Abstract Recent studies suggest that the traditional determinants of housing wealth are insufficient to explain its current inequality levels. Thus, they argue that efforts should focus on understanding institutional factors. From the perspective of complex adaptive systems, institutions are more than the ‘the rules of the game’, they also consider the interaction protocols or the ‘algorithm’ through which agents engage in socioeconomic activities. By viewing markets as complex adaptive systems, I develop a model that allows estimating how much housing wealth inequality is attributable to the market institution. It combines virtues from two different modeling traditions: (1) the microeconomic foundations from overlapping-generation models and (2) the explicit interaction protocols of agent-based models. Overall, the model generates prices and housing inequality endogenously and from bottom-up; without needing to impose assumptions about the aggregate behavior of the market (such as market equilibrium). It accounts for economic and institutional factors that are important to housing consumption decisions (e.g., wages, consumption of goods, non-labor income, government transfers, taxes, etc.). I calibrate the model with the British Wealth and Assets Survey at the level of each individual household (i.e., ~25 million agents). By performing counter-factual simulations that control for data heterogeneity, I estimate that, in the United Kingdom, the decentralized protocol interaction of the housing market contributes with one to two thirds of the Gini coefficient. I perform policy experiments and compare the outcomes between an expansion in the housing stock, a sales tax, and an inheritance tax. The results raise concerns about the limitations of traditional policies and call for a careful re-examination of housing wealth inequality.

中文翻译：

---

去中心化的市场和住房财富不平等的出现

摘要 最近的研究表明，住房财富的传统决定因素不足以解释其当前的不平等程度。因此，他们认为努力应该集中在理解制度因素上。从复杂适应系统的角度来看，制度不仅仅是“游戏规则”，它们还考虑了代理参与社会经济活动的交互协议或“算法”。通过将市场视为复杂的适应性系统，我开发了一个模型，可以估计市场制度造成的住房财富不平等程度。它结合了两种不同建模传统的优点：（1）重叠生成模型的微观经济基础和（2）基于代理的模型的显式交互协议。全面的，该模型自下而上地产生价格和住房不平等；无需对市场的总体行为（例如市场均衡）强加假设。它考虑了对住房消费决策很重要的经济和制度因素（例如，工资、商品消费、非劳动收入、政府转移支付、税收等）。我使用英国财富和资产调查在每个家庭（即约 2500 万代理）的水平校准模型。通过执行控制数据异质性的反事实模拟，我估计在英国，住房市场的分散协议交互贡献了基尼系数的一到三分之二。我进行政策实验并比较住房存量扩张之间的结果，销售税和遗产税。结果引发了对传统政策局限性的担忧，并呼吁仔细重新审视住房财富不平等。