Responding rapidly to epidemic outbreaks presents significant challenges, due to resource, capacity and time limitations. Anticipatory Action (AA) is a newly emerging strategy in the field of humanitarian aid, designed to preemptively address potential crises. By taking impact-reducing actions before a disaster strikes, AA seeks to minimize human loss. However, AA frameworks currently use static prepared-in-advance plans. As a result, AA is not sufficiently able to deal with the uncertainty levels in the onset and spread of epidemics. Effective epidemic control requires plans that can adapt to a constantly changing environment and incoming information, such as the number and location of suspected cases, weather forecasts and population movement, while balancing flexibility with an effective management approach. We show how the (DAPP) framework for decisionmaking under deep uncertainty can be adapted to enhance the common anticipatory action approach with flexibility and effective management for epidemic control. More specifically, we show how DAPP allows to take into account newly available information and change the strategy to minimize human loss. We illustrate it with a case study of cholera in Cameroon, for which the French, Netherlands, and Cameroon Red Cross, supported by EHESP, are developing an early action protocol and a model that assesses the cost-effectiveness of actions for different risk levels and external shocks. Our results suggest that DAPP increase flexibility and coordination in anticipatory action for epidemics and helps optimizing early action strategies, which could have larger implications for global disease control.

The rise in urban stresses has prompted the need for preparedness of urban areas to face precarious circumstances. Consequently, the concept of urban resilience has grown in popularity not only to tackle sudden shocks but also to face long-lasting socio-economic tensions. To implement the arrangements for resilience policy, literature suggests that social factors govern the resilience of urban areas. Communities in which the residents work together and have common goals have a stronger willingness to cooperate. To this end, social cohesion has been proven to be significant for subsistence in the event of a catastrophe. Cohesive communities protect residents against threats, care of others during hardships, and ultimately promote community resilience.


Balanced neighbourhood policies aim to strengthen the cohesion between citizens, communities, and social institutions departing from the assumption that social mix fosters social cohesion. Their goal is to increase the social mix of specific areas to avoid the clustering and segregation of disadvantaged households to, as a result, promote resilient actions. There is, however, literature that suggests that the anticipated effects are rather inconclusive and usually not achieved. Instead, balanced neighbourhood policies would promote segregation by forcing the displacement of groups of residents.


The issue arises whether balanced neighbourhoods trigger resilient actions that are pivotal in resilient communities. In other words, does neighbourhood balance increase resilient action of neighbourhood residents? We took a cross-sectional confirmatory approach to understand the social mechanism that triggers resilient action in balanced neighbourhoods based on Partial Least Squares—Structural Equation Modelling (PLS-SEM) and spatial econometrics. To this end, the research tests the assumptions on which the municipality of Rotterdam grounds their Resilience Strategy's balanced neighbourhood policy, the Woonvisie, using a 2019 public survey on social development. Rotterdam's definition of a balanced neighbourhood is defined from a set of conditions for the amount of houses in different house price segments. Therefore, the tested model is based on the grounds that geographically connected people become affected by their neighbourhood's balance to promote resilient action. As such, we use the willingness to help friends and neighbours to characterise informal support as resilient action. Here we show that balanced neighbourhoods are associated with less informal support: the higher the balance, the fewer residents are willing to help their friends and neighbours.


The results indicate that social cohesion is the social mechanism that triggers help between friends and relatives and fully acts as the mechanism for resilient actions triggered by the balance in a neighbourhood. From the multiple combinations of houses in different house price segments that the definition of balanced neighbourhoods allows, we distinguish two associations. On one hand, house price distributions which foment a reduction in polarisation (more middle-priced houses) are negatively associated with social cohesion. On the other hand, balanced neighbourhoods which foment polarisation (more low- and high-priced houses) are positively associated with social cohesion. This indicates that our results are in line with Putnam’s homophily principle, i.e. ‘birds of a feather flock together’. This outcome is opposite to the policy discourse of governments in favour of balanced neighbourhoods, including the municipality of Rotterdam, that mixed neighbourhoods foster social cohesion and therefore resilient action.


The testing of the theory is complemented in two ways. First, we show that social cohesion and informal support are not constrained by administrative boundaries, so the social perceptions and actions in nearby neighbourhoods affects the level of the other neighbourhoods. Second, we found no moderating effect of factors related to the demographics and the built environment that can promote or deter social interactions, and thus are aspects of what can be considered a resilient neighbourhood.


The analysis also shows that Rotterdam's definition of balance allows multiple and dispersed combinations of the amount of houses in each price segment, which can result in counterintuitive conceptions of balance. In addition, the results show apparently contradicting results of the relationship between balance and social cohesion depending on whether the distribution foments house price polarisation. As a result, we argue that the definition is under-specified and can be misleading.


Finally, only 2.1% of the possible balance distributions yielded an acceptable goodness-of-fit of our model. This could be indicative that the model needs to be reevaluated. We found that neighbourhood ethnic heterogeneity and house type heterogeneity are directly associated to social cohesion and informal support, respectively. Future research should elaborate on the theory on which the model is grounded and create coherence to the empirical relationships identified. In contrast, the few fitting distributions could otherwise indicate that that social cohesion and informal support cannot be explained by the balance in a neighbourhood and that the policy should be reevaluated. Under this second interpretation, the study has uncovered which are the balance distributions for the city that can actually show the alleged effects of balance.


Based on these findings, a policy advice is formulated. If the objective is to increase social cohesion and resilient actions, we discourage the municipality of Rotterdam to approach this by building balanced neighbourhoods. Nevertheless, we have seen that building social cohesion is a way to build social resilience, so recently developed city programs focused on the development of neighbourhood organisations, which not only provide a space for social cohesion but also to collect and share resources directly, are a step forward from the Woonvisie.

Communicative diseases have put a burden on societal well-being throughout history. The global advent of the COVID-19 virus made the world painstakingly aware of the impact a pandemic can have on all levels of society. To combat the spread and impact of a pathogen, policymakers turn to specialized tools to craft policy. One of these tools is mathematical modeling and simulation. By using the agent-based modeling paradigm, it is possible to evaluate the impact of policy measures on the viral spread and healthcare burden. An agent-based model always contains a set of agents. Agents are the digital representations of a studied entity, for example, a person, but agents can also be an organization or a vehicle. One of the characteristics of agent-based modeling is the reliance on data to create a valid agent population. However, data on the level of individual people, also known as microdata, can be difficult to acquire and work with due to GDPR regulations. Aggregated population data, also known as open data, may serve as a substitute for microdata. Both open and microdata may be used to synthesize a population of agents, this process is known as synthetic population generation. Although both of these forms of data can be used to synthesize an agent population, it is currently understudied how the results of the agent population synthesis process are affected by the type of used population data. This research project focuses on this gap in knowledge and systematically analyzes the differences between an open-data-based and a microdata-based agent population. The populations that are synthesized are part of the HERoS model. The HERoS model is an epidemiological agent-based model that maps the spread of COVID-19 in the city of The Hague. The agents in the HERoS model are digital individuals, each characterized by a set of attributes. These attributes define the behavior of an agent within the model. Examples of agent attributes in the HERoS model are age, household size, and social role. Two agent populations for the HERoS model are synthesized, one using a sample-free algorithm that uses open data, and one using an algorithm that converts microdata to an agent population. The differences between the agent populations are quantified by introducing two new metrics, the modified Freeman-Tukey statistic, and agent matching. It was found that the agent populations show similar characteristics on a city level, but show significant differences at the neighborhood level. Furthermore, it is shown that the quality of an agent population is both dependent on the number of attributes that are synthesized, as well as the type of attributes. Moreover, the relationship between the quality of an agent population and the outcome a model produces is also understudied. This relationship is investigated by using the open-data-based and microdata-based agent populations as input for the HERoS model. The model outcomes are then compared on a global and neighborhood level, using phase comparison and the quantification of the precision of the model. It is found that the effect of using open data instead of microdata is multifaceted. The model outcome curves show similar characteristics, albeit they differ in a numerical sense. Using open-data increases the number of hospitalizations, ICU occupants, and deaths in the HERoS model. Furthermore, the precision of the model decreases when open data is used. Finally, it observed that neighborhoods that significantly differ in input population also significantly differ in model outcomes, showing the importance of the quality of an agent population for smaller resolutions. Conclusively, this research project shows that open data can serve as a viable alternative to microdata when it comes to synthesizing agent populations for epidemiological agent-based models.

Over the last couple of decades, refugee flows have been telling us a story of increasing numbers that are unlikely to go down anytime soon. The region of Eastern Africa only is home to 2.7 million people finding refuge in refugee settlements that are not often associated with permanency. UNHCR handbooks and many competent authorities state that refugee camps are of temporary character and as a result we observe a significant similarity in post-disaster food assistance and food assistance in a quasi-permanent settlement. The subsequent disparity between the way of organising food distribution and the needs of the population is the main motivation for the development of Food ATMs. This novel method of food distribution is based on allowing beneficiaries to determine how food is withdrawn to suit their own lives. A key point on how beneficiaries will use the Food ATM is a population's perception of the Food ATM. For instance, alleged food shortages might move people to hoard food and consequently undermine whole-system performance. Assumed similarities in literature on panic-buying events suggest that a combination of behavioural factors is a key vulnerability. Since a system of Food ATMs has not been implemented before, a conceptual design was made by 1) acquiring information through expert consultation on the design of an individual Food ATM and 2) the creation of an optimisation model forming an optimal network of Food ATMs. In order to analyse the interaction between a system of Food ATMs and beneficiaries, the current set of Operation Research (OR) models with stochastic or game theory-based formulation of food demand had to be expanded. For this purpose, concepts from research on panic-buying and emotion-related social interaction were used. This research contributes with the union of these concepts in a geo-spatially explicit Agent-Based Model, replacing current OR models' top-down formulation of demand with an agent-based bottom-up formulation. This enabled 1) increasing the understanding of interaction between a supplier of recipient of food assistance and 2) analysis of proposed operational policy in the context developing the Food ATM. Following 2), this thesis provides policy recommendations that have been made based on subsequent quantitative and qualitative investigation of model results. This thesis represents a first step to investigate the effects of individual behavioural factors in the realistic context of improving food distribution in the Bidi Bidi refugee camp in Uganda. New insights taught the relevance for further research, most notably on the relation with the bigger humanitarian supply chain, information diffusion between beneficiaries and the strategy for a population's adoption of the Food ATM.

Migrants are moving in great numbers towards urban areas as a result of urbanization and drivers such as political-, economical- and climate crises. The influx of new people brings new challenges for cities and municipalities to provide for a suitable environment for new and existing citizens whilst preventing adverse effects to occur. Current literature describes the dynamics of the city using effects such as segregation, gentrification and urban decay. However, the changing character of urban areas is not linked with the decision-making behavior of citizens in current literature. Furthermore, the dynamics of urban areas are rarely observed from a meso-scale perspective. This research aims to describe the relation between migration patterns and changes to the fabric of the city. A conceptual model on the dynamics involved in observing decision-making of citizens and observing changes to the city fabric is made. Using real-world data of the city of The Hague in The Netherlands, a Data-Driven Agent-Based model is made in Netlogo. Data analysis on the outcomes of the model using Python provide a better understanding of the impact of migration on the changes to the city fabric. The model output describes the impact on the city caused by the influx of migrants. More migrants entering the city can drastically increase homelessness or the "pushing out" of citizen groups because of housing shortages. This is most prevalent for the poorest citizens, since social rent housing is first to become scarce when more people enter the city. An increase in income inequality or a decrease in average migrant income results in even more pronounced housing shortages. The model simulates 6 policy interventions for the municipality of The Hague. The model shows that all policies that increase housing options significantly increases well-being of citizens. However, the policies that try to improve safety and health do not show an increase in well-being of citizens. By means of historical validation the model is tested for validity. The model is able to describe much of the decision-making behavior of migrants and citizens. However, for certain ethnicities, the model is currently unable to describe the decision-making process in such a way that it can represent the observed real-world moving patterns. A framework has been presented that can be used to better explain the impact of migration on cities. Furthermore, the conceptual model can be applied to different aspects of city dynamics, for instance the transport or energy sector. The ABM shows that current housing options and expected influx of migrants become problematic within the coming 10 years. An indication of the usefulness of proposed policy interventions is presented. More research and surveys are needed to better understand these mechanics and their relation to changes to the fabric of the city. Additional research into the changes to the neighborhood amenities and businesses as a result of changes in citizen composition can aid policy making and understanding the developments of the ever-changing city landscape.