On-going research

Daniel Felsenstein and Yair Grinberger, Department of Geography, Hebrew University of Jerusalem

The COVID-19 crisis is likely to have far-reaching impacts on urban economies beyond those relating to public health. Businesses and local governments will have to react to changes in labor supply, productivity, demand for products and incomes. Thus, the decisions that policy makers make today need to be understood in the context of their long-term effects. Cities are complex systems in which it is hard to predict the output of any specific input. Urban governance therefore requires tools for making informed decisions. The proposed research  develops a computational model of long-term urban dynamics in the wake of the COVID-19 epidemic, providing a support tool for decision making which will allow assessing both the short-term (health-related) impacts of policy options and their long-term (economic and welfare-related) effects. The main objective of the research is to develop a simulation model that will act as a training tool for policy makers-makers and public officials. This tool will enable them to model different COVID-19 contagion scenarios and understand the implications of proposed interventions. Model outputs should thus be understood as demonstrative rather than predictive.

To this end we adapt an existing agent-based model of urban dynamics following a disaster. This was originally developed for simulating the urban outcomes of large-scale shocks such as earthquakes and missile attacks (see Grinberger and Felsenstein 2016, Grinberger, Lichter and Felsenstein 2017, Felsenstein and Grinberger 2020). Currently we are adapting the model to simulate the urban impacts of the COVID-19 pandemic.

In the proposed research we aim to focus mainly on the long-term aspect of the epidemic. We suggest addressing the following questions relating to Israeli cities:

• Under which contagion scenarios (i.e. epidemic conditions and outcomes) do long-term urban impacts emerge?
• How are different populations affected by the epidemic over both the short and long term and what are the welfare-related implications of this?
• What long-term impacts can emerge from the implementation of short-term policies?
• What are the conditions under which different short-term and long-term policy measures are most effective?
• Under which conditions do equilibria states emerge and what are the characteristics of these states? What are the spatial patterns of these states?

In the proposed research, simulations will be generated for different contagion scenarios representing varying levels of severity, starting with the no-policy intervention, i.e. the “herd immunity” scenario. These results will provide a baseline against which policy outcomes will be compared. The policy scenarios will include:

• Short-term policies:
  • Limiting mobility of populations in risk
  • Limiting general mobility of the population
  • Applying strict quarantine measures within contagion hotspots
  • Closing specific functions in the city, e.g. the largest commercial areas, schools
• Long-term policies:
  • Providing financial support (to the entire population / by household income)
  • Offering subsidies to local businesses (e.g. property-tax reduction; general or by status)
  • Subsidizing wages for local workers
  • Subsidizing housing costs (generally / by household income)
  • Removing restrictions on land use change to encourage quick adaptation to changes in demand for housing/commercial services

The simulation model allows for comparing baseline outcomes relating to the structure of the urban economy with results from the policy scenarios. Each simulation will produce a database representing snapshots of the state of the population and the urban environment, i.e. a synthetic census, which will facilitate flexible querying and visualization (spatial and a-spatial). This will allow producing outcomes relating to a slew of variables such as size of the local commercial stock, the profitability of businesses, average wages, unemployment rates, demand for labour, population size, average household income, demand for housing, house prices, and shifts in the spatial patterns for each of these variables (e.g. the formation of new commercial clusters or the migration of households by income group).

We expect that producing a valid model will promote more informed decision-making by urban governments via providing accessible outputs and visualizations relating to hypothetical yet realistic policy scenarios. These will include identifying efficient policies over both temporal and spatial scales thus constraining the spread of the virus, increasing the resilience of urban residents and the commercial stock and expediting the rejuvenation of economic activity and the relief of constraints.

Publication from this research related to short term policies:

Grinberger, A. Y., & Felsenstein, D. (2023). Agent-based simulation of COVID-19 containment measures: the case of lockdowns in cities. Letters in Spatial and Resource Sciences, 16(1), 10.

Other previeus related publications:

Grinberger, A.Y., & Felsenstein, D. (2016). Dynamic agent-based simulation of welfare effects of urban disaster. Computers, Environment and Urban Systems, 59, 129-141.

Grinberger, A.Y., Lichter, M., & Felsenstein, D. (2017). Dynamic agent based simulation of an urban disaster using synthetic big data. In P. Thakuria, N. Tilahun & M. Zellner (Eds.) Seeing cities through big data: Theory methods and applications in urban informatics (pp. 349-382). Springer: Cham.

Felsenstein D and Grinberger A.Y (2020).Cascading effects of a disaster on the labor market over the medium to long term, International Journal of Disaster Risk Reduction, 47, https://doi.org/10.1016/j.ijdrr.2020.101524