Roof Or Real Estate? An Agent-Based Model of Housing Affordability in the Netherlands
Introduction
Housing affordability is frequently misdiagnosed as a simple shortage of physical structures, yet recent economic modeling suggests the crisis is fundamentally monetary and financial in nature. This distinction is critical for policymakers, researchers, and housing professionals seeking effective solutions in developed economies. The academic paper "Roof or real estate? An agent-based model of housing affordability in The Netherlands" by Ruben Tarne and Dirk Bezemer provides a rigorous analysis of this phenomenon, challenging traditional neoclassical views that prioritize construction supply above all else.
By utilizing an agent-based model calibrated to Dutch data, the authors demonstrate that financial policies, such as loan-to-value caps and interest rate adjustments, can be as effective as massive construction boosts in tempering house price peaks.The Monetary Nature of Housing Affordability
Standard economic intuition often treats housing shortages as a physical quantity problem: too many people chasing too few homes. In this neoclassical view, money and finance are merely derivatives that facilitate demand, meaning that enduring housing affordability problems should not occur unless government regulation obstructs market adjustment. However, this perspective ignores the reality that capitalism is a monetary and financial system. Homes are not just durable consumption goods; they are investable, leveraged assets.
The authors ground their analysis in capital circulation theory, drawing from Marx, Keynes, Schumpeter, Kalecki, and Minsky. This theoretical framework posits that individually rational behavior can lead to aggregate outcomes like instability and crises.
In the context of real estate, over-accumulation of capital finds an outlet in land and housing assets, driving up prices and creating housing affordability deficits for asset users, such as renters and first-time buyers.
The document argues that a "housing shortage" is better understood as a shortfall in purchasing power relative to ask prices, rather than a deficit in physical units. This monetary definition of housing affordability highlights how credit expansion and investor money inflows can drive prices independently of physical supply trends.
Institutional Context of the Dutch Market
To understand the specific dynamics of the Dutch crisis since 2015, the paper details several institutional features that have constrained supply and fueled financial demand. A primary factor has been the decline of public building corporations, which historically provided affordable housing.
Following market liberalization in the 1980s and 1990s, these corporations were restricted, taxed, and forced to sell off stock, reducing their capacity to respond to the post-2015 housing affordability crisis. Consequently, the share of social rental properties has stagnated or fallen, moving corporation housing out of reach for many lower- and middle-income households.
Conversely, the role of private investors has expanded significantly. Wealthy investors, both natural persons and companies, have increased their portfolios, often purchasing properties without mortgages. Data indicates that the share of cash purchases doubled from 10% in 2006 to 20% in 2017, with even higher rates for apartments where first-time buyers typically compete.
This influx of private money is used primarily to trade housing and increase prices, not to build new supply. Furthermore, policy decisions between 2012 and 2022 liberalized the rental market and promoted foreign investment, redistributing housing units into the private sector and increasing rents. These institutional shifts underscore that housing affordability is influenced heavily by who holds the assets and how they are financed, not just by how many roofs exist.
Methodology: Agent-Based Modeling Approach
The study employs an agent-based model (ABM) to analyze the trade-offs between different policy interventions. Unlike econometric analyses that assume equilibrium, ABMs allow for the simulation of heterogeneous agents interacting in historical time, producing emergent properties like cycles and tipping points.
The model includes 10,000 households of four types: renters, owner-occupiers, buy-to-let (BTL) investors, and social housing residents. It also features a commercial bank that issues mortgage loans based on pro-cyclical loan-to-value (LTV) rules.
The model is calibrated using data from the 2017 wave of the European Central Bank’s Household Finance and Consumption Survey (HFCS) and validated against external data sources such as the OECD house price index and transaction records from the Dutch Land Registry.
While the model simplifies certain aspects, such as excluding interest-only loans and family support transfers, it successfully replicates key stylized facts of the Dutch housing market, including the long housing cycle of approximately 22 years and the pro-cyclical behavior of BTL investors.
This methodological rigor allows the authors to trace the transmission mechanisms of policy shocks in ways that traditional models cannot, providing deeper insights into housing affordability dynamics.
Simulation Results: Policy Trade-Offs
The core of the research involves simulating three types of exogenous shocks to estimate their effects on house price peaks: changes in housing supply, interest rates, and loan-to-value norms. The goal was to determine what magnitude of change is required to achieve a 10% reduction in the cyclical peak of the house price index.
The results reveal significant trade-offs. To achieve a 10% reduction in peak prices through supply-side measures alone, the ratio of private properties to households would need to increase from 69% to 74%. In real-world terms, this is equivalent to adding approximately 420,000 residential units to the Dutch housing stock. Given the spatial, administrative, and environmental constraints on construction, such a scale of building is challenging to achieve in the medium term.
Alternatively, the same 10% price reduction could be achieved by reducing the bank’s loan-to-value cap from 96.9% to 93.3%, or by increasing the mortgage interest rate from 4.0% to 5.4%. These findings suggest that financial and monetary policies are viable alternatives to supply responses in managing housing affordability.
The model highlights that banks’ pro-cyclical lending dynamics are central to understanding why construction alone may be less effective than anticipated. By constraining the credit available to investors, particularly BTL households who drive price peaks during upswings, financial policies can temper prices without requiring massive physical expansion.
Implications for Housing Affordability Policy
The document offers several policy recommendations based on its findings, emphasizing the need to restrain the inflow of money into the housing market for financial returns. First, the authors suggest implementing a tax on the value increase of residential real estate, akin to a Henry George land value tax.
This would discourage capital-gain-motivated investments without discouraging investment in housing as a dwelling. Second, addressing the inflow of investor money into short-term rentals, such as Airbnb, is crucial.
Municipalities have legal tools to restrict these practices, but they are often underused. Third, recapitalizing and strengthening housing corporations can help reorient the flow of money towards affordable housing for lower- and middle-income segments.
These recommendations align with the broader conclusion that housing affordability is not solely a construction issue. While increasing supply is beneficial, it is not the only lever available. Financial policies, such as LTV caps and interest rate adjustments, directly impact the purchasing power of different household types.
The model shows that BTL investors are more sensitive to these financial constraints than first-time buyers, making targeted macroprudential policies effective in cooling speculative demand. This approach recognizes that housing affordability is a monetary experience, defined by the balance between income, savings, and borrowing on one hand, and purchase prices and rents on the other.
Conclusion
The analysis presented in "Roof or real estate?" provides a compelling case for rethinking how we address the global housing affordability crisis. By shifting the focus from purely physical supply shortages to the monetary and financial drivers of house prices, the authors offer a more nuanced understanding of market dynamics.
The agent-based model demonstrates that financial policies can be powerful tools in reducing cyclical price peaks and improving access for first-time buyers. For researchers and policymakers, this underscores the importance of integrating financial regulation into housing strategy.
As the debate continues, the insights from this study remain invaluable for designing effective, equitable solutions to ensure that housing affordability is maintained in increasingly financialized markets. The document serves as a critical resource for anyone seeking to understand the complex interplay between credit, investment, and shelter in modern economies.
Also read: Integrated Urban Renewal in the Netherlands