks to examine the relations between the neighborhoods' diversity, in sense of the level of the neighborhood homogeneity/heterogeneity, and the probability to be infected with COVID-19. The paper asks whether people in heterogeneous environment may act differently, and in the case of global pandemic, such as COVID-19, if diversity would increase neighborhood resilience. As discussed above, cultural diversity provides new ways to adapt to change and promote learning and adaptation in groups (Coldin & Barthel, 2013; UNESCO, 2008) .

To empirically examine the relationship between the level of neighborhood diversity (i.e. share of UO in the neighborhood) and its infection rates, we designed a multistage comprehensive methodology that combines descriptive statistics and regression analysis.

The analysis is based on two data sets. 

First, we examined the relationship between the number of confirmed cases in a neighborhood and its UO rank (UOR), the non-ultra-orthodox religious Jews rank (RJR) and the Arab rank (AR) through a descriptive analysis. As we will see below,

only the UOR displayed a strong relation with the case count.

We then conducted a regression analysis at the neighborhood level, in which the outcome variable is a measure of the prevalence of confirmed cases in the neighborhood (either the infection rate or the case count), and the key explanatory variable is the neighborhood's UOR (Appendix. Summary Statistics). We make no assumption on the form of the relation between the UOR and the prevalence of confirmed cases, e.g. that it is linear. Rather, we estimate a flexible non-parametric regression which includes five binary variables indicating whether the neighborhood belongs to the five categories of the UOR. Formally, we estimated the regression:

UOR j = ∑ a k R j,k + b X j + e j J o u r n a l P r e -p r o o f where the index j signifies a neighborhood and the sum runs over the index k=1,2,3,4,5 which signifies the UOR categories. The variables R j,k are binary indicators (taking the values of 0/1) of whether a particular neighborhood j belongs to UOR rank k. The coefficients to be estimated, a k , reflect the average rate of infection (to be defined below) in the neighborhoods belonging to UOR category k. Estimating all five coefficients allows us to map the shape of the relation between UOR and prevalence flexibly without making any a-priori assumptions about its form.

The regression also includes control variables X, described below. The error terms are denoted by e j . Rather than assuming that the error terms are independent, we conservatively allow for the possibility that error terms representing neighborhoods within the same municipality are correlated. This should have the effect of increasing the standard errors of our estimates, making it "harder" to obtain statistically significant results.

We estimate two types of regressions. In the first, the outcome variable is the rate of infection, defined as the number of confirmed cases in a neighborhood per 100 residents (i.e. in percent). We estimate this relation through a OLS linear regression.

Since the actual number of cases in a neighborhood is a count variable with frequent zero values, we also estimate a more appropriate Poisson regression model in order to ensure our results are insensitive to the model specification. In this model, the outcome variable is the actual number of confirmed cases and the explanatory variables are the same as above. A Poisson regression assumes that the number of cases in a neighborhood follows a Poisson distribution with a mean value whose logarithm depends linearly on the explanatory variables.

If significant correlations are found between the UOR and the infection rate, it is of interest to assess which attributes of the UO population might be responsible for J o u r n a l P r e -p r o o f the association. We examine this question in two ways. First, we add the RJR and the AR of the neighborhood to the regression as additional explanatory variables. We then compare the strength of the relation between the prevalence of infection and these other rankings (i.e. their regression coefficients) to that of the UOR. To the extent that the relation between the infection prevalence and the UOR is driven by attributes shared by the UO or RJ communities (such as Jewish religious lifestyles) or between the UO or Arab communities (such as relative social isolation and mistrust of the government), one would expect the coefficients of the corresponding ranks to be of similar magnitude.

The second approach to this question is to include additional variables in the regression that might directly capture some of the attributes potentially driving the association between the UOR and the infection rate. For example, one might hypothesis that UO population is more likely to be infected because of its high density. If this were the case, controlling for the population density in the regression should reduce the strength of the association between UOR and the infection prevalence, i.e. lower the magnitude of the coefficients a k . We therefore also estimate regressions in which we progressively add additional control variables representing potential attributes of the UO society that may drive infection. These include the socio-economic rank of the neighborhood, its density (number of households per unit area ix ), the average household size, the technological index, and the fraction of the population which is above the age of 15.

Our empirical approach does not allow us to identify the causal impact of UOR on Covid-19 infection, but rather the correlation between these two variables.

To the extent that a significant correlation exists between these two variables that is robust to the inclusion of other correlates of UOR in the model, we interpret such a finding as suggestive that certain aspects of UO life cause Covid-19 infection to increase, but we remain aware of falling short of being able to conclusively prove the association is causal.

Our data covers a total population of 9,123,496, residing in 2383 neighborhoods. Table 2 presents the distribution of neighborhoods and people in Israel by the three neighborhood homogeneity ranks: the UO rank (UOR), the nonultra-orthodox religious Jews rank (RJR) and the Arab rank (AR).

Next, in Figure 2 , we compare the distribution of Israel's population and its total confirmed case count by each of the three neighborhood homogeneity ranks (separately for each rank).

The plots clearly show that the relationship between the case count and the Table 2 ).

Importantly, this plot also shows that the increasing relationship between the share of UO in the neighborhood and the infection rate is not linear, but convex, meaning that the rate of increase in the infection rate is itself increasing with the fraction of UO populations.

To see the significance of this, consider a simple model in which UO and non- individual is independent of the share of UO in the neighborhood, but rather that the probability of an UO individual becoming infected decreases significantly the more heterogeneous the neighborhood is. In other words, it suggests that UO who live in heterogeneous neighborhoods are less likely to be infected.

To examine the evidence for this more carefully, in Figure 3 , we again plot the average infection rates (black markers) in each UOR value against the UO share in the population (horizontal axis), but this time add their 95% confidence intervals (dotted black lines). We also plot (red triangles) the change from in the infection rate from each UOR category to the next, normalized for ease of visualization in units of 10%

of the share of UO in the neighborhood population, which we denote by D i . For example, the change in infection rates from UOR 4 to 5 is 0.82-0.45=0.37, and occurs over a change in the UO from 60% to 85%, so the change per 10% is D 5 =0.37/2.5=0.14.

If the probability of infection of UO were constant, one would expect D 2 =D 3 =D 4 =D 5. However, the plot suggests that the magnitude of D i in fact rises with the share of the orthodox population. Indeed, statistical t-tests reject equality of many of these coefficients, including that D 5 =D 3 (p<0.01) or that D 5 =D 2 (p<0.01), as well as that D 4 =D 2 (p=0.08), and are close to reject that D 5 =D 4 (p=0.15). This provides evidence against the hypothesis that the probability of infection for an UO individual is independent of the share of UO in the neighborhood, as discussed above.

In order to more carefully estimate the association between the UOR and the infection rate, we now turn to a regression analysis in which the outcome variable is a J o u r n a l P r e -p r o o f measure of the infection prevalence. We estimate two types of regression models: in the first, the outcome variable is the case count in a given neighborhood, and we estimate a Poisson regression. In the second, the outcome variable is the infection rate, and we estimate an OLS regression.

In Table 3 , we report estimations of Poisson regressions in which the outcome variable is the case count in a given neighborhood. To begin with, for brevity, we assume a linear relation between the UOR and the logarithm of the case count, and later relax this assumption. The table reports the incidence rate ratios, which can be interpreted as the factor by which the average case count is estimated to increase per one unit change in each of the explanatory variables.

In the first column of Table 3 , we report estimations of a basic model in which the only explanatory variables are the UOR and the population of the neighborhood (in thousands). The UOR is found to have a large and statistically significant effect on the case count: a one unit increase in the UOR is associated with about a doubling of the case count.

In subsequent columns of Tables 3, we add additional controls to the regression in order to test whether the association between UOR and the case count is driven by particular characteristics of the orthodox community. First, we control for the RJR and AR. The RJR captures the prevalence of the modern orthodox population in the neighborhood, who share similar religious practices, such as praying in groups several times a day. The Arab rank reflects the prevalence of Arab population in the neighborhood, which is often compared to the orthodox community in terms of social J o u r n a l P r e -p r o o f isolation and socio-economic situation. Adding these controls does not materially effect the estimated association between the UOR and the case count. The estimated associations between the case count and the RJR is much smaller in magnitude and is not statistically significant. The AR is, in fact, negatively associated with the case count, with a one unit increase in the AR associated with an about 11% decline in the case count (p=0.09). This suggests that neither religious practices or social isolation are likely to be driving the association between the UOR and infection prevalence.

Adding a control for the socio-economic rank of the neighborhood (Column 3)

does not have much of an effect the coefficient of the UOR, suggesting that the association between UOR and infection rates is also not driven by socio-economic conditions. Interestingly, the relation with the socio-economic rank is positive (p=0.06). The negative coefficient of the AR becomes smaller and insignificant, suggesting it was driven by socio-economic factors.

In Column 4, we add controls for two components of overall population density: the neighborhood density, and the mean size of the household, both of which are high in orthodox communities and could potentially be increasing infection rates.

While we find a positive and significant association between the case count and both indicators, as expected, the coefficient for UOR is slightly smaller, but remains large and significant, suggesting the association between the UOR and the case count is not driven by the density of the UO population.

In Column 5 we also control for technological literacy, finding an insignificant association on case count and little effect on the coefficient of interest. Finally, in Column 6 we control for the age composition of the population (the fraction of the population above the age of 15) to account for the possibility that the younger profile of the orthodox population might be somehow increasing infection rates. The results J o u r n a l P r e -p r o o f suggest that older populations display higher case counts, but the coefficient of the UOR is not significantly changed, suggesting the association between the UOR and the case count is also not driven by its age composition.

The regression estimates reported above assumed a linear relation between the UOR and the case count. We now relax these assumptions and estimate nonparametric associations which make no assumptions about the functional form. In Figure 4 , we plot the estimated coefficients of UOR categories 2-5, which represent the factor by which case count is estimated to increase relative to neighborhoods with UOR=1, along with 95% confidence intervals. The coefficients are plotted in different color for each of the models estimated in Columns 1-5 of Table 4 (labelled basic, other groups, socio-economic, density, technology, age, respectively).

The results indicate a clear rise in case count with the UOR. All models yield highly similar estimated coefficients, similarly to what was found in the linear model and reported in Table 4 . Relative to neighborhoods with UOR=1, those with UOR=2 are expected to have an almost double case count; those with UOR=3, a case count 3-4 times higher; those with UOR=4, 7-8 times higher; and those with UOR=5, 10-15

higher. Confidence intervals widen as the UOR goes above 3, likely reflecting the small number of neighborhoods in these categories.

For completeness, we now present parallel results to those presented in section 5.3.1 but derived from an OLS regression model in which the outcome variables is the infection rate (the number of cases divided by the total population). All other elements of the model are similar. In Table 4 

It was particularly interesting to find that, even when controlling for the density of the neighborhood and for household sizes, the impact of the UOR remains almost as strong and significant. This finding corresponds with the findings of a new study examining the effect of population density and connectivity in American metropolitan areas on the spread of COVID-19, which found that connectivity matters more than density (Hamidi et al., 2020) . An alternative way of thinking of this issue is the notion of the relationship between internal and external norms in closed communities. One of the biggest sources of social tension during the pandemic, and one that gained much media attention, was the lack of compliance among UO populations with mask-wearing and social distancing requirements-not just in Israel, but also in the large UO communities in New York and London. All of these communities suffered high COVID-19 infection rates ( Halbfinger, 2020; Holmes 2020; Sales, 2020 Antwerp were no higher than in the general population (Vanhamel et al., 2021) . This, J o u r n a l P r e -p r o o f focuses a spotlight on an important question in urban planning, regarding the creation or formation of insular neighborhoods with their own strongly defined social norms.

Most people adhere to social norms since deviating makes us feel odd or weird (Cialdini & Goldstein, 2004) or deviating makes us feel wrong and guilty and deviating threatens our relationships with our community (Crocker et al., 1998 implementing the urban-diversity approach in the construction of residential neighborhoods. As we have seen in the present case, the presence of neighbors from J o u r n a l P r e -p r o o f diverse backgrounds and lifestyles is a factor that reduces the likelihood of contracting COVID-19. We definitely are not arguing that the neighborhood's degree of heterogeneity is the only factor that affects the probability of contracting the disease, but our findings indicate a strong connection between the two that should not be ignored. The findings highlight the role of urban planning strategies in maintaining public health and their ability to reduce the likelihood and severity of epidemics.

Furthermore, considerable thought should be given in urban planning to the characteristics of communities, their social structures, and the formation of social norms within them.

v In May 5 th 2020, ended the general quarantine, which was implemented in Israel since mid-March. In retrospect, it was only the "first wave", but at that time it seemed like the end of the coronavirus pandemic in Israel. vi Points Location Intelligence website https://points.co.il/en/points-location-intelligence/ vii In the Ministry of Health the field is called "neighborhood name", while in the Points dataset, the field is called "EZ_NAME" viii A few neighborhoods were grouped by ministry of Health into aggregated observations (apparently, due to medical secrecy reasons, since separately those neighborhoods had a small coronavirus cases). For these observations, Points data was also aggregated using the comprising neighborhoods' populations as weights. ix Dunam is the unit of area measurement used in Israel. 1 Dunam = 0.25 Acre x It is worth noting that the UO community is largely law-abiding, and is also guided in this respect by the concept of "chilul Hashem," according to which it is important to follow accepted secular norms and laws in order not to damage the reputation of religious Judaism in the outside world and attract criticism. This is an important religious precept, and one that in most cases UO Jews would be very mindful of and extremely careful to uphold (Malvicki-Yaffe 2020). Yet the behavior observed during the COVID-19 pandemic was different.

J o u r n a l P r e -p r o o f Journal Pre-proof J o u r n a l P r e -p r o o f 

 The study examines the relationship between urban diversity and epidemiological resilience  The Study focuses on the ultra-Orthodox community in Israel, a relatively closed community  The Ultra-Orthodox community in Israel and elsewhere suffered from a high level of infections and mortality rates compared to the general population  the research examines the relationship between the level of neighborhood homogeneity with respect to the UO population and the probability of being infected by the coronavirus  The research findings indicate a nonlinear relationship between the neighborhood's UOR and the probability of infection. 

Coronavirus is revitalising the concept of community for the 21st century. The Conversation

Between the Individual and the Community: Residential Patterns of the Haredi Population in Jerusalem

Jerusalem of (foreign) gold: entrepreneurship and pattern-driven policy in a historic city

Revitalisation gone wrong: Mixed-income public housing redevelopment in Toronto's Don Mount Court

Coronavirus: Israel's Ultra-Orthodox Lockdown Challenge

Haredi Conceptions of Obligations and Rights: Polish Jewry, c. 1900-1939

Economics in the Time of COVID-19

Here, I'm not at ease": Anthropological perspectives on community resilience

Community Resilience: Toward an Integrated Approach

Urban design, regeneration and the entrepreneurial city

Uncharted: Revenge of the Suburbs-Suburbia was never as bad as anyone said it was. Now it's looking even better. The Atlantic

Coronavirus and Immigration Detention in Europe: The Short Summer of Abolitionism?

Orthodox Judaism

Jewish Political Theology: The Doctrine of "Da'at Torah" as a Case Study

64% of London's ultra-orthodox Jews infected with Covid. The Times

From survival to consolidation: Changes in Israeli Haredi society and its scholarly study

A voyage from religiousness to secularity and back: A glimpse into "Haredi" tourists

Israel in numbers on the Jewish New Year's Eve. Selected data from the annual tables

Measuring urban diversity of Songjiang New Town: A re-configuration of a Chinese suburb

The potential of "Urban Green Commons" in the resilience building of cities

Transitioning to resilience and sustainability in urban communities

Social influence: Social norms, conformity and compliance

Social stigma

Resilience: A Bridging Concept or a Dead End? Planning Theory and Practice

The War Within: Israel's Ultra-Orthodox

Effects of Covid-19 outbreak on environment and renewable energy sector. Environment, Development and Sustainability

Cities and diversity: should we want it? Can we plan for it

The return of urban renewal

The just city

Resilience and justice

Spatial Justice and Planning

The Jewish Enlightenment

Just urban futures? Exploring equity in "100 resilient cities

The Forces That Will Reshape American Cities. Bloomberg City Lab

The rise of the creative class

Growth and Segregation-The Ultra-Orthodox Community of Jerusalem

The Haredi (ultra-orthodox) society: Sources, trends and processes. Jerusalem: The Jerusalem Institute for Israel Studies

Ultra-Orthodox non-compliance rests on fear" The Times of Israel

Neighborhood social mix as a goal of housing policy: a theoretical analysis

The influence of COVID-19 on air quality in India: A boon or inutile

Global implication of bioaerosol in pandemic

Israelis are suffering due to coronavirus, and Haredim have been made the scapegoat. Washington Jewish Week

Placing the blame for Covid-19 in and on ultra-Orthodoxcommunities

Urban Hazard mitigation: Creating resilient cities

Mixed Use in Theory and Practice: Canadian Experience with Implementing a Planning Principle

Israel's Virus Czar Was Making Headway. Then He Tangled with a Key Netanyahu Ally

Virus Soars Among Ultra-Orthodox Jews as Many Flout Israel's Rules

Does Density Aggravate the COVID-19 Pandemic?

From central to marginal: The trajectory of Israel's publichousing policy

Can centralization, decentralization and welfare go together? The case of Massachusetts affordable housing policy

NYC Jewish schools to stay open as public education institutions close

2020) 1 in 73 haredi Orthodox Israelis over age 65 has died of COVIDreport

Is anybody home? The influence of ghost apartments on urban diversity in Tel-Aviv and Jerusalem

From New York to Milan, how coronavirus is hitting Jewish communities worldwide, Times of Israel May

The Attitude of the Ḥatam Sofer toward Moses Mendelssohn

Calls to Seal off Ultra-Orthodox Areas add to Israel's Virus Tensions. The Guardian

Westernization and Israelization within Israel's extreme Orthodox Haredi society

How Many Funerals Will Come Out of This One? The New York Times

Social implications of housing diversification in urban renewal: A review of recent literature

Translating resilience into the urban context: Past successes and future uncertainties in Tokyo

The nature of cities and the COVID-19 pandemic

Climate Change and Urban Resilience

Norm violators as threats and opportunities: The many faces of deviance in groups

How can we curb the spread of the coronavirus in the Haredi community" The times of Israel Retrieved

Elements of Modern Life or "Modern Ultra-Orthodoxy"? Numerical Assessment of Modernization Processes in Ultra-Orthodox Society

Haredim and Coronavirus: Policy Recommendations for Exit from Second Lockdown. The Israel Democracy Institute

Poor is pious": Distinctiveness threat increases glorification of poverty among the poor

Ultra-Orthodox Jewish Tourism: A Differential Passage Out of a Sociocultural Bubble to the "Open Space

Antivirus-built environment: Lessons learned from Covid-19 pandemic

Urban resilience for whom, what, when, where, and why? Urban Geography

Reactions to defiant deviants: Deliverance or defensiveness?

The socio-economic implications of the coronavirus pandemic (COVID-19): A review

Adaptive cyclic exit strategies from lockdown to suppress COVID-19 and allow economic activity

Understanding the emerging coronavirus: What it means for health security and infection prevention

A portrait of Jewish Americans. Pew Religion and Life Project

COVID-19, Mental Health, and Religious Coping Among American Orthodox Jews

The politics of resilience for planning: A cautionary note. Planning

Questioning the rhetoric of social mix: courteous community or hidden hostility?

Demographic trends in Israel

Housing market characteristics and the geographical distribution of the ultra-Orthodox population in Israel

Brooklyn's Orthodox neighborhoods have especially high rates of the coronavirus

Towards Cosmopolis: planning for multicultural cities

Israel's Leaders Clash With Ultraorthodox Over Covid-19 Lockdowns

One + One + One = A Lot: The Cumulative Effect of Israel's Flagship Urban Renewal Policy on Neighborhood Diversity

Orthodox Judaism as a risk factor of Covid-19 in Israel

The Ḥatam Sofer's Nuanced Attitude Towards Secular Learning

Growth and segregation: the ultra-Orthodox community of Jerusalem. Jerusalem Institute for Israel Studies

Space, heterogeneity, and everyday life: Ultra-orthodox heterotopia in Israel

Rome was not built in a day. Resilience and the eternal city: Insights for urban management

Plague on a Biblical Scale

Effect of socioeconomic and ethnic characteristics on COVID-19 infection: The case of the Ultra-Orthodox and the Arab communities in Israel

Transitioning to resilience and sustainability in urban communities

Will the Coronavirus Make the Suburbs Popular Again?

Is profane work an obstacle to salvation? The case of ultra-Orthodox (Haredi) Jews in contemporary Israel

Yeshiva fundamentalism: Piety, gender, and resistance in the ultra-Orthodox world

Fundamentalism's encounters with citizenship: the Haredim in Israel

Kanievsky's actions close to organized civil disobedience campaign" The Jerusalem Post

Enclosed residential neighborhoods in Israel: from landscapes of heritage and frontier enclaves to new gated communities

Public participation and local environmental planning: The collective action problem and the potential of social capital

Land use zoning and human diversity: Exploring the connection

Zoning and diversity in historical perspective

Contextualizing state-led gentrification: Goals of governing actors in generating neighborhood upgrading

The changing demographics of COVID-19

Understanding how communities respond to COVID-19: experiences from the Orthodox Jewish communities of Antwerp city

The influence of demographics and personality on COVID-19 coping in young adults

Technological Forecasting & Social Change

Combating COVID-19: health equity matters

Resilience, adaptability and transformability in social-ecological systems

Coronavirus pandemic: Numbers at a glance

Urban resilience: What does it mean in planning practice? Planning Theory and Practice

The Yiddish policemen's union by Michael Chabon

Islands of decay in seas of renewal: Housing policy and the resurgence of gentrification

Contours of urban diversity and coexistence

Justice and the Politics of Difference

Strategic housing plan for the Haredi population 2016-2035. Jerusalem: The Haredi Institute for Public Affairs and the Ministry of Construction and Housing

13% live in the Judea and Samaria District, mainly in the two UO municipalities Beitar Illit and Modi'in Illit; 19% live in the Tel Aviv District, most of them in the city of Bnei Brak; 16% live in the Central District, most of them in two UO municipalities, Elad and Kfar Chabad, and also in the city of Petah Tikva; 12% live in the Southern District, most of of them in the cities of Kiryat Malachi

ii Examples include two small municipalities, Beitar Illit and Modi'in Illit

Kiryat Ye'arim, a municipality in the Jerusalem District; and Elad, a city in the Central District. iii Such was the case of Ramat Beit Shemesh