The Macroeconomic Effects of Regulating Household Leverage: A Mesoeconometric Approach
Working Paper Series no. 1051. We develop a mesoeconometric framework to quantify the macroeconomic effects of policies that target household leverage. We focus on borrower-based macroprudential measures (BBMs), which regulate lending standards at loan origination. The approach first identifies lending standards shocks in a structural VAR-IV framework and then exploits heterogeneous borrower responses to BBMs to isolate the policy-induced component of those shocks. We apply this framework to France, where caps on debt-service-to-income (DSTI) ratios and loan maturities were introduced in 2019. We show that lending standards shocks have persistent and economically meaningful effects on housing and credit dynamics. By contrast, the BBM-induced component has statistically significant but economically moderate effects. In credit markets, the measure raises borrowing interest rates by around 0.15 to 0.20 percentage points and slows the growth of real outstanding housing credit by up to 0.8 percentage points between 2022 and 2023, a period during which outstanding credit fell by as much as 8%, mainly as a result of higher interest rates. The measure also reduces house price growth by 2 to 3 percentage points at its trough. We find no significant impact on real residential investment, household income, or real GDP. These results indicate that BBMs primarily affect the riskiest component of the housing credit market without generating broad macroeconomic spillovers, highlighting their role as targeted tools for enhancing financial resilience.
Macroeconomic effects of the borrower-based measures in France
Non-Technical Summary
Since the Global Financial Crisis, policymakers have paid increasing attention to household leverage as a potential source of financial instability. When housing credit expands too rapidly and lending conditions become too loose, households can become more vulnerable to adverse shocks, while the broader economy becomes more exposed to housing-market downturns. In France, these concerns led to the introduction of macroprudential measures in the form of a regulation on housing loan lending standards, also referred to as borrower-based measures (BBMs). First introduced as a recommendation in December 2019 and later made legally binding in 2022, these measures limit the debt-service-to-income (DSTI) ratio and the maturity of new housing loans (distinct from other types of BBMs such as loan-to-value caps), while allowing banks a flexibility margin.
This paper studies the macroeconomic effects of such measures. Estimating these effects is challenging because borrower-based measures are typically introduced in response to prevailing macro-financial conditions, which complicates the identification of their causal aggregate impact. To address this issue, the paper develops a new macroeconometric framework that exploits a central feature of borrower-based measures: their effects are concentrated at specific points of the housing loan lending standards distribution. Loans close to the regulatory thresholds are directly affected, whereas those sufficiently distant are not. This creates distinctive distributional shifts. The paper builds on this insight to propose a new two-step framework that first estimates the macroeconomic effects of changes in lending standards in a vector autoregression model, and then uses the specific distributional shifts induced by borrower-based measures to isolate their aggregate effects. We refer to this methodology as a mesoeconometric strategy, which lies at the intersection of macroeconometric and microeconometric analysis.
The paper first establishes that changes in housing loan lending standards have economically meaningful effects on the French housing and credit cycle. A tightening raises housing loan borrowing interest rates, slows housing credit growth, and lowers house price growth. These effects are persistent and sizable. By contrast, the impact on broader macroeconomic variables is more limited , with no clearly significant effects. In particular, the estimated responses of real GDP and household income remain small. This suggests that lending standards changes operate primarily through housing and housing credit markets rather than via large aggregate spillovers.
Building on these results, the paper then quantifies the specific contribution of borrower-based measures. The findings indicate that their effects are statistically significant but economically limited. The BBM-induced component of lending standards tightening increases housing loan borrowing interest rates by about 0.15 to 0.20 percentage points at its peak, slows housing credit growth by around 0.8 percentage points, and lowers house price growth by about 2 to 3 percentage points at its trough. The estimated contribution remains limited relative to the overall market slowdown observed over the period. At the same time, the paper finds no statistically significant effects on real residential investment, household income, or real GDP. Figure 1 illustrates these estimated effects.
Overall, the findings suggest that borrower-based macroprudential measures can enhance the resilience of housing credit markets without generating large adverse effects on aggregate activity. In the French case, the measures appear to have worked as intended as a targeted instrument affecting the riskiest segments of the housing credit market, while leaving broader macroeconomic dynamics largely unaffected. Beyond the French application, the paper also proposes a tractable framework for evaluating policies that act through targeted constraints in the distribution of credit.
Keywords: Macroprudential Policy, Lending Standards Shocks, Housing Market, Borrower-Based Measures
Codes JEL : E44, G21, G28
Download the full publication
Updated on the 6th of July 2026