- documents de travail
Registering land title is costly and time-consuming. This is a common picture all over the world. In France, the cost of a real estate transaction accounts approximately for 7.3 % of the property value and it takes on average 42 days to achieve it, according to the World Bank (2019). In Germany, the respective numbers are 6.6 % of the property value and 52 days. There are differences between countries due to many factors: regulations, taxes, insurance costs, and real estate agency fees. The paper focuses on transaction costs, leaving aside insurance costs and taxes. In this regard, the Blockchain is a candidate to replace or complement traditional land registration systems. The Blockchain, a particular type of distributed ledger technology (DLT), is a secure, transparent information storage and transmission technology which has numerous advantages, in terms of cost, duration and production of immutable information. The reduction of transaction costs through the diminishing role of trust intermediaries is a key proposition of the proponents of the blockchain. Yet, the Blockchain encounters numerous limits, both technical and legal, that might explain why the applications of this technology to real estates are scant. In the paper, we view the Blockchain as a cost-saver mechanism albeit one that can potentially be jeopardized by hacking. We look at the desirability of keeping traditional registration, in a context where blockchain registration is available. In our setup, there is a priori an infinite set of land titling systems, each characterized by its quality (namely, the probability for a given land that there is no forfeiture) and the unit transaction cost. We define an optimal land titling system as one that maximizes the aggregate expected value of land net of transaction costs (the expected value of a land is computed by taking into account the probability of forfeiture). Choosing a more secure land titling system increases the expected values of land but comes at a cost that can deter people to register their properties. Thus, there is a trade-off between security and the number of registered transactions. We find that under reasonable assumptions, resorting to the blockchain is optimal. Yet, it is never socially optimal to replace the traditional public registration by the Blockchain. This does not mean, however, that the traditional public registration system must not be adapted. We indeed show that the optimal quality of protection provided by traditional registration must be either sufficiently high (and higher than that of the blockchain), or low enough (and lower than that of the blockchain). To understand the thrust of this result, consider an economy where land values are often large. Introducing the Blockchain benefits owner of lands whose values are relatively low because it decreases the probability of forfeiture at a cost lower than that of the traditional land registration system. But since the owners of these lands are few and the values of the latter are low, using the Blockchain results in the mild increase in the aggregate social welfare. For the owners of lands whose values are large, it is better to use a high-quality traditional land system rather than the Blockchain. That is because, the increase in the expected value of lands with this system is higher than that obtained with the Blockchain, and still more than compensates the rise in the unit cost.